University of Southern California Dissertations and Theses

Recruitment of family physicians into rural California: Predictors and possibilities

(USC Thesis Other)

Recruitment of family physicians into rural California: Predictors and possibilities

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content RECRUITMENT OF FAMILY PHYSICIANS INTO RURAL CALIFORNIA: PREDICTORS AND POSSIBILITIES Copyright 2005 by Inger C. Nocella A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (PUBLIC ADMINISTRATION) December 2005 Inger C. Nocella Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UMI Number: 3220141 Copyright 2005 by Nocella, Inger C. All rights reserved. INFORMATION TO USERS The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleed-through, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion. ® UMI UMI Microform 3220141 Copyright 2006 by ProQuest Information and Learning Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code. ProQuest Information and Learning Company 300 North Zeeb Road P.O. Box 1346 Ann Arbor, Ml 48106-1346 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DEDICATION To Cali and Lexi - may you find your passion in life and be confident in your inherent gifts. Thank you for understanding and supporting mommy. You are beautiful inside and out. I love you. To my deceased father - this was for me, because of you. I had no choice - Sears long ago stopped selling popcorn, so this was the only other career alternative. May it bring some honor back to the name. To Joe - the wind beneath my wings, albeit sometimes turbulent. Thank you - you’ve earned this too. To Peter - thank you. I will forever be in awe of your brilliance and your passion. You are an amazing and talented individual. To Ric - what a ride. Thank you for the opportunity and for the belief in me - even when I didn’t have it myself. To my rural colleagues and friends - thank you for the education. You have each touch my life and taught me much. I’m indebted to that gift. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ACKNOWLEDGEMENTS California based Family Medicine Residency Program Directors - Thank you for your help, your data, and your passion. Coastal Research Group - Bill and Nancy Burnett - tiredless dedication. Thank you for all your support, encouragement, and for having kept a database like no other. J. Scott Chrisman - 1 value your maps - but value your friendship much more. Ricardo G. Hahn, M.D. Faculty and staff of the USC Department of Family Medicine Georgeann Novak, MD - How I wish I could clone you for every rural community - thanks for setting the standard. Peter Robertson, PhD LaVonna Blaire Lewis, PhD and Lyndee Knox, PhD - Thanks for being a great committee. Michael Cousineau, DPh - You got me going and got me done by convincing me to do what I know. Chris Feifer, DPh - my muse Pam Diamond, PhD - who single handedly got me through statistics - not an easy task. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TABLE OF CONTENTS DEDICATION ii ACKNOWLEDGEMENTS iii LIST OF TABLES vi LIST OF MAPS vii ABSTRACT viii CHAPTER 1 PROBLEM AND INTRODUCTION 1 Background and Overview 1 Graduate Medical Education Financing 13 Balanced Budget Act (BBA) of 1997 15 State of California Initiatives 17 2 REVIEW OF THE LITERATURE 21 The Medical School Literature 23 Medical School Characteristics 29 Literature on Individual Characteristics 34 Literature on Residency Program and Fellowship Characteristics 37 Community Characteristics 40 Research Questions 41 3 METHODOLOGY 43 Quantitative Analysis 43 Measures 52 Qualitative Analysis 73 4 RESULTS 74 Quantitative Results 74 Qualitative Analysis 95 Summary of Results 101 iv Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 5 CONCLUSIONS AND IMPLICATIONS 103 Summary and Conclusions 103 Residency Program Structure 105 Discussion 109 Implications 117 Limitations 121 Future Research 122 BIBLIOGRAPHY 124 APPENDICES 134 A RESIDENCY PROGRAM SURVEY 134 B RURAL HOSPITAL ADMINISTRATOR SURVEY 138 C MEDICAL SCHOOL DEAN SURVEY 142 D LETTER TO RURAL HOSPITAL ADMINISTRATORS 143 E LETTER TO MEDICAL SCHOOL LIBRARIANS 144 v Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF TABLES TABLE 1 IOM Model of Monitoring Access 6 2 Chart of String Variables 53 3 Chart of Variables with RUCA Scores 53 4 Chart of Medical School and Residency Independent Variables 53 5 Chart of Gender Variables 54 6 Visual of Methodology 55 7 Missing High School Cases 56 8 Missing Medical School Cases 58 9 Missing Values by Variable 66 10 Findings - Individual Characteristics 75 11 Findings - Medical School Characteristics 76 12 Findings - Residency Program Characteristics 76 13 Findings - Medical School Interactions 76 14 Findings - Residency Program Interactions 77 15 Production of Rural Vs. Urban Physicians by California Medical Schools 104 v i Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF MAPS MAP 1 HMO Penetration in California 2 Rural Urban Commuting Areas Map 3 Location of Database graduates in California Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ABSTRACT Purpose The impact of personal and medical school characteristics on a physician’s decision to practice in rural has been well documented, but little is known regarding the impact of residency training. This study looks at graduates of family medicine residency programs in California to determine what individual, medical school, residency program, and pipeline characteristics were significant in the practice decisions of family physicians. Method This study used a cross sectional design utilizing archival and survey data, including a sample of 1,410 graduates from California family medicine residency programs from 1998-2002. The database provided information on the dependent variable, practice location, and certain independent variables. Data for the remaining independent variables were collected from medical school and residency program surveys and other reference materials. Variables were coded as dichotomous and five logistic regressions were conducted. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Results Significant variables for specific categories are: graduating from a rural high school (p=.000), an admissions policy focusing on rural (p=.005), a specialized rural curriculum (p=.022), whether a residency program perceived itself as rural (p=.012) and if the residency program director trained or practiced in a rural area (p=.000). Along the pipeline, the same variables were significant, plus an additional two medical school variables. When compared as groups, the variables that impacted a family physician from rural differed from those from urban areas. Finally, when interaction terms were created between each dependent variable and high school location, no interaction effects occurred at the medical school level, yet three significant interactions occurred at the residency program level: required rural rotation (p=.023), elective rural rotation (p=.049) and program director who trained or practiced in rural (p=.011). Conclusion The importance of coming from rural areas as a predictor of a family physician’s practice decision is well documented. In order to address maldistribution in California, medical schools must recruit individuals from rural Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. areas and provide specialized rural education. And, despite the State and Federal policy interventions focusing on residency program structure, the results indicate that residency program culture and commitment to rural is what is matters in producing rural physicians. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 1 PROBLEM AND INTRODUCTION Background and Overview With 75% of its land mass designated as “rural,” and 3.8 million residents living in those rural areas, the State of California, despite popular perception, is a rural state (www. csrha. or elfactsheet. html. accessed April 3,2003). Like many states, however, California struggles with how to provide the necessary physician coverage in these rural areas. From the standpoint of physician recruitment in rural areas, efforts have been made, at the local, state, and federal levels, to address the issue of physician distribution (COGME, 1998,2000). Family physicians and family medicine residency programs are the primary focus of such interventions, possibly because it has been shown that the number one indicator of whether a physician will practice in a rural area is whether the physician is trained as a family physician (COGME, 2000) and because family physicians are more than twice as likely that physicians with other specialty training to select small rural areas for practice (Rosenblatt et al., 1992). With 80,197 non-military family physicians nationwide, and 9,421 in California, one might wonder why there is any difficulty in obtaining family physicians to practice in rural areas of the state. However, in a survey conducted by the American Academy of Family Physicians, of the 2,744 respondents in California, only 5% stated that their practice was located in a rural area (AMA, 2001; www.aafp.org/x759.xml. accessed May 16,2003). Urban areas 1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. of the state have approximately 460 residents per physician, as compared with the published data of 935 residents per physician in rural areas of the state (California 2002). This shortage of family physicians in rural areas represents a major health problem for California, and for the United States (COGME, 1998). In that same report, the Council on Graduate Medical Education (“COGME”) stated that “geographic maldistribution of health care providers and service is one of the most persistent characteristics of the US health care system” (COGME, 1998, pi 5). The shortage of primary care physicians and specifically family physicians in rural California, let alone rural America, impacts health outcomes (Shi, 1992; Starfield, 2004; Starfield, 2003; Macinko, 2003). There is a direct association between greater primary care to population ratios and overall mortality, including deaths from cancer, and heart disease as well as infant mortality rates including neonatal deaths and low birth rates (Shi, 1992). In a study of eighteen OECD countries over twenty eight years, Macinko et al found that a strong primary care system with practice characteristics that include geographic regulation, longitudinality, coordination and community orientation was associated with better population health outcomes (2003). Finally, Barbara Starfield has shown that strong primary care systems provides for equity in healthcare, that family physicians are better able to address care for individuals with co-morbidities, and that primary care is better than specialty care for population health outcomes (2003, 2004). 2 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The literature on health outcomes is more than the success or failure of a specific clinical intervention. Health outcomes research focuses not just on the patient’s clinical outcome or the population’s health status, but also on issues that are of the greatest concern to the patient or population, be they issues of satisfaction, quality of life, cost, convenience, the prevalence of chronic diseases, death rates, personal well-being and/or the ability to function. flittp://www. academvhealth.org/publications/healthoutcomesreport.html; accessed May 25, 2005). Health outcomes such as those listed above, are impacted by issues of: 1) access, which is often understood in terms of utilization or use of services; and 2) quality (Docteur, 1996). Aday and Anderson (1981) have delineated the term “access” into three categories: predisposing factors (e.g., income, education, etc.), enabling factors (e.g., presence of providers, insurance coverage, etc.), and an individual’s personal need for care. Each of the variables within these three categories impacts one’s utilization of healthcare services. However, it is not enough to look at whether health care services are used; we must also look at the issue of quality. Whether someone is able to get to and use a service is only part of the equation when it comes to outcomes. The other factor that impacts health outcomes is quality. Defining, let alone developing measures of “quality” that truly indicate quality based on the provider, patient, and payer perspectives is of current significant interest, importance, and emphasis at a national level. This is in part because the definition of quality varies depending on one’s vantage point. Providers, patients, payers and regulators each have a different perspective on the 3 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. definition of quality. From the providers’ perspective, “medical quality is the degree to which health care services, systems, and supplies for individuals and populations increase the likelihood for positive health outcomes” fhttp://www. acmq.org/policies/policiesland2.pdf. accessed May 29,2005). Similarly, according to the American Academy of Family Physicians (AAFP), quality care “should produce optimal health outcomes in the most cost-effective manner and result in high patient satisfaction... it is based on accepted principles of medical science and proficient use of appropriate technological and professional resources” (AAFP 2002) From the perspective of a health plan or payer, all too often quality comes down to issues of cost and measurement. Although the tools exist to measure quality, they are not uniformly applied on a national basis, nor do the information systems exist to support such routine measurement and monitoring (Schuster et al., 1998). Even if it did exist, at the level of the individual physician or mid level provider, there is little consensus as to whether that which is measured really is relevant and meaningful. For example, in an individual rural provider’s office, they may have 2,000 patients, that are either insured by one of five health plans, plus the cash paying patient population. When measuring the quality of treating a certain condition in this individual provider’s patient population, by health plan, there is a problem with small numbers, resulting in findings that are not relevant. Different health plans measure different things, depending on the scope of services they 4 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. cover and the information available to them, amongst other things (Ladenheim, 2000). During the past decade, there has been growing recognition of the patient’s perspective on quality. The concept of patient centered care - where the patient’s preferences, needs, culture, and values are all factored into the care in such a way that the patient is the source of control over their own health, has become one of the six key aims of the Institute of Medicine in their 2001 recommendations for health care improvement. Don Berwick, President of the Institute for Healthcare Improvement, provided a patient focused definition of quality in 2004 when he defined quality as “to have health care with no needless deaths, no needless pain or suffering, no unwanted waiting, no helplessness, and no waste” (IOM, 2004, 6). Consensus on the definition of quality is evolving, to the point that providers, patients and payers are seeing the IOM’s definition of quality as a unifying definition amongst the various perspectives (Sipkoff, 2004). The IOM defined quality, in 2001, as “the degree to which health services for individuals and populations increase the likelihood of desired health outcomes and are consistent with current professional knowledge” (IOM, 2001) Having a unifying definition of quality moves us closer to addressing the large gaps between the care that people should and actually do receive. “Americans tolerate more variation in the performance of their health care system than they would ever put up with in the performance of an airline, a computer 5 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. company, or even the manufacturer of their breakfast cereal. We have to change that” (Brook, 1997). Therefore, in looking at issues of access or utilization, one must also look at issues of quality. Together, access and quality impacts health outcomes both at individual and population levels. The Institute of Medicine (IOM) linked access and quality to outcomes in 1993 with the publishing of their Model for Monitoring Access. This model looks first at the barriers to achieving the desired health outcomes, with barriers breaking into three categories, those being structural, financial, and personal health access barriers (IOM, 1993). Next in the model is utilization, or the use of health services. Can individuals access providers for offices visits or procedures? Third in the IOM model is the category of mediators. Mediators represent certain quality indicators - such as whether the care was appropriate and efficient, as well as whether the patient adhered to the prescribed treatment regimen and whether the provider rendering the care provided quality of care. The final stage in the model is that of outcomes which as addressed earlier, includes many variables of health status and equity of services. Table 1 . Barriers' IOM Model of Monitoring Access y Mediators (Jutcome • Quality of providers • Health status • Appropriateness of • Equity of care services • Efficacy of treatment • Patient views • Patient adherence of care • Personal/family • Structural • Financial IOM, 1993 Visits Procedures Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Working backwards in this model, when in comes to rural health, there are many health outcomes that are worse than their urban or suburban counterparts. Research in rural health has demonstrated that people who live in rural areas tend to be poorer, older, and sicker (American College of Physicians, 1995; Hartley et al., 1994; Braden and Beauregard, 1994; Schur and Franco, 1999). We know that people residing in rural areas report more chronic diseases and see themselves as overall in poorer health than their urban counterparts (Ricketts, 2002). Chronic illnesses are more prevalent and death rates due to those chronic conditions are greater in rural areas, with death rates for children, young adults, and the elderly higher than those same populations residing in suburban areas (Eberhardt et al., 2001). In California, there is a higher death rate per 100,000 due to heart disease (245 vs. 211) and cancer (214 vs. 171) than in urban areas (California LAO, 2002). Rural residents are more likely to engage in risky health behaviors, with higher rates of alcohol use, smoking, and obesity (Eberhardt, et al., 2001), yet are more likely to experience delays in obtaining any medical care for these or other conditions due to being uninsured and other financial barriers (Schur and Franco, 1999) let alone other issues such as transportation. Ultimately, when it comes to health outcomes, “the most important consideration is whether people have the opportunity for a good outcome - especially in those instances in which medical care can make a difference. When those opportunities are systematically denied to groups in society because they face barriers to care, there is an access problem that needs to be addressed” (IOM, 2003,38), 7 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The issues that impact health outcomes are multi-factorial, even though they can be categorized as issues of quality and issues of access or utilization. As mentioned above, rural areas have a multitude of health outcome challenges as compared to their suburban and urban counterparts. This is due to having additional barriers (structural, financial, and personal) and less access to medical professionals for office visits, emergency care, specialty services, and procedures, be they preventive services, therapeutic services, or diagnostic services. In addition, rural areas have challenges in the mediating, or quality, factors that impact health outcomes. For example, if you have only one physician and one midlevel in a rural community of 5,000 people, how do these providers insure cultural competency, keep their medical skills up to date, monitor and evaluate the efficacy of their treatment plans, and conduct outreach for patient adherence along with seeing their daily patient case load? Information technology is one vehicle to assist with this, and yet rural areas, if they have access to broadband or DSL, and are ill equipped to fund, let alone purchase, install, and maintain information systems that can provide clinical decision support and population health management. A shortage of providers, although only one of the many factors that impacts health outcomes from both an access and a quality perspective, is a critical factor since providers are the backbone of any healthcare system. Because rural communities have a shortage of providers, providing and maintaining a sustainable level of healthcare has been a pervasive challenge (Rourke, 1997; Hutten-Czapski, 8 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1998; Tepper & Rourke, 1999; Ramsey et al., 2001). In fact, the loss of a provider can undermine an entire rural health system (Rosenblatt, 2004). A shortage of providers in rural areas represents a structural barrier that impacts the overall use of services as well as the quality mediators for health outcomes. A shortage of providers is one of many structural barriers that impact the utilization of healthcare services and hence individual health outcomes, as is transportation, and how the health care system is organized. The individual kinds of structural barriers are interrelated. For example, transportation is often a major issue in rural areas. It is not uncommon for rural residents to have to drive for 45 minutes to 3.5 hours to obtain labs or x-rays, let alone specialty services, assuming the individual has access to a car. If a certain service isn’t available in a region - either due to not having a provider with that skill set or due to the insurance company not being willing or able to contract with the local provider, care must then be rendered outside the region. A lack of providers, coupled with problems with transportation, only exacerbates the negative health outcomes. Rural areas are less likely to have formal safety net providers (Hartley & Gale, 2003). Provider availability and structural barriers are not the sole reason for poorer health outcomes in rural areas, but are a significant component of it. Another barrier that impacts the health outcomes of rural residents is financial in nature. Insurance coverage, reimbursement levels, and the extent of public financial support all impact the extent to which health care is available and accessible, and hence the individual and population health outcomes (IOM, 1993). 9 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The issues of health insurance, or lack thereof, and physician maldistribution are linked, since without insurance, physicians must provide care that is not reimbursed, which is a significant financial disincentive (COGME, 1998). In California, sixty percent of the rural population has an income below 300% of the poverty level, as compared to forty seven percent in the urban regions (CHIS, 2002). There are currently eleven rural counties with no penetration of managed care. All insurance is either publicly funded, or part of a PPO product that is provided by a large employer, such as the local government, state prison, or Bureau of Land Management, to the extent it exists in a given area. One in five rural Californians under 65 years old have no health insurance coverage and those that do often cannot afford health care services because of an inability to pay for high co-payments and deductibles or because of the limitations of their health benefits (CHIS, 2002; Hartley et al., 1994). Health care funding in general, let alone in rural regions, has become more challenged during the past ten years, with 76% of rural hospitals in California losing money on operations in 2000 (California) The last category of barriers to utilization of healthcare services is “personal” - meaning personal characteristics that impact one’s ability to avail themselves of medical visits or procedures, and hence impact their individual health status. Examples of personal barriers include cultural barriers, attitudes, poverty, and education (IOM, 1993). The impact of poverty and education on health outcomes is well documented (Wagstaff, 2002; Krieger et al., 2005). 10 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. F igu re 1 Many Flu rat Counties Lack HMOs Rural Counties and HMOs Number of KunOMr of Couravwida HMOt Counttcs ■ I N c* n o > n B One 12 \ : \ | Two or More 7 Map 1. HMO Penetration in California In California, more people are in poverty in rural areas than in urban areas (18% vs. 14%) and more people are unemployed in rural areas than in urban areas (9.3% vs. 6.8%) (California Legislative Analysts’ Office, 2002). Fewer rural than urban teens graduate high school and go to college, even though they are academically about the same, per standardized tests (USDA, Nov. 2003). Rural residents tend to hold dear values of individualism and self-reliance. As such they tend to seek support from informal networks and seek are from the formal healthcare system only when seriously ill (Casey, 2001). On the whole, individuals in rural have 1 1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. greater personal barriers to achieving health outcomes than their urban counterparts. All of these barriers to access impact the three fundamental indicators of poor quality in health care, those being “overuse, underuse, misuse” of health services. It is easy to see how health services could be underused or misused amidst the challenges and barriers that rural populations face. When one then considers Don Berwick’s definition of quality - “to have health care with no needless deaths, no needless pain or suffering, no unwanted waiting, no helplessness, and no waste” (IOM, 2004, 6), rural Californians are clearly far away from this. Key to addressing these barriers is the presence of sufficient health care providers who have the resources, skills, and knowledge to provide a quality healthcare environment for their rural communities for episodic, longitudinal, and preventive care. This study attempts to add further knowledge to the literature that exists to date on the individual, medical school, and residency program characteristics, as well as community dynamics, that either light a fire or nurture a spark in an individual to become a family physician in rural California, with the ultimate goal being to ameliorate some of the access and quality issues that rural Californians face every day and that impact their health status and health outcomes. Historically, there have been multiple policy interventions in an attempt to address the issue of physician shortage and maldistribution. Several initiatives have been put in place to potentially increase the number of rural physicians. 12 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Specifically at the Federal level, there was the Balanced Budget Act of 1997 (BBA) and its subsequent Balanced Budget Refinement Act of 1999 (BBRA) and at the State of California level there was the passage of the Song-Brown legislation in 1973 as well as the execution of a Memorandum of Understanding between the UC System and the Office of Statewide Health Planning and Development in 1993. The next section will first describe the history and methodology of GME and then describe the three above initiatives in more detail. Graduate Medical Education Financing With the passage of Medicare in 1965, the government made a policy decision to pay for medical education as well as patient care. It was seen as a vehicle to enhance the quality of patient care as well as possibly an interim step “until the community undertakes to bear such education costs in some other way, that a part of the net cost of such activities (including stipends of trainees, as well as compensation of teachers and other costs) should be borne to an appropriate extent by the hospital insurance program” [Medicare] (Committee on Ways and Means, 1965). Prior to this time, physicians rarely pursued additional training, and if they did, it was hospital based and they received room, board, and laundry services in addition to the training opportunity (Rich et al., 2002). As time and medical technology evolved, the need for specialization increased, so that during the time of World War II until the passage of Medicare, physicians participated in hospital based internships for which the hospitals paid, via reimbursement from 13 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. patient charges, for the costs of training. Medicare allowed for hospitals to be paid “reasonable costs” for the salaries and benefits of the residents, the faculty costs, and the administrative costs of the educational programs. This payment came to be known as Direct Graduate Medical Education (“DGME”) funds. Other payers did the same, reimbursing hospitals based on usual and customary charges. Both models - that of “reasonable costs” and “usual and customary charges” represent a model of reimbursement where the hospitals are paid based on their costs. In the healthcare industry, this is referred to as “cost based reimbursement”. Under cost based reimbursement there are no disincentives to adding more residents or providing more services to patients. If it is provided, it gets paid. Payment was rendered retrospectively and there were no limits or constraints on establishing or expanding residency programs from a financial standpoint. In 1982 the need for a second component to the GME financing mechanism was recognized by the Federal government in order to offset the additional costs hospitals incur due to having interns and residents involved in patient care. This component, Indirect Medical Education (IME), was institutionalized in 1984 as the Medicare system changed its overall patient care reimbursement model from one of retrospective cost based payment to that of prospective payment of set payments for specific diagnosis. This new model of payment by Medicare was coined Prospective Payment System (PPS) and was a significant reform of Medicare overall in an effort to control hospital costs. Under PPS, hospitals were now paid DGME on a “per resident amount” or “PRA”. A hospital’s PRA was determined 14 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. based on a calculation of: 1) the amount each specific hospital paid to the resident in salary, benefits, and malpractice, as well as the cost of faculty, educational administrative expense, and facility expense; 2) the number of full time equivalents (FTEs) interns and residents at the specific hospital; and 3) the number of inpatient days during the year that Medicare paid the hospital. Hospital PRAs varied widely, ranging from $10,000 per resident to $240,000 per resident (MPAC, 1998). The IME was also paid to each hospital, to the extent the resident trained within the hospital, to offset extra utilization expense of the hospital that occurred due to the presence of a teaching program. Concerned with costs and the projected surplus of physicians, Congress included significant changes to GME in the Balanced Budget Act of 1997. Balanced Budget Act of 1997 (BBA) The BBA was signed into law by President Clinton on August 5,1997 (COGME, March, 2000). It represented dramatic changes in the Medicare program, with an ultimate goal of balancing the federal budget by 2002. Although the BBA addressed many components of the healthcare system that is funded by Medicare, Graduate Medical Education (GME) experienced sweeping and dramatic changes. The stated objective of the GME changes was to address the issues of physician supply, specialty mix, and maldistribution by changing the incentives for training that existed at that time (COGME, March 2000). Specific changes were made to provide incentives for increased training in underserved areas as well as 15 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. for residency programs to develop rural tracks (COGME, March 2000), though the rural component could not be fully realized until the Balanced Budget Refinement Act (BBRA) was passed during 1999. Such Rural Training Tracks (RTT) have been demonstrated to be successful in the recruitment of residents to rural areas (Rosenthal, 1997; COGME, March 2000); however, some of the other incentives for residencies created by the BBA have not been demonstrated in the literature. The BBA “capped” all residency programs in the nation at their 1996 levels of resident FTEs, regardless of specialties. As such, hospitals would receive Graduate Medical Education (GME) payments for no more than their capped number of resident FTEs. Whatever number of residents existed on the hospitals’ 1996 cost report, with a few exceptions, represented the maximum number of resident FTEs that the hospital could be paid by Medicare. There were few means for a hospital to increase its cap. One way, which was refined through the BBRA, was for a hospital to start a RTT, where the first year of a resident’s training occurs in the sponsoring urban hospital and thus allows an increase to their resident cap, and the final two years occur at a rural hospital and allows them to increase their cap as well. This also has an effect of broadening the formal networks and alliances of urban hospitals, for in order for them to increase their cap they must reach out to a rural hospital and establish new relationships. There is no one best structure of such inter-organizational coordination, but rather it is a function of “fit”, taking into consideration the unique form (e.g.; governance, organization, resources), tasks, and context (e.g.; the number, diversity, and kinds of 16 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. organizations in the network) of the relationship (Alexander, 1995). In order to be successful, participants in such organizational collaboratives must work together to increase public value (Bardach, 1998; Austin, 2000). The participants must have aligned missions, strategies, values, and a clarity of purpose (Austin, 2000), as well as the incentive, willingness, ability, and capacity to collaborate (Robertson, 1998). This concept of creating external partnerships was also seen in the BBA provisions that allowed hospitals to be paid Indirect Medical Education (IME) funds for rotations that occurred in off the hospital property in non-hospital licensed space. Usually, this takes the form of ambulatory clinics or physician offices. For the first time, resident rotations that previously were unable to be compensated with IME funds could now be compensated. This provided incentives for the creation of new relationships to other ambulatory sites - both urban underserved and rural. Ultimately, these two components of the BBA were designed to target residency programs and incentivize them to form alliances in order to address the issue of physician maldistribution. State of California Initiatives The State of California has also enacted legislation to address the issue of physician maldistribution in the state. The most notable of this legislation is the Song-Brown Act, which was passed in September 1973. This act created a program which is still in existence today that helped to fund the training of family physicians and primary care physician assistants. One of its primary objectives is 17 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. to increase the number of family medicine residency program graduates who subsequently practice in underserved areas of the State, both urban and rural (http://www.oshpd.cahwnet. gov/pcrcd/manpo wer/songbrown/html). Today, there are 38 non-military residency programs in the State, of which 30 receive funding from Song-Brown (Lachica, 2003). In order to receive Song-Brown funding, programs must demonstrate a commitment to training family physicians who will practice in underserved communities via residency rotations in rural or urban Health Professions Shortage Areas (HPSA) or other similar underserved sites. This too resulted in residency programs being incentivised to develop new collaborative relationships with the underserved communities that were both formal and informal in nature. Finally, the State of California signed a Memorandum of Understanding, through its Office of Statewide Planning and Development (OSHPD), with the University of California in 1994. The purpose of this MOU was to increase the numbers of generalist physicians trained in UC sponsored or affiliated residency programs, with the goal of mitigating the maldistribution of physicians in the State. As part of the MOU, the UC system was to produce an annual report of their efforts to increase the training of generalists (MOU, 1994). These reports provided updates on the progress of the UC system towards achieving their goal of adding 311 new primary care residency positions while reducing the number of non primary care residency positions by 452 (UC System, OSHPD, 1994). 18 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. One would presume that the aforementioned federal and state initiatives were informed by the literature during their design and implementation. It is not altogether clear as to whether the literature was incorporated, or whether it is clear as to the success of these interventions in meeting their objectives. The literature describes a number of “predictors” of rural placement and practice for family physicians. Although being a family physician is the greatest predictor, what is it about the pipeline of becoming a family physician that contributes the most to their ultimate decision to become a rural family physician? Is it one’s rural background? Is it that you had a special opportunity to get into medical school because you came from a rural area? Was it that someone ignited a passion in you along the way? What is not known is the most effective mix of predictors, let alone, which predictors matter most when an individual enters the pipeline to rural practice. In addition, despite the fact that there is little in the literature that speaks to the effects of residency training on rural recruitment of family physicians, policy makers continue to intervene at the residency program level. Ultimately, the challenge is one of knowing what the right mix of predictor elements are in order to develop policies that will help to provide enough family physicians to effectively staff the rural populations in California. The shortage of family physicians attracted to rural practice is of even greater concern than in recent years. For the past twenty years, the forecasts on physician need have been that of surplus, which has trickled down to policy decisions impacting medical 19 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. schools and residency programs. Now, in 2003, the Council on Graduate Medical Education has reversed it position and is projecting physician shortages (COGME, 2003). Projections now call for a shortage of 20% of the needed physician workforce by 2020, as much as 200,000 physicians (Cooper, 2003). With the inertia going specifically towards scaling back the production of physicians during the past twenty years, this recent shift in forecast has left the United States and its medical schools and residency programs ill prepared to turn the train back towards the station (Cooper, 2004). Recognizing that it takes approximately a decade to train a physician, action must be taken immediately to increase the medical school and residency program capacity to address the shortage. Rural, as is also the case with inner city underserved populations, will feel the stress of the shortage at an even greater level, since as the shortfall of physicians grows, so will the economic incentives to recruit them. Urban and suburban areas have greater access to resources to attract providers in times of shortage than their rural counterparts, as was demonstrated during the current nursing shortage. What is already a difficult situation is about to become far worse. 20 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 2 REVIEW OF THE LITERATURE The process of becoming a family physician has many steps - steps that can be seen as a pipeline of education and training. This pipeline begins as early as childhood, going into high school, college, medical school and residency, eventually possibly resulting in a decision to practice in a rural community. This chapter will review the literature on various components of the pipeline as well as the pipeline itself. The literature that addresses the decision of a physician to practice in a rural area describes a number of predictive factors. These predictors fall into four general categories: 1) individual characteristics; 2) medical school characteristics; 3) residency characteristics; and 4) community characteristics. The amount and depth that is known about each of these categories varies significantly. Although the pipeline and its component parts are chronologically based (one grows up before one goes to medical school, and one goes to medical school before one starts a residency program), the literature on the components and the pipeline in aggregate will be reviewed in this chapter in a slightly different order, primarily based on the volume of literature and activity of each.. There is significant literature on medical schools and the characteristics of medical schools that produce rural physicians. In fact, much of the activity to address rural maldistribution has occurred within medical schools in certain states. Next, individual characteristics will be reviewed, since this category also has received 21 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. much attention in the literature. Following this, the literature on residency program characteristics and finally the aggregate pipeline will be discussed. This literature review of the characteristics of each key component of the pipeline as well as the pipeline itself will lead to research question and hypothesis as well as the independent variables chosen for this study. For more than thirty years, attempts have been made to address the issue of rural physician shortage, due to maldistribution and to a shortage of physicians altogether. Some of these interventions have been based on the literature, others have not. The literature on characteristics that predict a physician’s decision to practice in a rural area and the interventions themselves fall into some broad categories, namely: 1) efforts to attract rurally inclined individuals to medicine and prepare them for medical school; 2) efforts based within medical schools; 3) efforts oriented towards graduate medical education (e.g., residency and/or fellowships), 4) efforts to recruit physicians into rural practice; and 5) efforts to retain physicians in rural practice (Norris, 2000). These categories can be reduced to two categories, namely: 1) interventions that impact the decision of a physician to practice in a rural area; and 2) interventions that impact the decision of a physician to stay in a rural area. This study only addresses interventions that impact the decision of a physician to practice in a rural area. 22 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The Medical School Literature “The shortage of primary care physicians in rural areas as been one of the most intractable U.S. health policy problems of the past century” (Rabinowitz et al., 2001,1041). As such, there has been significant research on recruitment and retention of physicians into rural areas. Concern, both from a policy and from a research perspective, has existed through periods of anticipated physician shortages and surpluses. During the 1960’s there was concern about a looming shortage of physicians. From a policy perspective, 40 additional medical schools were established and medical school enrollment doubled to address the projected shortage (www. graham-center.org/xl45.xml accessed 11/5/2004). As the medical schools expanded, some, such as Jefferson Medical College, created special admissions and educational programs in an effort to increase not only the supply of family physicians, but the supply of family physicians that enter rural practice (Rabinowitz et al., 2001). Currently, there are at least forty seven U.S. allopathic medical schools that have special rural programs that recruit students from rural backgrounds who are interested in primary care and specifically family medicine, and offer focused training and rural exposure during the curriculum (Barzansky & Etzel, 2002). The degree to which they emphasize special admissions and curriculum, let alone in a comprehensive manner focused on rural production, varies significantly. Some of the schools better known for their emphasis on producing rural family physicians are described below. 23 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Jefferson Medical College - Pennsylvania Jefferson Medical College’s Physician Shortage Area Program (PSAP) is a long standing and well document admissions and educational program that was implemented in 1974 to address the need for family physicians in rural Pennsylvania. This is a comprehensive program that focuses on individuals who are from a small rural town and intend to practice rural family medicine. Their well published outcomes are impressive in that PSAP graduates: • Are more than 8 times as likely as their peers to become rural family physicians; • Have a retention rate of 87% after 5-10 years in practice, and; • Account for 21% of family physicians practicing in rural Pennsylvania who graduated from one of the state's 7 medical schools, even though they represent only 1% of graduates from those schools” (http://www. iefferson. edu/psap /home/index.cfin: accessed June 21,2005). PSAP applicants must meet the minimum admissions requirements of JMC, however by broadening the admissions requirements; individuals who might otherwise be rejected for admission are accepted. University of Illinois College of Medicine Steams et al. (2000) described a separate medical school track for those interested in rural practice that was developed at the University of Illinois College of Medicine at Rockford in 1993. This program, as described, with 39 physicians 24 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. having graduated, of which 82% have selected primary care residencies and 69% have entered family medicine residencies. Their model is also a comprehensive one, building off the knowledge that is currently available in the literature, leaning on Pathman and others to built a curriculum that emphasizes Community Oriented Primary Care, socio-cultural awareness (what it means to be rural), and role modeling how a rural physician can assimilate into a community, utilizing its existing resources to address health issues. They utilize information systems to support their network of rural sites and trainees, which represents an untapped opportunity, at this point in medical education and pipeline development. University of Minnesota The University of Minnesota offers a nine month elective for medical students in their third year who are interested in rural practice. Since its inception in 1971, the Rural Physician Assistants Program (RPAP) has had over 1,000 participants, with: (i) two out of three students practicing in Minnesota; (ii) two out of three students practicing in rural areas; and (iii) four out of five students practicing as primary care physicians (http://www.med.umn. edu/RPAP. accessed June 21,2005). The Duluth campus of the University of Minnesota provides a two year experience for students who are likely to pursue rural family medicine. During these two years, students shadow and may even live with rural family physicians on several occasions. 25 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. University of Washington The University of Washington School of Medicine is regarded as the best in the nation for teaching medical students about family medicine and rural medicine (U.S. News and World Report, 2003). UW students conduct their clinical training in small towns throughout Washington, Wyoming, Alaska, Montana, and Idaho with local, often rural, community physicians providing the training. The program, known as “WWAMI,” for the states represented in the partnership, was established more than thirty years ago and is based on a philosophy of publicly supported decentralized education. Each participating state designates a specific number of medical school slots, with those states and students paying no more than what students from Washington State would pay. This provides a vehicle for medical school education for states that have no freestanding medical schools. Those states provide the first year of education at their own state universities, with the students having half day clinical experiences in local physician offices. After the first year, students have access to a variety of rural related programs, including: (1) SPARX, which is an interdisciplinary effort to assist health professions students to select rural locations for their future practice; (2) the Rural/Underserved Opportunities Program (R/UOP), which provides first and second year students clinical experiences in rural and underserved sites; and (3) the WWAMI Rural Integrated Training Experience (WRITE), which provides a six month rural experience to third year medical students. Over the past thirty years, 61% of the students have stayed within the five state region to practice, nearly 50% have selected a primary 26 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. care specialty, and 20% have selected Health Professions Shortage Areas for their practice location (http: //www. uwmedicine. or g/Education/WWAMI; accessed June 29,2005). Mercer Medical School - Georgia Mercer Medical School was opened in 1982 with a clear focus to solve Georgia’s problem of getting primary care physicians where they are needed - specifically rural areas of Georgia (http://www.mercer.edu/publications/ Medicine/spr99/ ). There curriculum is focused completely on preparing Georgia residents to be physicians in rural and underserved areas of Georgia. East Tennessee State “The James H. Quillen College of Medicine at East Tennessee State University is ranked seventh in the nation for excellence in rural medicine education by U.S. News & World Report in its “Best Graduate Schools” 2006 edition” (http://com.etsu.edu/default.asp7V DOC ID=2087&V LANG EH ): accessed July 3,2005). The primary mission of Quillen is to educate primary care physicians who will practice in rural areas. This focused mission has resulted in approximately 60% of Quillen graduates ending up in primary care disciplines (http://com.etsu.edu/default.asp7V DOC ID= 1563&V LANG ID=0; accessed July 3,2005). 27 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. University of South Dakota In addition to a medical school admissions and curriculum focus that has resulted in the University of South Dakota being tied for 13th place for rural medicine (http://www.usd.edu/press/news/news.cfm7nid~419&uid=user; accessed July 3,2005), the University has taken an additional approach to addressing the supply issue of rural physicians, that being a baccalaureate/MD program (Yutrzenka & Amundson, 2004). The “Alumni Student Scholars Program” (ASSP) was developed by the medical school in a focused effort to identify, train, and retain promising individuals in South Dakota with a high likelihood of practicing rural medicine within that state. As of 2004, forty-two high school seniors have been selected as ASSP scholars, with the medical school having graduate eleven, out of which five have entered family medicine residency programs (Yutrzenka & Amundson, 2004). Because this program is relatively new (1996), and because it takes significant time to go through the pipeline, it is difficult to ascertain if this intervention resulted in more individuals entering the rural provider workforce in South Dakota as compared to other efforts or no efforts at all. Tulane University School of Medicine - Louisiana The Tulane Rural Medical Education (TRuMEd) program began in 2004 in order to facilitate the training of students in rural primary care practice. Louisiana spends the most Medicare dollars per beneficiary and yet has the lowest overall quality score as compared with all other states (Health Affairs, 2004). Thirty-two 28 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. percent of Louisiana’s population resides in rural areas, as compared to 24% nationwide (http://www.som.tulane.edu/departments/admissions/special.html; accessed June 21,2005). Louisiana educators and policymakers alike saw medical education as a vehicle for improving the problems with quality, outcomes and physician distribution. The TRuMEd program was modeled after the PSAP (http://www. contracostatimes.com/mld/cctimes/living/health/11559533.html; accessed June 21, 2005) and provides an overlapping strategy that encompasses an admissions process that encourages the selection of individuals most likely to be primary care physicians in rural areas. The program also provides advisors and role models to support the students during medical school, loan repayment support, and curriculum that includes rural preceptors and clinical experiences in rural communities. Osteopathic medical schools are another source of physician training, particularly for those interested in primary care. There are twenty accredited osteopathic colleges in the United States (http://www.osteopathic.org/index.cfm? PageID=sir college; accessed July 3,2005) many of which focus on training for rural communities. Medical School Characteristics Is medical school part of the pipeline ... a conduit to youth making a decision to attend a rural residency program, or does it serve as a control valve ... 29 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. through its admissions policies and curricular programs (Rabinowitz & Paynter, 2000)? The “pipeline” argument is one of providing experiences to students who have not yet made up their minds regarding their career in order to expose them, inspire them, and nurture them. The pipeline approach subscribes to the thinking that students will be retained where they train and that students can be, and have been, influenced by such exposure. The “control valve” argument rests heavily on the literature that demonstrates the significance of “growing up rural” as a predictor to eventual rural practice. Under this argument, individuals with personal characteristics that predispose them to rural practice will self select medical schools that mirror their values and mission. The logic then follows that if there are enough medical schools with rural oriented training and curriculum and that base their admissions requirements on these personal characteristics that predispose individuals to rural practice, and then there will be a control valve that can be loosened or tightened depending on societal need, by modifying admissions practices accordingly. The literature on medical school characteristics and their impact on physicians’ decisions to practice in rural areas addresses three sub-categories of factors, namely: 1) the admissions policies of the medical school; 2) the organizational and strategic characteristics of medical schools; and 3) the actual curriculum of medical schools, particularly the third and fourth year clerkship curriculum. 30 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Medical schools with admissions policies that call for purposeful recruitment of individuals who are interested in generalist careers and who grew up in a rural area are more likely to matriculate individuals who go into family practice, enter into a rural residency program, and practice in a rural area (Rabinowitz & Payntor, 2000; Brooks, 2002; Rabinowitz, et al., 1999; Rosenblatt, 1992; Stratton, et al., 1991). Rabinowitz (1999) presented data on a retrospective cohort study of 206 Physician Shortage Area Program (PSAP) graduates from Jefferson Medical College in Pittsburgh, Pennsylvania. PSAP began in 1974 in order to train family physicians for rural Pennsylvannia. Interestingly, PSAP is a program of a private, urban medical school. PSAP focuses heavily on selecting and admitting individuals who have personal characteristics that predispose them to practicing in rural areas, which for PSAP means individuals who grew up in a rural area, who are planning as an entering medical school students to specialize in family medicine, and who intend to practice is a rural community. In this study, PSAP graduates that were currently practicing family medicine in the rural and underserved regions of Pennsylvania were compared will: 1) all allopathic medical school graduates coming from one of Pennsylvania’s seven medical schools; 2) all US and international allopathic graduates; and 3) to their non-PSAP colleagues who graduated from Jefferson Medical College. Approximately 12-15 medical students graduate from the PSAP program each year, and so represent only 1% of the total number graduating from the state’s seven medical schools during that period. Nevertheless, the PSAP graduates account for 21% of the family 31 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. physicians practicing in rural Pennsylvania that graduated from the seven state schools. PSAP graduates were more likely than their non-PSAP colleagues from Jefferson Medical College to practice family medicine, let alone family medicine in a rural region. PSAP has significant impact on rural workforce issues in Pennsylvania. It is important to note, however, that the PSAP program’s focus on admitting individuals’ that are predisposed to rural practice, admits individuals who had not been accepted at any other medical school. Although these students met all of the admissions criteria for JMC, according to the AAMC Joint Acceptance Reports, more than two thirds of the PSAP graduates were not accepted at any other medical school (Rabinowitz, 1999). Although it has been argued that medical schools have no role in whether physicians practice in rural areas (Cohen, 1999), Rabinowitz and Paynter (2000) demonstrated that a strong institutional mission, a focus on primary care, and an admissions program that targets individuals with rural backgrounds and an interest in generalist care, were highly successful in matriculating physicians who then went on to practice in rural communities. Their study looked at outcomes of the PSAP program coupled with a review of literature on six other medical school programs with similar goals and strategies. All seven programs showed that medical schools can impact the production of rural physicians by serving both as a pipeline to the next stage of training, that being residency, as well as a control valve. 32 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Some studies have explored the characteristics of medical schools that matriculate a high percentage of rural physicians. Rosenblatt et al. (1992) found that medical schools that are publicly owned, produce a larger percentage of family physicians, are located in a rural state, and have smaller amounts of NIH funding are strongly associated with producing more physicians that practice in rural communities. This study was based on an analysis of the December, 1991 version of the American Medical Association (AMA) Physician Master file and looked at graduates of American medical schools during the years 1976-1985 and then analyzed against the characteristics of the medical schools. Interestingly, it was found that during this time twelve medical schools produced twenty five percent of physicians entering rural practice, with medical schools ranging in their production from 41.2% to 2.3% of the graduating class. Basco et al. (1998) found similar results - that those schools that are publicly owned and actively recruit individuals interested in becoming a generalist physician were more likely to graduate physicians who would then practice in rural communities. Rabinowitz and Payntor (2000) found that a strong institutional mission - meaning one that focuses on the recruitment, admission, training, and matriculation of generalist physicians desiring rural placement, to be a core feature of medical schools that are highly successful in matriculating rural physicians. Finally, whether or not a rural physician went to a medical school in that state was not associated with rural recruitment (Homer, et al 1993). 33 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The final sub-category of medical school factors is that of curriculum. Medical schools vary in their commitment to a curriculum in primary care, let alone rural primary care curriculum and clinical training. The focused curriculum offered at the University of Minnesota through its Rural Physician Associate Program, for example, has resulted in a 59% recruitment rate to rural areas (Verby, 1991). A specific curriculum, that focuses on community teaching and preceptorships in the third year of training results in far greater recruitment rates than those curricula that don’t include a specialized curriculum (Verby, 1991). Rabinowitz et al. (1993; 1999) focused on the importance of specialized curriculum in order to increase rural recruitment. Later, Rabinowitz and Payntor (2000), in their study of seven medical schools with a stated mission to increase the number of rural physicians, reiterated the importance of rural focused curricula, but also state that little is known about the independent effects of the various elements of such curricula. What is known, however, is that rural curriculum is not an independent predictor of rural recruitment and practice (Rabinowitz, 1993), in that the curricular experiences of the PSAP participants, although important, did not independently add to the likelihood of rural practice. Literature on Individual Characteristics Brooks et al. (2002) conducted a systematic review of the factors that were associated with family physicians’ decisions to practice in rural areas. For the category of “Individual Characteristics” they focused on six key articles, ranging in 34 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. dates of publication from 1991 to 1999. Using a variety of statistical methods, the six articles analyzed multiple “pre-medical school” variables, such as age at the time of entering medical school, gender, race, ethnicity, marital status, size of student’s undergraduate college, level of father’s education, growing up in a rural area and a student’s expressed intent, before medical school, to become a family physician (Brooks et al., 2002). Four of the articles demonstrated that rural upbringing was a significant, if not the most important factor, in rural recruitment (Stratton, et al., 1991; Fryer et al., 1997; Looney et al., 1998; Rabinowitz et al., 1999; Brooks et al., 2002). For example, Rabinowitz (1999) found in a study of 1,609 physicians of which 206 were in rural practice, that growing up in a rural area was the most predictive independent factor. Rabinowitz (1999) also found that there was only one other variable that was strongly associated with rural recruitment, that being the expressed interest of a student, prior to entering medical school, to be a family physician. Later, from a survey of 2,966 physicians who graduated from Jefferson Medical College during the academic year 1983-1984, Rabinowitz et al. (2000) found that there are four independent self reported predictors of providing care to underserved populations (rural and urban), those being: 1. Growing up in an underserved area; 2. Being a member of an underserved ethnic/minority group; 3. Having a strong interest in practicing in an underserved area before attending medical school; and 35 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4. Having participated in the National Health Service Corps Scholarship Program. Those with all four of the predictors were four times as likely to practice in a rural or urban underserved area than physicians with none of the predictor characteristics. The literature is mixed on the issue of gender. Some studies found that women are less likely than men to choose to practice in rural areas (West et al., 1996) and that those men who practice in rural areas tend to be slightly older (Homer et al., 1993). Two other studies (Rabinowitz et al., 1999; Looney et al., 1998) found that gender was not predictive of rural placement, nor were other individual characteristics such as age, race, marital status, site of undergraduate education, or grade point average. The role of gender as a predictor is inconclusive. Although the Council on Graduate Medical Education (COGME) (2000) has expressed concern about the fact that fewer women go to rural areas to practice and that there is a growing percentage of women then men are graduating from medical school, the literature to support gender being a predictive factor is inconclusive. With women now accounting for more than 50% of all medical students, this issue of gender as a predictor for rural practice is of concern, in that the fear is that as the percentage of women becoming physicians increases, it will become even more difficult to find physicians interested in mral practice. Hence, based on the literature concerning individual characteristics, the two critical predictors that have emerged are whether one was raised in a rural location 36 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. and whether one intends to be a family physician prior to entering medical school. In fact, Rabinowitz (1999) found that this combination predicted a 36% likelihood of practicing in a rural location, as compared to a 7% likelihood for those without these two characteristics. Literature on Residency Program and Fellowship Characteristics Even though the issue of surplus issue was addressed, or so it was thought, and medical schools created rural programs that varied in their approaches to producing rural physicians, there remained issues of maldistribution. Medical schools were recognized to be only one piece of the production pipeline. Residency training, particularly family medicine residency programs with rural training tracks were seen as a way to mitigate the issue of maldistribution, based on the thinking that physicians will choose to practice in environments similar to where they trained. In fact, the belief that providing training in the practical experiences of rural life through electives, training tracks, and emphasis is very ingrained in the thinking of academics as a positive enhancement to recruitment efforts (Brazeau 1990; Magnus 1993; Connor 1994; Fryer 1997; Bowman RC 1998). Of the 474 family medicine residency programs in this country, 29 have established separately accredited rural training tracks (www. graham- center. or g/x 14 5. xml. accessed 11/5/2004). The literature on the impact of family medicine residency programs - their curriculum, rotation sites, mission, etc. on rural recruitment is very sparse. This is 37 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. surprising since the major federal and state interventions to address the issues of physician maldistribution have occurred at the residency program level. Rural Training Tracks (RTT) have been demonstrated successful in the recruitment of residents to rural areas (Rosenthal, 1997; COGME, March 2000). However, the value of some of the other incentives for residencies created by the BBA has not been demonstrated in the literature. As described in Chapter 1, several initiatives have been put in place, both at the federal and state levels, to try to increase the number of rural physicians. Although these initiatives all focused at the residency program level, there is little literature to support that residency program training - its structure, tasks, and context, impacts recruitment to underserved areas. There was one study that looked at some of the characteristics of residency programs that are not RTTs that have some success in the placement of graduates in rural areas. Bowman and Penrod (1998) found that the residency programs which graduated more rural physicians tended to have more required rural and OB training months, full or partial rural missions, were located in states that were more rural, emphasized procedural training, such as flexible sigmoidoscopies, treadmills, and colposcopies, and had a program director designated as the rural contact. Nationwide, we know that 24% of graduates from Family Medicine residency programs are placed in rural areas (AAFP, 1999). In California, for reasons unknown, the rate is significantly less, at 5%. Ultimately, “The link between family practice residencies and rural practice is mostly indirect” (Brooks, 2002,796). As referenced above, there is literature on 38 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. interorganizational collaborations, alliances, and networks and how their structure, context, and tasks can result in optimum outcomes. What is not known is the best mix of task, structure and context to improve or increase rural recruitment, whether through an RTT or through a regular residency program with rotations in a rural facility. Fellowship training provides one to two years of optional additional clinical experience upon completion of residency. There are nine rural health fellowships in the United States offering a total of twenty positions each year (Acosta, 2000). Generally, their purpose is to provide additional training to graduated residents who wish to practice in rural area yet believe they need additional skills, competency, and confidence (Acosta, 2000). The fellowships differ in their curriculum, structure and admissions policies, as well as their outcomes. There is little literature on the outcomes of rural fellowships; however, Acosta (1999) found that a greater than 75% placement rate of rural fellowship grads into rural communities of less than 25,000 people, which is approximately the same recruitment rate for graduates of Rural Training Tracks (Rosenthal, et al.; 2000). None of these fellowships are located in California. Pipeline There is little in the literature looking at the pipeline and it is an area makes sense intuitively, however little is known. The lack of knowledge about the pipeline to rural practice has been recognized. For example, regarding studies on 39 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. medical school predictors, Pathman (1996) said that few take into account the “pre existing characteristics, interests and career plans of students as they vary across schools and programs, and between participants and non-participants of elective experiences”. So, despite there being significant literature on individual characteristics and medical school characteristics and some literature on residency program characteristics, there is virtually nothing that looks at all of the categories, and variables within each category, and compares them to each other. For example, is it more important to recruit people from rural areas or to have rural residency programs should a state want to address their maldistribution issue? This study provides a look at the rural pipeline in a way previously not accomplished, by utilizing a dataset of family medicine residency program graduates in one state and backing into their pipeline in order to determine what characteristics; individual, medical school, or residency, impact the decision to practice in a rural area. This study also aims to address some unanswered questions about residency training, namely does the structure or characteristics of a residency training program impact an individual’s decision to practice in a rural area. Community Characteristics The literature on the role a community can play to impact the successful recruitment of a physician to it is very sparse. In a case study of six rural communities, it was found that the perceived strength of the rural hospital, community, board and administrator were all important in the community’s success 40 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. in recruiting a physician (Coffman J 2002). In addition, issues of professional isolation and lack of opportunities for spouses and children have also been documented to impact recruitment (Crouse 1995; Forti 1995; WONCA 1995). Because community is an important, yet not well understood issue, it has been included as part of this investigation. In conclusion, the literature speaks to the importance of individual characteristics and medical school characteristics, but says very little about post graduate characteristics of residency training, despite federal and state policy interventions at that level, nor about the pipeline of training that intuitively would lead to a decision to practice in rural. The independent variables that comprise this study are all based on the factors that have historically been found significant in the literature. Research Questions Based on the literature reviewed, the following research question and hypotheses are proposed for this study: Question 1: Compared to all other determinants, which determinants along the pipeline will impact rural placement of family physicians in California? 41 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Hypothesis 1: Hypothesis 2 The individual characteristic of graduating from a rural highschool, as a proxy for growing up in a rural community, will be the most significant variable of all variables in the pipeline. Determinants which are residency related will have little impact in predicting rural placement. 42 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 3 METHODOLOGY In this chapter the methods used in the study will be discussed. This study used a cross sectional design in which archival and survey data were collected and analyzed in order to assess the predictors among California physicians of practicing in rural areas upon completion of residency training in family medicine. First, in the “Sample” section, there will be a review of the dataset obtained which provided the data on the dependent variable (urban vs. rural placement) as well as some of the independent variables (residency program name, medical school name, gender, and high school location). Next, there will be discussion of the means by which data was collected from the various residency programs, medical schools, and communities. Following that, a detailed review of each variable will be conducted under the “Measures” section. Next, a separate section of the data collection process and measures will occur for the qualitative component of this study. Finally, under “Analysis”, the process by which the quantitative and qualitative analysis were conducted will be detailed. Quantitative Analysis Sample and Data Collection The sample was limited to all family physicians who completed a California based Family Medicine residency program during the five year period of 43 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1998 to 2002 and whose information was included, in whole or in part, on the Coastal Research Group Database. Coastal Research Group is a California based non-profit organization whose mission is, in part, to conduct research in the academic discipline of family medicine and primary care health care delivery (www.coastalresearch.org. accessed 12.23.04). Coastal Research Group was formed in 1983 in part to house and maintain a database, now known as the “California Family Medicine Residency Graduate database” which began in 1981 from funds awarded by the American Academy of Family Practice (email communication, Burnett, 12.23.04). The database was constructed to permit longitudinal studies of the long-term impact of family practice residency training by gathering demographic information on every family physician who has graduated from a California-based family medicine residency program. The sample group included 1,410 family physicians. This time period of 1998 to 2002 was selected in order to provide a sufficient sample size to adequately address the research questions and to restrict the sample to those individuals who completed residency after the implementation of the Balanced Budget Act in 1997. The decision to restrict the sample to those who completed residency after the passage of the BB A was based on the structural and programmatic changes that could have occurred after the implementation of the Act and that are significantly different than how residency programs were allowed to be structured prior to 1997. As such, those who completed residency programs 44 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. prior to 1997 may well have different experiences that impact their recruitment than those who completed their programs after 1997. There are a variety of data fields collected in the database; however, for the purpose of this study, the Coastal Research Group provided the name of the residency program, the practice address, city, state and zip of the graduate, the medical school name, city, and state from which the graduate matriculated, the gender of the graduate, and the graduates high school name, city, state, and zip code. This information is obtained from the graduates via multiple means. The residency programs are each contacted annually to provide the information. To the extent it is unavailable, the graduates are contacted. The graduates also receive mailings requesting information updates every few years. Finally, the information is cross checked against the AMA Masterfile. This information on the 1,410 graduates was provided via Excel spreadsheet. Data Collection Procedures Data collection began with the submission of this proposal to the Coastal Research Group for approval by their Research Committee. Once approved, access was granted to their Family Practice Residency Graduate Database, which contained significant background information and practice location about all family physicians who have completed a non-military residency program in the State since 1970. 45 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In order to conduct an analysis of the training pipeline, it was necessary to develop a dataset by obtaining information on each of the independent variables within each category of determinants, namely: (i) individual characteristics; (ii) medical school characteristics; (iii) residency program characteristics; and (iv) community characteristics. Coastal Research Group initially provided the names of the medical schools and residency programs attended by the 1,410 graduates. The data was provided in Excel and then copied into SPSS. As Coastal Research Group verified and validated the information on additional cases, they forwarded updated Excel spreadsheets with practice location, high school location, and gender data. As additional cases and/or information were obtained the new Excel spreadsheets were compared to the initially obtained spreadsheet on a cell by cell basis, with any additions, deletions, or modifications being made, first into the master Excel spreadsheet and then copied into SPSS. For example, during the period of this research project, Coastal Research Group contacted the residency programs and individual graduates to obtain more information on high school, in an effort to have as complete a dataset as possible. In another situation, Coastal Research Group had retained certain years of graduates from specific residency programs until they had an opportunity to review and better complete that data with the residency program. Once Coastal Research Group forwarded the final version of the dataset, the data were organized by residency program and then by year of graduation from that residency program. 46 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Letters were sent to the medical school librarian at each school identified in the dataset, requesting copies of the course catalog spanning the five year period of 1990 to 1995, which represents the time during which individuals who completed residency training from 1998 to 2002 would have completed medical school. Similar letters were sent to all California medical schools as well. If the complete catalogs were not available for mailing, the librarians were requested to provide photocopies of the admissions policies and the curriculum sections of those catalogs. This turned out to be an unsuccessful approach. Although the librarians were very conscientious in their desire and effort to provide the information, the majority were either unable to find the course catalogs from those periods or the cost to photocopy was prohibitive, at times ranging as much as $700.00. Another approach was then taken to obtain the information. A letter and brief survey (see Appendix C) was sent to the Dean of each medical school (N = 104) requesting that they provide the following information for the period 1990-1995: 1 . Whether the word “rural” appeared in their mission statement; 2. Whether their admissions policies or practices had a stated objective of recruiting from rural communities; 3. Whether any of their admissions policies or practices had a stated objective of recruiting individuals interested in generalist medicine; 4. Whether they had a mandatory family medicine clerkship; 5. Whether they had a mandatory community clinic rotation; 6. Whether they had a mandatory rural clerkship rotation; 47 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 7. Whether they had a specialized, focused rural curriculum; 8. If their school was public; 9. What is their school’s percentage of NIH funding; and 10. What percent of their graduates entered family medicine residency training. Each of these questions assessed a variable that has been measured in other studies regarding medical school predictors of rural placement. Although there are 126 allopathic medical schools in the United States (www. aamc. org/medicalschools. html. accessed June 10,2003) and 20 osteopathic medical schools in the United States (www. aacom. org/about, accessed June 19, 2003), only the 105 medical schools that were attended by the 1410 graduates were queried. Of those queried, fifty (48%) completed and returned the survey. For the remaining 52% of the schools, a variety of publicly available information sources were utilized, including the schools’ websites such as the Association of American Medical Colleges (AAMC), the AAMC Medical School Admissions Requirements (MSAR), the National Resident Matching Program, the American Academy of Family Physicians, and the Society of Teachers in Family Medicine. MSAR (2005-2006) is published annually by the Association of American Medical Colleges and lists every U.S. and Canadian allopathic medical school and provides brief information on its mission, selection factors, and curriculum. AAMC has a curriculum database that is accessible only by the Academic Affairs Deans of a medical school. A request was made of the Associate Dean of Academic Affairs at the USC Keck School of Medicine, who in turn instructed his staff to query the 48 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. AAMC curriculum data base (CurrMIT) for any medical schools in the U.S. that self reported as a rural medical school. The AAMC web site was queried to obtain their data on the ranking and total NIH funding levels of the medical schools. The same web site also provides information on the number of individuals matriculating each year who are entering a family medicine residency program. Finally the American Academy of Family Physicians, the Society of Teachers of Family Medicine, the National Resident Matching Program, and an article by McPherson et al. (2004) were utilized to obtain information of the various schools and the percentage of their students who matched into Family Medicine residency programs. This was completed during the period of February through September, 2004. There are 38 non-military residency programs in the State of California (www.aafjp.org/residencies. accessed June 10,2003). Surveys (see Appendix A) were mailed to the program directors listed in the AAFP “Green Book” along with a letter explaining the nature of the research project and the consent to participate. A follow up letter and survey were sent to non-respondents. Follow up calls and emails were made by the principle investigator to the program directors and their residency program coordinators to encourage completion of the survey documents. Finally, the principal investigator obtained the completion of a few surveys during encounters with individual program directors during various state and national meetings. Ultimately a total of twenty three surveys were completed. The residency program directors’ survey queried multiple topics, such as mission, types 49 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. and length of training rotations, training and practice experience of the program director, capacity, willingness, and incentives to start a rural rotation, etc. For the remaining fifteen residency programs, data were obtained via the Accreditation Council of Graduate Medical Education website and the individual residency program websites. In addition, ACGME website information was utilized to validate information provided via program director survey response. Dependent Variable - Practice Location Information on the practice location of each graduate was obtained from the database. In order to classify each practice as either urban or rural however, one must begin by first defining what is rural. In the United States, multiple definitions of what connotes “rural” and “urban” exist, both at state and federal levels. These multiple definitions are inconsistent and conflicting, resulting in frustration to rural health entities as they try to argue for their “ruralness”, depending on which definition is being used by which entity. Definitions range from the dichotomous - in which the U.S. Office of Management and Budget defines rural as “anything that is not urban” to more of a continuum. In the late 1990’s, the Office of Rural Health Policy (ORHP) within the United States Health Resources and Services Administration contracted with the University of Washington to develop a new methodology to delineate “rural” and “urban” particularly as it pertained to ORHP funding initiatives. The RUCA index categorizes areas on an urban-rural continuum, looking at the 1990 census 50 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. tract data as well as 1998 zip codes and factoring in other measures of urbanization such as daily commuting patterns and population density (http://www.ruralhealth. ca.gov/pdf/RUCA%20Qverview.pdf, accessed 1.13.05; http://www.fammed. Washington, edu/wwamirhrc ). Communities are then ranked on a continuum of 1 to 10, with one being the most urban and ten being the most rural. A community with a score of 4 and higher is usually then defined as rural. The RUCA index is not a universally accepted method for delineating rural settings. In fact, rural healthcare institutions in California have advocated against the use of RUCA, stating that it is not an appropriate means for assessing “rurality” in the State (http://www.ruralhealth.ca. gov/pdf/RUCA%20Qverview.pdf). In California, a different definition of “rural is utilized within the State. The comparison of California’s definition to the national RUCA definition can be seen in Map 2. In addition, RUCA has not become the recognized method of delineating rurality for all federal and state agencies, but instead has become yet another definition of rural that is used sometimes for some purposes and excludes or includes a site that might be excluded or included with a different definition. Despite this, however, it provides a readily accessible means of delineating rurality for statistical analysis. RUCA scores for all zip codes in the United States, based on 1990 census data, can be readily obtained via the worldwide web (http://Chester, fammed. Washington, edu/ wwamirhrc/national. txt). All other methods were either not used nationwide, not easily obtainable, or not specific 51 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (e.g., at defines urban or rural at a county level) to delineate urban and rural placements for this project. Map 2 Rural Urban Commuting Areas Map Rural Urban Commuting A reas & Medical Service Study A reas 2P code Boundaries ( 4 Medical Service Study Area (MSSA) Rural Urban Commuting A n a s ! ■ J RUCA Codes t - 3 (Ineligible) Urban-Rural Definition [ J Urban Area fneSgibte) 1 RUCA Codes 4 and above (Eligible) ~ v - I T S itf O ” a v ['P P Rural Area (BigWe) iH t' 'c M W M I M M I M M I I n M W W tkwifiwtMwi IM IW IU tfC oM I U$.Cmw a<w»l8« GOTtywitim Measures Overview All measures were coded as dichotomous variables for purpose of logistic regression analysis. The following variables were provided from the initial dataset and entered into SPSS as string variables. 52 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 2. Chart of String Variables Name of Variable Label of Variable Type of Variable G radid Unique identification number for each graduate String res name Name of residency program String Med name Name of Medical School String Second, the RUCA index was queried for each high school and practice location zipcodes. The RUCA scores, ranging from 1 to 10 were entered into two numeric variables. Table 3. Chart of Variables with RUCA Scores Name of Variable Label of Variable Type of Variable pracruc RUCA Score for practice location zipcode Numeric HSJRUCA RUCA Score for highschool location zipcode Numeric Third, those two variables were recoded into dichotomous variables. Name of Variable Label of Variable Type of Variable Rural_p Rural or Urban practice location Numeric - Dichotomous variable RurHSRU Rural or Urban high school location Numeric - Dichotomous variable Fourth, the data obtained from medical schools, AAMC and residency programs were entered into SPSS as the following variables: Table 4. Chart of Medical School and Residency Dependent Variables Name of Variable Label of Variable Type of Variable Residency Level Variables req_rura Required rural rotation Numeric - dichotomous elec rur Elective rural rotation Numeric - dichotomous 53 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. req_ob Required obstetrics rotation of four months/blocks or greater Numeric - dichotomous Missionw “Rural” written into mission statement Numeric - dichotomous Missions Rural emphasis in survey response or ACGME documentation Numeric - dichotomous dirrural Program Director trained or practice in a rural area Numeric - dichotomous Medical School Level Variables SOM mis “Rural” written in school of medicine’s mission statement Numeric - dichotomous Adminrur Admissions policies specifically mentioning recruitment of individuals from rural areas Numeric - dichotomous Admin_gen Admissions policies specifically mentioning recruitment of individuals intending to practice primary care or family medicine Numeric - dichotomous ManFMC Mandatory family medicine clerkship Numeric - dichotomous ManCC Mandatory community clinic rotation Numeric - dichotomous Man rur Mandatory rural rotation Numeric - dichotomous Sp Rur Cu Specialized rural curriculum Numeric - dichotomous Sector Public or private sector medical school Numeric - dichotomous Fedfunds In top 40 of NIH funded medical schools Numeric - dichotomous PercFM Greater than 10% of medical school graduates entering family medicine residency training programs Numeric - dichotomous aamc_rur Listed as a rural medical school in CurrMIT Numeric - dichotomous Fifth, the gender data obtained from the Coastal Research Group database was entered into SPSS as the following variable: Table 5. Chart of Gender Variables Name of Variable Label of Variable Type of Variable Gender Numeric - dichotomous Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A visual representation of the methodology is provided below. Table 6 Visual of Methodology Methodology Gender Medical . School Practice Location Residency Program High School Database of 1,410 FM grads From California residency programs 1998-2002 .RU CA ) Survey p>ntent\ (survey) (C°n|ent) (ouCA V S V y Analyse V y Analysis/ " 50/112 Logistic Regressions Specific Detail on Each Variable Dependent Variable - Rural Practice The practice addresses for each graduate in the Coastal Research Group’s database was compared to the RUCA zip code list and each was assigned a RUCA value. RUCA values were entered into SPSS (“prac rur”) and then recoded into a new variable where cases with a RUCA of 4 or greater were coded as rural (“rural_p”). 55 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Independent Variables - Predictors of Rural Practice Individual Characteristic Variables The two individual characteristics that were measured were “growing up in a rural area” and gender. The database contained data on gender (n=1267) and location of where each individual, or case, went to high school. There were 595 cases (42.2%) that had high school data available. Table 7. Missing High School Cases HS avail Frequency Percent Valid Percent Cumulative Percent Valid No 815 57.8 57.8 57.8 Yes 595 42.2 42.2 100.0 Total 1410 100.0 100.0 Because of the low percentage of cases with high school data available, a Chi square test was conducted to determine the degree of confidence of the high school data. This will be discussed in greater detail in the Analysis section of this chapter. Growing up in a rural area can best be determined by the zip code of the high school from which a family physician graduated. The zip codes were available from the Family Medicine Residency Graduate Database. The zip codes were then given a RUCA value. Since the RUCA values are based on the 1990 census data, this provided a means of delineating rural versus urban based on census data that was closed to the years these individuals would have graduated 56 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. from high school. Gender, although not found as a conclusive determinant in the literature, was also included since the information was available in the database. Although an individual’s expressed desire, prior to medical school admission, to be a generalist physician is also a strong predictor of rural recruitment, this data is not available and so will not be measured for this proposal. Medical School Variables Medical School Information Of this sample group of approximately 1,410 family physicians, there is a subset (N=41) that are graduates of foreign medical schools (FMGs). For the purpose of this analysis, FMG includes those who matriculated from Canadian, Caribbean, Mexican medical schools as well. It is extremely difficult to obtain data, let alone reliable data, from non United States medical schools. Many family medicine residents and rural physicians are graduates of foreign medical schools, with estimates ranging in the neighborhood of 24% (Mick & Lee, 1997). Because of the delays and lack of reliable data experienced by the investigator in the past when trying to obtain original source verification from medical schools in foreign countries, it was decided to exclude those cases. In addition, 776 of the 1,410 cases, or 55%, had no medical school information available. 57 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 7. Missing Medical School Cases Medical School Available Frequency Percent Valid Percent Cumulative Percent Valid No 776 55.0 55.0 55.0 Yes 634 45.0 45.0 100.0 Total 1410 100.0 100.0 Because of the low percentage of cases with medical school data available, a chi square test was conducted to determine the degree of confidence of the high school data. This will be discussed in greater detail in the Analysis section of this chapter. As discussed in the literature review section, three categories of medical school variables serve as potential determinants of rural placement. Admissions Criteria A content analysis was completed using information obtained on the medical school websites and from the MSAR information about medical schools where the “Non-FMG grads” matriculated. A systematic search was conducted for references to a stated objective to recruit individuals interested in generalist medicine or to recruit individuals from a rural background. The medical schools, as a string variable, were recoded in SPSS into two new variables, “Admin rur” for those schools with a stated admissions objective of recruiting individuals from rural areas and “Admin gen” for those schools with a stated admissions objective of recmiting individuals interested in generalist medicine. Both variables were coded as a “1” if the school had stated admissions objectives of rural background or 58 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. generalist medicine, respectively. If there was no stated objective in either case, both variables were coded as a “0”. If no information was available, the medical school was coded with “SYSMIS” Medical School Characteristics The sector (private or public) of each medical school was obtained from U.S. News and World Report’s 2005 listing of America’s Best Graduate Schools and from survey respondents. Each U.S. medical school is profiled in that edition, and included in the profile is their governance/ownership status. All medical schools were recoded in SPSS into a new variable, “Sector”, with a 1 indicating public and a 0 indicating private. Medical schools were recoded as “SOMjmis” Medical schools were recoded into this variable, with a code of 0 indicating that the medical school had no rural content in its mission statement. Data on each school’s mission statement was obtained either from the survey response or from content analysis of their web site. If the word “rural” was used anywhere in the mission statement, the variable was coded as a “1”. Medical school admissions policies were also reviewed. Data was obtained from survey responses and the individual school web sites. Medical schools were recoded into two variables. The first, “Admin rur” was coded as a “0” for any medical school that did not have the word “rural” anywhere in their admissions policy. Schools with the word “rural” were coded as a “1”. The second, 59 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. “Admingen” was coded as a “0” for any medical school that did not have the words “generalist” or “primary care” in their admissions policy. Schools that had either of these were coded with a “1”. To obtain the percentage of medical school graduates who entered family medicine residency programs upon graduation, survey data was used or, in the absence of survey data, information was obtained from two sources. The first source was the individual web site of each medical school. The site was searched using key words “resident match,” “match day,” and “match results.” If match results were available, the total number of individuals participating in the match, for each year that data was available, were tallied as were the number of individuals in that group who matched with a family medicine program. A simple percentage of those matching into family medicine programs compared to all others matching that year were calculated. If multiple years of match data were available, the same process of obtaining a simple average will be done, and then the averages of the multiple years will be taken. Many medical school websites did not have any information on their resident match results, or they combined all primary care programs into one percentage. An additional resource, an article published in the journal Family Medicine, provided the average over the past three years of family medicine match results for the allopathic medical schools (McPherson et al., 2004). Medical schools were recoded in SPSS into a new variable “Perc FM” Those schools that graduated 10% of their senior class or less into family medicine 60 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. residency programs received a recode of “0”. Those that had more than 10% of their graduating class match into family medicine residency programs were recoded as a “ 1”. Those medical schools that had no available information were recoded as a “SYSMIS” Finally, the National Institutes of Health (NIH) website was queried for the ranking of U.S. medical schools by their amount of research funding (www. grants 1.nih.gov/grants/award/rank/medttl03.htm. accessed 8.26.04). Medical schools were recoded into a new variable “Fedfunds” based on the amount of federal research funds they receive from the NIH. Those medical schools that were in the top 40 of the NIH ranking were recoded as a “0”; those that were below the top 40 were recoded as a “1”. Curriculum Characteristics Survey responses and content analysis of multiple sources provided the data about medical school curriculum. In addition to the surveys, individual medical school websites, medical schools’ online course catalogs, the AAMC’s MSAR, the AAMC’s curriculum database “CurrMIT,” and U.S. News and World Report provided valuable material for content analysis. The surveys provided the primary source of data. To the extent that surveys were not available, the school’s web site was the primary source of data. If information on a specific variable was not available on the web site, then MSAR, CurrMIT, and U.S. News and World Report were queried to see if the data was available. Third and fourth year curriculum was 61 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. reviewed for mandatory versus voluntary family medicine clerkships, mandatory community rotations, and mandatory rural rotations and any specialized, focused rural curriculum. Based on the above, medical schools were recoded into four new variables. The first was “Man FMC” which indicated the presence of a mandatory family medicine clerkship in the third or fourth years of medical school training. Those medical schools that had a mandatory family medicine clerkship were recoded as a “1”; those that did not were recoded as a “0”. The second new variable was “Man CC” which represented mandatory community rotations during any of the four years of medical school. Those medical schools with mandatory community clinic rotations were recoded as a “1”; those that did not require community clinic rotations were recoded as a “0”. The third curricular component to be addressed through the creation of a new variable was the presence of a mandatory rural rotation during the four years of medical school. This new variable was designated “Man rur”. Medical schools that required students to have a rural rotation, regardless of the length of the rotation, were coded with a “1”. Those that did not were coded with a “0”. The last curricular variable was whether the medical school had any special rural curriculum. For example, there are some medical schools that have a special rural track for students, as part of a pipeline strategy to nurture and guide students into rural practice. This variable was designated “Sp rur Cu” and those medical schools that had such a curriculum were recoded with a “1”. Those that did not were recoded with a “0”. 62 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Post Doctoral - Residency Program Variables Residency Program Information There are currently thirty eight family medicine residency programs in California (http://www.acgme.org/adspublic, accessed 1.13.2005), in which one is sponsored by the Navy. Because of the unique characteristics and mission of a Navy sponsored program as compared to other hospital or university sponsored programs, the Navy program was not studied. A survey (see Appendix A) was sent to the 38 non-military residency programs in California. The survey focused on the following general categories of questions: • Mission. • Curriculum. o Higher than ACGME required numbers of Obstetrics rotations o Presence of a mandatory procedures rotation o Presence of a mandatory rural rotation • Impact of the Balanced Budget Act on the residency program structure, context, and tasks. • Presence of or desire for a rural training track. • Number of rural or underserved community rotation sites. • Nature of the relationship (values, mission, willingness, capacity, etc) with the rural or underserved community rotation sites. 63 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. • Tasks, structure, and context of the program. • The location (rural vs. urban) of the Family Practice Center. Of the twenty three programs that completed their surveys (60%), many did not answer all of the questions. For the remaining programs that did not return a survey, their website was reviewed for answers to these questions. The websites were able to provide all of the information about curriculum, mission, rural rotations, and locations. However, the web sites did not provide information for some of the less traditional questions, such as the impact of the Balanced Budget Act or the willingness or capacity of the program to develop a rural rotation. Because of this, not all questions were developed into variables. Residency programs were coded for seven different variables. The first was “missionw” which represented whether the word “rural” was written anywhere in their published mission statement. If “rural” was written in the mission statement, then the residency program was coded as a “1”. If it did not appear, it was coded as a “0”. The second variable was “missions”. During the analysis of the survey responses, there were residency programs that said they were “rural” even though rural was not in their mission statement, they were not located in a rural area, nor did they have a mandatory or elective rural rotation. For example, one program was not located in a rural area, did not have rural in its mission statement, and had no rural rotations. However, on the top of the survey, the Program Director wrote, “We are a rural program.” As similar examples arose, the need for a variable that addressed a residency programs perception of it as rural, based on what was written 64 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. throughout their survey or website, became apparent. Those family medicine residency programs that either stated, somewhere in their survey or in their website, that they were rural or rural focused, were coded with a “1”. Those that had no indication of a rural emphasis or interest were recoded with a “0”. The third residency program variable was “req_rura” which identified whether or not a residency program had a required rotation in a rural area. Length of the required rotation was not a factor; only whether or not one existed. Those programs that required a rural experience were coded as “1”; those that did not were coded as “0”. The fourth residency program variable was “elecrur”. This variable addressed whether the residency program had developed any rural sites as elective rotations for the students. Those programs that listed rural elective sites were coded as a “1” those that did not were coded as a “0”. The fifth residency program variable was “req_ob”. All family medicine programs are required by the ACGME to have obstetrical and gynecological rotations for their residents. It is a common belief amongst family physician educators that family physicians who go into rural areas require more Ob training and more procedures training. The ACGME requires family medicine programs to provide no less than two months of maternity (OB) education to every resident (http://www. acgme. org/acWebsite/ downloads/RRC progReq/120pr701 .pdf). Those programs that had four or more months (calendar) or blocks (4 weeks each) were coded as a “1”; those with fewer than four were coded as a “0”. Four months or blocks was determined to be the appropriate cutoff since this is more than what is required by the ACGME’s Family 65 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Medicine Residency Review Committee for obstetrical training in all family medicine training programs. The sixth residency program variable was “dirrural”. If the program director indicated, either on the website or in the survey, that they had ever practiced or trained in a rural environment, then the program was coded as a “1”. If there were no such indication, it was coded as a “0”. The seventh residency program variable was “BBA impt” - whether there had been any changes to the residency program as it pertained to a rural focus due to the passing of the Balanced Budget Act of 1997 or the Balanced Budget Refinement Act of 1999. Those programs that did note changes were coded with “1”; those that did not were coded with “0”. Missing Values The dataset had significant amounts of missing values for both the high school and medical school variables (57.8% and 55% respectively). Residency program information, in general, did not have sizable amounts of missing values. Table 9. Missing Values by Variable CATEGORY VARIABLE VALID MISSING PERCENTAGE of total cases missing Individual Gender 1267 143 10% High School 595 634 58% Medical School Rural in Mission Statement 634 776 55% Admissions policies seeking rural 634 776 55% Admissions policies seeking generalists 634 776 55% Mandatory family 634 776 55% 66 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. medicine clerkship Mandatory community clinic rotation 634 776 55% Mandatory rural rotation 634 776 55% Special focused rural curriculum 634 776 55% Sector 634 776 55% In top 40 ofNIH funded schools 634 776 55% Percent of grads entering FM residency 634 776 55% Residency Required rural rotation 1392 18 1% Elective rural rotation 1322 88 6% Greater or equal to 4 OB rotations 1392 18 1% Rural written in mission statement 1392 18 1% Rural emphasis 1392 18 1% Program director trained/practiced in rural 1182 228 16% Because of the amount of missing values for the high school and medical school variables, it was necessary to determine how confident the investigator could be about the generalizability of the data. The data set was divided into those cases for which high school data were available and those cases for which they were not, creating a new variable called “HSavail”. Chi-square tests were conducted on this variable against all other variables in the dataset, such as gender, residency program mission, medical school admission policies, etc. The same was done regarding medical school information. The dataset was recoded into a new variable (“MS avail”) delineating those cases where it was known what medical 67 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. school they attended and those cases where it was not known. Chi-square tests were conducted on this new variable against all other variables in the dataset. The Chi square tests conducted on the high school and medical school data were not significant and indicated that the missing values for those cases were not relevant and as such did not impact the overall generalizability of the data. Analysis The data were analyzed by conducting a logistic regression. Logistic regression was selected due to the binary nature of the data and because of the predictive nature of the research questions. In addition, because of the high number of cases that were missing either high school or medical school data, Chi square testing was conducted to determine the generalizability of the data. The dependent variable (rural vs. urban recruitment) was analyzed first within each category of independent variables: (i) Individual Characteristics; (ii) Medical School Characteristics; and (iii) Residency Program Characteristics. From this, the goal was to determine which individual variables within a specific category have a significant relationship with an individual’s decision to practice in a rural community. Next, a second regression was run, where all of the variables from the three categories were included in one regression. The goal of this regression was to determine which of the variables are significant predictors of a decision to practice in a rural community. A third regression including just the significant variables from the second regression was conducted to clarify even 68 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. further which variables along the pipeline of physician education were significant. Then, separate regressions were run both by category and for the entire pipeline, so as to compare the group that graduated from rural high schools to the group that graduated from urban high schools. Finally, dummy variables were created by multiplying the high school location variable against all other variables. Logistic regressions by category and along the pipeline were conducted of all variables and the dummy variables to determine any interaction effects. Individual Characteristics (Independent Variables! • High School located in a Rural or Urban region, as defined by the RUCA methodology. • Male or Female. Medical School Characteristics (Independent Variables) • Stated objective of recruiting individuals from rural areas. • Stated objective of recruiting individuals interested in generalist medicine. • Public versus private ownership. • Research ranking based on NIH funding. • Percentage of Family Physicians graduated. • Mission statement that references rural. • Presence of mandatory family medicine clerkship. • Presence of mandatory community clinic rotation. • Presence of mandatory rural rotation. 69 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. • Presence of specialized, focused rural curriculum. Post Graduate - Residency Program Characteristics (Independent Variables) • Number of Obstetrics rotations. • Presence of a mandatory rural rotation. • Presence of an elective rural rotation. • Mission statement that has the word rural written in it. • Mission and culture that emphasizes rural. • Program director that trained or practiced in a rural area. • Stated changes to the program due to the Balanced Budget Act and Balanced Budget Refinement Act. Qualitative Analysis Sample and Data Gathering A multi-methods approach was utilized for the qualitative component of this project. Data regarding rural communities was obtained from a focus group of rural hospital administrators that occurred in March, 2003 and from a survey that was sent to 69 rural hospital administrators in the State of California. A focus group of rural hospital administrators was conducted in Fish Camp, California in collaboration with the California State Rural Health Association in March, 2003. The purpose of this focus group was to obtain information from rural administrators regarding what curricular components they view as necessary for training family physicians who will practice in rural areas. During the course of a 70 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ninety minute discussion, five sets of questions were answered by ten participants. All responses were typed into a Word file by the Director of Public Relations for the California State Rural Health Association. Content analysis was conducted and key words and concepts tabulated. Key themes were identified based on the tabulated results. The intent of the focus group and the surveys was to gamer additional descriptive information about rural communities in California through the hospital administrators of those communities. Of particular interest were the desired knowledge, skills, and capabilities of rural family physicians, the budgets available and avenues used for recmitment, the challenges of the communities in recruitment and with their providers in general, and the extent to which the communities are networked with Family Medicine residency programs and/or academic institutions. Measures Five groups of questions were asked of the ten focus group participants. 1) What are the qualities and skills of the “ideal” rural physician? 2) What curriculum pieces (give examples) are essential for a rural family medicine physician to develop during a training? 3) Should the curriculum be static (everyone graduates with the same skills, e.g., OB, surgical) or allow for flexibility? If flexibility is important, how should it be customized? 71 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4) What non-medical skills are important in a rural physician? (e.g., leadership, community outreach, administrative, etc.) Should training in such areas be part of the curriculum? 5) Let’s talk about rural training. What are the necessary characteristics of a rural training site? How much training should be in a rural site as compared to a large urban academic site? Multiple rural sites? Size? Location? Does it matter if the trainee is employed at that location? What do you think are the pros and cons of a rural family medicine training program for California? Second, a survey (See Appendix B) was mailed to the sixty nine rural hospital CEOs in California. Follow up calls were made and a second mailing was conducted. Twenty-eight hospital CEOs completed the surveys and the responses to the following questions were tabulated: 1) D o you need more physicians? 2) Is recruitment o f primary care physicians difficult for your facility? 3) What strategies have you used in the past to recruit? 4) What has proven successful for you? 5) Does your facility have a relationship with an academic health center? a. If so, which one and what is the nature o f that relationship? 6) Does your facility have a relationship with a family medicine residency program? a. If so, which one? What is the nature o f the relationship? 72 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Analysis Rural Community Data Information regarding the rural communities in California obtained through rural hospital administrator surveys and focus groups was utilized to describe the factors and issues in rural communities in California that impact physician’s decisions to practice in rural communities as well as the issues, challenges and resources of rural communities in their efforts to recruit. In addition, it was used to describe any gaps between the training of physicians, as it occurs today, and the needs and expectations of the rural communities. This will complement and enhance the quantitative analysis. 73 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 4 RESULTS The process of understanding the relationship of each variable to an individual’s decision to practice in a rural area required five separate logistic regression analyses, described within this chapter. After presenting the results from these regressions, the qualitative analysis of focus group and survey data are also discussed. First, however, a map of California representing the practice location decisions of the graduates is provided below. Map 3 Location of Database graduates in California California Medical Service Study Areas Census 2000 Configuration WMn. Rural and FrorMar P —IgnaMona 74 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Quantitative Results A total of five regressions were conducted. The first set of regressions looked at the variables within each broader category, those being individual characteristics, medical school characteristics, and residency program characteristics, to determine the significant variables within each category. The second set of regressions looked at the pipeline, analyzing all of the variables against each other. The third set of regressions were identical to second with the exception that only the significant variables from regression two was included in regression three. The fourth set of regressions was a comparison of those who graduated from rural high schools and those who graduated from urban high schools. The fifth set of regressions looked for interaction effects between high school, as the proxy for growing up rural or urban, and all variables within each specific category as well as the pipeline. Summary tables comparing the outcomes of regressions one through four may be found in Tables 3-7. Table 10. Findings - Individual Characteristics INDIVIDUAL CHARACTERISTICS 1st Regression Pipeline Regressions Pipeline of Only Significant Variables from 1s t Regression % Of Cases Included 39.9% 28.6% 30.3% Significant Variables Graduating from a rural high school (p=.000) Graduating from a rural high school (p=.000) Graduating from a rural high school (p=.000) 75 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 11. Findings - Medical School Characteristics MEDICAL SCHOOL 1st Regression Pipeline Regressions Pipeline of Only Significant Variables from 1s t Regression % Of Cases Included 47.8% 28.6% 30.3% Significant Variables - R ural Admissions Policies (p=.005) - Special R ural Curriculum (p=.022) - Sector (p=.026) - R ural Admissions Policies (p=.005) - Special Rural Curriculum (p=.022) - Mandatory Community Clinic Rotation (p=,019) - Percent of graduates entering FM> 10% (p=.005) Table 12. Findings - Residency Program Characteristics RESIDENCY PROGRAM 1st Regression Pipeline Regressions Pipeline of Only Significant Variables from 1st Regression % Of Cases Included 71.1% 28.6% 30.3% Significant Variables “Stateed” rural mission (p=. 012) Director trained or practiced in a rural area (p=.000) - “Stated mission of rural (p=. 017) - Director who trained/ practiced in a rural area (p=.036) Table 13. Findings - Medical School Interactions MEDICAL SCHOOL Regression of all Medical School Variables plus High School AND dummy variables for all medical school variables plus high school % Of Cases Included 34.5% Significant Variables - Sector (p=. 041) NO INTERACTION EFFECT Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 14. Findings - Residency Program Interactions RESIDENCY PROGRAM Regression of all Residency Program Variables plus High School AND dummy variables for all residency program variables plus high school % Of Cases Included 34.6% Significant Variables HS * Required Rural Rotation (p=.023) HS * Elective Rural Rotation (p= 049) HS * Director trained/practiced in rural (p= 011) Required OB (p=.024) LOTS OF INTERACTION The first set of regressions resulted in the following results for the three categories of independent variables: Individual Characteristics Regression 39.9% of the cases were included in the logistic regression of independent variables of high school and gender data against the dependent variable of rural vs. urban practice location. The only individual characteristic that was significant was whether the physician graduated from a high school located in an area with RUCA scores of 4 or greater, indicating a rural area, with a Wald significance of .000 and an odds ratio of 5.148. The odds ratio represents the ratio of the odds of an event occurring in one group to the odds of it occurring in another group, with a ratio of “1” indicating greater likeliness in the first group to a second group and an odds ratio of less than “1” indicating that the event is less likely in the first group. This odds ratio indicates that if you are a family physician who matriculated from a rural 77 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. high school you are five times more likely to be practicing in a rural area. Nevertheless, the Cox and Snell and Nagelkerke R Square tests indicated a weak relationship, explaining only 3.8 to 7.6% of the variance. Even if we use the most liberal definition of the Wald significance of .1 due to the conservative nature of Wald, the Gender variable is still not significant with a result o f. 197. Case Processing Summary Unweighted Cases(a) N Percent Selected Cases Included in Analysis 562 39.9 Missing Cases 848 60.1 Total 1410 100.0 Unselected Cases 0 .0 Total 1410 100.0 Model Summary Step -2 Log likelihood Cox & Snell R Square Nagelkerke R Square 1 372.555(a) .038 .076 Variables in the Equation B S.E. Wald df Sig. Exp(B) Gender -.363 .281 1.662 1 .197 .696 Rur HS RU 1.639 .347 22.318 1 .000 5,148 Constant -2.138 .192 124.121 1 .000 .118 78 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Medical School Characteristics Regression 47.8 percent of the cases were included in this regression of nine independent variables and one dependent variable. There is a weak relationship with only 2.5 to 5.1 percent of the variance accounted for. Of the independent variables, three were significant with a Wald of less than .05, those being “Admissions policies seeking rural” (.005 and odds ratio of .293), whether the medical school is public or private (.022 and odds ratio of 2.045) and whether the medical school is in the top forty of federally funded research institutions (.026 and odds ratio of 2.045). This indicates, although weakly, that individuals who attend medical schools who admit individuals from rural areas are only slightly more likely to end up practicing in rural areas; that public medical schools are two times more likely to graduate family physicians who later are recruited into rural areas, and that medical schools that are not in the top 40 of NIH funded medical schools are also two times more likely to graduate family physicians who later are recruited into rural areas. Case Processing Summary Unweighted Cases N Percent Selected Cases Included in Analysis 674 47.8 Missing Cases 736 52.2 Total 1410 100.0 Unselected Cases 0 .0 Total 1410 100.0 79 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Model Summary Step -2 Log likelihood Cox & Snell R Square Nagelkerke R Square 1 449.179(a) .025 .051 Variables in the Equation B S.E. Wald df Sig. Exp(B) Step 1(a) SOMmis -1.286 1.144 1.264 1 .261 .276 Adminjrur -1.229 .441 7.777 1 .005 .293 Admin^gen .740 .435 2.898 1 .089 2.096 ManJFMC -.142 .430 .109 1 .741 .867 ManCC .399 .381 1.097 1 .295 1.491 Manrur .373 .562 .439 1 .507 1.451 SpRurCu .792 .345 5.275 1 .022 2.207 Sector .716 .321 4.955 1 .026 2.045 Fed_funds .178 .319 .311 1 .577 1.195 PercFM .068 .044 2.428 1 .119 1.071 Constant -2.844 .509 31.178 1 .000 .058 Residency Program Characteristics 71.1 percent of cases were included in the analysis of residency program variables. Six of the seven residency program variables were included in the regression. This is due to the number of cases that have missing data on the impact of the Balanced Budget Act. When that variable is included in the analysis, the number of cases included drops dramatically to 33.8% and skews the statistical outcomes. Looking at the six remaining variables, only two were Wald significant, those being whether the residency program perceived themselves as having a rural emphasis (.012; odds ration = 1.969) and whether the residency program director 80 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ever practiced or trained in a rural area (.000; odds ratio = 2.189). As with the other regressions, there is a weak relationship, with a Cox and Snell and Nagelkerke explaining 4.6 to 8.5% of the variance. Case Processing Summary Unweighted Cases N Percent Selected Cases Included in Analysis 1003 71.1 Missing Cases 407 28.9 Total 1410 100.0 Unselected Cases 0 .0 Total 1410 100.0 Model Summary Step -2 Log likelihood Cox & Snell R Square Nagelkerke R Square 1 741.062(a) .046 .085 Variables in the Equation B S.E. Wald df Si*. Exp(B) reqjura -.069 .321 .046 1 .831 .934 elecrur .277 .239 1.341 1 .247 1.319 missions .678 .269 6.361 1 .012 1.969 mission .300 .286 1.098 1 .295 1.350 w dirrural .783 .222 12.464 1 .000 2.189 req_ob .181 .226 .644 1 .422 1.199 Constant -2.873 .227 159.735 1 .000 .057 Second Set of Regressions A second set of regressions was conducted running all eighteen independent variables and the dependent variable in order to determine which individual variables are significant in the overall pipeline of developing physicians, regardless of the category they may be in. When this regression was run, only 28.6% of the 81 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. cases were included. However, there is a stronger relationship in this regression, with the Cox and Snell and Nagelkerke explaining 12.3 to 23.3% of the variance. With all of the variables in the equation, there are six variables that have a significant Wald. For the individual categories, only the physician graduating from a rural high school (.000, odds ratio = 5.267) was significant from a pipeline perspective. Within the medical school characteristics, there were four significant variables within the pipeline, those being if the medical school they graduated from has a stated objective of recruiting individuals from rural areas (.005, odds ratio = .193), if the medical school has a mandatory community clinic rotation (.019, odds ratio = 3.016) if the medical school has a special rural curriculum (.022, odds ratio = 2.935) and if the percentage of individuals who graduate from the medical school who match into family medicine residency programs (.005, odds ratio = 1.174). Finally, within the pipeline of training at the residency program level, the only variable that was significant was if the residency program perceived itself to have a rural emphasis (.017, odds ratio = 4.388). Case Processing Summary Unweighted Cases N Percent Selected Cases Included in Analysis 403 28.6 Missing Cases 1007 71.4 Total 1410 100.0 Unselected Cases 0 .0 Total 1410 100.0 82 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Model Summary Step -2 Log likelihood Cox & Snell R Square Nagelkerke R Square 1 249.278(a) .123 .233 Variables in the Equation B S.E. Wald df Si*. Exp(B) req_rura -1.272 .782 2.649 1 .104 .280 elecjrur .483 .462 1.095 1 .295 1.621 missions 1.479 .622 5.655 1 .017 4.388 missionw .607 .532 1.303 1 .254 1.835 dirrural .396 .449 .776 1 .378 1.485 recLpb -1.011 .522 3.743 1 .053 .364 Gender -.364 .353 1.061 1 .303 .695 RurHSRU 1.661 .466 12.735 1 .000 5.267 SOMjmis -21.360 17052.258 .000 1 .999 .000 Admin_rur -1.648 .587 7.880 1 .005 .193 Admin_gen .708 .558 1.614 1 .204 2.031 ManJFMC -.489 .519 .885 1 .347 .614 ManCC 1.104 .472 5.482 1 .019 3.016 M an_rur -.065 .727 .008 1 .929 .937 Sp_Rur_Cu 1.077 .469 5.282 1 .022 2.935 Sector .701 .439 2.550 1 .110 2.015 Fed_funds -.197 .422 .218 1 .641 .821 Perc_FM .160 .058 7.756 1 .005 1.174 Constant -3.199 .722 19.636 1 .000 .041 Third Set of Regressions Working from the results of the second set of regressions where all independent variables and the dependent variables were entered into the regression with no consideration of the broader categories, we next took just the six significant variables from that first run and conducted another regression. 34.5% of the cases were included with 5.5% to 10.9% of the variance being explained based on the Cox and Snell and Nagelkerke R Squares. Four of the six continued to be Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. significant, those being whether the family doctor graduated high school (.000, odds ration=4.164), if there was a mandatory community clinic rotation at the medical school (.014, odds ratio=2.479), if the medical school matched greater than ten percent of its graduating class into family medicine residency programs(.020, odds ratio =1.115) and if the residency program perceives itself as rural (.036, odds ratio=1.879). Case Processing Summary Unweighted Cases N Percent Selected Cases Included in Analysis 1004 71.2 Missing Cases 406 28.8 Total 1410 100.0 Unselected Cases 0 .0 Total 1410 100.0 Model Summary Step -2 Log likelihood Cox & Snell R Square Nagelkerke R Square 1 315.892(a) .055 .109 Variables in the Equation B S.E. Wald df Sig. E*p(B) missions .631 .302 4.377 1 .036 1.879 Rur HS R U 1.427 .392 13.251 1 .000 4.164 Adminrur -.616 .431 2.038 1 .153 .540 M anC C .908 .371 5.981 1 .014 2.479 SpR urC u .601 .394 2.326 1 .127 1.823 PercFM .109 .047 5.439 1 .020 1.115 Constant -3.119 .346 81.331 1 .000 .044 84 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Fourth Set of Regressions - Comparison of Those Who Are from Rural Areas to Those from Urban Areas Next, a set of regressions was conducted to compare those who came from rural high schools to those who came from urban high schools. First, regressions were conducted looking at the variables within each of the medical school and residency program categories, comparing those from rural areas to those from urban areas. For those who went to urban high schools, 81.2% of the cases were included in the medical school regression, with sector being the sole significant variable at the .05 level. There were three variables that were within the .05 to . 10 range, those being an admissions policy focusing on rural (p=.078), mandatory community clinic rotations (p=.081) and greater than 10% of graduates entering family medicine residencies (p=.094). For those who went to rural high schools, 89.4% of the cases were included in the medical school regression, with no variables being significant at the .05 level. Having greater than ten percent of graduates entering family medicine residencies was approaching significance (p=.052). It is important to note that for those who went to rural high schools, the medical schools that graduate a higher percentage of family physicians was almost significant. 85 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Medical School Level Variables Case Processing Summary - Urban Areas Unweighted Cases N Percent Selected Cases Included in Analysis 445 81.2 Missing Cases 103 18.8 Total 548 100.0 Unselected Cases 0 .0 Total 548 100.0 Note: Total cases is 1410, however for those regressions were one category of variables (eg medical school) were selected, the total n was less than 1410 due to missing values. Model Summary Step -2 Log likelihood Cox & Snell R Square Nagelkerke R Square 1 265.895(a) | .027 .059 Variables in the Equation B S.E. Wald Df Sig. Exp(B) SOMmis -19.832 20089.177 .000 1 .999 .000 Admin_rur -1.021 .580 3.097 1 .078 .360 Admm^gen .293 .576 .259 1 .611 1.340 ManJFMC -.475 .510 .865 1 .352 .622 ManCC .862 .493 3.054 1 .081 2.367 Man_rur .040 .834 .002 1 .962 1.041 SpR urCu .642 .463 1.923 1 .166 1.900 Sector .873 .427 4.187 1 .041 2.394 Fedfunds .234 .415 .319 1 .572 1.264 PercFM .093 .055 2.807 1 .094 1.097 Constant -2.886 .600 23.166 1 .000 .056 86 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Case Processing Summary - Rural areas Unweighted Cases(a) N Percent Selected Cases Included in Analysis 42 89.4 Missing Cases 5 10.6 Total 47 100.0 Unselected Cases 0 .0 Total 47 100.0 Model Summary Step -2 Log likelihood Cox & Snell R Square Nagelkerke R Square 1 43.258(a) .187 .264 Variables in the Equation B S.E. Wald Df Sig. Exp(B) SOM_mis -21.286 40192.970 .000 1 1.000 .000 Admin rur -2.260 1.425 2.516 1 .113 .104 Admin_gen 1.287 1.303 .975 1 .323 3.623 ManJFMC -.983 1.579 .388 1 .533 .374 ManCC .713 1.178 .366 1 .545 2.040 Manrur .156 1.396 .012 1 .911 1.169 SpR urC u -.003 1.310 .000 1 .998 .997 Sector .967 1.446 .448 1 .503 2.631 Fedfunds -.422 1.126 .141 1 .708 .656 PercFM .319 .164 3.788 1 .052 1.375 Constant -.763 1.750 .190 1 .663 .466 Residency Program Level Variables With the category of residency program variables, 81.8% of the cases for those graduating from urban areas and 85.1% of those graduating from rural areas were included in the analysis. For those from urban areas, only one variable, having four or more OB rotations, was significant. For those individuals coming 87 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. from rural areas, multiple variables were significant, with some having very high odds ratios. Required rural rotation was significant (.008, odds ratio = .000), elective rural rotation was significant (.030, odds ratio = 38.664), a more subjective mission of rural, not written in a program’s mission statement was significant (.046, odds ratio = 79.827) and having a program director that trained or practiced in a rural area was significant (.023, odds ratio = .011). In addition, having a mission statement with the word “rural” in it was nearly significant (p=.078). For those who graduate from rural high schools, completing a residency program that has more obstetrics rotations than required by the RRC was not significant. What is particularly interesting is the odds ratio, with an odds ratio of 1 implying that the event is equally likely in both groups; an odds ratio greater than one implying that the event is more likely in the first group; and an odds ratio less than one implying that the event is less likely in the first group (http://www.cmh.edu/stats/definitions/or.htm, accessed October 9,2005). Based on this, if you graduated from a rural high school and completed a family medicine residency program with an elective rural rotation, s/he is almost thirty nine times more likely to end up in rural practice. Similarly, if the residency program has a non-written mission of rural and s/he came from a rural background, s/he is almost eighty times more likely to end up practicing in a rural area. 88 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Case Processing Summary Rural High School Unweighted Cases N Percent urban Selected Cases Included in Analysis 448 81.8 Missing Cases 100 18.2 Total 548 100.0 Unselected Cases 0 .0 Total 548 100.0 rural Selected Cases Included in Analysis 40 85.1 Missing Cases 7 14.9 Total 47 100.0 Unselected Cases 0 .0 Total 47 100.0 Model Summary Rural High School Step -2 Log likelihood Cox & Snell R Square Nagelkerke R Square urban 1 273.245(a) .051 .105 rural 1 38.510(b) .283 .389 Variables in the Equation Rural High School B S.E. Wald df Sig. E*P(B) urban req_rura -.895 .789 1.287 1 .257 .408 elecrur .230 .422 .297 1 .586 1.259 missions .935 .606 2.383 1 .123 2.547 missionw .829 .552 2.255 1 .133 2.291 dirrural .639 .420 2.313 1 .128 1.895 req_ob -1.001 .444 5.077 1 .024 .368 Constant -2.721 .377 51.952 1 .000 .066 rural req_rura -8.004 3.036 6.952 1 .008 .000 elecrur 3.655 1.684 4.710 1 .030 38.664 missions 4.380 2.196 3.978 1 .046 79.827 missionw 3.085 1.748 3.114 1 .078 21.864 dirrural -4.539 1.992 5.192 1 .023 .011 req_ob -1.570 1.542 1.036 1 .309 .208 Constant -1.938 1.324 2.145 1 .143 .144 Second, a set of regressions was conducted to compare the two groups includes all of the variables in the pipeline. When all of the variables are analyzed Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. against each other to reflect the educational pipeline, 67% of the urban high school graduate cases were included and 76.6% of the rural graduate cases were included. However, the results for the rural graduates were inconclusive. For the urban graduates there are significant variables in the educational pipeline. Two of the significant variables in the pipeline are at the medical school level: special rural curriculum (.028, odds ratio = 3.031) and percentage of graduates entering family medicine residency (.014, odds ratio =1.158), with one medical school variable being right at the cusp of significance, that being if there were a mandatory community clinic experience (.051, odds ratio = 2.762). One variable, having a residency program director who practice or trained in a rural area (.050, odds ratio = 2.634) was significant at the residency program level. Pipeline Variables Case Processing Summary - Urban areas Unweighted Cases N Percent Selected Cases(a) Included in Analysis 367 67.0 Missing Cases 181 33.0 Total 548 100.0 Unselected Cases 0 .0 Total 548 100.0 Model Summary Step -2 Log likelihood Cox & Snell R Square Nagelkerke R Square 1 212.868(a) .082 .169 90 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Variables in the Equation B S.E. Wald df S i* Exp(B) req_rura -.558 .846 .435 1 .510 .573 elecrur .040 .502 .006 1 .937 1.041 missions 1.100 .682 2.603 1 .107 3.004 missionw .399 .617 .418 1 .518 1.490 dirrural .969 .495 3.837 1 .050 2.634 reqob -.885 .553 2.557 1 .110 .413 Gender -.306 .380 .648 1 .421 .737 SOMmis -19.694 21424.779 .000 1 .999 .000 Adminjrur -1.156 .644 3.223 1 .073 .315 Admin_jen .342 .629 .296 .586 1.408 ManFMC -.354 .548 .418 1 .518 .702 ManCC 1.016 .520 3.817 1 .051 2.762 Manrur -.039 .927 .002 1 .967 .962 Sp_Rur_Cu 1.109 .503 4.858 1 .028 3.031 Sector .649 .463 1.964 1 .161 1.914 Fedfunds -.082 .450 .033 1 .856 .921 PercFM .147 .060 6.087 1 .014 1.158 Constant -3.351 .776 18.638 1 .000 .035 Case Processing Summary - Rural Areas Unweighted Cases N Percent Selected Cases(a) Included in Analysis 36 76.6 Missing Cases 11 23.4 Total 47 100.0 Unselected Cases 0 .0 Total 47 100.0 Model Summary Step -2 Log likelihood Cox & Snell R Square Nagelkerke R Square 1 2.773(a) .698 .969 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Variables in the Equation B S.E. Wald df % Exp(B) req_rura elec rur -130.733 47464.517 .000 1 .998 .000 2507067270966 missions 74.602 17390.147 .000 1 .997 5150000000000 0000000.000 2182352553994 67.555 45590.571 .000 1 .999 9790000000000 0000.000 missionw 93.248 20549.302 .000 1 .996 3.141719644007 336E+40 dir_rural -94.204 41422.587 .000 1 .998 .000 req_ob -30.143 26392.446 .000 1 .999 .000 Gender -18.925 11956.651 .000 1 .999 .000 B S.E. Wald df Sig. Exp(B) SOMjmis .228 78434.430 .000 1 1.000 1.256 Admin_rur -58.140 52526.246 .000 1 .999 .000 Adminjjen 1.640 62518.658 .000 1 1.000 5.155 ManFMC -29.395 36892.703 .000 .999 .000 Man_CC 35.377 28349.134 .000 1 .999 2313163304205 014.000 Manrur -16.122 18585.107 .000 1 .999 .000 Sp_Rur_Cu -36.536 20404.913 .000 1 .999 .000 Sector 13.349 9482.696 .000 1 .999 626989.890 Fed_funds .117 15821.688 .000 1 1.000 1.124 Perc_FM 5.043 3348.558 .000 1 .999 154.888 Constant -7.061 35112.805 .000 1 1.000 .001 Fifth Set of Regressions In an effort to further understand the dynamic occurring with those who came from rural areas and those that came from urban areas, a fifth set of regressions were conducted at the medical school and residency program levels where interaction terms were created to look for an interaction effect. Interaction effect is defined as “the differing effect of one independent variable on the dependent variable, depending on the particular level of another independent variable” (Cozby, 1997,314). In this study, we are looking at the differing effect Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. of each of the categorical variables depending on urban or rural upbringing. To do so, dummy variables were created by multiplying RUR_HS_RU with each of the independent variables in a given category. For this regression, the original independent variables were run in a logistic regression for the medical school category and for the residency program category. Medical School Independent Variables and Interaction Terms Case Processing Summary Unweighted Cases N Percent Selected Cases Included in Analysis 487 34.5 Missing Cases 923 65.5 Total 1410 100.0 Unselected Cases 0 .0 Total 1410 100.0 Variables in the Equation B S.E. Wald D f Sig. Exp(B) Rur HS RU by SOM mis -1.454 44933.839 .000 1 1.000 .234 Rur HS RU by Adminrur -1.239 1.538 .649 1 .421 .290 Rur HS RU by Admin_gen .994 1.425 .487 1 .485 2.703 Rur HS RUby Man FMC -.509 1.659 .094 1 .759 .601 Rur HS RU by Man CC -.149 1.277 .014 1 .907 .862 Rur HS RU by Man rur .116 1.626 .005 1 .943 1.123 Rur HS RU by SpRurCu -.645 1.389 .216 1 .642 .524 Rur HS RU by Sector .095 1.508 .004 1 .950 1.099 Rur HS RU by Fedfunds -.656 1.200 .299 1 .584 .519 Rur HS RU by Perc_FM .226 .173 1.705 1 .192 1.253 RurHSRU 2.123 1.850 1.316 1 .251 8.355 SOMmis -19.832 20089.177 .000 1 .999 .000 93 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Adminjrur -1.021 .580 3.097 1 .078 .360 Admingen .293 .576 .259 1 .611 1.340 ManJFMC -.475 .510 .865 1 .352 .622 ManCC .862 .493 3.054 1 .081 2.367 Man_rur .040 .834 .002 1 .962 1.041 SpRurCu .642 .463 1.923 1 .166 1.900 Sector .873 .427 4.187 1 .041 2.394 Fedfunds .234 .415 .319 1 .572 1.264 PercFM .093 .055 2.807 1 .094 1.097 Constant -2.886 .600 23.166 1 .000 .056 34.5% of the cases were included for the medical school category, resulting in one significant variable, that being Sector (.041, odds ratio = 2,394). This was not one of the interaction terms; hence, there are no significant interactions due to growing up rural or urban that occurs at the medical school level. Residency Program Variables and Dummy Variables Case Processing Summary Unweighted Cases N Percent Selected Cases Included in Analysis 488 34.6 Missing Cases 922 65.4 Total 1410 100.0 Unselected Cases 0 .0 Total 1410 100.0 Variables in the Equation B S.E. Wald df Sig. Exp(B) Rur_HS_RU by req_rura -7.109 3.137 5.137 1 .023 .001 Rur_HS_RU by elec rur 3.425 1.736 3.890 1 .049 30.714 Rur_HS_RU by missions 3.445 2.278 2.287 1 .130 31.344 Rur_HS_RU by missionw 2.256 1.833 1.515 1 .218 9.545 Rur_HS_RU by dirjrural -5.178 2.036 6.469 1 .011 .006 94 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Rur_HS_RU by req_ob -.569 1.605 .126 1 .723 .566 req_rura -.895 .789 1.287 1 .257 .408 elec_rur .230 .422 .297 1 .586 1.259 missions .935 .606 2.383 1 .123 2.547 missionw .829 .552 2.255 1 .133 2.291 dir_rural .639 .420 2.313 1 .128 1.895 req_ob -1.001 .444 5.077 1 .024 .368 Rur_HS_RU .782 1.376 .323 1 .570 2.187 Constant -2.721 .377 51.952 1 .000 .066 34.6% of the cases were included for the regression on the residency program category variables. Unlike the medical school category, here we see three of the interaction terms exhibiting significant interactions, those being the interaction of high school location and a required rural rotation (.023, odds ratio = .001), an elective rural rotation (.049, odds ratio=30.714) and if the director trained or practiced in rural (.024, odds ratio = .368). Required obstetrics rotation, although not an interaction, was also significant at .024. Qualitative Analysis Context - Community Characteristics Does the rural community itself play a part in the recruitment of rural physicians? Coffman et al. (2002), through a series of six case studies of rural communities in California, found that certain characteristics of the local healthcare entity influenced the decision of practitioners to practice in that rural community. Specifically noted were characteristics such as the financial solvency of the recruiting entity, the perceived competence of the administrator and the board of 95 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. directors, the presence of a support network of other primary care providers, the proximity to referral hospitals and specialists, and finally, relationships with academic health centers and primary care residency programs. There are other community level characteristics that may impact a practitioner’s decision to practice in that rural community. For example, can they receive loan repayment either through the State or the National Health Services Corp by practicing in that community? Can the community help with the loan repayment or offer loan forgiveness? Pathman et al. (2000) found no relationship between level of debt and a physician’s decision to practice in a rural versus and urban location, though this finding was contradictory to findings of Mantovani (1976) and others. Surveys and consent forms were mailed (see Appendix B), with a cover letter explaining the research project, to the administrators of all sixty-nine rural hospitals in California. A follow up mailing was completed also. Twenty eight hospitals completed the survey. This survey focused on the following general categories of questions: • Structure of the facility • Perception of the facility by community, board and staff • Perception of the administration by community, board, and staff • Perceived financial status • Perceived physician need in community • Recruitment strategies utilized in the past 96 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. • Ability and willingness to assist with loan repayment • Proximity to referral hospitals, specialists, and other primary care physicians • Relationships with academic health centers or training programs • Compensation paid to physicians In addition, a focus group of rural hospital CEOs was conducted at a state rural health meeting in March, 2003 in regard to what rural hospital administrators think should be components of family physician training in order to prepare them for rural practice. Residency Program Surveys Of the twenty three California based residency programs that responded to the survey, there were certain themes. The vast majority (n=21) have no interest in creating a rural training track or component. The most common reason given for this was that it was outside the mission of the Sponsoring Institution, with the “sponsoring institution” being defined by the ACGME as “the institution that assumes the ultimate responsibility for a program of GME” (http://www.acgme. org/acWebsite/GME info/gme_sect2.asp#gloss, accessed 1.28.05). The only respondents that answered positively that they had some of the capacity, ability, willingness, and incentive to create a rural program were the two programs that already have some sort of rural program going on. Uniformly, the balance did not see themselves having the interest of the residents or their sponsoring institutions, 97 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the knowledge of the federal regulations on how to establish such a curricular component, the funds, or the incentives. Finally, amongst the residency program survey responses, there was frequently a mismatch between the objective categories (e.g., rural in the mission statement, mandatory rural rotations, etc.) and the subjective categories (e.g., the program director’s opinion and what they emphasized in their survey response. Hospital Surveys The twenty eight hospital CEOs who responded to the survey had a couple of common themes and issues. The most significant finding from the survey data was the virtually non-existent relationships in California between family medicine training programs and rural communities. Of the twenty eight, only three stated that they had a relationship with a family medicine training program. Of those, one facility had a relationship with an osteopathic family medicine training program that is not accredited by the ACGME. It was not much better for the rural hospitals regarding any relationship with an academic institution. Of the 28 respondents, only 12 had any relationship with a California medical school. Of those 12, the nature of the relationship for eight of them was solely or primarily one of a transfer agreement between the two hospitals for the efficient transfer of patients needing a higher level of care than what is available in a rural emergency room. A transfer agreement is a weak relationship - a contract on file for purpose of defining the capabilities, capacities, and expectations between two facilities in situations where 98 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. a patient’s acuity is beyond the scope of care available at the referring facility. Rural hospitals, as all hospitals, must have such agreements on file with any other hospital they routinely transfer patients to. As such, a transfer agreement between a rural hospital and an academic health center is probably one of many such agreements on file at the rural hospital The remaining four facilities that had a relationship with an academic health center did so primarily for telemedicine and continuing medical education purposes. Other interesting findings from the surveys were that more than 50% of the respondents have difficulty recruiting primary care physicians, and that to do so, the facilities primarily rely on search firms and income guarantee arrangements - which minimize the financial risk for the provider. Recruitment budgets varied, but the number one response was that there were no funds budgeted for recruitment. Foeus Group Findings Ten individuals participated in a focus group in Yosemite, California in March, 2003. The purpose of the focus group was to determine what rural hospital administrators and community leaders want in the training of family physicians that may end up in rural areas for practice. When designing curriculum for rural physician training programs, it is not uncommon to hear, or read, that the curriculum needs to be heavy in obstetrical and procedural experiences in order to suitably prepare the family physician for future rural practice. This tends to be the perspective and belief of the academic family physician. However, what is 99 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. unknown is what the actual rural community, particularly the hospital CEO who looks to this individual to provide ER services, skilled nursing services, outpatient services and/or inpatient services, needs the physician to know. The responses of the participants were enlightening, at times supporting the traditional academic family physician perspective but often contradicting it. They wanted their doctors to be trained in obstetrics and procedural skills, but not to the exclusion of interpersonal skills, personality traits, community engagement skills, and business skills. They expressed that every rural community is different, with different needs from the population. Because of this, some communities have no need for a family physician to do OB, let alone be cesarean section capable. However, what universally they all needed were physicians who were well trained in behavioral medicine, since most rural communities do not have access to psychiatrists and/or psychologists. They wanted people who had clinical skills, but more importantly could show respect. As they put it, “patients don’t understand “standard of care” but they do understand ‘respect’. They need physicians who are open, honest, creative, independent, flexible, and willing to participate in a team approach to patient care and community well being. Team building skills, community organizing skills, an understanding and belief in the principles of community oriented primary care, and cultural competency were all high on their list of requisite training - just as high if not higher than OB and procedural skills training. Ultimately, they emphasized the need to construct new models of training. They want to “grow their own” or if not that, identify individuals 100 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. interested in rural, “plunge them (into the rural environment rather than the urban academic environment), give them support and train them in the rural community”. This is in sharp contrast to what was seen in California’s existing residency programs as it pertains to their rural rotations, where required rural rotations vary in length and structure. Three residency programs reported that they required rural rotations in their survey response. The length of these required rotations ranged from 4 days to one month to a longitudinal family practice center experience. With the exception of the longitudinal family practice center experience, where the outpatient clinic is located in a rural area, neither the four-day experience or the one month experience represents anything one could call a “plunge” into rural practice. Summary of Results Multiple regressions provided the opportunity to look at each category of predictor, the pipeline, a comparison of those who came from rural vs. those who came from rural areas, and an understanding of the extent of interactions between where a family physician originated and the independent variables within their broad categories. The one of the two original two hypotheses proved true, that being the individual characteristic of graduation from a rural high school, as a proxy for growing up in a rural community, were the most significant variable of all variables in the pipeline. More so than any other variable, high school proved to be significant across the multiple regressions, resulting to additional analysis being conducted to determine the extent of interaction of this variable with the other 101 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. independent variables and whether the significant independent variables are different for those who came from rural areas versus those who came from urban areas. The second hypothesis did not prove to be true in that there were some residency related variables which did impact a physician’s decision to practice in rural areas. However, the residency variables that impacted a physician’s decision to practice in a rural area were not the variables that would have been affected by the historic federal and state policy interventions, but the variables that speak to the behaviors and attributes of the program director of a residency program. In other words, our historic federal and state policy interventions to address maldistribution by focusing at the residency program level have been misguided, trying to address structural issues rather than issues of passion as demonstrated by program leadership and the self perception of the program. 102 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CHAPTER 5 CONCLUSIONS AND IMPLICATIONS Summary and Conclusions Forty five percent of rural residents in California live in Health Professions Shortage Areas (HPSAs) (Coffman, et al.; 2002) - a statistic that mirrors the twenty eight survey responses of the State’s rural hospital administrators where fifteen stated that they experience difficulty in recruiting primary care physicians and twenty stated that they would be needing one or more primary care physicians in the next five years. HPSAs are designated based on a number of factors, one of which is the number of primary care physicians in the census tract. In rural California, there is nearly half as many physicians per resident as in urban and in some counties, the primary care physician to population rations run anywhere from 1:2,500 up to certain frontier regions that are 1:5,000 (Hill, 2002; CRHPC, 2003). In the modified words of Apollo 13 pilot James Lovell, “Sacramento, we have a problem.” Our current strategies, if one could call them that, are not fixing the problem. For the five year period studied, 12.5% of the family physicians for which data was available went into rural practice - and not necessarily in California. Being that family physicians are the specialty most likely to go into rural practice, the percentage of other disciplines training in California that go into 103 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. rural practice will be significantly less. Supply is not meeting demand. And, as the projected physician shortage materializes (COGME, 2003; Cooper 2004), the shortage in rural will only get worse. For example, detailed below is a chart of the medical schools in California and their numbers of graduates, comparing those who chose to practice in rural with those who chose to practice in urban. Table 15. Production of Rural Vs. Urban Physicians by the California Medical Schools California Medical Schools: Rural and Urban Production S 1 R u ra l a n d U rb a n This study attempted to understand two things: (1) What variables matter when compared to all the variables that have been demonstrated to matter along the pipeline; and (2) What about residency programs can predict rural placement. 104 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission Along the way, the study also tried to show that communities have an important and usually overlooked role in this pipeline. All three, residency program structure, the pipeline, and communities, will be addressed in this chapter. Residency Program Structure The ACGME requires many things of family medicine training programs. Physical plant, funding, medical staff support, curricular components and patient volumes. The Balanced Budget Act of 1997 and the Balanced Budget Refinement Act of 1999 had a stated goal of trying to fix the maldistribution of physicians by creating certain incentives for developing rural training tracks. The legislation provided for hospitals to be able to increase the number of residents they got paid for if they were to create a rural training track. This was not enough of an incentive for any of the residency programs surveyed as not one created a rural track, let alone did it because of the BBA. The State of California past the Song-Brown Family Physician Training Act to address the shortage of Family Physicians in the State by allocating funds to training programs that put a higher priority on that issue (California Health and Safety Code Section 128200-128240). Based on the logistic regression of residency program variables, the ACGME, the Federal government, and the State of California could have had more of an impact on the shortage of family physicians in rural areas by requiring and funding “passion.” The analysis showed that structure - such as the number of obstetrics rotations and 105 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. whether there was a mandatory rural rotation, in general did not matter. What mattered was whether a residency program “saw” itself as rural. It didn’t matter if rural was written in the mission statement. What mattered was if the program was in some way focused on rural issues, believed in addressing the rural workforce shortage, or otherwise possessed a culture that was committed to rural practice. This happened in programs where the program director had either trained or practiced in a rural area. Who is better to be passionate about rural than someone who trained there? Once the analysis was broken down to compare those from rural backgrounds to those from urban backgrounds, structure started to matter. For “urbans,” having more OB rotations is predictive, yet is the only factor that is predictive against all other residency program variables, and is not predictive within the pipeline of urban medical student training. In terms of residency programs, there are four things that this study adds to the existing body of knowledge. Being that there is little in the literature on variables at the residency program level and their impact on rural placement, this is exciting. The first thing discovered is that contrary to common belief, structure is not relevant as a predictor of rural placement. This is contrary to how our system currently works and the State and Federal policy interventions that have been implemented. Second, it shows us that culture and commitment - possibly driven by the passion of the residency program director who trained in rural, matters. Third, and possibly the most controversial, this study shows us that if a focus on residency training to address rural workforce needs is only a viable option if a focus is made on 106 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. recruiting individuals who come from rural areas, since the fourth set of logistic regressions indicated that for those from urban areas, only increased obstetrics at the residency program variables had any relationship with rural practice decisions, but for those from rural areas the variables did have multiple significant relationships. Fourth, this study opens a new door for future research in order better understand the behavior and attributes of residency program directors that interact with those from rural high schools to nurture and inspire them to practice in rural areas. Medical School On the whole, rural physicians are family physicians (COGME, 2000). In order to understand the issue of rural physician recruitment we must also understand how the pipeline produces, or doesn’t produce, family physicians. Unlike the other medical disciplines, family physicians are more likely to have grown up in an inner city or rural area and have made the decision to be a family physician before entering medical school. They tend to be highly influenced by their personal values. Other specialties are less influenced by personal values, with their career decisions being influence by financial considerations and/or medical school experiences (Xu, et al., 1996). In other words, we should be able to tell, prior to medical school admission, which individuals will most likely become family physicians and possibly rural physicians. This may explain why there was only one significant medical school variable when all significant variables were put 107 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. into the logistic regression. Schools that match a higher percentage of individuals into family medicine residency programs may have a reputation of being a family medicine supportive academic environment. Such a reputation may be more valuable to attracting college grads with the desire to be a family physician than other variables that were initially significant, such as admissions policies with the word rural in it or a mandatory community clinic rotation. Unfortunately, there is a declining number of people, interested in family medicine, who are entering medical school. From 1998 to 2003, there was a significant decline in students going into family medicine, from 16% to 10.4% (Newton, 2003). With fewer individuals going into family medicine, the physician shortage in rural areas will only increase. Complicating this issue is the Hopwood Case, which prohibits giving admission preferences to minority students. The impact to individuals from rural areas is unknown, with many medical schools now eliminating any blanket procedures for any population or demographic group (Basco, 2003). We know from this research study that physicians who graduated from rural high schools are five times more likely to practice in rural areas, any policy that limits special consideration will have negative impact to the rural workforce. This study show us that medical school served as a control valve, not a conduit, for channeling physicians into rural areas. 108 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Community - The Value of the Voices With a family doctor being five times more likely in this study to end up in a rural area if they came from a rural area, it becomes clear that the rural community represents the two pieces of bread in this sandwich - the beginning and the end. Yet, historically, the community voice is absent, leaving us with a slice of meat and some cheese. We treat modem education like a factory - filling our students with technical information along the assembly line. Assuming one agrees that it is ok to educate like an assembly line, then one must also apply assembly line thinking to it. Saturn doesn’t design a car without first understanding what their customers’ want Mini Cooper allows its customers to “build their own” online, check on the status of their car as it is moving down the assembly line like a proud parent would, and then receive the car that meets the customer’s specific needs and expectations. Juran and Deming gave birth to the total quality management movement and one of the early adopters was the automotive industry. A focus on the customer is a core principle of TQM and companies, like Mini Cooper, have made an art of keeping the eye on the customer despite the assembly line nature of the industry. Discussion Starr (1982) wrote how the medical profession has elevated itself above the simple marketplace rales of the exchange of goods and services and that in doing so determines its success, its quality, and its outcomes by the voice of peers, and 109 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. not by the voice of customers. The practice of medicine evolved into a sovereign profession in part due to the role of markets. Medicine, like other disciplines, has a self-perception of being above markets and the simple exchange of goods and services. Because it is a discipline that transcends markets, it requires a higher standard of conduct - a code of conduct established by peers, not by patients (Starr, 1982). The concept of “peer to peer” decision making and regulation is pervasive within the culture and systems of medicine. Whether studying referral patterns or peer review, governance rules or legal structures, the clear message throughout the business and practice of medicine is that this is a process by physicians where the key decisions are made by physicians, with a view that a customer of the profession is not the market, not the patient, but the physician. The profession emereged a focus not on the individual patient, but on the individual physician. This is not to imply that the individual physician does not care about the patients. On the contrary, the majority of physicians demonstrate strong commitments to their patients. It is the profession, which represents the collective interests of all the physician members, that has lost site of who the customer so that it can maintain its place outside the market. Overtime the market has evolved into a threat to the profession (Starr, 1982; Ranade, 1998; Daniels, 1984). The market represents competition and loss of control to the profession. This fear of the market has galvanized and bonded the profession together, leaving the patients and the communities not as customers, but as resources to achieve the profession’s means of self-preservation. The outcome of this is a gap between what the communities 110 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. need and what the medical schools and residency programs are providing. California needs to hold a mirror up to itself and revisit its social contract since our medical school training model does not reflect our population needs. The same holds true for the educational arm of the professions - medical schools. The current dynamic in academic medicine is an emphasis not on education, not on clinical service, but on biomedical research at the expense of humanism (LeBaron, 2004). Robert Bellah, in his book Habits of the Heart, speaks of his vision for academics, focusing on the need for universities to evolve to a place that it is concerned with the whole of society, both historical and philosophical. He speaks about the value of a holistic approach to education that would not only include technical training, but historical and philosophical learning and an on-going public dialogue. The rise of the “research university,” has resulted in technical and professional skills education being emphasized above the creation of “men of learning.” No longer does the university president teach a “capstone” course during the senior year to tie together all that was learned and put it in the context of what is good for society. Instead, now we have the specialization of academics, which is an individualistic approach. How do we expect to cure society’s ills if are physicians are only taught technical skills? Of course healthcare will be fragmented when physicians are provided only technical training in the absence of a larger societal message and common moral underpinnings. Students may enter medical school wanting to save the world, but as referenced earlier, medical 111 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. schools train them to be individualistic technicians, not men and women of learning. At the residency program level, the current dynamic is about survival. Since July, 2000,41 family medicine residency programs have closed (www.agcme.org. accessed 1.15.2005). Of those, four were closed for reasons other than “voluntary withdrawal.” Over 10% of those programs that were voluntarily withdrawn were based in California. Residency programs have been facing significant financial pressure due to declining hospital margins at the same time that the need for family medicine programs increase in their importance as key access points to the health care system (Pugno, et al., 2000). Therefore, it shouldn’t be a surprise that residency program directors stated in their survey responses that a rural rotation or track would not be considered because in was beyond the mission of their sponsoring institution - sponsoring institutions that are, on the whole, located in urban areas. Short of a clear and convincing return on investment, teaching hospitals cannot afford to develop new programs that may benefit society but not their EBITA. No where, at the medical school or the residency program level, is the focus on producing the types of physicians that the communities need. Instead, it is on survival - either of the individual institution or of the profession of the whole. Sesame Street’s character, the Cookie Monster, has a mission statement of “Rule the World, Get lots of cookies, Eat the cookies, Get more cookies” (Nutzworld.com). Not all that different from the residency program and medical 112 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. school unstated mission of survival. And yet, we saw in this study that it was those family physicians that came from residency programs with a different unstated mission - that of a rural emphasis, that predicted rural placement. The National Dairy Development Board in Gujarat, India, is the largest producer of milk in the world. However, the company does not see itself in the dairy business. Instead, it is in “the society building business” (In Senge, et al.; 2004) They see business not as a goal, but as means to building a better society that is just, fair, and empowering of those with less. When we look at the pipeline for producing physicians, let alone rural physicians, it is about producing more MD/PhDs so that the schools can rank higher on the NIH sweeps. It is about laying low and minimizing costs so that the hospital’s bottom line is not impacted too negatively by the presence of a teaching program. Its about return on investment. It is about getting more cookies. Most of us probably don’t wake up in the morning thinking about return on investment. Its not what makes our hearts sing. It is not what gives us a sense of purpose. Yet, the belief in and tenets of capitalism is a fundamental assumption of our society. Edgar Schein, in his 1990 study on organizational culture, told us that the artifacts - the symbols and structure and systems of an organization were formed by our values, which in turn were formed by our societal assumptions. The values of the U.S. healthcare system are well documented, and mirror the values of the U.S. society. They are professional autonomy, patient autonomy, consumer sovereignty, patient advocacy, expectation of the highest quality of care, and expectation that patients get access to care (Priester, 1992). The core assumptions that provide the 113 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. foundation for our American culture can be seen as a collective ideology of beliefs about how the world operates, what outcomes should be achieved, and how to achieve them. Within a society, these ideologies are bom out of the society’s political and economic systems (Simons, 1997). In America, our economic systems support a rational, utility maximizing ideology, which argues that human nature is driven by individual welfare and self-interest, with moral behavior being a function of individual self-interest. Our political ideology is one of individual rights, freedom, and autonomy (Bellah, 1991). In addition, our political ideology is one of representative democracy as compared to a participative democratic process. In other words, the policy decisions in America are made by an elected few as compared to a participative process of all concerned citizens. Based on Schein’s model, these ideologies - the belief that people are self-interested, the belief that moral behavior is a subset of self-interest, the belief that autonomy, individual rights and freedom are inviolate, and the belief that full participation would be destabilizing to the government, are what comprise the American society’s values and artifacts or symbols of healthcare. McGregor argued that “if people are treated consistently in terms of certain basic assumptions, they eventually come to behave according to those assumptions in order to make their world stable and predictable” (Schein, 1992). These ideologies have consistently been at the core of American society and therefore represent the roots of the society’s values and artifacts. Based on these values it is then no surprise that the American system of healthcare, let alone the rural workforce pipeline is fragmented and out of 114 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. alignment. Healthcare access implies an ideology of “social solidarity” - the concept that individual interests are best served by collective action; in other words, cooperation (Karsten, 1995). Instead, our society evaluates in terms of economics - the value of dollars from bottom line, short-term perspective (Flower, 1997). Individual wealth and personal achievement - financial and human capital - are dominant values in the American society. It is about getting more cookies. These institutions - the “rules of the game” (Ostrom 1994) endure because our economic model keeps reinforcing it. No one has an interest in changing our institutions because it would change in economic equilibrium (Rothstein, 1996). “Institutions are not created to be socially efficient, rather to serve the interests of those with the bargaining power to devise new rules” (North, 1990). Therefore, until society is ready to change its assumptions and shift its values, probably resulting in a different economic model, we will not see fundamental change in the institution of the physician education pipeline. However, that does not mean that all is lost. Institutions change for three reasons (Goodin, 1996), either due to accident or unforeseen circumstances, evolutionary reasons, or intentional design by strategic agents. Just like the Cookie Monster recently went through his own identity crisis, now realizing that cookies must be a “sometimes food,” over time, our societal values and assumptions will shift, or evolve, and as they do we will see a new system of medical education develop. Thereby, a group of people who share a concern about the system as it exists today can come together, design a new system, and implement a change. Together, they can form a community of change 115 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. agents who will work with all of the entities in the pipeline - from community to high school to medical school to residency and back to community, to build a more efficient and effective system - one that reflects values of equity, fairness, empowerment, and humanism. They can bring balance back into medical training, so that rural areas are no longer underserved and so that medical schools have a well rounded focus of sophisticated science and humanism. We have created this system. We maintain this system. We passively concur with the inequities of this system. And we can change this system. But it will take passion, commitment, and leadership. So what do we do? “The whole world is a community...what divides us into warring camps is precisely the lack of this awareness ” M . Scott Peck M. Scott Peck says there are five building blocks for developing communities, those being developing an overarching vision, creating a psychology of abundance, having the capability for vulnerability, creating opportunities to learn from each other, and developing a cadre of leaders and a depth of leadership. Based on the results of this study, Peck’s community building blocks and the research on rural workforce that has come before this, the following recommendations are being made. 116 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. General Recommendations • Gather together the leadership from all the pipeline components -from community through community, and including other key entities such as local, state and federal government representatives. • Together commit that there is no option but to change, or at least acknowledge that addressing societal need by producing rural physicians is a “sometimes food.” • Create a better pipeline - with clear purpose and principles grounded in conviction. • Get the pipeline talking. It is no surprise that rural areas are struggling with workforce when they have no relationships with the residency programs. Help the communities to understand the training programs and the training programs to understand and listen to the communities. Find mechanisms to bridge all of the pipeline components. • The new model is based on culture, commitment and passion. Recruit residency program directors who are from rural areas and have trained or practiced in rural areas to fan the passion of those that enter their programs. • The focus of the pipeline is not on “filling a barrel” with skills, but on “lighting a fire” (Senge et al., 2004) with the students and then nurturing and supporting that fire and passion. 117 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. • Consider models of rural locum tenens support by medical schools and residency programs, looking at the model developed at the University of New Mexico Medical Center. (Larsen et al., 1999). • Identify rural high school students with an interest in family medicine early on. Nurture and support them through the pipeline, reserving slots for them in family medicine friendly medical schools and in family medicine residency programs that are passionate about rural health. • Develop measurable outcomes for the medical schools and residency programs as it pertains to the components and outcomes of the rural pipeline. Align the incentives with the needs and the outcomes. • Identify vehicles for decentralizing education into the rural areas, through the creation of satellite medical schools in rural areas. This will not only address the workforce shortage projections through the creation of additional medical schools (Cooper, 2004) but will also have economic development impact on the rural areas and create a stronger rural pipeline. Similar models have occurred in Austrailia, as part of their rural solution. • We measure what is important to us. If recruitment and retention of rural physicians is important, than greater effort must go into developing a comprehensive, publicly accessible database • Recognize that this is not dichotomy of rural versus urban. We are all rural. We may not live in a RUCA of 4 or greater, but we are connected as we eat the strawberries grown in the rural community, use the energy harnessed in the 118 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. rural region, drink the water and build houses with the timber grown and cut in the rural region. We are connected. • Remember that our solutions of old may no longer work in our new world. Be open to new models, new solutions, and new ways of seeing • Infuse passion, light fires, and foster alignment and balance. • Nurture intention and align it with education • Enhance Graduate Medical Education funding and simply the process and formula for reimbursement. Current GME policies have not provided the requisite incentives to address maldistribution, in part because of its focus on structure. Instead, it would be helpful to simplify the calculations, rules, and auditing processes, increase the funding for residency training, and focus the funding on creating programs with cultures and commitments to rural training and placement. For Medical Schools • The new model of training focuses on meeting the health needs of the communities, the State, and the nation - not on NIH research funding, return on investment, or survival. It focuses on building society. • Diversity and cultural competency is not just about race, ethnicity, arid gender. It is also about “place”. Rural is not small urban. It is a different culture based on different norms and values. Those from rural areas are currently underrepresented in our medical schools. “Place” should be one factor that is 119 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. addressed in recruitment strategies of medical schools to promote diversity and better reflect the fabric of California. For Residency Programs • Recruit program directors that have trained or practiced in rural areas. Better yet, recruit those that were raised in rural areas. • Reach out to rural communities. Include them in developing training models, curriculum, and determining educational outcomes. Solicit the input of the rural communities in residency design and curriculum. For Rural Hospital Administrators • Experientially, it has been the PFs observation that rural administrators and rural physicians struggle to understand each other. Their agendas, let alone values and approaches, are different and too often conflicting. Rural administrators, albeit under significant pressure, must try to focus equally on building their health care community as they do on “return on investment”. Doing so - seeing a bigger picture than finance - will help to bridge the chasm between providers and administration. For Rural Communities • There is great value to your collective voice. Don’t lose sight of that value. Demand it be heard. The rural community is the beginning and end of this training process. The purpose of medical school and residency is to meet your needs. 120 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. • Network. Rural communities too often have very limited networks with weak relationships. This was demonstrated in the survey responses as well. Resources flow through networks, so focus on expanding the network of your community. Limitations The limitations of this study are as follows: 1. It is extremely difficult to obtain data, let alone reliable data, from non United States medical schools. Many residents and rural physicians are graduates of foreign medical schools (Mick & Lee, 1997). However, based on the delays and lack of data experienced by the investigator in the past when trying to obtain original source verification from medical schools in foreign countries, it was decided to collect all other data on as it pertains to California residents and not collect the data on their medical schools. When doing the logistical regression, the data on foreign medical graduates (FMGs) will be utilized for all determinants except medical school. The data will then be rerun for all determinants, eliminating the FMG dataset. This limitation may prove helpful by shedding light on the impact of foreign medical graduates on rural placement. 2. There are no formal Rural Training Tracks in California. Because of this, certain elements of RTTs that also exist in residency programs (e.g., number 121 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. of months of obstetrics training, rural clinical experience, procedural training, etc) will be used as a proxy to analyze this issue. 3. The Balanced Budget Act (BBA) represented sweeping reform for Medicare and significantly impacted Graduate Medical Education (GME). Because of the significance of this legislation, the study will begin looking at graduates of California residency programs since 1997 - the year that the BBA was implemented. No prior-years data will be utilized. 4. Although studies have shown that the expressed interest of a student towards family and/or generalist medicine to be a significant predictor of future rural placement, both as an individual characteristic and as a medical school admissions policy, there is no reliable means of obtaining this data. The sample being studied of 1,410 physicians would need to be surveyed and accurately recall their intent from seven to twelve years prior. Therefore, it was determined to eliminate this variable in the analysis. Gender, however, will be included in the analysis in the Individual Characteristics category due to the lack of conclusive literature on this determinant. Future Research 1 . Future research could be conducted looking at how this analysis would differ if the various definitions of rural were used to code practice sites and high school locations. The answer to this would provide valuable information at 122 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the Federal level as the issue of which definition should be used continues to be discussed. 2. Survey the 1,410 graduates in the database as to their values, including spirituality factors, to determine which personal values predict rural practice. 3. A social network analysis of residency programs should be developed to analyze structural patterns and outcomes. This could provide a “best practice” model of training. 4. A survey of residency program directors to better understand their attributes and behaviors that result in an interaction with their residents that nurtures and inspires them to choose rural practice. 123 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. BIBLIOGRAPHY Retrieved June 21,2005, from http://www.jefferson.edu/psap/home/index.cfm. Retrieved June 21, 2005, from http://www.med.umn.edu/RPAP. Retrieved June 29, 2005, from http://www.uwmedicine.org/Education/WWAMI. Retrieved June 29,2005, from http://www.mercer.edu/publications/Medicine/spr99/).. Retrieved July 3,2005, from http://com.etsu.edu/default.asp?V_DOC_ID=2087&V_LANG_ID: =0. Retrieved July 3,2005, from http://www.usd. edu/press/news/news. cfm?nid=419&uid=user. Retrieved June 21,2005, from http://www.som.tulane.edu/departments/admissions/special.html. Retrieved June 21,2005, from http://www.contracostatimes.com/mld/cctimes/living/health/11559533.html. Retrieved July 3,2005, from http://www. osteopathic.org/ index. cfm?PageID=sir_college. Retrieved June 10, 2005, from www.aamc.org/medicalschools.html. Retrieved June 19,2005, from www.aacom.org/about. Retrieved June 10,2003, from www.aafp.org/residencies Retrieved January 13, 2005, from http://www.ruralhealth.ca.gov/pdf/RUCA%200verview.pdf. Retrieved January 13, 2005, from http://www.fammed.washington.edu/wwamirhrc. Retrieved January, 2005, from http://www.ruralhealth.ca.gov/pdf/RUCA%200verview.pdf. 2005, from (http://chester.fammed.washington.edu/wwamirhrc/national.txt. 124 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Retrieved August 26, 2004, from www. grants 1. nih. gov/grants/award/rank/medtt 103. htm. Retrieved January 13, 2005, from http://www.acgme.org/adspublic. Retrieved January 13, 2005, from http://www. acgme.org/acWebsite/downloads/RRC_progReq/120pr701 .pdf. Retrieved January 28,2005, from http://www.acgme.Org/acWebsite/GME_info/gme_sect2.asp#gloss. (2003). America's Best Graduate Schools 2003. U S News and World Report. (2005). Americaa€™s Best Graduate Schools 2005. U.S. News and World Report AAFP. Retrieved May 16,2003, from www.aafp.org/x759.xml. AcademyHealth Retrieved May 25,2005, from http://www.academyhealth.org/publications/healthoutcomesreport.html. Acosta, D. (2000). "Impact of rural training on physician workforce: The role of post residency education." Journal of Rural Health 16: 254-61. Aday L A (1981). "Equity of Access to Medical Care: A conceptual and empirical overview." Medical Care 19(12 Supplement): 4-27. Alexander, E. (1995). How organizations act together: Interorganizational coordination in theory and practice. Amsterdam, Overseas Publishers Association. American College of Physicians (1995). "Rural Primary Care." Annals of Internal Medicine 122(5): 380-90. American Council of Medical Quality "Definition of Medical Quality." Retrieved May 29,2005, from (http://www.acmq.org/policies/policiesland2.pdf. Austin, J. (2000). The collaboration challenge. San Francisco, Jossey Bass. Bardach, E. (1998). Getting agencies to work together: The practice and theory of managerial craftsmanship Washington DC, Brookings Institute Press. Barzansky, B. P. E., Sylvia I. (2002). "Educational Programs in US Medical Schools, 2001-2002." JAMA 289(9): 1067-1072. 125 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Basco WT, B. S., Duggan AK, Wilson MH. (1998). "Associations between primary care oriented practices in medical school admissions and the practice intentions of matriculates." Academic Medicine 73: 1207-10. Bellah R, (1991). The Good Society. New York, Vintage Books. Bowman RC (1998). "Family practice residency programs and the graduation of rural family physicians." Family Medicine 30: 288-92. Braden J.(1994). Health Status and Access to Care of Rural and Urban Populations. AHCPR Pub No 94-0031, National Medical Expenditure Research Findings 18, AHCPR. Brazeau, N., Potts, MJ, & Hickner, JM. (1990). "The upper peninsula program: A successful model for increasing primary care physicians in rural areas." Family Medicine 22(5): 350-355. Brook, R. (1997). "Managed Care is Not the Problem, Quality Is." JAMA 278(19): 1612-1624. Brooks R, W. M., Mardon R, Lewis M, and Clawson A (2002). "The Roles of Nature and Nurture in the Recruitment and Retention of Primary Care Physicians in Rural Areas: A Review of the Literature." Academic Medicine 77: 790-798. Burnett, W. (2004) Personal Communication. K. Nocella. California State Rural Health Association "Fact Sheet." Retrieved April 3,2003, from www.csrha.org/factsheet.html. Casey, M. (2001). "State HMO Accreditation and External Quality Review Requirements: Implications for HMOs Serving Rural Areas." Journal of Rural Health 17(1): 40-52. Casey, M. C., K. Klinger, J (2001). "Are rural residents less likely to obtain recommended preventative health care services?" American Journal of Preventative Medicine 21(3): 188. Coffman J, R. E., Grumbach K (2002). CP AC Issue Brief: Improving recruitment and retention of primary care practitioners in rural California, Regents of the University of California. COGME (1998). Tenth Report: Physician Distribution and Health Care Challenges in Rural and Inner-City Areas U. D. o. H. a. H. Services. 126 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. COGME (2000). Resource Paper Compendium: Update on the Physician Workforce. U. S. Department of Health and Human Services. COGME (2000). Resource Paper: The Effects of the Balanced Budget Act of 1997 on Graduate Medical Education U. S. Department of Health and Human Services. Cohen, J. (1999). "Consensus Statement on the Physician Workforce." 2004, from http://wwfw.aamc.org/meded/edres/workforc/aamc.htm. Cohen, J. (1999). "Why doctors dont always go where they’re needed." Academic Medicine 73: 1277. Connor, R., Hillson, SD, & Kralewski, JE. (1994). "Association between rural hospital's residencies and recruitment and retention of physicians." Academic Medicine 69(6): 483-488. Cooper, R. (2004). "Weighing the Evidence for Expanding Physician Supply." Ann Intern Med 141(9): 705-14. Cozby, P. (1997). Methods in Behavioral Research, Sixth Edition. Mountain View, Ca., Mayfield Publishing Company. Crouse, B. (1995). "Recruitment and retention of family physicians." Minnesota Medicine 78(10): 29-32. Daniels, N. (1984). "Understanding Physician Power: A review of the Social Transformation of American Medicine." Philosophy and Public Affairs 13(4): 383- 357. Docteur ER, Gold M (1996). "Shifting the Paradigm: Monitoring Access in Medicare Managed Care." Health Care Financ Rev 17(4): 5-21. Eberhardt MS, I. D., Makuc DM, Pamuk ER, Freid VM, Harper SB, Schoenom CA, Xia H (2001). Urban and Rural Chartbook. Health United States 2001, Health, United States. Washington DC, U.S. Government Printing Office. Flower, J. (1997). "Beyond Economics: Healthy Communities and Healthy Economies." National Civic Review 86: 53-59. 127 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Forti, E., Martin, KE, Jones, RL, & Herman, JM (1995). "Factors influencing retention of rural Pennsylvania family physicians." Journal of the American Board of Family Practice 8(6): 469-74. Fryer, G., Stine, C, Vojir, C, & Miller, M. (1997). "Predictors and profiles of rural versus urban family practice." Family Medicine 29(2): 115-118. Fryer GE, S. C., Vojir C, Miller M. (1997). "Predictors and profiles of rural versus urban family practice." Family Medicine 29: 115-118. Hartley D, G. J. (2003). Rural Health Care Safety Nets. Tools for Monitoring the Health Care Safety Net. B. J. Wienick RM, Agency for Healthcare Research and Quality. Hartley D, Q. L., Lurie N (1994). "Urban and rural differences in health insurance and access to care." Rural Health 10(2): 98-108. Homer RD, S. G., Ricketts TC (1993). "Preliminary evidence on retention rates of primary care physicians in rural and urban areas." Medical Care 31: 640-8. Hutten-Czapski, P. (1998). "Rural Incentive Programs: A failing report card." Canadian Journal of Rural Medicine 3(4): 242-247. IOM, Institute of Medicine. (1993). Access to Health Care in America. Washington D.C., National Academies Press. IOM, Institute of Medicine. (2001). Crossing the Quality Chasm: A New Health System for the 21st Century. Washington, DC, National Academies Press. IOM, Institute of Medicine. (2003). Health Professions Education: A Bridge to Quality. Washington DC, National Academies Press. IOM, Institute of Medicine. (2004). The 1st Annual Crossing the Quality Chasm Summit: A focus on communities. Washington D.C., National Academies Press. Karsten, S. (1995). "Health care: Private Good vs. Public Good." The American Journal of Economics and Sociology 54: 122-44. Krieger N, C. J., Waterman PD, Rehkopf DH, Subramanian SV. (2005). "Painting a truer picture of US socioeconomic and racial/ethnic health inequalities: The Public Health Disparities Geocoding Project." Am J of Public Health 95: 312-323. 128 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Lachica, M. (2003). Personal Communication K. Nocella, OSHPD: Information about Song-Brown Funding. Ladenheim K, L. E., Ritter CS (2000). State Purchasing for Health Care Quality: Quality and Accountability Systems in Four States, National Conference of State Legislators. Larsen LC, D. D., Garland JL, et al (1999). "Relief, recruitment, and retention at the University of New Mexico Medical Center." Academic Medicine 74(1): S136-40. LeBaron, S. (2004). "Can the Future of Medicine Be Saved from the Success of Science?" Academic Medicine 79(7): 661-665. Legislative Analyst Office. (2002). HMOs and Rural California. State of California. Looney SW, B. R., Gagel JR, Pentecost MW (1998). "Which medical school applicants will become generalists or rural based physicians? ." Kentucky Medical Journal 96: 189-93. Macinko J, S. B., Shi L (2003). "The Contribution of Primary Care Systems to Health Outcomes within Organizations for Economic Cooperation and Development (OECD) Countires, 1970-1988." Health Services Research 38(3): 831. Magnus, J., & Tollan, A. (1993). "Rural doctor recruitment: Does medical education in rural districts recruit doctors to rural areas? ." Medical Education 27: 250-253. Mantovani RE, G. T., Johnson DG (1976). Medical student indebtedness and career plans, 1974-75. H. R. A. Public Health Service, Bureau of Health Manpower. Washington DC, US Government Printing Office. Mick SS (1997). "The safety net role of international medical graduates." Health Affairs: 141-50. Newton DA (2003). "Trends in Career Choice by US Medical School Graduates " JAMA 290(9): 1179-1182. North, D. (1990). Institutions, Institutional Change and Economic Performance. New York, Cambridge University Press. 129 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Pathman, D. (1996). "Medical Education and Physician Career Choices: are we taking credit beyond our due? Academic Medicine 71(9): 963-968. Pathman DE, K. T., King TS, Spaulding C, Taylor DH (2000). "Medical Training Debt and Service Commitments: The rural consequences." Journal of Rural Health 16: 264-272. Peck, M. S. (1987). The Different Drum. Community Making and Peace. New York, Simon & Schuster. Priester, R. (1992). "A Values Framework for Health System Reform." Health Affairs 11(1): 84. Pugno, P. e. a. (2000). "Entry of U.S. Medical School Graduates into Family Practice Residencies: 1999-2000 and 3-year Summary." Family Medicine 32(8). Rabinowitz, H. (1993). "Recruitment, retention and follow-up of graduates of a program to increase the number of family physicians in rural and underserved areas." New England Journal of Medicine 328: 934-9. Rabinowitz, H. (1993). "Recruitment, retention, and follow-up of graduates of a program to increase the number of family physicians in rural and underserved areas." New England Journal of Medicine(April): 934-989. Rabinowitz, H., MD; James J. Diamond, PhD; Fred W. Markham, MD; Nina P. Paynter, BS (2001). "Critical Factors for Designing Programs to Increase the Supply and Retention of Rural Primary Care Physicians "JAMA 286: 1041-1048. Rabinowitz HK, D. I., Hojat M, Hazelwood CE (1999). "Demographic, educational and economic factors related to recruitment and retention of physicians in rural Pennsylvania." Journal of Rural Health 15: 212-8. Rabinowitz HK, P., NP (2000). "The role of the medical school in rural graduate medical education: pipeline or control valve? ." Journal of Rural Health 16: 249- 253. Ramsey PG, C. J., Hunt DD, Marshal SG, Wenrich MD (2001). "From Concept to Culture: The WWAMI Program at the University of Washington School of Medicine." Academic Medicine 76(8): 765-775. Ranade, W. (1998). Markets and Health Care: A comparative analysis, Harlow. 130 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Ricketts III, T. C. (2002). Arguing for Rural Health in Medicare: A Progressive Rhetoric for Rural America. Chapel Hill, Cecil G Sheps Center for Health Services Research, The University of North Carolina at Chapel Hill. Robert Graham, C. Retrieved November 5,2004, from www.graham- center.org/xl45.xml. Robert Neelly Bellah, W. M. S., Steven M Tipton, Arm Swidler, Richard P Madsen (1985). Habits of the Heart. Berkeley and Los Angeles, California, University of California Press. Robertson, P. (1998). Interorganizational relationships: Key issues for integrated services. Universities and communities: remaking professional and interprofessional education for the next century J. M. a. S. Einbinder. Westport, Ct., Praeger Publishers. Rosenblatt, R. (2004). "A View From the Periphery - Health Care in Rural Americal." New England Journal of Medicine 351(11): 1049. Rosenblatt RA, W. M., Culien TJ, et al. (1992). "Which medical schools produce rural physicians? ." JAMA 268: 1559-65. Rosenthal TC, M. M., Osbome J (1997). "One-two residency tracks in family practice: are they getting the job done? ." Family Medicine 30: 90-3. Rosenthal TC, M. M., Anderson G. (2000). "Rural residency tracks in family practice: Graduate outcomes." Family Medicine 32:174-7. Rourke, J. (1997). "In search of a definition of rural." Canadian Journal of Rural Medicine 2(3): 113-115. Schein, E. (1992). "Organizational Culture and Leadership." Schur C, F. S. (1999). Access to Health Care. Rural Health in the United States. T. C. Ricketts III. New York, Oxford University Press. Senge PM, e. a. (2004). Presence: Human Purpose and the Field of the Future, Society for Organizational Learning Shi, L. (1992). "The relationship between primary care and life chances." J Health Care Poor Underserved Fall 3(2): 321-35. 131 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Shuster ME, M. E., Brook RH (1998). "How Good is the Quality of Health Care in the United States." Milbank Quarterly 76(4): 517-563. Simons, T. a. I., P. (1997). "Organization and Ideology: Kibbutzim and Hired Labor, 1951-1965." Administrative Science Quarterly 42: 784-813. Sipkoff, M. (2004). "The New Consensus Favoring IOM's Definition of Quality." Managed Care. Starfield, B. M. M. F. (2003). Primary Health Care Reform in Contemporary Health Care Systems. Conference on Implementation of Primary Care Reform: Canadian and International Perspectives, Kingston, Ontario. Starfield, B. M. M. F. (2004). Worldwide Issues in Primary Care. 17th World Conference of Family Doctors, Orlando, Florida. Starr, P. (1982). Social Transformation of American Medicine: The Rise of a Sovereign Profession and the Making of a Vast Industry, Basic Books. Steams JA, S. M., Glassner M, Londo RA (2000). "Illlinois RMED: A Comprehensive Program to Improve the Supply of Rural Family Physicians." Family Medicine 32(1): 17-21. Stratton TD, G. J., Ludtke RL, Fickensher KM (1991). "Effects of an expanded medical curriculum on the number of graduates practicing in a rural state." Academic Medicine 66: 101-5. Tepper JD, R. J. (1999). "Recruiting Rural Docs: Ending a Sisyphean Task." Canadian Medical Association Journal 160(8): 1173-1174. University of California, OSHPD (1994). Memorandum of Understanding, State of California. Verby JE, N. J., Andresent SA, Swentko WM (1991). "Changing the medical school curriculum to improve patient access to primary care." JAMA 266: 110-3. Wagstaff, A. (2002). "Poverty and Health Sector Inequalities." Bulletin of the World Health Organization 80(2): 97-105. West PA, N. T., Gore EJ (1996). "The demographic and temporal patterns of residency trained physicians: University of Washington Family Practice Residency Network." Journal of the American Board of Family Practice 9: 100-108. 132 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. WONCA (1995). Policy on training for rural practice. E. b. t. W. W. C. Meeting. Xu G, V. J., Barzansky B, Hojat M, Diamond J, Silenzio VMB (1996). "Comparisons Among Three Types of Generalist Physicians: Personal Characteristics, Medical School Experiences, Financial Aid, and Other Factors Influencing Career Choice " Advances in Health Sciences Education 1(3): 197- 207. Yutrzenka GJ, A. L. (2004). "USDSM Alumni Student Scholars Program: Successes of the First Decade." South Dakota(November): 491-498. 133 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX A SURVEY TO RESIDENCY PROGRAMS Name of Residency Program _____________________ Name of Person completing this survey _____________________ Title of Person completing this survey _____________________ What is your mission statement? Is rural training part of your mission ? Yes____ No___ Do you have a required rural rotation? Yes____ No___ How long is the rural rotation? Months Where are the rural rotations? Please list names of facilities and cities they are in Do you have elective rural rotations? Yes_____ No____ If so, how many residents do the rural rotation each year? ______ How long is the rural rotation?__________________________ ______ Months Where are the rural rotations? Please list names of facilities and cities they are in: Do you have formal agreements with the rural sites? Yes_____ No How many months of Obstetrics do you require? Months 134 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Do you emphasize procedural training? If yes, how? Yes No What changes, if any, were made to your program (its structure or content) due to the passage and implementation of the Balanced Budget Act of 1997? Do you have an ACGME recognized rural training track? Yes No If yes, what was the incentive to create it? __________________________ What were the obstacles you faced in creating it? What are the obstacles you face in maintaining it? Where do you see as its opportunities and growth during the next 5-10 years? How would you describe the relationship between the rural site and the residency program? If not, are you interested in creating a rural training track? Yes No. If yes, what do you see as the barriers to accomplishing this? Time Money Knowledge of the federal regulations Interest of the residents Other (Please provide detail)________________________ 135 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. If no, why not? Beyond the mission of the residency program Beyond the mission of the Sponsoring Institution Program Director not interested in rural Other (Please provide detail)__________________________ Has the Program Director completed any of his/her training in rural areas? Yes No_____ If yes, for how many months?_______ Has the Balanced Budget Act of 1997 provided any incentives to your program to change its structure, rotation sites, or create a rural track or rotation? Yes No_____ If yes, what occurred? Does your program have the ability (e.g., Knowledge, skills, and staffing) to develop/maintain a rural training track? Yes_____ No___ If not, what is lacking? Does your program have the capacity (e.g., mechanisms for participative discussion, collaborative relationships with sites, or existing informal relationships) to develop/maintain a rural training track? Yes________ No________ If not, what is lacking? Does your program have any incentives to develop/maintain a rural training track? Yes________ No________ What are they? How many ambulatory community sites (urban, rural, and suburban) do your residents rotate at? (e.g., non-family practice center ambulatory sites such as community clinics, FQHCs, etc.)._________ 136 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. How many of these do you have formal legal agreements with that govern the terms of the rotation? How are decisions made between the ambulatory community site and the residency program? Have the number of ambulatory community sites changed since 1997-1998, with the passage of the BBA?____________________________________ Yes_____ No____ If so, from what to what? Are there regular forums for dialogue between the ambulatory community site and the program?__________________________________________ Yes_____ No___ Do the ambulatory community sites and the program share.... An aligned mission? Yes_____ No_ Aligned strategies? Yes_____ No_ Aligned values? Yes_____ No_ Clarity of purpose?______________________________ Yes_____ No_ Has the relationship with the ambulatory community sites and the residency program increased public value?__________________________________Yes_____ No_____ If so, how?__________________________________________________________ 137 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX B SURVEY TO RURAL HOSPITAL ADMINISTRATORS Name of Facility _____________________________ Name of Person completing this survey _____________________________ Title of Person completing this survey? _____________________________ How long have you been in this role? _________ Years Months Does your facility have a hospital based rural health clinic? Yes No_ Are there any other rural health clinics in your community? Yes No_ If so, how far away?____________________________________________________ Are they hospital based or investor/provider owned?________________________ Are there any FQHCs in your community? Yes_____ No_ If so, how far away?_________________________________________________ __ Are there any private primary care medical offices in your community? Yes No_ If so, how far away?_____________________________________________________ Are there any Indian health clinics in your community? Yes No_ If so, how far away?_____________________________________________________ Does your community perceive your facility as being financially solvent? Yes No Why? Does your facility directly employ or contract with primary care physicians? For your ER?__________________________________________ Yes_____ No_ Contract______ ? Employee_____ ? For your RHC?________________________________________ Yes_____ No_ Contract______ ? Employee_____ ? For your SNF?_________________________________________ Yes_____ No_ Contract______ ? Employee_____ ? 138 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. For your inpatieut (hospitalist) services? Yes_____ No Contract ?_________________________ Employee_____? How many primary care physicians do you employ/contract with? ______ For how many FT E s?______ How many of these are Family Physicians? ______ Are any of them graduates from foreign medical schools?______ Are you happy with your current arrangement? Yes No If not, how would you change it? __________________________________ Do you need more physicians? Yes_____ No_____ How many FTEs do you currently have and currently need? Have Additional needed_____ How many do you project you will need five years from now?_____ Are the other RHCs, FQHCs, etc in your community fully staffed with primary care physicians? Yes_____ No_____ If not, how many more primary care physician FTEs are needed now? _____ Five years from now? _____ Does the community perceive the board of directors of your facility as competent? Not competent 1 2 3 4 5 Extremely Competent Does the medical staff view the board of directors of your facility as competent? Not competent 1 2 3 4 5 Extremely Competent Do you as the administrator view the board of directors of your facility as competent? Not competent 1 2 3 4 5 Extremely Competent Does your staff view the board of directors of your facility as competent? Not competent 1 2 3 4 5 Extremely Competent 139 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Does the community perceive the hospital administration as competent? Not competent 1 2 3 4 5 Extremely Competent Does the medical staff view the hospital administration as competent? Not competent 1 2 3 4 5 Extremely Competent Does the board of directors view the hospital administration as competent? Not competent 1 2 3 4 5 Extremely Competent Does your staff view the hospital administration as competent? Not competent 1 2 3 4 5 Extremely Competent Are the primary care physicians in your community supportive of new physicians who come to practice in your community? Yes No____ How far away is the nearest referral hospital? Miles_______ Hours Minutes_____ Is it a community hospital or a tertiary hospital? Does your facility have a relationship with an academic health center? Yes No____ If so, which one and what is the nature of that relationship? Does your facility have a relationship with a family medicine residency program? Yes No If so, which one? What is the nature of the relationship? Are vou aualified as a State physician loan repavment site? Yes No Are you qualified as a Federal physician loan repayment site? Yes No Would vour facility or communitv ever offer navment of their medical school loans as additional compensation to attract/recruit a primary care phvsician? Yes No Is recruitment of primary care physicians difficult for your facility? Yes No 140 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. What strategies have you used in the past to recruit? What has proven successful for you?_______ What salary are you offering to primary care physicians that you are recruiting? What other compensation (% benefits, loan repayment, CM E time, travel dollars, moving expenses, etc) have you/would you offer?_________________________ What is your annual budget for recruitment? 141 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX C LETTER TO RESIDENCY PROGRAM DIRECTORS Date Name Title Residency Program Name Address City, State Zip Dear This survey is being sent to you as part of a research study to better understand the factors that predict a family physician’s recruitment into rural communities in California. The study will look at a variety of the traditional determinants that impact rural recruitment; at the individual level, the medical school level, the residency program level, the fellowship level, and the community level, in an effort to understand what determinants have the greatest significance and most importantly, in an effort to develop policies that will help to provide enough family physicians to effectively staff the rural populations in California. Your input is critical to the success of this study. I would greatly appreciate your attention to and completion of this survey, so that it can be returned to me by __________in the enclosed envelope. If you prefer, I will gladly conduct a telephone interview to complete the survey. Enclosed you will also find a consent form. Please read and sign the consent form and return it with your completed survey. If you have questions, I can be reached at 626-457-4283. Thank you for your support. Kiki C. Nocella, MHA, PhD (Cand) Enel: Residency Program Director Survey Consent Form Return envelope 142 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX D LETTER TO RURAL HOSPITAL ADMINISTRATORS Date Name Title Hospital Name Address City, State Zip Dear This survey is being sent to you as part of a research study to better understand the factors that predict a family physician’s recruitment into rural communities in California. The study will look at a variety of the traditional determinants that impact rural recruitment; at the individual level, the medical school level, the residency program level, the fellowship level, and the community level, in an effort to understand what determinants have the greatest significance and most importantly, in an effort to develop policies that will help to provide enough family physicians to effectively staff the rural populations in California. Your input is critical to the success of this study. I would greatly appreciate your attention to and completion of this survey, so that it can be returned to me by __________ in the enclosed envelope. If you prefer, I will gladly conduct a telephone interview to complete the survey. Enclosed you will also find a consent form. Please read and sign the consent form and return it with your completed survey. If you have questions, I can be reached at 626-457-4283. Thank you for your support. Kiki C. Nocella, MHA, PhD (Cand) Enel: Rural Administrator Survey Consent Form Return envelope 143 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. APPENDIX E LETTER TO MEDICAL SCHOOL LIBRARIANS Date Name Title Medical School Address City, State, Zip Dear This request is being sent to you as part o f a research study to better understand the factors that predict a family physician’s recruitment into rural communities in California. The study will look at a variety o f the traditional determinants that impact rural recruitment; at the individual level, the medical school level, the residency program level, the fellowship level, and the community level, in an effort to understand what determinants have the greatest significance and most importantly, in an effort to develop policies that w ill help to provide enough family physicians to effectively staff the rural populations in California. In order to successfully complete this study, a content analysis o f the course catalogs for your medical school must be conducted. This study looks at approximately 1,625 individuals who graduated from California family medicine residency programs for a five year period. Each o f their medical schools, if they are in the United States, is receiving the same letter to obtain the medical school catalogs. For____________ , we need to obtain the catalogs for the academic years 19 through 19 . Your help is critical to the success o f this study. I would greatly appreciate your attention to request so that it can be returned to me b y . I recognize that there may be photocopying and/or postage expense. If so, please include an invoice with the catalogs. Catalogs can either be originals or photocopies. Please keep in mind that this is currently not a funded study, so the funds to reimburse, although available, are minimal. Enclosed you will also find a consent form. Please read and sign the consent form and return it with the catalog materials. If you have questions, I can be reached at 626-457-4283. Thank you for your support. Kiki C. Nocella, MHA, PhD (Cand) Enel: Consent Form Return envelope 144 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Physician profiling and clinical pathways: Combining the tools to change physician resource utilization

PDF

The effects of case mix, hospital competition, and managed care penetration on hospital costs and revenues

PDF

Nurse executive leadership and the relationship to organizational commitment among nurses

PDF

Variations in physician practice patterns for eye care under the National Health Insurance of Taiwan

PDF

The effects of Medi -Cal managed care on access to mammography screening for women residing in Los Angeles County

PDF

Performance management in health care in Iceland

PDF

Toward a theory of networks and network effectiveness: An analytical framework for the study of multilevel networks in health care service delivery

PDF

A quantitative and qualitative evaluation of the least burdensome provisions of the Food and Drug Administration Modernization Act of 1997

PDF

The extent and effectiveness of nursing home regulation in the 50 states

PDF

Regionalization of fire protection and emergency medical aid services: A comparative case study analysis of economic and socio-political impacts

PDF

Health care finance reform and market failure at King/Drew Medical Center, Los Angeles: An analysis of pediatric social outcomes in an urban public safety -net hospital

PDF

Gambling with public health: How government officials brought Los Angeles County to the brink of disaster

PDF

An analysis of health risk selection and quality of care under Medicare fee -for -service and Medicare managed care health care systems

PDF

Out -of -pocket health expenditures by older adults in relation to age, race, and insurance

PDF

Does procedural fairness matter in standardization? An examination of a drug standardization process in hospitals and its impact on hospital effectiveness

PDF

Service evaluation and quality improvement of a geriatric evaluation and management program

PDF

Using network perspective to examine the organization of community -based elder care systems across four communities

PDF

Effects of a formulary expansion of the use of SSRIs and health care services by depressed patients in the California Medicaid program

PDF

Physician adherence to national hypertension guidelines in an elderly Medicaid population

PDF

The 1996 Health Insurance Portability and Accountability Act (HIPAA) and the 1997 Balanced Budget Act (BBA): Implications for organizational change and delivery of care in cardiology group practices

Asset Metadata

Creator Nocella, Inger C. (author)

Core Title Recruitment of family physicians into rural California: Predictors and possibilities

Degree Doctor of Philosophy

Degree Program Public Administration

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag health sciences, health care management,OAI-PMH Harvest,Political Science, public administration

Language English

Contributor Digitized by ProQuest (provenance)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-613977

Unique identifier UC11341111

Identifier 3220141.pdf (filename),usctheses-c16-613977 (legacy record id)

Legacy Identifier 3220141.pdf

Dmrecord 613977

Document Type Dissertation

Rights Nocella, Inger C.

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA