University of Southern California Dissertations and Theses

Effects of a formulary expansion on the use of atypical antipsychotics and health care services by patients with schizophrenia in the California Medicaid Program

(USC Thesis Other)

Effects of a formulary expansion on the use of atypical antipsychotics and health care services by patients with schizophrenia in the California Medicaid Program

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content EFFECTS OF A FORMULARY EXPANSION ON THE USE OF ATYPICAL
ANTIPSYCHOTICS AND HEALTH CARE SERVICES BY PATIENTS WITH
SCHIZOPHRENIA IN THE CALIFORNIA MEDICAID PROGRAM

by
Parvez Mulani

------------------------------------------------------------------------------------

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
(PHARMACEUTICAL ECONOMICS AND POLICY)

May 2009

Copyright 2009 Parvez Mulani
ii

Table of Contents
List of Tables Iii
List of Figures V
Abstract vi
Chapter 1. Introduction
1.1 Background
1.2 Policy Context for Adding Olanzapine, Risperidone, and
Quetiapine to the Medi-Cal Drug Formulary
1.3 Research Questions and Pertinence to Medi-Cal
1.4 Statistical Challenges on Assessing Formulary Policy Impact
1
2
6
7
10
Chapter 2. Data and Primary Outcomes
2.1 Data
2.2 Outcome Measures
2.3 Treatment Episode as the Unit of Observation
17
17
19
24
Chapter 3. Statistical Methods
3.1 Descriptive Statistics
3.2 Multivariate Regression Analyses
3.3 Propensity Score Analyses

30
30
31
33
Chapter 4. Formulary Policy Impact on Atypical Antipsychotic Use and
Health Care Costs
4.1 Results
4.2 Discussion
4.3 Conclusions
40
40
86
93
References 94

iii

List of Tables

Table 1. Patient Treatment Episode Characteristics by Formulary
Access Period: All Patients

44
Table 2. Patient Characteristics by Formulary Access Period 45
Table 3. Health Care Costs Prior to Treatment by Formulary Access
Period

46
Table 4. Health Care Costs in First Post-Treatment Year by
Formulary Access Period

50
Table 5. Health Care Costs in First Post-Treatment Year by
Formulary Access Period: Ambulatory Re-starts with
Schizophrenia

53
Table 6. Summary of Patient Outcomes Over Time by Initial
Therapy: Ambulatory Restart Patients with Schizophrenia

56
Table 7. Conventional verses Second Generation Antipsychotic Use
Over Time By Race: Ambulatory Restart Patients with
Schizophrenia

57
Table 8. Impact of Open Access on Health Care Costs: All Episodes 60
Table 9. Characteristics of Patients Re-Starting Drug Therapy 61
Table 10. Comparison of Health Care Costs and Persistency in
Transition and Open Access Period Versus Closed Access
Period: Re-Start Episodes

63
Table 11. Characteristics of Patients Switching or Augmenting
Current Drug Therapy

65
Table 12. Comparison of Health Care Costs and Persistency in
Transition and Open Access Period Versus Closed Access
Period: Switching/ Augmenting Episodes

66
Table 13. Characteristics of Patients Following Matching for Patients
Re-Starting Drug Therapy: All Patients
69
iv

Table 14. Characteristics of Patients Following Matching for Patients
Re-Starting Drug Therapy: Patients With Schizophrenia

70
Table 15. Characteristics of Patients Following Matching for Patients
Re-starting Drug Therapy: Ambulatory Patients With
Schizophrenia

71
Table 16. Characteristics of Patients Following Matching for Patients
Re-Starting Drug Therapy: All Patients

72
Table 17. Characteristics of Patients Following Matching for Patients
Re-Starting Drug Therapy: Patients With Schizophrenia

73
Table 18. Characteristics of Patients Following Matching for Patients
Re-starting Drug Therapy: Ambulatory Patients With
Schizophrenia

74
Table 19. Comparison of Health Care Costs Using Propensity Score
Method: Restarting Episodes

75
Table 20. Characteristics of Patients Following Matching for Patients
Switching Drug Therapy: All Patients

78
Table 21. Characteristics of Patients Following Matching for Patients
Switching Drug Therapy: Patients With Schizophrenia

79
Table 22. Characteristics of Patients Following Matching for Patients
Switching Drug Therapy: Ambulatory Patients With
Schizophrenia

80
Table 23. Characteristics of Patients Following Matching for Patients
Switching Drug Therapy: All Patients
81
Table 24. Characteristics of Patients Following Matching for Patients
Switching Drug Therapy: Patients With Schizophrenia

82
Table 25. Characteristics of Patients Following Matching for Patients
Switching Drug Therapy: Ambulatory Patients With
Schizophrenia

83
Table 26. Comparison of Health Care Costs Using Propensity Score
Method: Switching Episodes

84
v

List of Figures

Figure 1. Episodes of Antipsychotic Drug Therapy Initiated per Month
by Episode Type: First Use, Switch, Re-start

40
Figure 2. Episodes of Antipsychotic Drug Therapy Initiated per Month
by Type of Drug Used as Initial Therapy

41

vi

Abstract

In October 1997, the California Medicaid Program (Medi-Cal) added atypical
antipsychotics to its formulary to facilitate the substitution of the atypical antipsychotics
for older medications as clinically warranted, especially in minority patients thought to be
particularly at risk for poor outcomes using older medications. Moreover, it was expected
that the overall use of antipsychotics would increase as patients who experienced sub-
optimal outcomes prior to the formulary expansion would again seek treatment once new
options were available. The formulary expansion did significantly alter the clinical
treatment decision process, resulting in an immediate but temporary increase in the
number of patients initiating antipsychotic therapy, many with a recent
institutionalization, who restarted drug therapy using the new antipsychotics. There were
significant changes in the characteristics of patients using antipsychotic medications. The
likelihood of minority patients i.e. African American’s gaining access to atypical
antipsychotics improved substantially. Persistence on initial antipsychotic decreased and
total health care costs increased following open access. However the magnitude of the
increase in costs was not uniform across all patient types. Program administrators must
use caution when evaluating the impact of unrestricted access on drug therapy outcomes
and treatment costs given the changes in the characteristics of patients seeking treatment
1

Chapter 1. Introduction

Concern has mounted that restricted formularies and prior-authorization policies
implemented in many Medicaid programs may compromise the quality of care and the
health of vulnerable populations, particularly poor patients with severe mental disorders
and the chronically ill elderly. For example, decades of clinical research and experience
have documented that, despite the effectiveness of many medications in treating chronic
debilitating conditions, a lack of compliance with drug therapy has been associated with
increased admissions to hospitals and nursing homes. Logically, then, prior-authorization
policies that restrict access to newer, more effective medications may decrease patient
compliance and increase the rate of adverse clinical outcomes and the health care costs.
Not all studies support the concern that prior-authorization policies adversely
affect the cost-effectiveness of drug therapy. For example, Smalley (1995) and White
(1997) found that prior-authorization restrictions for nonsteroidal anti-inflammatory
drugs (NSAIDs) and antimicrobials did not have an adverse impact on clinical and
economic outcomes pertaining to the Medicaid population. However, we need to be
cautious when generalizing these results to the other disease states like schizophrenia.
Prior-authorization policies in California Medicaid Program (Medi-Cal) restricted
the use of olanzapine and risperidone to the treatment of refractory patients with
schizophrenia before October 1997. The addition of olanzapine, risperidone, and
2

quetiapine to Medi-Cal’s formulary at that time created an opportunity for researchers to
observe the effectiveness of formulary inclusion. This issue is of particular concern to the
Medi-Cal program as olanzapine and risperidone became the two most costly drugs
within a few years after being included in the formulary. (Stahl 2006)
It has been hypothesized that adding olanzapine, risperidone, and quetiapine to
Medi-Cal’s formulary would improve the effectiveness of medication therapy and reduce
the health care costs associated with treating schizophrenia. This study investigates the
effects of this formulary expansion on the treatment patients with schizophrenia received
using the Medi-Cal paid claims data. Specifically, we examine the impact of formulary
expansion on initial antipsychotic selection, medication use patterns, health care
utilization patterns, and health care costs for Medi-Cal patients with schizophrenia.
1.1 Background
1.1.1 Brief History of Medicaid Formularies
Medicaid, the largest public payer of drug benefits in the United States, uses
federal and state funds to pay for inpatient and outpatient medical services and
prescription drugs for primarily the poor, the elderly, permanently disabled persons, and
recipients of Aid to Families with Dependent Children (AFDC). The accelerated rise in
both the prices and use rates of pharmaceuticals has increased drug benefit expenditures
and prompted many states to restrict the benefits in Medicaid programs. Restrictive drug
regulations have also been motivated by the perception of frequent inappropriate use,
3

such as overprescribing by physicians, abuse by patients, and prescribing for
inappropriate indication (Kunin, 1990). For example, Medicaid recipients are often
perceived as potential abusers of psychoactive drugs, which may explain the triplicate
prescription-monitoring programs for benzodiazepines in three states. On the other hand,
well-designed studies have documented substantial underutilization of psychoactive
agents such as antidepressants, primarily due to non-compliance (McCombs, 1990).
State legislatures and policy makers have often implemented drug formularies
with little empirical evidence about their true consequences. Whether this policy reduces
expenditures for specific drugs without causing the unwanted substitution of other drugs
(e.g., a barbiturate for a benzodiazepine) or medical services (e.g., additional office visits
to identify an effective approved drug) has been largely unstudied.
1.1.2 Cost Sharing and Prescription Limits
Prescription drug cost sharing in state drug benefit programs can manifest either
as copayments (typically $1 to $3 per prescription), or prescription limits (the number or
total value of prescriptions reimbursed in a specified time frame). Prescription limits are
the most severe form of cost sharing for low-income populations. In Medicaid, a
prescription limit (or cap) may be as low as three prescriptions per recipient per month
and is often accompanied by a strict one-month supply limit. A policy maker may not
view a reimbursement cap as cost sharing, but, from the perspective of Medicaid
4

recipients reimbursement caps require them to bear the full cost for any prescription over
the limit.
The effects of a three-prescription-per-month limit (three-drug cap) were analyzed
in a study of the New Hampshire Medicaid program (Soumerai, 1987). The three-drug
cap caused a precipitous, sustained reduction in prescriptions (including effective and
essential drugs) per patient per month. The limits were found to increase the risk of
institutionalization in nursing homes for frail, low-income, elderly patients and may have
increased total Medicaid costs (Soumerai, 1991). The three-drug cap on the use of
psychotropic drugs was found to increase the use of acute mental health services among
low-income, noninstitutionalized patients with schizophrenia (Soumerai, 1994).
Similarly, a policy change from a six-drug cap to a five-drug cap in the state of Georgia
reduced the prescriptions reimbursed by Medicaid, but increased prescriptions paid for
out-of pocket by patients, resulting in an overall decreased total prescriptions filled
(Martin, 1996).
1.1.3 Administrative Restrictions on Prescribing of Specific Drugs
1.1.3.1 Drug Category Exclusions
Medicaid programs are allowed to eliminate reimbursement for certain classes of
products. For example, concern about potential abuse and overuse led five states to stop
reimbursing for benzodiazepine sedatives/hypnotics in 1990. However, these states did
pay for more potent and potentially toxic sedatives such as barbiturates, presenting
5

potential problems of inappropriate drug substitution. The Oregon Medicaid program
eliminated reimbursement for over-the-counter (OTC) medications. This policy was
successful in reducing drug costs to the state with limited evidence for substitution of
more expensive prescription-only products (Zechnich, 1998). Florida’s policy of
restricting Medicaid reimbursement for anti-ulcer drugs was associated with a substantial
reduction in outpatient anti-ulcer drug utilization without any significant increase in the
rate of hospitalization for peptic disorder-related conditions (Cromwell, 1999).
1.1.3.2 Formularies and Prior Authorization
Formularies provide the foundation for guiding clinicians in choosing the safest,
most effective agents for treating particular medical problems. “Positive” formularies
have been described as a list of reimbursable agents based on their superior safety,
efficacy, or cost-effectiveness. In contrast, “negative” formularies reimburse all marketed
drugs, except excluded from payment mainly due to price concerns.
Formularies are often confused with prior-authorization requirements for off
formulary use. As of 1992, Medicaid programs in 34 states had implemented prior-
authorization programs in conjunction with formulary restrictions (Gondek, 1994).
In theory, prior authorization provides a method to limit access of costly, newly
introduced, and/or potentially toxic drugs to recipients who truly need them and eliminate
their use when less expensive or safer alternative could be used. A prior-authorization
policy involving non-generic NSAIDs in the Tennessee Medicaid program reduced
6

expenditures as a result of increased use of generic NSAIDs and a 19% decrease in
overall NSAID use. There was no concomitant increase in Medicaid expenditures for
other medical care. Therefore, the prior-authorization requirements might be highly cost-
effective with regard to NSAIDs (Smalley, 1995).
However, the findings of this study of NSAIDs may not be directly applicable to
antipsychotic medication. Although patients’ response to individual NSAIDs may vary
for reasons that are as yet poorly understood, there is limited evidence of systematic
differences in the NSAIDs mechanisms of action, efficacy, or safety. In contrast, the
complex pharmacological properties of each antipsychotic drug led to substantial
differences in both efficacy and safety (Worrel, 2000; Citrome, 1997).
Requirements to change the medications of patients with well-controlled
schizophrenia might entail substantial hazards because of the lesser efficacy of the
alternative medications or interruptions to drug therapy. Even if medication substitutions
could be made without risk to the patient, the additional visits to physician (e.g. for new
dose titration) could cancel out the saving from drug cost. In addition, prior-authorization
requirements might substantially influence the initial antipsychotic selection.
1.2 Policy Context for Adding Olanzapine, Risperidone, and Quetiapine to
the Medi-Cal Drug Formulary
California’s Medicaid (Medi-Cal) program required prior-authorization for all
atypical antipsychotics prior to October 1997. Olanzapine, risperidone, and quetiapine
7

were added to Medi-Cal’s formulary at that time based on available data documenting the
efficacy and effectiveness of atypical antipsychotics relative to conventional/typical
antipsychotics. Since the addition of these three antipsychotics to the Medi-Cal
formulary, the Medi-Cal beneficiaries with schizophrenia have had unrestricted access to
these three atypical antipsychotics as the initial pharmacotherapy. This study evaluates
the impact of this formulary expansion on the treatment of patients with schizophrenia.
It should be noted that intended and unexpected consequences may result from
any formulary expansion. Policy makers are well advised to evaluate the effectiveness of
a formulary expansion initiative because open access may alter drug use patterns,
including the possibility of inappropriate use of the newly added drugs after the
formulary expansion.
1.3 Research Questions and Pertinence to Medi-Cal
In October 1997, Medi-Cal removed the prior-authorization requirement on the
use of second generation/atypical antipsychotics (risperidone, olanzapine) and allowed
quetiapine to be used without restriction when it was approved by the FDA in October
1997. This formulary expansion was driven by several factors. First, the superior safety
and efficacy profile of atypical antipsychotics to conventional antipsychotics was
documented in both clinical and naturalistic research studies (Worrel, 2000; Citrome,
1997). Second, there was an increasing body of published literature advocating the use of
atypical antipsychotics as the first line treatment for schizophrenia rather than limiting
8

their use to patients who have failed first generation antipsychotics (PORT
recommendations 1996, McCombs 1999, 2000). Third, Medi-Cal officials expressed
concern about the under-use of newer antipsychotics in minority populations. Differences
in the metabolism rates of psychotropic drugs across racial groups suggest that a high
proportion of minority patients are at increased risk for serious side effects with the first
generation antipsychotics and thus increased concerns about restricting access to the
second generation/atypical antipsychotics (Lawson (2000)). By revoking the Treatment
Authorization Request (TAR) requirement on the atypical antipsychotics, Medi-Cal had
expected to facilitate the use of these agents by minority patients with schizophrenia.
Finally, several studies found significant evidence that first generation antipsychotics
were not meeting the therapeutic needs of Medi-Cal patients with schizophrenia. Over
24% of patients studied did not use any medication to treat their illness for over one year;
17% of patients delayed therapy for over two years (McCombs, et al., 1999; McCombs,
Luo, Johnstone and Shi, 2000). Data also showed that patients who received drug
therapy did not incur lower health care cost than untreated patients. Of the patients who
used an antipsychotic medication within the first year, 24% delayed therapy for more that
30 days (McCombs, et al., 1999). Such delays in therapy were associated with increased
health care costs of $9,832, primarily due to significantly higher expenditures for acute
hospital care ($3,456), ambulatory services ($2,878) and nursing home care ($2,064)
(McCombs, Nichol, Johnstone, et at, 2000). For those patients with no delay in therapy,
47% switched or augmented their original antipsychotic medication within one year
(McCombs, et al., 1999). This switching behavior was associated with significantly
9

higher health care costs ($8,516), again due primarily to higher costs for acute hospital
care, ambulatory services, and nursing home care (McCombs, Nichol, Johnstone, et al.,
2000). McCombs, Luo, Johnstone and Shi (2000) found that the costs associated with
these dysfunctional drug use patterns associated with conventional antipsychotic
medications continued to increase into the second year of treatment. Delays in drug
therapy were associated with an increase in post-treatment cost of $10,511, while
switches in therapy were associated with an increase in total cost of $14,263 over two
years.
Based on this body of literature, advocates for the mentally ill, legislators, and
regulators believed that many Medi-Cal recipients were not receiving the most effective
drug therapies for schizophrenia, with particular concern for California’s large minority
communities. Program administrators concluded that an essential need existed to justify
an expansion of the formulary to include the atypical antipsychotics despite their higher
acquisition costs. Specifically, Medi-Cal officials believed that adding atypicals to the
formulary would have several beneficial effects including reducing non-compliance and
under-treatment, especially among African-Americans, and improving the quality of
schizophrenia care. They further hoped that the higher quality of care would at least
partially offset the greater acquisition costs of the medications.
As expected, after the formulary expansion, Medi-Cal experienced a rapid
increase in the drug budgets. Specifically, olanzapine and risperidone became the two
most costly drugs covered by Medi-Cal after their addition to the formulary. In the first 6
10

months of 1999, expenditures for these two medications amounted to over $83 million for
nearly 10 million days of therapy (California's Medical Assistance Program, 1999). As
with any organization, Medi-Cal is interested in establishing the return on their
investment. Three issues need to be addressed to answer this concern:
1. What was the impact of the formulary expansion on drug use patterns such as
days of continuous drug therapy?
2. What was the impact of the formulary expansion on the costs of treating Medi-Cal
patients with schizophrenia? Were the increased drug costs offset by savings on
hospitalization and acute care costs?
3. Did the formulary expansion improve the access to atypical antipsychotics by the
minority patients with schizophrenia?

1.4 Statistical Challenges on Assessing Formulary Policy Impact
All research studies designed to estimate the impact of government policy
initiatives over time face three important statistical challenges. First, the analysis must
take into account exogenous historical factors that may have confounded the investigated
outcomes during the study period (i.e. 1994-2000). Historical factors are a common threat
to the internal validity of all studies that measure the impact of policy changes without a
parallel control population that was not affected by the policy (Stolk P, Schneeweiss et al
11

2008). Second, the analysis must make an effort to disentangle the multitude of ways that
the policy initiative may have affected the system under study. Of particular importance
in policy analysis are the possible unexpected effects that must be explicitly accounted
for in the analysis. Finally, if alternative treatment options are to be compared, the
analysis must account for possible treatment selection bias. In the case of a formulary
expansion, the treatment itself is affected by greater access to new therapies. Since each
of these factors may play an important role in the analysis, the specifications of the
statistical models must be carefully constructed to attempt to control for these factors.
1.4.1 Historical Factors Affecting Medi-Cal Program
During the study period the Medi-Cal program experienced significant exogenous
changes, some of which were unrelated to the treatment of schizophrenia or the formulary
expansion. First, an improving economy in the mid 1990’s provided greater opportunities
for able-bodied welfare recipients to return to work. At the same time, welfare reform
limited the time that able-bodied recipients could receive welfare and Medi-Cal coverage.
As a result, the Medi-Cal program experienced a steady decline in the number of eligible
recipients over the study period. Second, a concurrent health policy initiative was
implemented aimed at enrolling non-disabled Medi-Cal recipients into managed care
programs. This initiative decreased the total number of Medi-Cal recipients who received
services in the fee-for-service section, from which the data for this study were derived.
However, both factors were unlikely to have a significant impact on Medi-Cal recipients
12

with schizophrenia or on the characteristics of the patient population seeking treatment
over the course of this study.
1.4.2 Formulary Expansion and Clinical Decision Process
The expansion of the Medi-Cal formulary to include olanzapine, risperidone, and
quetiapine was primarily intended to encourage substitution of these medications for
conventional antipsychotics in the treatment of schizophrenia. However, the impact of
these potential substitution effects on treatment outcomes and costs depend on the
characteristics of the patient population initiating treatment with atypicals before and
after the formulary expansion.
The dynamics of the clinical decision to initiate therapy using an atypical
antipsychotic in response to formulary expansion can be very complicated. Specifically,
the following three patient populations may respond differently to the formulary changes:
1.4.2.1 Patients Switching/Augmenting Drugs [Switching/Augmenting episodes]
McCombs et al (1999) found that nearly half of all treated patients used a second
antipsychotic medication during the first year of initiating drug therapy, probably because
of intolerance to the side effects or exacerbation of symptoms (refractory patients). These
drug therapy episodes in which a patient changes his antipsychotic medication while
being on an active antipsychotic can be defined as switching episodes.
13

Prior to the formulary expansion, treatment authorization required that physician
validate the necessity of prescribing atypical antipsychotics, generally by documenting
two prior conventional antipsychotic episodes. As a result, patients switching
antipsychotic medications prior to formulary expansion were likely to have severe side
effects or exacerbation of symptoms with conventional antipsychotics. Once the prior
authorization was lifted, physicians may have changed the clinical criteria for switching a
patient from conventional to atypical antipsychotics. Specifically, patient may be
switched to an atypical antipsychotic without experiencing severe side effects or an acute
exacerbation of symptoms. As a result, “switchers” in the post formulary expansion
period may consist of two types of patients:
1. Patients who meet the pre-formulary expansion criteria for switching to an
atypical antipsychotic (“Acute Switchers”); typically these patients have
higher pre-switch health care costs.
2. Patients who did not meet the pre-expansion criteria for switchinged to
atypical and was switched to achieve better outcomes (“Non-acute
Switchers”); typically these patients have lower pre-switch health care
costs.

The expected impact of formulary expansion on the population of patient
observed to switch medications is quite clear since the “nonacute” switchers would have
14

been added to “acute switchers”. This may result in the following differences in drug
therapy outcomes and post-treatment costs for the “switcher” patient episodes after the
formulary expansion:
1. The likelihood of completing therapy in the switcher episodes treated with the
atypical antipsychotics would be reduced in the post expansion period as these
products inherit a disproportionate share of less severely ill patients.
2. Total post-treatment costs for switching episodes treated using conventional drugs
would increase in the post-expansion period compared with the pre-expansion
period due to the use of atypical antipsychotics by the less ill patients.
A good proxy for distinguishing the “Nonacute Switchers” and “Acute Switchers”
was to examine the pre-switch health care utilization. It is possible that the “Nonacute
Switcher,” being relatively healthy, may have less health care costs in the pre-switch
period than the “Acute Switcher” who was sicker and had higher health care costs in the
pre-switch period. However, it is very difficult to completely distinguish them as the
physicians may have used other criteria than the one mentioned above in switching
patients from typical to atypical antipsychotics, which is subject to change after the
formulary expansion.
1.4.2.2 Patients Restarting Treatment After A Break In Treatment [“Restarter”
Episodes]
15

In addition to the switching of patients on active therapy, the formulary expansion
may have affected the decision by patients to “restart” antipsychotic drug therapy.
McCombs et al (1999),Croghan et al (1999), and McGlashan (1988) found that most
patients with schizophrenia use the medications intermittently. This implies that there is a
sizable pool of patients who limit the use of the medications to periods of acute
symptoms punctuated by extended periods when antipsychotic medications are
discontinued against medical advice. The patients with extended periods of no drug
therapy could be from the one the following subgroups:
1. Delayed Switchers: Treatment refractory patients and patients who may have been
discouraged from seeking treatment in the pre-expansion period due to the side
effects of conventional antipsychotics and prior-authorization restriction on
atypical antipsychotics.
2. Restarters on the same antipsychotic: Patients in full or partial remission being
taken care of by their families who may have withdrawn from therapy with
antipsychotics.
Prior to formulary expansion, both groups of patients probably restart
antipsychotic therapy only after acute exacerbation of symptoms. Once the prior
authorization restriction on atypical antipsychotics was lifted, the clinical
criteria/symptoms which would cause a patient to restart drug therapy may have changed.
Even patients in full or partial remission may seek to restart therapy with an atypical
antipsychotic medication.
16

The impact of the formulary expansion on the characteristics of patients in the
restarting group depended upon which of the subpopulations was dominant:
1. If the treatment refractory patients constituted the majority of the restarter
population, the average pre-expansion drug therapy cost of treating a patient with
schizophrenia who restarted drug therapy would increase in the post-expansion
period.
2. If the patients in full or partial remission dominated this population, the
compliance and average pre–expansion drug therapy costs would decrease for
episodes started in the post-expansion period. The post-expansion drug therapy
prescription costs would increase, where as the post–expansion drug therapy
health care costs could either increase or decrease.

The overall effects could be determined based on pre drug therapy costs.
However, it will be impossible to determine to which group an individual patient belongs.
Based on the discussions above, the first step in the analysis of the impact of
formulary expansion was to define patient episodes as “Switching”, “Augmenting” and
“Restarting (Delayed Switchers or Restarters on the same drug),” followed by utilizing
appropriate statistical methods to control for confounding variables.
17

Chapter 2. Data and Primary Outcomes
2.1 Data
2.1.1 Data
The Medi-Cal program maintains a longitudinal database of paid claims as long as
the recipient is eligible for the program. This database contains membership information
at the individual member level, combined with institutional claims at the claim level,
professional service claims at the service level, and prescription drug services at the drug
level. Data include the type of services, date of service, amount billed, amount paid, and
units (days) of service. Prescription drug claims identify the specific product dispensed,
quantity, days supplied, and the date the prescription was filled.
The amount paid by Medi-Cal understates the total payments incurred by elderly
and disabled Medi-Cal recipients who are dually eligible for Medicare, because Medicare
is the primary payer for these patients. To adjust for this under-reporting of payments,
Medi-Cal payments for ambulatory services covered under Part B of Medicare were used
to estimate total ambulatory care costs in both pre-treatment and post-treatment periods
for all patients with reported dual eligibility or over 65 years of age. This algorithm is
based on the amount of the deductible that Medi-Cal pays before Medicare coverage is
applied and the coinsurance rate specified by Medicare (20%). Actual Medi-Cal
expenditures for outpatient services was used for patients under the age of 65 years who
were not identified as dually eligible for Medi-Cal and Medicare. All reported non-
institutional expenditures were adjusted to 2000 dollars using historical Medi-Cal-
18

specific fee schedule adjustments by type of service prior to the estimation of missing
Medicare payment. Actual Medi-Cal payments were used for services not covered by
Medicare such as drug costs.
The under-reporting of the total cost for institutional services was bridged by
multiplying the days of hospital, skilled nursing facility (SNF) and intermediate care
facilities (ICF) care by per diem cost estimates for these services. Hospital days were
assigned a cost of $1,032 per day (California’s Medical Assistance Program, 1998), while
SNF and ICF costs per day were valued at $270 (Health Care Financing Administration
1996). To assure consistency, the costs for hospital services and nursing home stays were
estimated using this methodology for not only Medicare eligible patients but all patients
in the study.
2.1.2 Inclusion Criteria for Study Sample
The complete Medi-Cal data set for the period (Jan 1994- Aug 2000) was used to
create patient episodes for patients with schizophrenia. An episode indicated any change
in antipsychotic drug therapy. The first episode for any patient was labeled as a ‘new’
episode. The subsequent episode for a particular patient was labeled as a ‘restarting’ or
‘switching’ episode. If a change in drug therapy occurred while the patient was on active
therapy, the subsequent episode was labeled as a ‘switching’ episode; if the change
occurred while the patient was not on active therapy, that episode was labeled as a
‘restarting’ episode. For each patient episode data were divided into 6 months pre
19

treatment and 12 months post treatment period, based on the date on which the
prescription was filled. Services were categorized by the type of service. The patient
episodes were included if they met the following criteria:
(1) More than 6 months of pre-treatment data and a minimum of one year of post-
treatment data were available. This requirement would give us sufficient
information on pre-treatment period for modeling the choice and outcomes and
the one year post-treatment information was necessary as we analyzed outcomes
with a one-year window.
(2) The patient was not institutionalized in a nursing home within 30 days from
the start of therapy, i.e., he or she was not a long term care patient while initiating
drug therapy. Institutionalized patients would have had serious conditions, which
might substantially influence the costs. Institutionalized patients would have drug
compliance monitored by the facility, thus there might be an artificial correlation
between compliance and use of long term care facilities.

2.2 Outcome Measures
To analyze the impact of formulary expansion on the treatment of patients with
schizophrenia in Medi-Cal, the following outcome measures were used:
1. Drug utilization patterns:
20

• Days of continuous drug therapy, often referred to as time to all cause
discontinuation (TTAD)
2. The health care costs in the post treatment period, broken down by type of service:
• Psychiatric hospitalization costs
• Acute care hospitalization costs
• Ambulatory care costs
• Drug costs
• Long term care costs
3. The proportion of minority (African-American) patients with schizophrenia utilizing
atypical antipsychotics.
The rationale for choosing these endpoints is discussed below:
1. Days of continuous therapy:
The substantial burden of schizophrenia at the individual, family, and community
levels has been well documented. Also well established is the importance of
antipsychotic medications for symptom control in the acute phase of an episode and for
relapse prevention in the subsequent stabilization and maintenance phases.
21

Less widely appreciated are the potential benefits of continuous or uninterrupted
treatment with the same antipsychotic agent. Research has shown that discontinuations or
prolonged interruptions in antipsychotic medication treatment is linked to increases in
psychotic symptoms, hospitalizations, emergency room visits, and homelessness. In
general, persistence of antipsychotic treatment for schizophrenia appears to be far less
optimal than treatment for other diseases. For example, Weiden et al (1995) found that,
among discharged patients with schizophrenia, approximately 50 % had discontinued
pharmacotherapy after one year, and 75% had discontinued by two years post discharge.
Interruptions in antipsychotic treatment can be influenced by many personal, clinical,
social, economic and logistical factors. Frequently, the lack of persistence is assumed to
directly reflect an individual’s conscious decision not to comply with or adhere to
prescribed treatment. Indeed, a lack of adherence to planned or prescribed treatment by
many with schizophrenia, often a symptom of the disease, represents a significant
obstacle to their sustained recovery.
The efficacy and tolerability of an antipsychotic regimen for a specific patient are
critical determinants of medication persistence. For example, if a high dose of one agent
is necessary to achieve symptom control but cannot be tolerated by the patient (e.g. due
to extrapyramidal side effects (EPS)), the physician is likely to switch the patient to a
different agent. An unfortunate but likely alternative scenario is that the patient simply
stops taking the medication without appropriate consultation with the physician. In either
case, the goal of medication persistence has not been achieved.
22

Data revealing treatment persistence associated with different antipsychotic agents
can have important clinical and economic implications. Extended pharmacotherapy is
recommended in published guidelines for treating schizophrenia. Moreover, antipsychotic
medication persistence and avoidance of drug switches have been empirically linked to
better longer-term outcomes (Weiden et al 1995). Therefore, measures of persistence and
medication switches can serve as a marker for treatment response. As such, this marker
may provide an additional informational tool for assessing the overall value of particular
antipsychotic agents.

2. Costs:
Spending of the Medi-Cal program has more than doubled in the past decade,
increasing from $11.9 billion in FY 1992 to $24.1 billion in FY 2001 (www.medi-
cal.org). Approximately $700 million in Medi-Cal expenditure is spent on mental health
services annually in California. The average mental health expenditure per Medi-Cal
beneficiary in FY 1999 was $2,027, excluding payments for state psychiatric care. This
reflects an increase of 56% since 1992, placing increasing fiscal pressure on the state
budget (Mental health fact sheet, www.medi-cal.org). Since their addition to the
formulary in fall of 1997, risperidone and olanzapine have become the two most costly
drugs covered by Medi-Cal in terms of pre-rebate expenditures. In the first 6 months of
1999, expenditures for these two medications amounted to over $83 million for nearly 10
million patient-days of therapy (California’s Medical Assistance Program (1999)). As
23

with any organization, Medi-Cal was interested in establishing the return on their
investment in these two drugs. Specifically, the question was: Are the procurement costs
for these drugs offset by any savings in reduced health care use, especially acute hospital
care and psychiatric hospitalizations? To facilitate Medi-Cal’s evaluation of its decision
to expand the formulary, this study stratified and analyzed the health care costs by type of
service.
Using costs as outcome measures in paid claims databases raises several
methodological concerns. Some patients may not utilize services for variety of reasons
unrelated to their health status (e.g., loss of eligibility, incarceration). In this study, only
patients with eligibility at-least six months pre-drug therapy and 12 months post-drug
therapy in Medi-Cal data were included (the choice of the 12 months follow up period is
justified later in this document). This somewhat controlled for factors such as lost
eligibility.

3. The proportion of minority (African-American) patients with schizophrenia
utilizing atypical antipsychotics:
One of the major factors that led to the formulary expansion was the concern that the
minority patients had less access to newer antipsychotic medications, despite the fact
that the minority patients are at a higher risk of developing side effects with the
conventional antipsychotics than Caucasian patients. To assess the effectiveness of the
formulary expansion policy, the proportion of minority patients receiving atypical
24

antipsychotics after the formulary expansion could be a direct measure of the
effectiveness of the stated policy objective.
4. Duration of follow up:
This analysis was limited to the patient drug therapy episodes with 12 months of
post-drug therapy data. There were several reasons for choosing a 12-months follow up
period. First, expert consensus guidelines clearly point out that the most appropriate
method for evaluating the cost-effectiveness of medications is analyze the total average
yearly direct costs of all psychiatric care on alternative medications (ECG 16). The PORT
recommendations indicate that once a patient has achieved symptom relief, the patient
should be maintained on that particular antipsychotic for at-least one year (PORT
recommendation 11). Second, the choice of a longer follow-up period will significantly
decrease the sample size. Third, from a policy maker’s perspective, annual costs are
easier to provide fiscally meaningful interpretation of outcomes.
2.3 Treatment Episode as the Unit of Observation
Several previous studies on Medi-Cal claims data influenced the design of this
research. First, a significant proportion of Medi-Cal patients with schizophrenia
apparently use conventional antipsychotics periodically to treat acute episodes of positive
(i.e., psychotic) symptoms, then stop drug therapy to avoid the frequent and significant
side effects associated with these drugs. If this was the case, a pool of previously treated
patients exists at any point in time. Second, a significant proportion (47%) of treated
25

patients switch therapies within 1 year (McCombs et al, 1999). Patients who either re-
start antipsychotic therapy or switch therapies appear to consume significant amount of
health care, probably because of an acute exacerbation of positive symptoms (re-starters)
or treatment failure (switchers), primarily derived from hospital and nursing home care.
Therefore, it is reasonable to assume that open access to second-generation antipsychotic
medications would have induced previously treated patients to re-start antipsychotic drug
therapy using these medications. Similarly, patients experiencing side effects with their
current therapy might also have been induced to switch to the second-generation
medications once they became available. This potential ‘access effect’ would be indicated
by a short-term increase in the frequency of re-starting and switching episodes as the pool
of these patients dwindled over time. At the same time, there is a steady population of
patients who were taking conventional antipsychotics continuously and, as soon as the
second generation antipsychotics became available without prior authorization, were
switched to the second generation antipsychotics, which led to a substitution effect.
Moreover, these hypothesized ‘access effect’ patients should have exhibited a different
pattern of health care use than patients re-starting or switching antipsychotic drug
therapies in the period prior to the formulary expansion. First, access-effect patients
would, by definition, use the newly added medications. Second, these patients would
likely incur lower health care costs prior to re-starting or switching therapies as these
changes in therapy may not be associated with an exacerbation of positive symptoms or
treatment failure. Finally, and most important, these access effect patients might incur
lower post-treatment costs because of improved health status when they initiated another
26

treatment episode. If this was the case, the formulary change would be associated with
lower post treatment costs due to patients being drawn back into drug therapy.
Any proposed research design must try to disentangle the health care cost
implications of the access effect from the true effect of the policy change on treatment
costs because of the medication substitution. Consider first a research design in which the
unit of observation is a single patient and compares each patient’s pattern of health care
use before and after the formulary expansion. If the access effect was present, patients re-
starting drug therapy or switching medications immediately after the formulary expansion
would have appeared to incur increased costs, especially in drug and physician office
visit costs. Such an increase in costs would not necessarily indicate that the treatment of
schizophrenia had become more costly after the formulary expansion. Even an
‘interrupted time series’ approach that drops the months immediately following the
formulary expansion from the analysis would only aggravate the confounding effect due
to disease progression and patient aging.
Of course, one alternative to controlling for the statistical problems associated
with having each patient serve as their own control would be to compare Medi-Cal
patients to Medicaid patients from a similar state that did not undergo a similar formulary
change. The internal validity of this approach can be problematic. First, it would be
difficult to find a state comparable to California in terms of how patients with
schizophrenia were treated during this time period. Second, most state Medicaid
programs made the second-generation antipsychotic medications available immediately
27

upon FDA approval. Therefore, a state-to-state comparison would measure the extent to
which California’s treatment patterns ‘caught up with’ other states. Finally, comparable
data from another Medicaid program were not readily available.
Given the limitations of a patient-based time series approach, this study used an
episode of drug treatment as the unit of analysis. Specifically, an episode of care was
defined each time a patient with schizophrenia changed antipsychotic medications. This
approach is consistent with the fact that most patients with schizophrenia use
antipsychotic medications periodically and change medications frequently. Moreover, an
episode of treatment provides an opportunity to better define similar units of observation
(episodes) before and after the formulary expansion. Specifically, the data available for
analysis can be used to characterize the patient’s antipsychotic use history prior to each
episode of re-starting therapy or switches in therapy.
An episode-based analysis also provided a more valid method for controlling for
the potential ‘access effect’ that may appear immediately after the formulary expansion in
October 1997. Instead, episodes of treatment initiated in the period immediately
following the formulary expansion could be analyzed separately.
Three types of episodes of treatment were defined.
• A re-starting episode was defined when a patient switched therapy after a break of
more than 15 days from their previous antipsychotic medication.
28

• A switching episode was defined as a change in drug therapy within 15 days of
the end of the previous antipsychotic episode. Switching episodes included both
true switches in therapy and augmentation of the original drug therapy.
Augmentation was defined as occurrences when the patient made at-least one
additional purchase of the original drug after starting the second drug.
• Finally, a new episode of therapy referred to the occurrence when a patient started
drug therapy without any previous antipsychotic medication in his or her paid
claims history. However, caution must be exercised when considering results for
‘new’ patient episodes as it is likely that many of these patients had been treated
with antipsychotics prior to becoming eligible for Medi-Cal and their prior
antipsycotic utilization would not be captured in this data set. Given these
limitations, separate analyses on the new episodes were not conducted.
The index date for re-starting and switching episodes of drug therapy was set based
on the month in which the change took place. Three intervals were analyzed: the 6
months prior to the index month, the month in which the drug therapy change was
undertaken, and a one-year post-treatment period. For new patients, the index date was
set on the day on which the patient filled their initial prescription for their initial
antipsychotic medication. For new episodes, only the 6-month pre- and 12-month post-
treatment periods were defined based on the index date. Episodes of drug therapy were
excluded from the analysis if less than 6 months of pre-treatment or 12 months of post-
treatment data were available for analysis.
29

It was possible that most patients had multiple re-starting and/or switching drug
therapy episodes. Moreover, new, re-starting and switching episodes might overlap and
the one-year post-treatment period for an earlier episode might be included in the pre-
treatment period of a later episode. If the patient switched or re-started therapy early in
the post-treatment period of the preceding episode, it was possible for a significant
portion of the post-treatment periods for both episodes to overlap.
The analysis developed a set of independent and dependent variables to capture
the extent to which episodes overlap. For example, changes in therapy during the post-
treatment period were used as an outcome variable in this analysis. Similarly, the number
of prior episodes and data related to the prior treatment episode were used as independent
variables in the analysis of re-starting and switching episodes. These data included the
class of medication used (e.g., depot, clozapine), prior initial dosing in chlorpromazine
equivalents, and duration of therapy.
Structuring the data set using episodes as the unit of observation introduced issues
related to non-independence of observations and autocorrelation. Although the ordinary
least square (OLS) estimators remained unbiased and consistent in the presence of
autocorrelation, they were no longer efficient. As result, the usual t and F tests of
significance might miss significant effects that could have existed. This “conservative”
form of bias was against finding significant effects and was therefore less problematic
than the bias of increasing the likelihood of finding significant but spurious results. To
30

address this concern, the heteroscedasticity-consistent matrix developed by White (White
1980) was used to adjust for estimated standard errors for clustering of episodes.
Chapter 3. Statistical Methods
3.1 Descriptive Statistics
Initiation of Episodes over Time: A simple time-series of the number of episodes
of care initiated per month was created to determine whether the access expansion
increased the incidence of episodes of drug therapy immediately after October 1997. A
second time-series was generated to compare conventional antipsychotics to the second-
generation antipsychotics, as the access effect should be limited to second-generation
antipsychotic medications. Conversely, the substitution effect was expected to reduce the
use of conventional antipsychotics in the open-access period. Finally, a time series was
generated for each type of episode. These time series are presented graphically.
Changes in Patient Characteristics over Time: Descriptive statistics and simple
univariate statistical tests (chi-square, ANOVA) were used to compare patient
characteristics across three access time periods: closed access period (1994 – October
1997), transition period (October 1997 – May 1998), open access period (May 1998 –
Feb 1999).
Patient Outcomes: Medi-Cal administrators requested information on whether the
investment in second-generation antipsychotics generated better use patterns, and
reductions in the cost of other services. The use of a second antipsychotic medication
31

(switching or augmentation) and the duration of use during the one-year post-index
period were compared across the three access periods using ANOVA and chi-square
statistical tests. The impact of open access on treatment costs was measured by
comparing post-index costs per patient across the three access periods using ANOVA.
3.2 Multivariate Regression Analyses
OLS models were used for the components of health care costs (ambulatory care,
prescription drugs, psychiatric hospitals, acute hospitals, nursing homes, psychologists,
hospice care, home health, and other services) as well as total cost and total cost net of
drug cost and duration of use of the antipsychotics. Separate models were used by type
of episode (switching/augmentation and re-starts) and for all patients, patients with a
primary diagnosis of schizophrenia and non-institutionalized patients with a primary
diagnosis of schizophrenia.
The presence of multiple episodes of treatment observations for some patients
violated the assumption of independence across observations for OLS regression. While
OLS estimates for model parameters remained unbiased under these conditions, the OLS
standard error estimates might be biased. Statistical estimation techniques were used to
adjust for estimated standard errors for serial correlation and heteroscedasticity (White
1980).
A primary mental health diagnosis was assigned to each episode of care based on
a hierarchical ranking of the patient’s mental health diagnoses reported during the 6
32

months immediately prior to starting an episode of antipsychotic drug therapy (pre-index
period). The hierarchical order of the diagnoses was determined based on the extent to
which antipsychotic medications played a role in treating each disease state. A patient
was coded as having primary mental health diagnosis of schizophrenia if an ICD-9-CM
diagnostic code of 295.00-295.99 appeared at any time during the 6-month pre-index
period, regardless of the presence of other mental health diagnoses. Other diagnoses
were assigned based on the importance of antipsychotic medications in treating each
disorder. The order of these diagnoses were: non-schizophrenic psychosis (291.00-
294.99), bipolar disorder (296.00-296.19, 296.40-296.89), depression (296.20-296.39,
300.40-300.49), other affective disorders (296.90-296.00), anxiety (297.00-297.99,
300.00-300.99), substance abuse (303.00-305.99), personality disorder (301.00-301.99),
dementia (290.00-290.99), other mental health diagnoses (299.00-299.99, 302.00-302.99,
306.00-314.99, 316.00-316.99). A patient was recorded as not having a mental health
diagnosis if none of the previous ICD-9 codes was recorded in the pre-index period.
Over 80 variables were used in the cost models to control for differences over
time in the patient population. These variables included age, race, gender, urban or rural
residence (versus mixed counties), disability status, the mix of medical diagnoses
recorded in the prior 6 months, the mental health diagnosis preceding the episode, and
prior use of health care by type of service in both the index month and the pre-treatment 6
months. Data describing the patient’s prior use of antipsychotic drugs were also used as
independent variables, including the number of previous treatment episodes, type of
33

drug(s) used, and the chlorpromazine equivalent dose in the episode immediately
preceding treatment.
3.3 Propensity Score Analysis
3.3 1 Introduction
In a randomized experiment, the randomization of subjects to different treatments
ensures that, on average, there should be no systematic differences in observed or
unobserved covariates between patients assigned to the different treatments. However, in
a non-randomized observational study, investigators have no control over treatment
assignment; therefore, direct comparisons of outcomes from the treatment groups may be
misleading. This difficulty may be partially avoided if information on measured
covariates is incorporated into the study design such as through matched sampling.
Traditional methods of adjustment like matching have limitations because they can only
accommodate a small number of covariates for adjustment.
Rosenbaum and Rubin (1983) demonstrated that the propensity score is a
balancing score and can be used in observational studies to reduce bias through the
adjustment methods mentioned above. The propensity score for an individual is defined
as the conditional probability of being treated given the individual’s covariates. The
propensity score can be used via matching to balance the covariates in the compared
groups and thus reduce bias. Propensity scores provide a scalar summary of the covariate
information and can incorporate a large number of covariates, which plagues
34

conventional stratification. The propensity score for an individual is formally defined as
the probability of being treated conditional on the individual’s covariate values.
Intuitively, the propensity score is a measure of the likelihood that a person would have
been treated using only their covariate scores.
Propensity score matching is usually presented as a way to mimic randomization -
- if a treated and an untreated patient truly have the same propensity for getting treated,
the patient who actually received treatment can be thought of as randomly chosen from
the matched individual. Although patients cannot in principle be randomized to different
time periods (e.g., pre- vs. post-policy), the method of propensity score matching is still
conceptually valid in this setting. When time period (e.g., pre- vs. post- policy) is
considered as a treatment, the same assumptions are required for an observational study
comparing treatments in parallel cohorts.
In this study, patients in the pre- and post-policy expansion cohorts might differ in
terms of important baseline characteristics that could affect their outcomes. To isolate the
effect of the policy from the effects of differences in baseline characteristics between the
pre- and post-policy periods, the pre- and post-policy cohorts were matched to ensure that
baseline characteristics were well balanced between the two groups. In particular, each
post-policy patient was matched to a pre-policy patient with similar baseline
characteristics using a propensity score matching method. The propensity score for each
patient represented their probability, given their baseline characteristics, of being selected
into the post-policy vs. the pre-policy cohort. Patients with similar propensity scores
35

tended to have similar baseline characteristics on average. Therefore, matching patients
with similar propensity scores could balance baseline characteristics between the cohorts.
Mathematical Specification
With complete data, Rosenbaum and Rubin (1983) introduced the propensity
score for subject I(I=1,…N) as the conditional probability of assignment to a particular
treatment (Z
i
=1) versus (Z
i
=0) given a vector of observed covariates x
i
:
e(x
i
) = pr(Z
i
= 1 | X
i
= x
i
)
Where it is assumed that, given the X’s, the Zi are independent:
Pr( Z
1
= z
1
, …, Z= z
N
| X
1
= x
1
, …., X
N
= x
N
) = ∏
i=1

n
e(x
i
)
Zi
{ 1 – e(x
i
)}
1- Zi

The propensity score is the ‘coarsest function’ of the covariates that is a balancing score,
where a balancing score, b(X), is defined as , a function of the observed covariates X
such that the conditional distribution of X given b(X) is the same for treated (Z=1) and
control (Z=0) patients. For a specific value of the propensity score, the difference
between the treatment and control means for all units with that value of propensity score
is an unbiased estimate of the average treatment effect at that propensity score, if the
treatment assignment is strongly ignorable given the covariates. Treatment assignment is
considered strongly ignorable if the treatment assignment, Z, and the response Y, are
known to be conditionally independent given the covariates, X (that is, when Y ⊥ Z | X),
that is, there are no unobserved differences between the treated and the control subjects.
36

Thus, matching on the propensity score tends to produce unbiased estimates of the
treatment effect when treatment assignment is strongly ignorable.
When covariates contain no missing data, the propensity score can be estimated
using discriminant analysis or logistic regression. Both of these techniques lead to
estimates of probabilities of treatment assignment conditional on observed covariates.
The observed covariates are assumed to have a multivariate normal distribution
(conditional on Z) when discriminant analysis is used, where as this assumption is not
needed for logistic regression.
The mathematical specification described above can be easily applied to the
research question on hand, i.e. assessment the impact of formulary expansion on the
health care costs and duration of antipsychotic use. The treatment (Z
i
=1) versus (Z
i
=0)
specified above would be the formulary expansion. So the propensity score would be the
likelihood of initiating drug therapy in the closed access period as opposed to the
transition period or open access period. The covariates Xi will be the set of 80 variables
specified earlier, including age, race, gender, urban or rural residence (versus mixed
counties), disability status, the mix of medical diagnoses recorded in the prior 6 months,
the mental health diagnosis preceding the episode, and prior use of health care by type of
service in both the index month and the prior 6 months. Data describing a patient’s prior
use of antipsychotic drugs were also used as independent variables, including the number
of previous treatment episodes, type of drug(s) used and the chlorpromazine equivalent
37

dose in the episode immediately preceding treatment. The outcome variable Y would be
health care costs and the duration of use of antipsychotics.
There are several ways in which the propensity score can be used (D’Agostino
(1998)). In this study it was applied to matching.
3.3.2 Applications
As an alternative to the OLS regression model, the impact of formulary expansion
on health care costs was also evaluated by a propensity score matching method. First,
unique patient episodes were selected. This was necessary because the multiple-episode
structure could result in multiple episodes of the same patient being allocated to different
matched pairs, leading to the violation of the independence assumption required by the
(propensity score) PS method. As a result, one episode was included for each patient in
the analysis. Thus, one episode from each patient was randomly selected and included in
the analysis. This approach would not affect the expected outcomes, but would reduce
the statistical power. However, given the large sample size available for the analyses, this
should not be an issue as evidenced by the statistical significance in the PS result tables.
Second, these episodes were divided into 3 groups based on the period of their
observation, i.e. closed access period, transition period, and open access period. Third,
logistic regression models were used to predict the likelihood of a patient being in the
closed access period (reference group) versus the transition and open access periods using
the same set of 80 variables used in the OLS model. This predicted probability was
38

defined as his or her propensity score. Fourth, each patient in the open access period was
matched 1:1 to a patient in the close access period on the basis of their propensity scores,
baseline demographics (age, gender, race), and the number of previous treatment
episodes. Similar matching was conducted for transition period.
Finally, Wilcoxon signed rank test was used to compare the impact of open access
on each health care cost component and total days of continuous therapy among the
matched patients groups. Wilcoxon signed rank test was used because it is a
nonparametric test and does not require the assumption of normal distribution of outcome
variables.
The propensity score analyses attempted to further address the issue of selection
bias and provide intuitive interpretation of the results. In terms of selection bias, this
method matched patients on more than 80 variables ranging from demographics,
comorbidities, and concomitant medications to prior health care costs, mental health
diagnoses, and prior antipsychotic utilization. Thus, the patients selected for the analyses
were very similar; the only difference was the period in which they initiated the episodes.
After matching these patients on all the attributes, unobserved heterogeneity in the groups
not accounted by the observed variables was very unlikely to exist.
Regarding the ease of interpretation, multivariate analyses typically compare
adjusted means (mean costs, mean days of therapy), where as propensity score analysis
compares mean costs/drug therapy outcomes, which could be easily interpreted by
various stake holders.
39

Propensity score matching facilitates simultaneous comparison of multiple
outcomes, making it straightforward to interpret cost differences between close access
period and open access period with parallel differences in the cost components.
Furthermore, because the two periods being compared were balanced following
matching, the adjustment procedure was transparent and accessible to a wide audience.

40

Chapter 4. Impact of Formulary Expansion on Atypical Antipsychotic
Use and Health Care Costs
4.1 Results
4.1.1 Descriptive Analyses
Figure 1
Episodes of Antipsychotic Drug Therapy Initiated per
Month by Episode Type: First Use, Switch, Re-Start
0
1000
2000
3000
4000
5000
6000
7000
1994M06
1994M10
1995M02
1995M06
1995M10
1996M02
1996M06
1996M10
1997M02
1997M06
1997M10
1998M02
1998M06
1998M10
1999M02
New Switch Restart TOTAL

The Initiation of Episodes over Time:
Figure 1 displays the time trend of episode initiation by type of episode. A clear
increase in the number of episodes initiated per month occurred in October 1997, the first
month of the transition period under open access. This increase was consistent with the
“access effect” hypothesis, as the total number of episodes would not have increased
41

simply due to substitutions of medications. Moreover, for patients re-starting therapy, the
access effect appeared to be only temporary, as the rate of re-started episodes actually
dropped below levels experienced in the closed access period. The access effect for new
patients and patients who switched or augmented their antipsychotic drug therapy
returned to their levels in the closed access period within 6 months. The stability of first
use episodes before and after the 6 month transition period may indicate that the criteria
for antipsychotic prescribing did not undergo any permanent change from physicians
taking advantage of the more benign side effect profile of these medications to treat less
severely ill patients.
Figure 2
Episodes of Antipsychotic Drug Therapy Initiated per
Month by Type of Drug Used as Initial Therapy
0
1000
2000
3000
4000
5000
6000
7000
1994M06
1994M10
1995M02
1995M06
1995M10
1996M02
1996M06
1996M10
1997M02
1997M06
1997M10
1998M02
1998M06
1998M10
1999M02
Olanzapine Risperidone Quetiapine Typicals Total

Figure 2 presents the time trend data for olanzapine, risperidone, quetiapine, and
conventional antipsychotic medications. The formulary expansion resulted in an
42

immediate increase in the use of olanzapine and risperidone, while quetiapine’s
availability coincided with the Medi-Cal formulary expansion and its use grew steadily
thereafter. Use of conventional antipsychotics dropped significantly, but not enough to
offset the increased use of second-generation antipsychotics. Specifically, the number of
episodes started per month for conventional antipsychotics dropped by 639 (-26%) in the
transition period, and by 1388 (-57%) in the open access period. Episodes started per
month for the second-generation antipsychotics increased by 2214 (+237%) in the 6-
month transition period, but fell to an increase of 876 (+94%) under open access. Thus,
in the long term, the monthly incidence of treatment initiation decreased by 512 (-15%)
under open access relative to closed access. This result may simply reflect the selection
criteria for the study that required a minimum of 12 months of post-treatment data, a
criterion that is less likely to be confirmed as the end of the data period approached.
Changes in Patient Characteristics over Time:
Tables 1-3 present data on the changes over time in the characteristics of patients
initiating episodes of antipsychotic treatment. Open access resulted in a very rapid
uptake of second-generation antipsychotics after October 1997 as conventional
antipsychotic use dropped from 85.9% of the episodes in the closed access period to only
36.6% in the open access period. Changes over time in the mix of episode types were
consistent with both temporary and permanent access effects. Under closed access, new
episodes represented just over half of all episodes of treatment (52.7%), followed by
restarted episodes (36.2%) and switching/augmentation episodes (11.1%). During the
43

transition period, re-started episodes grew to nearly 55% of all episodes, but then
subsided to a level approaching that documented in the closed access period (41.7% vs.
36.2%). This observation was consistent with the hypothesized temporary access effect
that suggested that persons dissatisfied with previously available treatment would re-enter
the system to try the new therapy. Conversely, first-use episodes declined as a proportion
of all episodes in the transition period, then rebounded to a level in the open access
period (47%) that approached the proportion of first-use episodes under closed access
(52.7%).
44

Table 1. Patient Treatment Episode Characteristics by Formulary Access Period:
All Patients
Variable
Closed Access
Period
Transition Period
(10/97-5/98)
Open Access
Period
N=151,938 N=33,288 N=30,660
Initial Medication Used N(%)
3

Typical antipsychotic 130,547 (85.9%) 12,233 (36.7%) 11,227 (36.6%)
Olanzapine 6,786 (4.5%) 12,285 (36.9%) 9,667 (31.5%)
Risperidone 14,605 (9.6%) 8,125 (24.4%) 7,645 (24.9%)
Quetipine 0 (0%) 645 (1.9%) 2,121 (6.9%)
Episode type
3

1st use 80,041 (52.7%) 11,251 (33.8%) 14,388 (46.9%)
Restart 55,073 (36.2%) 18,214 (54.7%) 12,788 (41.7%)
Switch or augmentation 16,824 (11.1%) 3,823 (11.5%) 3,484 (11.4%)
# days between episodes (restarts only)
3
599 995 751
No MH Diagnosis (%)
3
34.8 39.8 51.7
Mental Health Diagnoses among those
with any MH Dx (%)
3

Schizophrenia 46.4 44.5 31.6
Non-Schizophrenic Psychosis 11.5 12.9 16.2
Bipolar 6.2 8.0 9.6
Depression 14.3 14.0 11.9
Other Affective 0.2 0.2 0.3
Anxiety 8.7 6.4 10.1
Substance Abuse 1.9 1.9 3.8
Personality Disorder 0.4 0.4 0.4
Dementia 1.2 1.8 2.4
Other MH Diagnosis 9.2 9.8 14.2
1
p<0.05;
2
p<0.01;
3
p<0.0001

45

Table 2. Patient Characteristics by Formulary Access Period
Variable Closed Access Period
Transition Period
(10/97-5/98)
Open Access
Period
N=151,938 N=33,288 N=30,660
Demographics
Age (in years at episode start)
3
44.8 44.6 45.1
Age Categories
3

< 25 7.7 10.0 12.4
25-35 20.9 19.0 18.5
35-55 37.0 35.4 31.3
55-65 18.7 20.3 19.4
>= 65 15.8 15.3 18.4
Male Gender
3
44.0 46.3 44.6
Aid Category
3

Blind 0.8 0.7 0.6
Old Age Assistance 8.5 8.6 11.5
Disabled 68.3 66.3 55.7
AFDC 17.1 17.3 23.4
Urban residence
3
73.5 75.0 73.6
Rural residence
3
1.8 1.9 2.3
Race
3

White 48.4 49.0 46.6
Black 15.7 14.3 14.8
Hispanic 6.4 5.9 7.2
Other 29.5 30.8 31.4
1
p<0.05;
2
p<0.01;
3
p<0.0001
46

Table 3. Health Care Costs Prior to Treatment by Formulary Access Period
Variable Closed Access Period
Transition Period
(10/97-5/98)
Open Access Period
N=151,938 N=33,288 N=30,660
Health Care Costs (Prior 6 months)
Ambulatory Care
3
2166 2543 2426
Drug
3
561 848 982
Acute Hospitalization (1=yes)
3
2.9% 2.6% 2.2%
Acute hospital services
3
273 154 138
Psych Hospitalization (1=yes)
3
0.05% 0.2% 0.2%
Psychiatric hospital services
3
4.5 29 25
LTC (1=yes)
3
2.9% 5.8% 4.8%
Nursing home costs
3
1100 2354 1842
Psychologist
3
14 15 6
Home Health
3
27 50 59
Hospice
2
4 17 1
Other
3
81 77 115
Net costs
3, 4
3670 5240 4613
Total costs (6 months prior)
3
4232 6088 5595
Health Care Costs (Index Month)
Ambulatory Care 569 535 559
Drug
3
180 348 371
Acute Hospitalization (1=yes)
3
1.3% 0.9% 1.0%
Acute hospital services
3
78 37 44
Psych. Hospitalization (1=yes)
3
0.4% 1.1% 1.1%
Psychiatric hospital services
3
3 9 7
LTC (1=yes)
3
2.6% 5.3% 4.4%
Nursing home costs
3
205 414 334
Psychologist
3
3 3 1
Home Health
3
6 7 11
Hospice
1
1 5 1
Other
3
19 16 25
Net costs
3, 4
885 1026 987
Total Costs (Index month)
3
1065 1374 1357
1
p<0.05;
2
p<0.01;
3
p<0.0001
4
Net cost = total costs – drug costs
47

In the closed access period, schizophrenia was the most common diagnosis among
patients with a mental health-related diagnosis (46.4%), followed by depression (14.3%),
non-schizophrenic psychoses (11.5%), anxiety (8.7%), and bipolar disorders (6.2%). In
the open-access period, a diagnosis of schizophrenia was recorded for only 31.6% of all
episodes with a mental disorder diagnosis. Diagnoses for nonschizophrenic psychoses
(16.2%), anxiety (10.1%), bipolar disorder (9.6%), substance abuse (3.8%), dementia
(2.4%) and other mental disorders (14.2%) also increased in the open access period. The
likelihood that a patient would start antipsychotic drug therapy without any mental health
diagnosis recorded in the prior 6 months increased from 34.8% under closed access to
51.7% under open access. One possible explanation is that open access induced patients
who had been stabilized and not taking medications to return for drug therapy after a long
period of no therapy. For example, the average gap between treatment episodes for
ambulatory patients restarting antipsychotic therapy increased from 599 days under
closed access to 995 days during the transition period, then dropped to 751 days average
under open access. A second hypothesis is that physicians may have been less likely to
record the potentially stigmatizing diagnosis of schizophrenia under open access, as it
was no longer required for access to the second generation antipsychotics.
Patient demographic characteristics are presented in Table 2. The mean age
increased from 44.8 to 45.1 years in the open access period. However, this modest
increase in average age does not reflect the near doubling of patients under age 25 (7.7%
to 12.4%) or the growth in the proportion of patients over age 65 (15.8% to 18.4%) under
48

open access. The proportion of the patient population classified as disabled decreased
significantly from 68.3% to 55.7%, offset mostly by an increase in AFDC (renamed
“TANF” in July, 1997) eligible patients from 17.1% to 23.4%. Hispanic patients
increased from 6.4% to 7.2% of users of antipsychotic drugs, accompanied by a reduction
in the proportion of white patients from 48.4% to 46.6%. The proportion of African
American patients also dropped from 15.7% to 14.8%.
Table 3 presents data on the use of health care services by type of service in the 6-
month pre-index period and the index month. For each access period, prior use of health
care services increased significantly in the transition period, primarily because of a
significant increase in monthly nursing home costs from $1,100 under closed access
period to $2,354 in the transition period. The proportion of patients with prior use of
nursing home services increased from 2.9% to 5.8% in the transition period, then dropped
to 4.7% under open access. Similar increases in nursing home costs and likelihood of use
were also evident in the month in which the episode of care was initiated.
The increased use of antipsychotic medications by patients in nursing homes is
consistent with the hypothesis of a temporary access effect. Consulting pharmacists and
physicians who supervised nursing home care might have re-started atypical
antipsychotics for patients with a history of responding poorly to conventional
antipsychotics soon after the formulary expansion. Nursing home patients may have
been particularly sensitive to the side effects associated with conventional antipsychotic
medications, thus increasing the demand for alternative medications disproportionately in
49

the frail elderly. The greater severity of illness and co-morbidity that might be expected
among nursing home patients may also explain the increased prior use of psychiatric
inpatient services after the closed access period.
Patient Outcomes:
The average duration of uninterrupted therapy during the closed access period
was 69.1 days (Table 4). Only 6.5% of patients used their initial drug for more than 360
consecutive days. This result is consistent with previous results (McCombs 1999) that
found an 11.3% one-year compliance rate across all medications for Medi-Cal patients
with schizophrenia. However, days of continuous therapy and one-year compliance
increased significantly in the transition period, possibly due to the influx of patients re-
starting therapy, often from the nursing home environment (98.6 days; 9.2% compliant
over 1 year). However, during the open access period, duration of therapy fell below the
average in the closed access period (61.7 days versus 69.1 days, respectively).
50

Table 4. Health Care Costs in First Post-Treatment Year by Formulary Access
Period
Variable
Closed Access
Period
Transition Period
(10/97-5/98)
Open Access
Period
N=151,938 N=33,288 N=30,660
Drug Therapy Outcomes
Total days of therapy on initial drug
3
69.1 98.6 61.7
One year of uninterrupted therapy
3
6.5% 9.2% 2.4%
Health Care Costs
Ambulatory care
2
4244 4307 3958
Drug
3
1830 3420 3196
Acute hospitalization (1=yes)
3
12.1% 8.2% 8.9%
Acute hospital services
3
1153 568 694
Psychiatric hospitalization (1=yes)
3
0.4% 0.7% 3.5%
Psychiatric hospital services
3
41 123 458
LTC (1=yes)
3
4.2% 6.9% 6.1%
Nursing home costs
3
2577 5034 4152
Psychologist
3
28 10 3
Home Health
3
68 85 162
Hospice 18 19 14
Other
3
182 203 248
Net costs
3, 4
8313 10352 9689
Total Costs (1 year post)
3
10144 13772 12886
Relative change in costs per month
Net costs +13.3% -1.3% +5.0%
Total costs +19.5% +13.1% +15.2%
1
p<0.05;
2
p<0.01;
3
p<0.0001
4
Net cost = total costs – drug costs
51

The absolute cost data over time in Table 4 suggests that open access increased
costs. The average total cost of the first post-index year increased from $10,114 per
patient in the closed access period to $13,772 per patient in the transition period, then
decreased to $12,886 per patient in the open-access period. However, pre-index costs
also increased over time, as indicated in Table 4. It remains unclear whether the increase
in absolute costs over time can be attributed to open access. Specifically, the majority of
the observed increase in post-index costs under open access was due to a significant
increase in average nursing home cost in the 12-month post index period over time
($2,577; $5,034; and $4,152 for closed, transition, and open access periods, respectively).
Such increases are expected given the influx of nursing home patients under open access.
Simple methods were used to adjust for post-index costs for changes in prior use
over time. The data in the last two rows of Table 4 are based on the ratio of average
monthly costs in the 12-month post-index period to the cost per month in the 6-month
pre-index period. For example, a ratio of 1.50 would indicate a 50% increase in monthly
costs in the post-index period relative to the average level of prior use. Similarly, a ratio
of 0.50 represents a 50% decrease in costs in the post-index period. On average, patients
initiating treatment in the closed access period experienced an increase of 19.5% in total
cost per month in the post-index period relative to the average monthly costs in the pre-
index period. Costs net of drug costs increased 13.3%. Patients initiating treatment in
the open access period experienced a small increase in net cost per month of 5 percent,
primarily due to a smaller increase in nursing home cost. These smaller increases in
52

relative costs were offset by increases in drug costs, resulting in a 15.2 percent increase in
total cost per month under open access, down from 19.5 percent under closed access
Therefore, the conclusions derived from the relative monthly cost data over time lead to
the opposite conclusion from the comparison of unadjusted absolute cost over time.
Specifically, open access appears to decrease post-index costs per month.
Further clarification of the association between open access and patient outcomes
can be gained by focusing the analysis on more homogeneous subgroups of patients. To
avoid the potentially confounding effects of differences in nursing home use, diagnoses,
and episode type, we analyzed ambulatory patients who restarted therapy and had a
diagnosis of schizophrenia in the prior 6 months. The patient outcome data for this sub-
group are provided in Table 5.
53

Table 5. Health Care Costs in First Post-Treatment Year by Formulary Access
Period: Ambulatory Re-starts with Schizophrenia
Variable
Closed Access
Period
Transition Period
(10/97-5/98)
Open Access
Period
N=24,622 N=5,683 N=2,504
Drug Therapy Outcomes
Total days of therapy on initial drug
3
79.0 110.0 65.5
One year of uninterrupted therapy
3
7.5% 12.0% 2.2%
Added 2
nd
Antipsychotic within one year 41.2% 38.2% 41.1%
Health Care Costs
Ambulatory Care
2
9814 10478 11864
Drug
3
2908 4672 4360
Acute hospitalization (1=yes)
3
15.3% 10.5% 14.2%
Acute hospital services
3
1811 748 1045
Psych. hospitalization (1=yes)
3
0.2% 0.4% 5.7%
Psychiatric hospital services 38 101 654
LTC (1=yes) 1.8% 2.3% 2.0%
Nursing home costs 351 418 413
Psychologist
3
49 16 3
Home Health
2
48 92 82
Hospice
2
4 0 22
Other
3
168 227 293
Net costs
4
12284 12080 14377
Total Costs (1 year post)
3
15193 16752 18738
Gap in claims, prior 6 mos.
3
14.9% 19.1% 23.6%
Gap in claims, 1 year post
3
78.7% 76.0% 39.9%
Relative change in costs per month
Net costs -4.6% -15.2% -10.9%
Total costs +3.2% +1.4% -0.1%
1
p<0.05;
2
p<0.01;
3
p<0.0001
4
Net cost = total costs – drug costs

In general, the results from this more homogeneous sub-population did not vary
substantially from the general results. The duration of therapy and the likelihood that a
patient achieved 360 days of uninterrupted therapy on their initial medication increased
from 79 days and 7.5% under closed access to 110 days and 12% during the transition
period, respectively, but fell to 65.5 days and 2.2% in the open access period. The
absolute costs in the 12-month post-index period increased under open access, primarily
due to higher drug costs.
54

The pattern of pre- to post-index changes in monthly costs also parallels the full
population results. Ambulatory patients with schizophrenia who re-started antipsychotic
drug therapy in the closed access period experienced a 3.2% increase in total cost per
month from their pre-index to post-index periods, as compared with a 19.5% increase for
all patients. As described before, much of this increase was for prescription drugs, as the
total amount of other costs per month actually decreased –4.6% (+13.3% for all patients).
The post-treatment cost profile for re-started episodes was significantly less expensive in
the open access period. The average net monthly post-index cost decreased 10.9% and
were sufficient to offset increased drug costs, resulting monthly costs in the post-
treatment period remaining virtually unchanged relative to the prior 6 months.
Two additional and important questions arose from the results in Table 5. First,
does the decline in duration in therapy under open access indicate that the substitution of
second-generation antipsychotics for conventional drugs was not effective? Second, was
the reduction in total post-index cost per month due to changes in population
characteristics or due to the substitution of second-generation drugs for conventional
antipsychotics? This latter question is particularly relevant given the reduction in average
duration under open access.
The data in Table 6 summarize patient outcomes for restarted episodes by
ambulatory patients with schizophrenia, broken down by time period and initial drug.
Several results are of interest. First, all three second-generation antipsychotic
medications exhibited notably longer duration of therapy than conventional
55

antipsychotics. Second, all four classes of antipsychotics exhibited a decrease in average
duration of therapy in the open access period. This suggests that the decline in duration
under open access was related primarily to changes in the characteristics of the patient
population rather than the drugs’ characteristics. Third, the change in the average cost in
the post-index year relative to the 6 pre-index months improved for olanzapine,
risperidone, and conventional antipsychotics under open access. Under closed access,
relative costs had increased. This result was likely explained by the changes in the
characteristics of patients, especially in light of reduced duration of therapy.
56

Table 6. Summary of Patient Outcomes Over Time by Initial Therapy: Ambulatory
Restart Patients with Schizophrenia

PATIENT OUTCOMES
Closed
Access
Period
1
Transition
Period
(10/97-5/98)
1
Open Access
Period
1
p-Value
Conventional Antipsychotics
N=17,831 N=1,429 N=768
Duration on Initial Rx 72 82 51 0.0001
Duration : All Drugs 104 114 62 0.0001
1 Year of Uninterrupted Therapy 5.6% 6.2% 1.0% 0.0001
Added 2
nd
Antipsychotic within one year 44.0% 50.2% 49.4% 0.0001
% change in net costs
2
-2.6% -18.2% -8.2%
% change in total costs +2.2% -6.9% -1.0%
Olanzapine
N=2,843 N=2,869 N=886
Duration on Initial Rx 110 125 74 0.0001
Duration: All Drugs 168 154.4 83 0.0001
1 Year of Uninterrupted Therapy 16.0% 15.2% 2.9% 0.0001
Added 2
nd
Antipsychotic within one year 29.0% 30.4% 32.96% 0.2822
% change in net costs
2
-7.2% -15.1% -14.3%
% change in total costs +17.9% +4.8% +1.5%
Risperidone
N=3,946 N=1,160 N=514
Duration on Initial Rx 88 109 68 0.0001
Duration: All Drugs 134 146 84 0.0001
1 Year of Uninterrupted Therapy 9.8% 11.9% 2.9% 0.0001
Added 2
nd
Antipsychotic within one year 37.6% 42.5% 42.0% 0.0018
% change in net costs
2
-14.0% -8.8% -13.6%
% change in total costs -0.1% +8.2% -3.5%
Quetiapine
N=0 N=222 N=336
Duration on Initial Rx - 105 72 0.0001
Duration: All Drugs 130 82 0.0001
1 Year of Uninterrupted Therapy - 9.5% 1.8% 0.0001
Added 2
nd
Antipsychotic within one year - 37.4% 42.3% 0.4883
% change in net costs
2
- -25.2% +4.1%
% change in total costs - -7.9% +3.3%
1
p<0.0001 for all comparisons of duration of initial therapy, completion rates and switch rates across drugs
within each of the three time periods.
2
Net cost = total costs – drug costs
57

Open access greatly expanded access to second generation antipsychotics for
African American patients, with smaller increases for other minorities, as indicated in
Table 7. Under closed access, African American patients were less likely than white
patients to receive second generation antipsychotics (22% vs. 30%). Under open access,
this situation was reversed, with African American patients being more likely than white
patients to be prescribed second generation antipsychotics (72.7% vs. 67.7%). Other
racial minorities saw similar changes, although the disparities versus white patients were
not as great. Quetiapine, which only became available during this study’s transition
period, remained more likely to be prescribed for white patients during the open access
period (14.2%) compared with 11.6% of African Americans and 13.5% of other races.
Table 7. Conventional verses Second Generation Antipsychotic Use Over Time By
Race: Ambulatory Restart Patients with Schizophrenia
Closed Access Period
1
White Black Other
Conventional AP 70.0% 78.1% 73.4%
2nd Generation AP 30.0% 21.9% 26.6%
Olanzapine 12.6% 8.2% 11.4%
Risperidone 17.4% 13.2% 15.2%
Quetiapine n/a n/a n/a
Transition Period (10/97-3/98)
1
White Black Other
Conventional AP 24.0% 27.7% 25.6%
2nd Generation AP 76.0% 72.3% 74.4%
Olanzapine 51.1% 48.0% 51.0%
Risperidone 20.3% 21.5% 20.0%
Quetiapine 4.6% 2.8% 3.4%
Open Access Period
1
White Black Other
Conventional AP 32.3% 27.3% 30.3%
2nd Generation AP 67.7% 72.7% 69.7%
Olanzapine 33.9% 37.5% 36.2%
Risperidone 19.6% 23.6% 20.0%
Quetiapine 14.2% 11.6% 13.5%
1
p<0.0001 for the cross-tabulation of race by drug in each of the three time periods.
58

All Patient Episodes:
Comparison of all episodes is listed in Table 8. The results were similar to the
restarting episodes. Increased costs and decreased treatment persistence were consistently
seen across all patient types. Ambulatory care costs were the major drivers for the cost
differences.
Patients Re-starting Antipsychotic Drug Therapy
The characteristics of patients re-starting drug therapy are presented in Table 9.
Open access patients were slightly older, on average, than patients with re-start episodes
under closed access. However, the small change in the mean masks the increase in the
proportion of patients younger than 25 (from 4.9% to 8.7%) and over 65 (from 10.4% to
14.6%). The proportion of patients classified as disabled decreased significantly from
80.8% under closed access to 62.7% under open access. This change corresponded well
with the increase in the proportion of patients with no mental health diagnosis in the 6
months prior to re-starting therapy from 27.3% to 50.5%, mainly due to a decrease in
reported diagnoses of schizophrenia (47.3% under closed access, 20.8% under open
access). This increase in missing mental health-related diagnoses in the 6 months prior to
initiation of drug therapy was consistent with the significant increase in days off therapy
from an average of 366 days under closed access to 607 days in the transition period and
453 days under open access. However, despite the apparent reduction of reported mental
illness in patients restarting therapy, the total cost for health care in the 6 months prior to
restarting therapy increased from $7,106 under closed access to $8,058 under open
59

access. This increase was due primarily to the increase in prior use of nursing home
services and prescription drugs. The cost of acute hospital care decreased significantly in
the open access period.
60

Table 8. Impact of Open Access on Health Care Costs: All Episodes
Component of Cost All Patients Patient with
Schizophrenia
Ambulatory
Patients with
Schizophrenia
Number of Episodes N=21,6003 N=59,612 N=56,776
Number of Patients N=14,2364 N=34,236 N=32,577
Transition Period
Total days therapy on intial therapy 19*** 23*** 19***
Total days of therapy: All Drugs 21*** 25*** 21***
Ambulatory care $599*** $1171*** $1265***
Prescription drugs $537*** $795*** $786***
Psychologist -$18*** -$32*** -$31***
Home health -$14** -$4*** -$3***
Hospice -$13 -$25*** -$6***
Acute hospital -$300** -$673** -$633**
Psychiatric hospital $25*** $45*** $47***
Nursing home care $155*** $379*** $122***
Other services $41*** $51*** $53***
Net costs $475*** $913*** $816***
TOTAL COSTS $1011*** $1707*** $1602***
Open Access Period
Total days therapy on initial therapy -8*** -10*** -11***
Total days of therapy: All Drugs -9*** -13*** -14***
Ambulatory care $894*** $1353*** $1501***
Prescription drugs $320*** $428*** $430***
Psychologist -$15*** -$35*** -$35***
Home health $38** $43 $43*
Hospice -$8 -$5 $9
Acute hospital -$260** -$896** -$852*
Psychiatric hospital $353*** $459*** $471***
Nursing home care $138 $46 $66
Other services $28*** $47*** $42***
Net costs $1167*** $1012*** $1249***
TOTAL COSTS $1487*** $1440*** $1678***
*p<0.05; **p<0.01; ***p<0.0001 by ANOVA or Chi-Square
61

Table 9. Characteristics of Patients Re-Starting Drug Therapy
Patient Characteristic Closed Access
N=55,131
Transition Period
N=18,237
Open Access
N=12,810
Age (years at episode start)*** 43.2 +/-16.1 44.8 +/-17.7 44.5 +/-19.6
Age (Categories)***
<25 4.9% 6.0% 8.7%
25-35 24.0% 20.2% 21.5%
35-55 42.6% 40.0% 35.5%
55-65 18.2% 21.4% 19.8%
65+ 10.4% 12.5% 14.6%
Gender (% male)*** 49.6% 47.1% 45.8%
Aid category***
Blind 0.6% 0.5% 0.5%
Old age assistance 5.2% 6.4% 8.8%
Disabled 80.8% 75.1% 62.7%
AFDC 13.5% 18.0% 28.0%
Days off drug therapy*** 336 days +/-290 607 days +/-.434 453 days +/- 387
Primary diagnosis***
No MH diagnosis 27.3% 38.6% 50.5%
Schizophrenia 47.3% 33.8% 20.8%
Non-schizophrenic
psychoses
6.5% 6.6% 6.7%
Bipolar disorder 4.3% 4.9% 5.2%
Depression 6.7% 6.7% 4.9%
Anxiety 2.8% 3.0% 3.5%
All other MH disorders 5.1% 6.4% 8.4%
Prior use of health care (6 mon
prior to index month)

Ambulatory care $3252 +/-13443 $3061 +/-11998 $3022 +/-11677
Prescription drugs*** $769 +/-977 $955 +/-1339 $1151 +/-1985
Used acute hospital (%)*** 6.4% 4.0% 4.3%
Acute hospital costs*** $588 +/-3441 $234 +/-1946 $252 +/-1996
Used psychiatric
hospital(%)***
0.1% 0.3% 0.4%
Psychiatric hospital costs** $9 +/-446 $36 +/-1636 $49 +/-1997
Used nursing home care (%)*** 6.4% 8.7% 9.0%
LTC costs*** $2372 +/-9887 $3509 +/-12104 $3429 +/-11832
All other costs*** $116 +/-79 $138 +/-72 $155 +/-110
TOTAL COSTS*** $7106 +/-17511 $7933 +/-17242 $8058 +/-16938
*p<0.05; **p<0.01; ***p<0.0001 by ANOVA or Chi-Square test.
62

The estimated impact of open access on health care costs for re-started episodes is
summarized in Table 10. Separate estimates were generated for the transition and the
open access periods relative to closed access in order to exclude the transitory access
effect on cost. This analysis focuses on the open access results.
Open access was associated with significant increase in health care costs over the
first post-treatment year, ranging from $2,006 for all patients re-starting drug therapy
(p<0.0001) to $2,974 for non-institutionalized patients with a schizophrenia diagnosis in
the six months prior to re-starting therapy. Most of the increase in costs was due to
increased use of ambulatory care and psychiatric hospital services. Prescription drug
costs also increased significantly. The drug therapy outcomes such as persistence on the
initial therapy and subsequent therapies did not change significantly.
63

Table 10. Comparison of Health Care Costs and Persistency in Transition and Open
Access Period Versus Closed Access Period: Re-Start Episodes
Component of Cost All Patients Patient with
Schizophrenia
Ambulatory Patients
with Schizophrenia
Number of Episodes N=86,165 N=34,886 N=32,787
Number of Patients N=59,486 N=23,763 N=22,281
Transition Period Estimate (SE) Estimate (SE) Estimate (SE)
Total days therapy on initial
therapy
32 (2)*** 32 (4)*** 31 (4)***
Total days of therapy: All Drugs 34 (3)*** 36 (4)*** 34 (4)***
Ambulatory care $1170(160)*** $2197(404)*** $2350(421)***
Prescription drugs $621(33)*** $722(49)*** $717(51)***
Psychologist -$18(1)*** -$30(2)*** -$29(2)***
Home health -$17(6)** -$15(10) -$13(10)
Hospice -$2(5) -$5(4) -$6(3.8)
Acute hospital -$127(43)** -$291(93)** -$246(94)**
Psychiatric hospital $4(24) $32(52) $33(55)
Nursing home care $215(58)*** $398(100)*** $70(56)
Other services $46(5)*** $64(9)*** $66(9)***
Net costs $1272(181)*** $2350(437)*** $2224(445)***
TOTAL COSTS $1893(184)*** $3073(439)*** $2941(447)***
Open Access Period
Total days therapy on initial
therapy
2 (2) 2 (4) 2 (4)
Total days of therapy: All Drugs -5 (3) -2 (4) -2 (5)
Ambulatory care $1218(172)*** $1997(576)*** $2242(603)***
Prescription drugs $364(42)*** $394(71)*** $404(75)***
Psychologist -$16(1)*** -$34(2)*** -$34(2)***
Home health $24(8)** $24(12) $31(13)*
Hospice $3(7) $17(21) $19(23)
Acute hospital -$138(54)** -$366(127)** -$304(127)*
Psychiatric hospital $373(45)*** $502(92)*** $513(98)***
Nursing home care $139(75) $36(153) $51(82)
Other services $39(7)*** $58(13)*** $53(13)***
Net costs $1642(203)*** $2235(622)*** $2570(634)***
TOTAL COSTS $2006(208)*** $2629(627)*** $2974(640)***
*p<0.05; **p<0.01; ***p<0.0001 by ANOVA or Chi-Square
64

Patients Switching or Augmenting Current Therapy:
The descriptive statistics for patients switching therapies or augmenting their
current antipsychotic drug regimen are presented in Table 11. Changes in population
characteristics paralleled closely those observed for patients re-starting antipsychotic
drug therapy. Open access shifted the age distribution towards younger and elderly
patients. The proportion of patients who were classified as disabled decreased under
open access as did the proportion of patients with a prior recorded diagnosis of
schizophrenia. As with patients re-starting therapy, prior use of prescription drugs and
nursing home care increased under open access, while prior use of acute hospital services
decreased significantly in patients with switching, or augmenting episodes. Total costs in
the prior 6 months for the switching/augmentation population increased from $6,805
under closed access to $8,605 under open access.
The estimated impact of open access on health care costs for switching/
augmentation episodes is summarized in Table 12. The estimated impact of open access
was quite different from the corresponding estimates for patients re-starting therapy.
Open access was associated with a small but not statistically significant decrease in total
costs for patients with schizophrenia who switched or augmented their drug therapy in
the open access period, primarily due to reductions in the costs for ambulatory care and
acute hospital services, which were sufficient to offset increased costs for prescription
drugs and psychiatric hospital care.
65

Table 11. Characteristics of Patients Switching or Augmenting Current Drug
Therapy
Patient Characteristic Closed Access
N=16,839
Transition Period
N=3,823
Open Access
N=3,492
Age (years at episode start)* 41.7+/-15.6 42.3+/-17.7 41.2+/-19.3
Age (Categories)***
<25 6.1% 9.1% 12.9%
25-35 26.1% 24.5% 25.0%
35-55 42.0% 36.0% 33.5%
55-65 17.6% 19.3% 16.8%
65+ 8.3% 11.1% 11.7%
Gender (% male)*** 53.1% 51.0% 49.3%
Aid category***
Blind 0.5% 0.7% 0.5%
Old age assistance 3.8% 5.8% 8.0%
Disabled 81.9% 72.0% 61.3%
AFDC 13.8% 21.5% 30.4%
Primary diagnosis***
No MH diagnosis 20.8% 30.9% 41.7%
Schizophrenia 55.7% 41.5% 27.7%
Non-schizophrenic psychoses 6.8% 8.2% 9.8%
Bipolar disorder 4.1% 5.0% 5.5%
Depression 5.6% 6.1% 3.9%
Anxiety 2.4% 2.0% 3.8%
All other MH disorders 4.7% 6.3% 7.6%
Prior use of health care (6 mon
prior to index month)

Ambulatory care $3423+/-13847 $3945+/-14264 $3423+/-12789
Prescription drugs*** $872+/-1058 $1304+/-1469 $1599+/-1719
Used acute hospital (%)*** 4.6% 3.0% 2.6%
Acute hospital costs*** $450+/-3357 $176+/-1560 $202+/-2255
Used psychiatric hospital
(%)***
0.1% 0.4% 0.3%
Psychiatric hospital costs** $5+/-345 $45+/-1438 $36+/-905
Used nursing home care (%)*** 5.1% 8.1% 8.6%
LTC costs*** $1927+/-8990 $3235+/-11664 $3187+/-11422
All other costs*** $128+/-228 $219+/-272 $158+/-320
TOTAL COSTS*** $6805+/-17175 $8924+/-18776 $8605+/-17245
1Year Post Total Health Costs $15363 +/-30546 $17618 +/-30626 $17227 +/-29460
*p<0.05; **p<0.01; ***p<0.0001 by ANOVA or Chi-Square
66

Table 12. Comparison of Health Care Costs and Persistency in Transition and Open
Access Period Versus Closed Access Period: Switching/ Augmenting Episodes
Component of Cost All Patients Patient with
Schizophrenia
Ambulatory
Patients with
Schizophrenia
Number of Episodes N=24,154 N=11,934 N=11,309
Number of Patients N=19139 N=9,120 N=8,616
Transition Period Estimate (SE) Estimate (SE) Estimate (SE)
Total days therapy on initial
therapy
29 (1)*** 29 (2)*** 25 (2)***
Total days of therapy: All Drugs 30 (1)*** 30 (2)*** 26 (2)***
Ambulatory care -$63(338) -$342(731) -$243(772)
Prescription drugs $935(62)*** $1077(102)*** $1085(106)***
Psychologist -$26(2)*** -$36(4)*** -$36(4)***
Home health -$3(14) $14(25) $19(27)
Hospice -$24(13) -$33(24) -$3(4)
Acute hospital -$589(84)*** -$859(159)*** -$826(168)***
Psychiatric hospital $108(40)** $41(32) $44(35)
Nursing home care $154(133) $435(206)* $251(136)
Other services $24(12)* $14(15) $23(16)
Net costs -$421(383) -$766(789) -$774(816)
TOTAL COSTS $514(391) $311(800) $311(827)
Open Access Period
Total days therapy on initial
therapy
-8 (1)*** -7 (2)*** -10 (2)***
Total days of therapy: All Drugs -9 (1)*** -10 (2)*** -12 (2)***
Ambulatory care $187(363) -$1040(919) -$968(979)
Prescription drugs $531(73)*** $396(127)** $374(133)**
Psychologist -$25(2)*** -$40(4)*** -$40(4)***
Home health $38(20) $99(33)** $87(31)**
Hospice -$10(12) -$29(16) -$4(5)
Acute hospital -$479(99)*** -$1022(194)*** -$984(203)***
Psychiatric hospital $399(73)*** $594(156)*** $638(167)***
Nursing home care $177(142) $188(255) $159(130)
Other services $18(11) $4(19) $8(19)
Net costs $304(420) -$1247(1012) -$1104(1044)
TOTAL COSTS $835(428) -$851(1022) -$731(1055)
*p<0.05; **p<0.01; ***p<0.0001
67

4.1.2 Propensity Score Analyses
Patients Re-starting Antipsychotic Drug Therapy
Following patient matching, the comparisons of key demographic variables,
diagnoses, and prior health care costs between closed access and transition period, and
between closed access and open access are presented in Tables 13 through 18. None of
the variables were found to be significantly different between the matched groups,
demonstrating a successful matching process.
The comparisons in 12-month post-treatment costs (broken down by service type)
and total days of therapy between closed access, transition, and open access periods are
presented in Table 19. Both the transition period and open access period were associated
with increase in health care costs.
In the transition period, the total annual health care costs increased from $1,766
for all restarting patients to $3,777 for all restarting episodes of ambulatory patients’ with
schizophrenia. Ambulatory care cost was the major driver for these differences. There
was a small but statistically significant increase in total days of continuous therapy
ranging from 15.4 days for all restarting patients to 7.8 days for all restarting ambulatory
patients with schizophrenia.
The open access period was also associated with an increase in total healthcare
cost ranging from $1,327 for all patients re-starting drug therapy (p<0.0001) to $4,760
for patients with a schizophrenia diagnosis in the six months prior to re-starting therapy
68

(p<0.0001). Ambulatory care accounted for a majority of the increase in costs. However,
unlike the transition period, the today days of continuous therapy decreased in the open
access period from –14.4 days for all patients to –24.1 days for ambulatory patients with
schizophrenia. These differences were statistically significant.
69

Table 13. Characteristics of Patients Following Matching for Patients Re-Starting
Drug Therapy: All Patients

Baseline Characteristic Closed Period Transition Period
N = 10,038 N = 10,038
Age (years at episode start) 44.8+/-17.8 44.9+/-18.1
Age (Categories)
<20 6.9 % 6.9 %
20-34 20.6 % 20.6 %
35-49 38.1 % 38.1 %
50-64 20.7 % 20.7 %
65+ 13.8 % 13.8 %
Gender (% male) 47.6 % 47.6 %
Race
White 50.8 % 50.8 %
Black 4.3 % 4.3 %
Hispanic 14.4 % 14.4 %
Other
2

30.5 % 30.5 %
Aid category
Blind 0.6% 0.6%
Old age assistance 7.3% 7.6%
Disabled 74.3% 73.9%
AFDC 11.4% 12.0%
Days off drug therapy 548.0+/-349.9 538.9+/-404.5
Primary diagnosis
No MH diagnosis 32.7% 39.5%
Schizophrenia 41.8% 29.9%
Non-schizophrenic psychoses 6.7% 7.4%
Bipolar disorder 3.8% 4.8%
Depression 6.3% 7.2%
Anxiety 2.6% 3.3%
All other MH disorders 6.0% 7.9%
Prior use of health care (6
months prior to index month)

Ambulatory care $3,868+/-$14,451 $4,949+/-$17,789
Prescription drugs $799+/-$1,193 $826+/-$1,200
Used acute hospital (%) 4.5% 4.1%
Acute hospital costs $295+/-$1,872 $256+/-$2,031
Used psychiatric hospital (%) 0.1% 0.3%
Psychiatric hospital costs $20+/-$872 $14+/-$361
Used nursing home care (%) 8.9% 9.2%
LTC costs $3,541+/-$12,138 $3,634+/-$12,237
All other costs
$73+/-$254 $73+/-$247
TOTAL COSTS $7,515+/-$18,172 $7,449+/-$15,544

70

Table 14. Characteristics of Patients Following Matching for Patients Re-Starting
Drug Therapy: Patients With Schizophrenia
Baseline Characteristics Closed Period Transition Period
N = 3,108 N = 3,108
Age (years at episode start) 42.9 +/- 14.2 42.8 +/- 14.5
Age (Categories)
<20 3.5 % 3.5 %
20-34 23.4 % 23.4 %
35-49 44.8 % 44.8 %
50-64 21.0 % 21.0 %
65+ 7.3 % 7.3 %
Gender (% male) 55.1 % 55.1 %
Race
White 53.7 % 53.7 %
Black 1.8 % 1.8 %
Hispanic 17.6 % 17.6 %
Other
2

26.8 % 26.8 %
Aid category
Blind 0.5% 0.3%
Old age assistance 2.7% 2.2%
Disabled 86.5% 85.3%
AFDC 5.3% 7.1%
Days off drug therapy 559.5 +/- 360.8 566.5 +/- 419.6
Primary diagnosis
No MH diagnosis . .
Schizophrenia 100.0% 100.0%
Non-schizophrenic psychoses . .
Bipolar disorder . .
Depression . .
Anxiety . .
All other MH disorders . .
Prior use of health care (6 months
prior to index month)

Ambulatory care $7,098 +/- $23,458 $10,214 +/- $28,695
Prescription drugs $989 +/- $1,124 $945 +/- $1,218
Used acute hospital (%) 5.3% 5.0%
Acute hospital costs $373 +/- $2,268 $327 +/- $2,205
Used psychiatric hospital (%) 0.2% 0.1%
Psychiatric hospital costs $19 +/- $455 $12 +/- $430
Used nursing home care (%) 7.1% 7.9%
LTC costs $2,842 +/- $11,097 $3,027 +/- $11,277
All other costs
$85 +/- $279 $88 +/- $236
TOTAL COSTS $9,538 +/- $23,533 $10,063 +/- $19,180
71

Table 15. Characteristics of Patients Following Matching for Patients Re-starting
Drug Therapy: Ambulatory Patients With Schizophrenia
Baseline Characteristics Closed Period Transition Period
N = 2,872 N = 2,872
Age (years at episode start) 41.5 +/- 13.0 41.4 +/- 13.4
Age (Categories)
<20 3.7 % 3.7 %
20-34 24.9 % 24.9 %
35-49 47.1 % 47.1 %
50-64 20.0 % 20.0 %
65+ 4.4 % 4.4 %
Gender (% male) 56.2 % 56.2 %
Race
White 52.9 % 52.9 %
Black 1.8 % 1.8 %
Hispanic 18.0 % 18.0 %
Other
2

27.2 % 27.2 %
Aid category
Blind 0.3% 0.2%
Old age assistance 1.0% 0.7%
Disabled 87.5% 86.7%
AFDC 5.8% 7.5%
Days off drug therapy 563.2 +/- 360.5 566.0 +/- 418.2
Primary diagnosis
No MH diagnosis . .
Schizophrenia 100.0% 100.0%
Non-schizophrenic psychoses . .
Bipolar disorder . .
Depression . .
Anxiety . .
All other MH disorders . .
Prior use of health care (6
months prior to index month)

Ambulatory care $7,095 +/- $22,675 $10,487 +/- $29,224
Prescription drugs $970 +/- $1,140 $942 +/- $1,235
Used acute hospital (%) 5.0% 4.6%
Acute hospital costs $358 +/- $2,252 $300 +/- $2,158
Used psychiatric hospital (%) 0.3% 0.1%
Psychiatric hospital costs $45 +/- $1,399 $12 +/- $447
Used nursing home care (%) . .
LTC costs $0 +/- $0 $0 +/- $0
All other costs
$82 +/- $263 $84 +/- $222
TOTAL COSTS $6,721 +/- $20,908 $7,382 +/- $17,200
72

Table 16. Characteristics of Patients Following Matching for Patients Re-Starting
Drug Therapy: All Patients
Baseline Characteristic Closed Period Open Period
N = 9,153 N = 9,153
Age (years at episode start) 45.1 +/- 18.9 45.4 +/- 20.0
Age (Categories)
<20 7.8 % 7.8 %
20-34 20.9 % 20.9 %
35-49 35.6 % 35.6 %
50-64 19.8 % 19.8 %
65+ 15.9 % 15.9 %
Gender (% male) 46.6 % 46.6 %
Race
White 48.3 % 48.3 %
Black 5.0 % 5.0 %
Hispanic 14.7 % 14.7 %
Other
2

32.1 % 32.1 %
Aid category
Blind 0.5% 0.6%
Old age assistance 9.6% 9.5%
Disabled 64.4% 66.0%
AFDC 16.8% 16.7%
Days off drug therapy 471.7 +/- 341.3 457.4 +/- 387.7
Primary diagnosis
No MH diagnosis 33.7% 52.6%
Schizophrenia 38.4% 18.7%
Non-schizophrenic psychoses 6.8% 6.9%
Bipolar disorder 4.2% 4.9%
Depression 7.5% 4.8%
Anxiety 3.0% 3.4%
All other MH disorders 6.4% 8.8%
Prior use of health care (6
months prior to index month)

Ambulatory care $3,726 +/- $15,105 $4,232 +/- $15,453
Prescription drugs $877 +/- $1,190 $934 +/- $1,269
Used acute hospital (%) 4.6% 4.0%
Acute hospital costs $279 +/- $1,772 $262 +/- $2,157
Used psychiatric hospital (%) 0.1% 0.3%
Psychiatric hospital costs $15 +/- $574 $46 +/- $2,181
Used nursing home care (%) 9.4% 9.6%
LTC costs $3,644 +/- $12,196 $3,683 +/- $12,231
All other costs
$98 +/- $394 $94 +/- $321
TOTAL COSTS $7,578 +/- $18,700 $7,664 +/- $15,834
73

Table 17. Characteristics of Patients Following Matching for Patients Re-Starting
Drug Therapy: Patients With Schizophrenia
Baseline Characteristic Closed Period Open Period
N = 1,731 N = 1,731
Age (years at episode start) 41.0 +/- 14.1 40.8 +/- 14.5
Age (Categories)
<20 3.8 % 3.8 %
20-34 30.3 % 30.3 %
35-49 41.6 % 41.6 %
50-64 18.0 % 18.0 %
65+ 6.4 % 6.4 %
Gender (% male) 57.0 % 57.0 %
Race
White 51.5 % 51.5 %
Black 3.0 % 3.0 %
Hispanic 18.6 % 18.6 %
Other
2

27.0 % 27.0 %
Aid category
Blind 0.3% 0.7%
Old age assistance 2.5% 2.1%
Disabled 79.7% 79.3%
AFDC 10.1% 9.9%
Days off drug therapy 467.3 +/- 344.0 446.8 +/- 381.3
Primary diagnosis
No MH diagnosis . .
Schizophrenia 100.0% 100.0%
Non-schizophrenic psychoses . .
Bipolar disorder . .
Depression . .
Anxiety . .
All other MH disorders . .
Prior use of health care (6 months
prior to index month)

Ambulatory care $7,071 +/- $23,430 $10,238 +/- $25,490
Prescription drugs $1,091 +/- $1,228 $1,065 +/- $1,267
Used acute hospital (%) 5.3% 5.4%
Acute hospital costs $311 +/- $1,813 $373 +/- $3,011
Used psychiatric hospital (%) 0.2% 0.1%
Psychiatric hospital costs $19 +/- $542 $3 +/- $127
Used nursing home care (%) 5.8% 7.0%
LTC costs $2,251 +/- $9,714 $2,532 +/- $10,278
All other costs
$104 +/- $390 $128 +/- $318
TOTAL COSTS $9,960 +/- $26,728 $10,829 +/- $20,577

74

Table 18. Characteristics of Patients Following Matching for Patients Re-starting
Drug Therapy: Ambulatory Patients With Schizophrenia
Baseline Characteristic Closed Period Open Period
N = 1,599 N = 1,599
Age (years at episode start) 39.8 +/- 13.0 39.5 +/- 13.1
Age (Categories)
<20 3.9 % 3.9 %
20-34 31.7 % 31.7 %
35-49 43.3 % 43.3 %
50-64 17.3 % 17.3 %
65+ 3.7 % 3.7 %
Gender (% male) 56.7 % 56.7 %
Race
White 50.1 % 50.1 %
Black 2.8 % 2.8 %
Hispanic 19.6 % 19.6 %
Other
2

27.6 % 27.6 %
Aid category
Blind 0.3% 0.7%
Old age assistance 1.1% 0.9%
Disabled 80.1% 79.9%
AFDC 10.8% 10.3%
Days off drug therapy 466.2 +/- 344.4 444.6 +/- 383.5
Primary diagnosis
No MH diagnosis . .
Schizophrenia 100.0% 100.0%
Non-schizophrenic psychoses . .
Bipolar disorder . .
Depression . .
Anxiety . .
All other MH disorders . .
Prior use of health care (6
months prior to index month)

Ambulatory care $7,047 +/- $23,707 $10,552 +/- $26,206
Prescription drugs $1,070 +/- $1,210 $1,040 +/- $1,271
Used acute hospital (%) 5.0% 4.6%
Acute hospital costs $289 +/- $1,671 $232 +/- $2,045
Used psychiatric hospital (%) 0.2% 0.1%
Psychiatric hospital costs $20 +/- $564 $3 +/- $132
Used nursing home care (%) . .
LTC costs $0 +/- $0 $0 +/- $0
All other costs
$99 +/- $324 $119 +/- $296
TOTAL COSTS $7,864 +/- $26,028 $8,422 +/- $18,831
75

Table 19. Comparison of Health Care Costs Using Propensity Score Method:
Restarting Episodes
All Patients
Patients With
Schizophrenia
Ambulatory Patients
With Schizophrenia
Transition Period
Number of Episodes
Close Access Period 55,131 26,054 24,614
Transition Period 18,237 6,169 5,677
Number of Patients
Close Access Period 36,840 15,655 14,652
Transition Period 13,990 4,271 3,881
Number of Matched Patients
1
10,038 3,108 2,872
Component of Cost
Ambulatory care $1,081** $3,115*** $3,392***
Prescription drugs $527*** $416*** $424***
Psychologist -$18*** -$28*** -$26***
Home Health -$26 -$17 -$40
Hospice -$1 -$15 -$16
Acute hospital -$163** -$97 -$65
Psychiatric hospital -$14 $102 $52
Nursing home care $341 $464 -$22
Other services $40 $78*** $78***
Net costs $1,239 $3,603*** $3,353***
TOTAL COSTS $1,766*** $4,019*** $3,777***
Therapy Duration
Total days of continuous therapy 15.4*** 13.6*** 7.8**
Open Access Period
Number of Episodes
Close Access Period 55,131 26,054 24,614
Open Access Period 12,810 2,669 2,501
Number of Patients
Close Access Period 38,579 16,512 15,455
Open Access Period 10,666 2,014 1,870
Number of Matched Patients
1
9,153 1,731 1,599
Component of Cost
Ambulatory care $506 $3,167*** $3,504***
Prescription drugs $346*** $345** $356**
Psychologist -$16*** -$26*** -$26***
Home Health $35** $38* $64**
76

Table 19, Continued.

Hospice -$2 $32 $34
Acute hospital -$252*** -$135 -$123
Psychiatric hospital $375*** $436*** $479***
Nursing home care $312 $812 $141
Other services $24 $91*** $85***
Net costs $981** $4,415*** $4,159***
TOTAL COSTS $1,327*** $4,760*** $4,515***
Therapy Duration
Total days of continuous therapy -14.4*** -21.3** -24.1***

* p<0.05; ** p<0.01; *** p<0.0001. p-values using Wilcoxon signed rank test.
77

Patients Switching or Augmenting Current Therapy
The comparisons among patients across periods in demographics, 6 months pre-
index health care cost, and mental health diagnoses after matching are listed in Tables 20
through 25. Similar to the results for restarting episodes, none of the comparisons
indicated significant imbalance between any two groups.
The comparison of post-treatment costs and drug utilization (days of continuous
therapy) are presented in Table 26. In the transition period, a pattern similar to that
observed in restarting patients was seen: There was increase in health care costs (albeit of
a smaller magnitude) and total days of continuous therapy. Acute hospitalizations and
ambulatory care were the significant contributors to the increase in total health care costs.
In the open access period, there was an increase in total health care cost ranging
from $99 for all switching/augmenting patients to $3149 for all ambulatory
switching/augmenting patients with schizophrenia. These differences were statistically
significant. There was no statistically significant changes in the total number of days of
continuous therapy in the open access period compared with closed access period.
78

Table 20. Characteristics of Patients Following Matching for Patients Switching
Drug Therapy: All Patients
Baseline Characteristic Closed Period Open Period
N = 2,167 N = 2,167
Age (years at episode start) 42.4 +/- 18.1 42.5 +/- 18.9
Age (Categories)
<20 9.8 % 9.8 %
20-34 24.4 % 24.4 %
35-49 36.7 % 36.7 %
50-64 17.2 % 17.2 %
65+ 12.0 % 12.0 %
Gender (% male) 49.0 % 49.0 %
Race
White 54.6 % 54.6 %
Black 3.9 % 3.9 %
Hispanic 12.4 % 12.4 %
Other
2

29.2 % 29.2 %
Aid category
Blind 0.4% 0.6%
Old age assistance 8.4% 7.6%
Disabled 66.3% 67.5%
AFDC 16.0% 16.4%
Primary diagnosis
No MH diagnosis 23.2% 43.7%
Schizophrenia 48.3% 25.7%
Non-schizophrenic psychoses 8.4% 10.0%
Bipolar disorder 4.2% 5.0%
Depression 6.7% 4.0%
Anxiety 2.4% 4.2%
All other MH disorders 6.8% 7.4%
Prior use of health care (6 months
prior to index month)

Ambulatory care $3,797 +/- $12,983 $4,271 +/- $15,037
Prescription drugs $1,150 +/- $1,549 $1,171 +/- $1,221
Used acute hospital (%) 3.0% 2.5%
Acute hospital costs $173 +/- $1,454 $188 +/- $2,010
Used psychiatric hospital (%) . 0.1%
Psychiatric hospital costs $0 +/- $0 $9 +/- $305
Used nursing home care (%) 8.5% 7.8%
LTC costs $3,359 +/- $11,757 $2,819 +/- $10,736
All other costs
$92 +/- $306 $91 +/- $274
TOTAL COSTS $8,045 +/- $20,195 $7,505 +/- $15,470
79

Table 21. Characteristics of Patients Following Matching for Patients Switching
Drug Therapy: Patients With Schizophrenia
Baseline Characteristic Closed Period Open Period
N = 546 N = 546

Age (years at episode start) 40.2 +/- 13.3 39.9 +/- 13.8
Age (Categories)
<20 3.1 % 3.1 %
20-34 34.1 % 34.1 %
35-49 41.0 % 41.0 %
50-64 17.8 % 17.8 %
65+ 4.0 % 4.0 %
Gender (% male) 59.2 % 59.2 %
Race
White 55.3 % 55.3 %
Black 3.1 % 3.1 %
Hispanic 14.3 % 14.3 %
Other
2

27.3 % 27.3 %
Aid category
Blind 0.4% 0.4%
Old age assistance 1.5% 1.3%
Disabled 81.1% 80.6%
AFDC 10.1% 9.3%
Primary diagnosis
No MH diagnosis . .
Schizophrenia 100.0% 100.0%
Non-schizophrenic psychoses . .
Bipolar disorder . .
Depression . .
Anxiety . .
All other MH disorders . .
Prior use of health care (6
months prior to index month)

Ambulatory care $6,042 +/- $18,669 $8,331 +/- $24,297
Prescription drugs $1,308 +/- $1,383 $1,326 +/- $1,288
Used acute hospital (%) 4.6% 2.9%
Acute hospital costs $291 +/- $1,833 $145 +/- $1,354
Used psychiatric hospital (%) . .
Psychiatric hospital costs $0 +/- $0 $0 +/- $0
Used nursing home care (%) 7.9% 6.0%
LTC costs $2,789 +/- $10,498 $1,995 +/- $9,162
All other costs
$95 +/- $226 $91 +/- $187
TOTAL COSTS $10,802 +/- $26,597 $11,400 +/- $20,728

80

Table 22. Characteristics of Patients Following Matching for Patients Switching
Drug Therapy: Ambulatory Patients With Schizophrenia

Baseline Characteristic Closed Period Open Period
N = 508 N = 508
Age (years at episode start) 38.8 +/- 11.8 38.2 +/- 12.2
Age (Categories)
<20 3.4 % 3.4 %
20-34 36.4 % 36.4 %
35-49 42.9 % 42.9 %
50-64 16.1 % 16.1 %
65+ 1.2 % 1.2 %
Gender (% male) 60.2 % 60.2 %
Race
White 53.5 % 53.5 %
Black 3.4 % 3.4 %
Hispanic 14.6 % 14.6 %
Other
2

28.5 % 28.5 %
Aid category
Blind 0.2% 0.2%
Old age assistance 0.2% 0.2%
Disabled 81.7% 81.9%
AFDC 11.2% 9.4%
Primary diagnosis
No MH diagnosis . .
Schizophrenia 100.0% 100.0%
Non-schizophrenic psychoses . .
Bipolar disorder . .
Depression . .
Anxiety . .
All other MH disorders . .
Prior use of health care (6
months prior to index month)

Ambulatory care $6,470 +/- $19,426 $8,828 +/- $25,002
Prescription drugs $1,295 +/- $1,343 $1,287 +/- $1,269
Used acute hospital (%) 3.9% 2.2%
Acute hospital costs $280 +/- $1,868 $110 +/- $1,319
Used psychiatric hospital (%) . .
Psychiatric hospital costs $0 +/- $0 $0 +/- $0
Used nursing home care (%) . .
LTC costs $0 +/- $0 $0 +/- $0
All other costs
$92 +/- $200 $92 +/- $189
TOTAL COSTS $8,277 +/- $25,684 $9,466 +/- $18,830

81

Table 23. Characteristics of Patients Following Matching for Patients Switching
Drug Therapy: All Patients
Baseline Characteristic Closed Period Transition Period
N = 2,617 N = 2,617
Age (years at episode start) 42.8 +/- 17.3 42.8 +/- 17.5
Age (Categories)
<20 8.1 % 8.1 %
20-34 23.3 % 23.3 %
35-49 38.1 % 38.1 %
50-64 19.1 % 19.1 %
65+ 11.3 % 11.3 %
Gender (% male) 50.7 % 50.7 %
Race
White 56.1 % 56.1 %
Black 3.6 % 3.6 %
Hispanic 12.8 % 12.8 %
Other
2

27.5 % 27.5 %
Aid category
Blind 0.7% 0.6%
Old age assistance 6.4% 5.7%
Disabled 72.9% 74.4%
AFDC 12.8% 12.9%
Primary diagnosis
No MH diagnosis 23.6% 32.6%
Schizophrenia 51.5% 37.6%
Non-schizophrenic psychoses 7.0% 8.8%
Bipolar disorder 3.1% 5.2%
Depression 6.3% 6.8%
Anxiety 2.4% 2.0%
All other MH disorders 6.1% 7.1%
Prior use of health care (6
months prior to index month)

Ambulatory care $4,654 +/- $16,329 $4,394 +/- $14,515
Prescription drugs $1,017 +/- $1,122 $1,075 +/- $1,152
Used acute hospital (%) 3.3% 2.9%
Acute hospital costs $237 +/- $1,799 $179 +/- $1,515
Used psychiatric hospital (%) 0.1% 0.3%
Psychiatric hospital costs $4 +/- $150 $14 +/- $294
Used nursing home care (%) 7.9% 7.9%
LTC costs $3,241 +/- $11,656 $3,064 +/- $11,331
All other costs
$77 +/- $210 $81 +/- $270
TOTAL COSTS $7,486 +/- $16,729 $7,743 +/- $16,239

82

Table 24. Characteristics of Patients Following Matching for Patients Switching
Drug Therapy: Patients With Schizophrenia
Baseline Characteristic Closed Period Transition Period
N = 966 N = 966
Age (years at episode start) 41.6 +/- 13.6 41.9 +/- 13.9
Age (Categories)
<20 3.3 % 3.3 %
20-34 27.5 % 27.5 %
35-49 42.3 % 42.3 %
50-64 21.0 % 21.0 %
65+ 5.8 % 5.8 %
Gender (% male) 57.4 % 57.4 %
Race
White 59.7 % 59.7 %
Black 2.3 % 2.3 %
Hispanic 13.6 % 13.6 %
Other
2

24.4 % 24.4 %
Aid category
Blind 0.5% 0.4%
Old age assistance 2.6% 1.1%
Disabled 83.4% 86.3%
AFDC 8.1% 6.4%
Primary diagnosis
No MH diagnosis . .
Schizophrenia 100.0% 100.0%
Non-schizophrenic psychoses . .
Bipolar disorder . .
Depression . .
Anxiety . .
All other MH disorders . .
Prior use of health care (6
months prior to index month)

Ambulatory care $6,930 +/- $20,353 $7,783 +/- $20,296
Prescription drugs $1,250 +/- $1,244 $1,304 +/- $1,326
Used acute hospital (%) 4.1% 3.8%
Acute hospital costs $327 +/- $2,047 $222 +/- $1,630
Used psychiatric hospital (%) 0.1% 0.1%
Psychiatric hospital costs $1 +/- $34 $7 +/- $203
Used nursing home care (%) 6.5% 7.9%
LTC costs $2,396 +/- $9,957 $2,824 +/- $10,778
All other costs
$89 +/- $236 $102 +/- $273
TOTAL COSTS $8,619 +/- $19,031 $11,170 +/- $20,843
83

Table 25. Characteristics of Patients Following Matching for Patients Switching
Drug Therapy: Ambulatory Patients With Schizophrenia

Baseline Characteristic Closed Period Transition Period
N = 888 N = 888
Age (years at episode start) 40.5 +/- 12.6 40.7 +/- 12.9
Age (Categories)
<20 3.5 % 3.5 %
20-34 29.1 % 29.1 %
35-49 43.9 % 43.9 %
50-64 19.9 % 19.9 %
65+ 3.6 % 3.6 %
Gender (% male) 58.7 % 58.7 %
Race
White 59.0 % 59.0 %
Black 2.3 % 2.3 %
Hispanic 14.3 % 14.3 %
Other
2

24.4 % 24.4 %
Aid category
Blind 0.5% 0.3%
Old age assistance 0.7% 0.3%
Disabled 84.8% 86.7%
AFDC 8.3% 7.2%
Primary diagnosis
No MH diagnosis . .
Schizophrenia 100.0% 100.0%
Non-schizophrenic psychoses . .
Bipolar disorder . .
Depression . .
Anxiety . .
All other MH disorders . .
Prior use of health care (6
months prior to index month)

Ambulatory care $7,030 +/- $20,761 $7,936 +/- $20,500
Prescription drugs $1,277 +/- $1,268 $1,280 +/- $1,288
Used acute hospital (%) 3.9% 3.3%
Acute hospital costs $305 +/- $1,917 $192 +/- $1,535
Used psychiatric hospital (%) 0.1% 0.1%
Psychiatric hospital costs $1 +/- $35 $7 +/- $212
Used nursing home care (%) . .
LTC costs $0 +/- $0 $0 +/- $0
All other costs
$80 +/- $187 $98 +/- $260
TOTAL COSTS $6,419 +/- $17,516 $8,495 +/- $18,516
84

Table 26. Comparison of Health Care Costs Using Propensity Score Method:
Switching Episodes
All Patients
Patients With
Schizophrenia
Ambulatory Patients
With Schizophrenia
Transition Period

Number of Episodes

Close Access Period
16,839 9,382 8,950
Transition Period
3,823 1,585 1,456
Number of Patients

Close Access Period
13,226 6,937 6,588
Transition Period
3,193 1,226 1,118
Number of Matched Patients
1

2,617 966 888
Component of Cost

Ambulatory care
-$259 $853 $906
Prescription drugs
$760*** $919*** $834***
Psychologist
-$26*** -$30*** -$28***
Home Health
-$52 $19 $13
Hospice
-$8 -$17 $1
Acute hospital
-$534*** -$1,129*** -$1,084***
Psychiatric hospital
$67** $12 $13
Nursing home care
$96 $679 $212
Other services
$22 $41* $58*
Net costs
-$694 $429 $92
TOTAL COSTS
$66** $1,347* $926
Therapy Duration

Total days of continuous therapy
28.0*** 24.3*** 24.1***
Open Access Period

Number of Episodes

Close Access Period
16,839 9,382 8,950
Open Access Period
3,492 967 903
Number of Patients

Close Access Period
13,371 7,026 6,670
Open Access Period
2,979 769 718
Number of Matched Patients
1

2,167 546 508
Component of Cost

Ambulatory care
$473 $2,289* $2,358*
Prescription drugs
$498*** $799*** $796***
Psychologist
-$25*** -$31*** -$33***
Home Health
$38 $121 $58
85

Table 26, Continued.

Hospice
$4 $0** $0**
Acute hospital
-$649*** -$1,030* -$1,051*
Psychiatric hospital
$431*** $791*** $821***
Nursing home care
-$697 -$1,211 $157
Other services
$25 $44 $43
Net costs
-$399 $972* $2,353**
TOTAL COSTS
$99** $1,771** $3,149**
Therapy Duration

Total days of continuous therapy -8.5 -11.6 -12.1
* p<0.05; ** p<0.01; *** p<0.0001. p-values using Wilcoxon signed rank test.
86

4.2 Discussion
4.2.1 Impact of the Formulary Expansion on Treated Patient Population and
Outcomes
This study traced the impact of providing open access to second-generation
antipsychotic medications under the Medi-Cal formulary. This policy change triggered
a rapid but temporary influx of patients restarting antipsychotic therapy. This access
effect indicated a rush to meet the pent-up demand for an alternative treatment to
conventional antipsychotics.
Several changes in the characteristics of the patient population seen in the
analyses were consistent with the access effect hypothesis. First, the increase in the
number of episodes initiated per month was largely due to patients restarting therapy.
Moreover, the average gap between treatment episodes for ambulatory patients restarting
antipsychotic therapy increased from 270 days under closed access to 644 days during
the transition period, then dropped to 388 days under open access. The increased days
off therapy for patients restarting therapy may explain the drop in the proportion of
patients with a schizophrenia diagnosis in the transition and open-access period.
Specifically, the mental health status of these patients may have been relatively stable,
resulting in a lower rate of use of services for which a mental health diagnosis was
recorded.
87

A second explanation for the observed change in the patient diagnostic mix is
that physicians may have been less likely to record the potentially stigmatizing diagnosis
of schizophrenia under open access since it was no longer required for access to second-
generation antipsychotics.
The increased use of antipsychotic medications by patients in nursing homes is
also consistent with the hypothesis of a temporary access effect. Open access made it
easier for physicians to prescribe second-generation in the nursing home environment as
prescription drugs used during nursing home days not covered by Medicare are paid for
by Medi-Cal and recorded in the prescription paid claims database. Consulting
pharmacists and physicians were more inclined to restart antipsychotics for patients with
a history of responding poorly to conventional antipsychotics soon after the second-
generation medications were available. Nursing home patients may have been
particularly sensitive to the side effects associated with conventional antipsychotic
medications, thus increasing the demand for alternative medications disproportionately
in the frail elderly.
Patients treated in psychiatric hospitals and psychiatric units of community
hospitals frequently had open access to second-generation antipsychotics. However, to
continue to use these medications on an outpatient basis required that their community-
based physician obtain prior authorization. Anecdotal evidence suggested that prior
authorization (before October 1997) resulted in many physicians switching patients back
to conventional antipsychotics on discharge. As a result, patients often discontinued
88

drug therapy on their own after discharge based on their previous experience with these
older medications. If this phenomenon was common, open access may have made it
easier for physicians and patients to continue the use of second-generation
antipsychotics in the ambulatory setting, thus increasing the likelihood that patients
treated under open access would have a history of prior institutionalizations.
The trends toward increased use of the newer antipsychotic medications by
nursing home patients and patients discharged after a psychiatric admission continued
beyond the temporary transition period. The increased use of the newer medications by
these high-risk patients represents the desired substitution effects envisioned by
policymakers. Of note, the goal of improving access of atypical antipsychotics to
minorities was clearly accomplished.
As noted in the earlier sections, steps were taken to evaluate the impact of
formulary access on health care costs and persistence of therapy. The biggest challenge
to address this issue was the change in the mix of patients using drug therapy following
formulary expansion. Patients were classified in terms of drug therapy episodes (i.e.
switching/augmenting, restarting) and multivariate regression models were used to
control for potential confounders. Propensity score was also used to match homogenous
groups.
For most part, these analyses indicated that there was an increase in total health
care costs for all the patient types after open access. However, the increase was much
higher for patients restarting drug therapy as opposed to those switching drug therapies.
89

For example, the total annual health care cost for all switching patients was $99
compared with $1,327 for all restarting episodes (Tables 24, 25). This could be because
patients involved in switching episodes were by definition on active therapy and were
inherently more symptomatically stable compared with patients involved in restarting
episodes who had been off therapy for a significant amount of time. It is also clear that
restarting episodes constituted a significant portion of all episodes (41.7%) compared
with switching episodes (11.4%) (Table 1). Thus, the aggregate impact on total cost of
the formulary expansion was driven primarily by the cost changes in the restarting
episodes.
Persistence with the initial therapy in the open access period was significantly
lower compared with the closed access period. Even among those using the atypical
antipsychotics the persistence was no better, or slightly worse, than persistence with
initial therapy among those treated with conventional antipsychotics under closed access
(Table 6). It may be because open access made it easier to change drugs, thus reducing
duration of therapy on the initial medication used in the treatment episode.
The proportion of patients using a second antipsychotic within 1 year did
increase significantly under open access across all medications (Table 6). Of note, the
persistence for all episodes improved (Tables 8, 9, 11, 24, 25) in the transition period
compared with closed access period. These findings are consistent regardless of the
methodology used. The explanation may be that, during the transition period, clinicians
90

may have effectively matched medications to patients and this behavior may have
changed over time, leading to suboptimal persistence in the open access period.
4.2.2 Methodological issues
In terms of methodology, there could have been alternative methods to address
potential selection bias in addition to propensity score, such as instrumental variable and
sample selection models. The challenge of applying instrumental variable methodology
in these settings is to identify a variable that is strongly correlated with the time-period
variable (open access, transition, and closed access periods) and not a predictor of the
outcome variables such as cost and continuous drug use. There is no variable in the
dataset that would fit this criterion. Sample selection model suffers from the same
drawback, albeit to slightly lesser extent than the instrumental variable method, because
they require certain variables in the treatment regression that do not belong to the
outcome regression. Other sample selection models rely heavily and are extremely
sensitive to the assumption of joint distribution of the error terms between the outcome
regression and the treatment regression. This assumption is empirically untestable, but
given the lack of any variables that could be potentially exclusive to the outcome (cost
and drug use patterns) regression and not to the treatment (closed access/transition/open
access), it is difficult to support any assumption of joint distribution of the error term.
These limitations make propensity score methodology, which does not require these
stringent assumptions, a reasonable approach.
91

4.2.3 Policy Implications of Adding Atypical Antipsychotics to the Medi-Cal
Formulary
Increases in the cost of drug therapy have led to the design of many management
techniques aimed at reducing cost by restricting physicians’ prescribing autonomy. Many
Medicaid programs have established formularies and often require prior authorization for
nonlisted drugs or therapeutic substitution of a listed agent when a nonlisted drug is
prescribed. Drug formularies in Medicaid programs are also used to leverage
pharmaceutical companies to sell their products at discounted price in exchange for
formulary status. Unfortunately, the state legislatures, managed care plans, and health
policy makers have often implemented formularies with little empirical evidence about
their true impact.
Previous studies reported increase in total resource use when restrictions were
placed on the number of monthly reimbursement for prescriptions in Medicaid programs
or formulary restrictions on expensive drugs (Soumerai, 1987; Soumerai, 1991;
Soumerai, 1994; Kozma, 1990; Moore, 1992, Dranove, 1989; Moore, 1993). The
findings in this study are opposite to what was observed in the prior studies. However
these results are not inconsistent with recent findings from the Clinical Antipsychotic
Trials of Intervention Effectiveness (CATIE) study (Lieberman 2005, Stroup 2006,
McEvoy 2006, Rosenheck 2006) and a study using the same Medi-Cal data (Chen
2008), which indicated that patients with schizophrenia frequently do not achieve stable,
long term drug therapy regardless of the specific drug used. It is not surprising,
92

however, the formulary expansion of atypical antipsychotics in Medi-Cal has a
collective impact that resulted in some unanticipated global changes in overall
antipsychotic use.
4.2.4 Limitations
Some aspects of this study may be considered limitations and need to be
addressed. First, this study only analyzed the direct health care costs incurred by patients
on antipsychotic therapy], which might be appropriate from the perspective of Medi-Cal.
From the societal perspective, in addition to direct costs, the indirect costs of
schizophrenia could be larger than the direct health care costs, as demonstrated by data of
economic burden of schizophrenia (Tedios, 1993). This study did not have the
information on potentially significant benefits such as quality of life improvement,
functional improvement, and work productivity. In addition, this study did not analyze
the long-term treatment (>12 months) outcomes such as symptom relapses and episode
recurrence.
Finally, claims or billing data rarely contain the clinical information on clinical
severity or functional outcomes necessary to detect or adjust for potential biases. Data
compiled for billing or claims purposes are often limited by significant errors and
omissions in clinical important areas (Jollis, 1993). Prospectively collecting the
information in a “real word” practice setting can provide more rigorous answers to such
policy issues as when to begin atypical antipsychotic treatment, which agents brings
93

about the best clinical, functional, and economic outcomes, and how the effectiveness of
atypical antipsychotic treatment changes with formulary policy change.
4.3 Conclusions
Providing unrestricted access to second-generation antipsychotics had a
multifaceted and complex impact on the Medi-Cal program. Two of the added
medications quickly became the most expensive drugs covered by the Medi-Cal program
in terms of total expenditures. Our data indicate that the average monthly cost of
treating patients prescribed antipsychotics increased dramatically across all types of
services. However, a more nuanced analysis indicated that the magnitude of the increase
was not uniform across all patient types. This study also found that, persistence on the
initial medication prescribed decreased under open access. In addition, this analysis
found that open access caused high-cost, previously treated patients to restart
antipsychotic drug therapy using the newer products, especially patients who were
institutionalized in a nursing home within the prior 6 months. Finally, this policy change
clearly achieved one of its objectives for expanding the access of new and more
effective antipsychotics to minorities.

94

References
Alvir JMJ, Lieberman JA, Safferman AZ et al. Clozapine induced agranulocytosis:
incidence and risk factors in the United States. N Engl J Med;329;162-7;1993
American Psychiatric Association. Practice guideline for the treatment of patients with
schizophrenia. Am J Psychiatry 1997; 154(suppl.4)

Anderson C. Measuring what works in health care. Science 263: 1080-1082, 1994
Andreasen NC, Flaum M. Schizophrenia; the characterstic symptoms. Schizophr Bull
17(1) ; 27-49; 1991

Anette-Hanson M. Pharmacoeconomics of schizophrenia in the 21st Century. J Clin
Psychiatry 1999; 60 (suppl 1):26-27

Anthony WA, Cohen MR, Vitalo R. The measurement of rehabilitative outcome.
Schizophrenia Bulletin. 4;365-83;1978

Barnow B, Gain G, Goldberger, A. Issues in the analysis of selectivity bias. In
Stromsdorger E, and Farkas G eds., Evaluation Studies Review Annual, vol 5, Bevlerly
Hills: Sage Publications, 1981

Belsley D, Edwin K, Welsch R. Regression diagnostics: Identifying influential data and
sources of collinearity, New york: John Wiley and sons, 1980

Ben-Akvia M, Lerman SR. Discrete choice analysis: theory and application to travel
demand. Cambridge (MA): MIT Press, 1985

Bengelsdorf H, Church JO, Kaye RA, et al. The cost effectiveness of crisis intervention:
admission diversion savings can offset the high cost of service. J Nerv Ment Dis: 181
(12); 757-62; 1993

Berndt ER Miceli R et al. Comparing SSRI treatment costs for depressiion using
retrospective claims data: the role of non-random selection and skewed data. Manuscript.
1999

Bierer J. Therapeutic social clubs. London: HK Lewis. 1948

Brown GW, Birley JLT, Wing JK. Influence of family life on the course of schizophrenic
disorders: a replication. Br J Psychiatry: 121:241-58; 1972

Buckley PF. Treatment of schizophrenia: let's talk dollars and sense. AM J Managed Care
1998; 4:369-383.
95

California's Medical Assistance Program, Medi-Cal top 100 drug use report:1999 Year-
to-Date through June 1999.

Cameron DE. The day hospital; experimental forms of hospitalization for patients. Mod
Hospital. 69(3); 60-2; 1947

Chen L, McCombs JS, Park JP. Duration of antipsychotic drug therapy in real-world
practice: A comparison with CATIE trial results. Value in Health 11(3) 487-496. 2008

Cheung HK. Schizophrenics fully remitted on antipsychotics for 3-5 years: to stop or
continue drugs? Br J Psychiatry ; 138;490-4;1981

Christison GW, Kirsch DF, Wyatt RJ. When symptoms persist: choosing among
alternative somatic treatments for schizophrenia. Schizophr Bull: 17(1);217-45;1991

Croghan TW, Johnstone BM, Bueshing DP, and Kessler RC. Information needs for
medication coverage decisions in a state Medicaid program. Medical Care 1999; 37
(4):AS24-AS31.

Cromwell DM Bass EB et al. Can restrictions on reimbursement for anti-ulcer drugs
decrease Medicaid pharmacy costs without increasing hospitalizations? Health Service
Research 33:1993-1610, 1999

Crown WH Obenchain RL et al. The application of sample selection models to outcomes
research: the case of evaluating the effects of antidepressant therapy on resource
utilization. Statistics in Medicine 17: 1943-58, 1998

Cutting J. Outcome of schizophrenia: overview. In: Kerr TA. Shaith P. editors.
Contemporary issues in schizophrenia. Washington. DC : American Psychiatric Press.
1986:433-40

D’Agostino RB Jr. Tutorial in biostatistics propensity score methods for bias reduction in
the comparison of a treatment to a non-randomized control group. Statistics in Medicine
17: 2265-2281, 1998

Danzon, PM, Lillard LA. The resolution of medical malpractice claims-modeling the
bargaining process. R-2792-ICJ, The Institute for Civil Justice, Rand Corporation, Santa
Monica

Dickey B., Normand ST., Norton EC., Azeni H., Fisher W., Altaffer F. Managing the
care of shizophrenia: lessons from a 4 year Massachusetts Medicaid Study. Archives of
General Psychiatry 1996;53:945-952.
96

Drake RE, Wallach MA, Teague GB et al. Housing instability and homelessness among
rural schizophrenic patients. Am J Psychiatry 27;405-10;1993

Drannove D. Medicaid drug formulary restrictions. Journal of Law Economics 32: 143-
162, 1989

Duan N. Choosing between the sample-selection model and the multi-part model. J Bus
Econ Stat 2: 283-289, 1984

Dzhagarov M. Experience in organizing a half hospital for mental patients. Neuropathol
Psikhatria ; 6;137-47; 1937

Eaton WW, Bilker W, Haro JM et al. Long term course of hospitalization for
schizophrenia: II Change with passage of time. Schizophrenia Bulletin; 18 (2); 229-41;
1992

Emsley RA, Roberts MC et al. Ethnicity and treatment response in schizophrenia : A
comparison of 3 ethnic groups. J clin Psychiatry;63:9-14, 2002

Fenton WS. Longitudinal course and outcome of schizophrenia. In: Moscarelli M., Ruff
A., Sartorius N., Eds. Handbook of mental health economics and health policy -
Schizophrenia. New York, NY., John Wiley & Sons, Vol 1:1996.

Gibson PJ, Loosbrock DL. Pretreatment patient differcnces : choice of drug therapy
within schizophrenia ISPOR 6
th
Annual International Meeting, VA 2001.

Gondek K. Prescription drug payment policy: past present, and future. Health Care
Finance Rev 15:1-7, 1994

Gujarati DN. Basic Econometrics McGraw Hill Publications Third Edition ; 400-449;
1995

Gunderson JG, Frank AF, Katz HM et al. Effects of psychotherapy in schizophrenia: II.
Comparative outcome of two forms of treatment. Schizophr Bull; 10 (4); 564-98; 1984
Handbook of mental health economics and health policy - Schizophrenia. New York,
NY., John Wiley & Sons, Vol 1:1996

Harding CM, Brooks GW, Ashikaga T. et al. The Vermont longitudinal study of persons
with severe mental illness: II. Long term outcome for subjects who retrospectively met
DSM III criteria for schizophrenia. AM J Psychiatry ; 144 (6); 727-35;1987

97

Hay JW, Leu R et al. Ordinary least Squares and sample-selection models of the demand
for medical care. Journal of Business and Economic Statistics 5: 499-506, 1987

Hay JW. Treatment selection bias in a simultaneous logit-OLS model: physician specialty
choice and specialty income. Dissertation, Yale University

Heckman JJ, Smith J. Assessing the case for social experiments. Journal of Economic
Perspectives 9(2); 85-110, 1995

Heckman JJ. Treatment selection as a specification error. Econometrica 47(1): 153-161,
1979

Herz MI, Endicott J, Spitzer RL et al. Day versus inpatient hospitalization: a controlled
study. Am J Psychiatry: 127 (10); 1371-82; 1971

Hylan TR, Crown WH. et al. Tricyclic antidepressant and selective serotonin reuptake
inhibitors antidepressant selection and health care costs in the naturalistic setting: a
multivariate analysis. Journal of Affective Disorders. 47(1-3): 71-9, 1998

Jeste DV., Akiko J., Lindamer LA., Lacro JP. Medical comorbidity in schizophrenia.
Schizophrenia. Bulletin 1996; 22:413-430.

Jollis JG, Ancukiewicz M, DeLong ER, Pryor DB, Muhlbaier LH, Mark DB.
Discordance of databases designed for claims payment versus clinical information
systems. Annals of Intern Medicine 119(8):844-850, 1993

Judd LL. The clinical course of unipolar major depressive disorders. Arch Gen
Psychiatry 54: 989-991, 1997

Keith SJ., Regier DA, Rae DS. Schizophrenic disorders. In: Robins LN, Regier DA, eds.
Psychiatric disorders in America: the epidemiological Catchment area study. New York,
Toronto, The Free Press, 1991.

Keller MB Kocsis JH et al. Maintenance phase efficacy of sertraline for chronic
depression. JAMA 280:1665-72, 1998

Kozma CM Reeder CE Lingle EW. Expanding Medicaid drug formulary coverage:
effects on utilization of related services. Medical Care 28(10): 963-977, 1990

Kraepelin E. Dementia praecox and paraphrenia (1919). Translated by RM Barclay. New
York: Robert E Kreiger Publishing Company. 1971

98

Kunin CM, Jahansen AM et al. Report of a symposium on use and abuse of antibiotics
worldwide. Review of Infect Dis 12:12, 1990

Kupfer DM Frank E et al. Five-year outcome for maintenance therapies in recurrent
depression. Arch Gen Psychiatry 1992; 49: 769-773

Lamb RH. What did we really expect from deinstitutionalization ? Hosp Community
Psychiatry 32;105-9;1981

Lawson WB. Issues in pharmacotherapy for african americans. In: Ruiz P, ed. Ethnicity
and psychopharmacology. Review of psychiatry, vol 19. Washington, DC: American
psychiatric press: 37-53: 2000

Lee L. Generalized econometric models with selectivity. Econometrica 51: 507-12, 1983

Lee LF. The computation of opportunity costs in polychotomous choice models with
treatment selection. Review of Economics and Statistics 77(3): 421-435, 1995

Leff JP. Schizophrenia and sensitivity to the environment. Schizophr Bull 2;566-74;1976

Leung SF and Yu S. On the Choice between sample selection and two-part models. J
Econometrics 72:197-299,1996

Linn MW, Caffey EM, Klett J et al. Day treatment and psychotropic drugs in the
aftercare of schizophrenic patients: a veterans administration cooperative study. Arch of
Gen Psychiatry ;36;1055-66; 1979

Lieberman JA, Stroup TS, McEvoy JP, et al. Effectiveness of antipsychotic drugs in
patients with chronic schizophrenia. N Engl J Med 2005;353:1209-23

Stroup TS, Lieberman JA, McEvoy JP, et al. Effectiveness of olanzapine, quetiapine,
risperidone and ziprasidone in patients with chronic schizophrenia following
discontinuation of a previous atypical antipsychotic. Am J Psychiatry 2006;163:611-22

Martin BC and McMillan JA. The impact of implementing a more restrictive prescription
limit on Medicaid recipients: effects on cost, therapy, and out-of-pocket expeditures. Med
Care 34:686-701, 1996

Martin BC, Stephen-Miller L. Expenditures for treating schizophrenia: a population-
based study of Georgia Medicaid recipients. Schizophrenia Bulletin 1998 24(3):479-488.

99

Mccombs JS, Luo M, Johnstone BM, Shi J. THe use of conventional antipsychotic
medication for patients with schizophrenia in a MEdicaid population: therapeutic and
cost outcomes over two years. Value in Health 2000 :3(3) 222-231.

McCombs JS, Mulani PM. Impact of atypical antipsychotics in California Medicaid
population ISPOR 6
th
Annual International Meeting, VA 2001

McCombs JS, Nichol MB, Johnstone BM, Stimmel GL, Shi J, Smith R. Use patterns for
conventional antipsychotic medications in Medicaid patients with Schizophrenia. J. Clin.
Psych. 1999; 60(suppl 19): 5-11.

McCombs JS, Nichol MB, Johnstone BM, Stimmel GL, Shi J, Smith R. Antipsychotic
drug patterns and the cost of treating schizophrenia. Psychiatric services 2000 51:525-
527.

McEvoy JP, Liberman JA, Stroup TS, et al. Effectiveness of clozapine versus olanzapine,
quetiapine and risperidone in patients with chronic schizophrenia who did not respond to
prior atypical antipsychotic treatment Am J Psychiatry 2006;163:600-10

McFarland BH. Cost-effectiveness considerations for managed care systems: Treating
depression in primary care. Am. J. Med. 97(suppl) 60:47S-58S, 1994

McGlashan TH. A selective review of recent North American long-term followup studies
of schizophrenia. Schizophrenia. Bulletin. 1988; 14:515-542.

Moore WJ, Newman RJ. US Medicaid drug formularies: Do they work?
Pharmacoeconomics 1(suppl 1): 28-31, 1992

Moore WJ, Newman RJ. Drug formulary restrictions as a cost-containment policy in
Medicaid programs. Journal of Law Economics 36: 79-97, 1993

New York State Office of Mental Health (NYSOMH) Mental Health Statistics Unit.
Statewide chartbook of mental health information. Albany : NYSOMH, 1994

Nicholas AK, Impact of Newer Antipsychotics on Outcomes in Schizophrenia. Clinical
Therapeutics Vol 19. No1, 1997

Nightengal BS, Crumly JM, Liao J, et al. Economic Outcomes of Antipsychotic Agents
in Medicaid Population: Traditional Agents Vs Risperidone. Psychopharmacology
Bulletin 34(3):373-382,1998.

North CS, Current concepts of schizophrenia. Compr Ther 15(3);8-21 1989

100

Rascati KL, Johnsrud M, et al. Olanzapine versus risperidone in the treatment of
schizophrenia : a comparison of costs among texas medicaid patients ISPOR 6
th
Annual
International Meeting, VA 2001

Regier DA, Burke JD, Manderscheid RW et al. The chronically mentally ill in primary
care. Psychol Med 15;265-273; 1985

Rice DP., Miller LS. The economic burden of schizophrenia conceptual and
methodological issues, and cost estimates. In Moscarelli M., Ruff A., Sartorius N., Eds.

Robbins, LN et al Lifetime prevalence of specific psychiatric disorders in three sitesArch
Gen Psychiatry 1984;41;949-58

Rosenbaum PR, Rubin DB. The centrol role of the propensity score in observational
studies for causal effects. Biometrica 70(1): 41-55, 1983

Rosenbaum PR, Rubin DB. Reducing bias in observational studies using subclassification
on the propensity score. Journal of the American Statistical Association 79: 516-524,
1984

Rosenbaum PR. Discussing hidden bias in observational studies. Annals of Intern
Medical 115: 901-905, 1991

Rosenheck RA, Leslie DL, Sindelar J, et al. Cost – Effectiveness of second generation
antipsychotics and perphenazine in a randomized trial of treatment for chronic
schizophrenia. Am J Psychiatry 2006;163:2080-9

Rothman D. Discovery of the asylum: social order and disorder in the new republic.
Boston: Little Brown and Co: 1971

Rubin DB. Estimating causal effects from large data sets using propensity scores. Annals
of Internal Medicine. 127(8S): 757-763, 1997

Russo P, Kirkpatrick B. Deficit syndrome and drug use patterns in a community based
schizophrenia care ISPOR 6
th
Annual International Meeting, VA 2001.

Sartorius N, et al. Early manifestations and first contact incidence of schizophrenia in
different cultures: a preliminary report on the evaluation of the WHO collaborative study
in determinants of outcome of severe mental disorders. Psychol Med ; 16: 909-28;1986

Shermock KM, Fuller MA, et al Health care utilization and costs in schizophrenic patient
taking risperidone versus olanzapine in a veterans administration population. ISPOR 6
th

Annual International Meeting, VA 2001.
101

Smalley WE Griffin MR et al. Effect of a prior-authorization requirement on the use of
nonsteroidal antiinflammatory drugs by medicaid patients. The New Engl J Med
332(24):1412-17,1995

Souetre E. Economics evaluation in schizophrenia. Neuropsychobiology 1997;35:67-69.
Soumerai SB, Avorn D et al. Payment restriction for prescription drugs under medicaid:
effects on therapy, cost, and equity. N Engl J Med 317:550-6, 1987

Soumerai SB, McLaughlin J et al. Effects of limiting medicaid drug-reimbursement
benefits on the use of psychotropic agents and acute mental health services by patients
with schizophrenia. N Engl J Med 331:650-5, 1994

Soumerai SB, Ross-Degnan D et al. Effects of medicaid drug-payment limits on
admission to hospitals and nursing homes. N Engl J Med 325:1072-7, 1991

Stein LI, Test MA. Alternative to mental hospital: Conceptual model, treatment program,
and clinical evaluation. Arch Gen Psychiatry: 37 (4) ; 392-7; 1980

Stahl S, Grady M. High cost use of second generation antipsychotics under California
Medicaid Program Psychiatric Services 57:127–129, 2006

Stolk P, Schneeweiss S et al Impact analysis of the discontinuation of reimbursement: the
case of oral contraceptives. Contraception 2008;78(5):399-404

Stroup AL and Manderscheid RW. The development of the state mental hospital system
in the United States. J of the Washington Academy of Sciences 1988; 78(1): 59-68.

Stroup TS, Lieberman JA, McEvoy JP, et al. Effectiveness of olanzapine, quetiapine,
risperidone and ziprasidone in patients with chronic schizophrenia following
discontinuation of a previous atypical antipsychotic. Am J Psychiatry 2006;163:611-22

Susser E., Streuning EL, Conover S. Psychiatric problems in homeless men: lifetime
psychosis, substance abuse, and current distress in new arrivals at New York City
Shelters. Arch Gen Psychiatry ; 46;845-50;1989

Talbott JA., Golman HH., Ross L. Schizophrenia: an economic perspective. Psychoiatr
Ann 1987; 17:577-579.

Tediosi F, Parazzini F, Garattini L. Cost of schizophrenia studies: a methodological
review HEPAC 2000-1:14-19 2000

102

Torrey EF. A viral-anatomical explanation of schizophrenia. Schizophrenia Bulletin
1991:17 (1): 15-8

Turner JC, Tenhoor WJ. The NIMH Community support program: pilot approach to a
needed social reform. Schizophr Bull; 4;319-48; 1978

Van Putten T, Marder SR Wirshing WC, et al. Surreptitious non-compliance with oral
fluphenazine in a voluntary inpatient population (letter). Arch Gen Psychiatry ;
47(8);786-7;1990

Vaughn CE. Leff JP. The influence of family and social factors on the course of
psychiatric illness: a comparison of schizophrenic and depressed neurotic patients. Br J
Psychiatry ; 129;125-37; 1976

Vella F. Estimating models with treatment selection bias: a survey. The Journal of
Human Resources 33(1): 127-169, 1997

Vella F. Generating conditional expectations from models with treatment selection bias.
Economics Letter 28(1): 97-103, 1988

Viale G et al. Impact of Risperidone on the Use of Mental Health Care Resources.
Psychiatric Services Vol 48. No.9 1997.

Weiden PJ, Olfson M: Cost of relapse in schizophrenia. Schizophrenia Bulletin 21: 419-
429, 1995

White AC, Atmar RL et al. Effects of requiring prior authorization for selected
antimicrobials: expenditures, susceptibilities, and clinical outcomes. Clin Infect Dis
25:230-9, 1997

White H. A heteroscedasticity-consistent matrix estimator and a direct test for
heteroscedasticity. Econometrica 1980;48:817-38

Worrel JA, Marken PA, Beckman SE, Ruether VL. Atypical antipsychotic agents: A
critical review. Am J Health-Syst Pharm: 57; 238-58: 2000

Worrel JA, Markin PA et al. Atypical antipsychotic agents: A critical review. Am J
Health System Pharm 57: 238-258, 2000

Zechnich AD and Greenlick M et al. Elimination of over-the-counter medication
coverage in the Oregon Medicaid population: the impact on program costs and drug use.
Med Care 36:1283-94, 1998

103

Zhao Z. A cost analysis of olanzapine versus risperidone in the treatment of uncontrolled
schizophrenia ISPOR 6
th
Annual International Meeting, VA 2001.

Zhao Z. A retrospective economic evaluation of olanzapine versus risperidone in the
treatment of schizophrenia ISPOR 6
th
Annual International Meeting, VA 2001.

Abstract (if available)

Abstract In October 1997, the California Medicaid Program (Medi-Cal) added atypical antipsychotics to its formulary to facilitate the substitution of the atypical antipsychotics for older medications as clinically warranted, especially in minority patients thought to be particularly at risk for poor outcomes using older medications. Moreover, it was expected that the overall use of antipsychotics would increase as patients who experienced suboptimal outcomes prior to the formulary expansion would again seek treatment once new options were available. The formulary expansion did significantly alter the clinical treatment decision process, resulting in an immediate but temporary increase in the number of patients initiating antipsychotic therapy, many with a recent institutionalization, who restarted drug therapy using the new antipsychotics. There were significant changes in the characteristics of patients using antipsychotic medications. The likelihood of minority patients i.e. African American's gaining access to atypical antipsychotics improved substantially. Persistence on initial antipsychotic decreased and total health care costs increased following open access. However the magnitude of the increase in costs was not uniform across all patient types. Program administrators must use caution when evaluating the impact of unrestricted access on drug therapy outcomes and treatment costs given the changes in the characteristics of patients seeking treatment.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Assessment of the impact of second-generation antipscyhotics in Medi-Cal patients with bipolar disorder using panel data fixed effect models

PDF

Three essays on comparative economic and clinical outcomes of antipsychotics using retrospective claims data

PDF

The development of the prognostic propensity score: an introduction to a method to identify optimal treatment according to individual tailoring variables when heterogeneous treatment effects are ...

PDF

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

PDF

Three essays on estimating the effects of government programs and policies on health care among disadvantaged population

PDF

Pharmacotherapy among patients with type 2 diabetes

PDF

Impact of pharmacy-based transitional care on healthcare utilization and costs

PDF

Understanding primary nonadherence to medications and its associated healthcare outcomes: a retrospective analysis of electronic medical records in an integrated healthcare setting

PDF

The impact of Patient-Centered Medical Home on a managed Medicaid plan

PDF

Impact of medication persistence on clinical and health care cost outcomes in patients with major depressive disorder using retrospective claims data

PDF

The impact of treatment decisions and adherence on outcomes in small hereditary disease populations

PDF

An evaluation of risk predictors for disease progression in Veteran Affairs patients with chronic hepatitis C

PDF

Crohn’s disease: health outcomes and resource utilization in the biologic era

PDF

Investigating racial and ethnic disparities in patient experiences with care and health services use following colorectal cancer diagnosis among older adults with comorbid chronic conditions

PDF

The value of novel antihyperlipidemic treatments in the U.S. healthcare system: Reducing the burden of cardiovascular diseases and filling the gap of low adherence in statins

PDF

A series of longitudinal analyses of patient reported outcomes to further the understanding of care-management of comorbid diabetes and depression in a safety-net healthcare system

PDF

Acculturation team-based clinical program: pilot program to address acculturative stress and mental health in the Latino community

PDF

Investigating a physiological pathway for the effect of guided imagery on insulin resistance

Asset Metadata

Creator Mulani, Parvez (author)

Core Title Effects of a formulary expansion on the use of atypical antipsychotics and health care services by patients with schizophrenia in the California Medicaid Program

School School of Pharmacy

Degree Doctor of Philosophy

Degree Program Pharmaceutical Economics

Publication Date 04/27/2009

Defense Date 03/18/2009

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

Tag antipsychotics,California Medicaid,formulary expansion,Medicaid,OAI-PMH Harvest,schizophrenia

Place Name California (states)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor McCombs, Jeffrey S. (committee chair), Graddy, Elizabeth A. (committee member), Nichol, Michael B. (committee member)

Creator Email parvez.mulani@abbott.com,parvez_mulani@hotmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m2121

Unique identifier UC1106136

Identifier etd-MULANI-2812 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-223786 (legacy record id),usctheses-m2121 (legacy record id)

Legacy Identifier etd-MULANI-2812.pdf

Dmrecord 223786

Document Type Dissertation

Rights Mulani, Parvez

Type texts

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

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu