Developing a case report form for assessing the risk factors for cervicitis among women in rural El Salvador

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content DEVELOPING A CASE REPORT FORM FOR ASSESSING THE RISK
FACTORS FOR CERVICITIS AMONG WOMEN IN RURAL EL SALVADOR
by
Carol Ann Davis
A Thesis Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(APPLIED BIOSTATISTICS AND EPIDEMIOLOGY)
August 2005
Copyright 2005 Carol Ann Davis
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
UMI Number: 1430387
INFORMATION TO USERS
The quality of this reproduction is dependent upon the quality of the copy
submitted. Broken or indistinct print, colored or poor quality illustrations and
photographs, print bleed-through, substandard margins, and improper
alignment can adversely affect reproduction.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if unauthorized
copyright material had to be removed, a note will indicate the deletion.
®
UMI
UMI Microform 1430387
Copyright 2006 by ProQuest Information and Learning Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
ProQuest Information and Learning Company
300 North Zeeb Road
P.O. Box 1346
Ann Arbor, Ml 48106-1346
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
ACKNOWLEDGEMENTS
I would like to acknowledge the principal investigator of the pilot study, Dr. Miriam
Cremer, for allowing me to use the data collected as a part of this research. I would
also like to thank Dr. Cremer for her energy and enthusiasm in helping me to
develop this project.
I would like to thank my thesis committee members, Dr. Susan Groshen and Dr. Juan
Felix for their insight and valuable advice on this thesis. I would especially like to
thank my thesis advisor and mentor, Dr. Victoria Cortessis. I have come a long way
in my four years at USC, and I could have never done it without her support,
guidance and friendship.
Thank you to my family for supporting me all the way from Massachusetts as I
pursue my career. To my fiance: thanks for always being there to support and
encourage me.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
TABLE OF CONTENTS
ACKNOWLEDGEMENTS.............................................................................................ii
LIST OF TABLES AND FIGURES..............................................................................iv
ABBREVIATIONS.......................................................................................................... v
ABSTRACT.....................................................................................................................vi
CHAPTER ONE: INTRODUCTION.............................................................................. 1
Literature Review and Hypotheses............................................................................ 1
Pilot Study in El Salvador.........................................................................................13
CHAPTER TWO: METHODS.......................................................................................18
CHAPTER THREE: RESULTS.................................................................................... 21
Statistical Analyses...................................................................................................21
Case Report Form.....................................................................................................27
Demographic and Anthropometric Factors............................................................31
Reproductive and Pregnancy History..................................................................... 34
History of Hormone Use...........................................................................................35
Medical History........................................................................................................ 36
Sexual History.......................................................................................................... 39
Infection History.......................................................................................................41
CHAPTER FOUR: DISCUSSION................................................................................45
REFERENCES................................................................................................................52
APPENDIX......................................................................................................................55
iii
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
LIST OF TABLES AND FIGURES
Table 1: Self-reported demographic and reproductive characteristics of women
with and without cervicitis as measured by biopsy.......................................................21
Table 2: Self-reported demographic and reproductive characteristics of women
over severity of cervicitis as measured by biopsy........................................................22
Table 3: Associations between cervicitis (cervicitis vs. normal) and selected
covariates......................................................................................................................... 24
Table 4: Association between cervicitis as measured by biopsy and the results of
four screening procedures..............................................................................................26
Table 5: Association between inflammation as measured by Pap smear and the
results of four screening procedures.............................................................................27
Figure 1: Case report form..............................................................................................28
iv
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
ABBREVIATIONS
BV Bacterial Vaginosis
CIN Cervical Intraepithelial Neoplasia
CMV Cytomegalovirus
HIV Human Immunodeficiency Virus
HPV Human Papillomavirus
HSV-2 Herpes Simplex Virus-2
IUD Intrauterine Device
LEEP Loop Electrosurgical Excision Procedure
OC Oral Contraceptive
PID Pelvic Inflammatory Disease
SES Socioeconomic Status
SIL Squamous Intraepithelial Lesion
STI Sexually Transmitted Infection
STD Sexually Transmitted Disease
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
ABSTRACT
Cervicitis is a common gynecologic condition that may lead to serious
complications, including cervical cancer. Rates of cervicitis and cervical cancer are
high in Latin America, which make this a promising region for investigating the
association between these conditions. In January 2004, a pilot study was conducted
in which 510 Salvadorian women were evaluated for cervicitis. This report presents
the associations between factors collected by questionnaire and the presence of
cervicitis in these women, and a revised questionnaire that was designed using the
results of this analysis and the published literature. Goals of future research will be
to use the questionnaire to investigate the risk factors for cervicitis among women in
El Salvador, and to study the association between cervicitis and cervical cancer.
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
CHAPTER ONE: INTRODUCTION
Literature Review and Hypotheses
Cervicitis is a common condition among women, affecting almost half of all
women at some point in their lives. It is defined as inflammation of the uterine
cervix and varies in both severity and duration of symptoms. When clinical
symptoms of the condition are present, the term mucopurulent cervicitis is often used
to describe it. Mucopurulent cervicitis is defined as the presence of an inflamed
endocervix, green/yellow discharge containing greater than 30 polymorphonuclear
neutrophils per 400 x microscopic field, swelling and redness in the zone of ectopy
and easily induced cervical bleeding (Branham et al. 1984, Sweet 1998). However,
the majority of cases of cervicitis are asymptomatic, and as a result they often remain
undiagnosed (Sweet 1998).
Since cervicitis is not a reportable condition, its true prevalence is unknown.
However, studies have shown that there is variation in the prevalence of the
condition across different populations. In a cross-sectional study conducted in the
United States by Critchlow et al., the prevalence of cervicitis was 32% among a
population of college students and 45% among women attending a sexually
transmitted disease (STD) clinic (Critchlow et al. 1995). In contrast, a study
conducted by Keshavarz et al. found the prevalence of cervicitis among nomadic
tribes in Iran ranged from 71% to 88% (Keshavarz et al. 2001). Overall, the high
prevalence of this condition is thought to be a reflection of the many factors involved
in its etiology.
1
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
The major etiologic factors associated with cervicitis are infectious agents.
In general, the cervix is highly vulnerable to infection due to its anatomic location
near the entry to the reproductive tract (Keith et al. 1989). As a result, many types of
bacteria that are commonly found inside the vagina, including staphylococcus and
streptococcus bacteria, may migrate to the cervix and cause inflammation of the
cervical tissue (Keith et al. 1989).
Aside from bacteria normally present in the vagina, sexually transmitted
agents are a major source of cervical inflammation. Bacterial sexually transmitted
infections such as Chlamydia trachomatis and Neisseria gonorrhoeae are two of the
most frequent causes of infectious cervicitis. Many women infected with either
bacterium develop the clinical symptoms of mucopurulent cervicitis. A study
conducted by Paavonen et al. found that C. trachomatis had a strong association with
the diagnosis of mucopurulent cervicitis among a group of women examined at a
clinic in Seattle, Washington (Paavonen et al. 1986). These bacteria are known to
target the columnar epithelium of the endocervix, which is much more vulnerable to
bacterial infection than other cell types in the cervix (Sweet 1998). Infection with
these bacteria is known to not only have acute effects on cervical tissue, but also to
produce chronic inflammation. In particular, Chlamydia has been consistently
shown to induce chronic cervical infections (Paavonen et al. 1986). If untreated,
chronic infection and inflammation may lead to significant complications, including
pelvic inflammatory disease, adverse pregnancy outcomes and cervical carcinoma.
These complications will be discussed later in further detail.
2
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Some viruses transmitted through sexual contact are also associated with
cervicitis. Herpes Simplex Virus-2 (HSV-2) is an agent known to infect the cervix
by targeting the squamous epithelium of the exocervix (Sweet 1998). HSV-2
produces characteristic lesions (or “sores”) that are often painful and may result in
inflammation of the cervical area. Cytomegalovirus (CMV) has also been linked
with cervicitis. A study conducted by Critchlow et al. found that infection with
CMV was an independent risk factor for cervicitis (Critchlow et al. 1995). In
addition, infection with the Human Papillomavirus (HPV) may produce cervical
inflammation; though research has shown that cervicitis associated with HPV is
much more common and more severe when there is co-infection with Chlamydia,
Gonorrhea or HSV-2 (Castle et al. 2003). HPV infection may produce genital warts
that appear on the outside of the genital area or inside the lower genital tract, and
these lesions may result in some degree of inflammation (Sweet 1998). However,
HPV often remains dormant with no obvious symptoms, and there no evidence that
suggests HPV causes inflammation while in this state (Castle et al. 2003). Though
HPV is not an important cause of cervicitis, it plays a key role in the etiology of
cervical cancer, which will be described later in further detail.
Other agents known to induce cervicitis include some fungi, protozoa and
parasites. Trichomonas vaginalis is a protozoon that has been recognized as a cause
of cervicitis for many years (Sweet 1998). Infection with T. vaginalis has been
shown to produce intense inflammation in the cervix. Additionally, bacterial
vaginosis (BV) is a condition that has been associated with cervicitis. BV is a mild
3
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
infection that is caused by an overgrowth of certain bacteria that are normally present
in the vagina. It is characterized by a decrease in vaginal acidity and may produce
vaginal discharge, odor, itching and pain (Keshavarz et al. 2001). It is often
associated with sexual activity, but also may occur in women who are not sexually
active. Studies have found that many women with BV also have concurrent cervical
inflammation (Sweet 1998). Overall, cervicitis is most often induced by bacteria or
viruses, though other agents may play a more minor role in its etiology.
When no infection can be identified, cervicitis may be attributed to non-
infectious agents. Examples of physical or chemical agents that may induce acute
inflammation are: physical trauma to the cervix, spermicides, douches, and radiation
(Sweet 1998). Additionally, evidence suggests that sexual intercourse may introduce
some degree of inflammation. First, intercourse may cause abrasions or irritation to
the cervix, which may lead to an inflammatory response. Second, as postulated by
Keith et al., inflammation may result from an immunologic response to coitally
transmitted antigens, particularly those of sperm. Since the sperm produced by each
sexual partner are unique, contact with a new sexual partner may result in an
immunologic response that leads to cervicitis (Keith et al. 1989). Thus, it is
important to consider both non-infectious and infectious agents in determining the
cause of cervicitis.
Since the majority of cases of cervicitis are caused by sexually transmitted
infections (STIs), the additional observed risk factors for cervicitis are highly
correlated with the risk factors for many STIs. Many studies have found that a major
4
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
risk factor for STIs is young age (particularly being under age 25), and this is largely
attributed to the fact that younger women are more sexually active than older women
(La Montagne et al. 2004, Ghanem et al. 2005). In accordance with this finding,
studies by Marrazzo et al. and Wilson et al. both found that women younger than age
25 had a higher risk of cervical inflammation as compared to older women
(Marrazzo et al. 2002, Wilson et al. 1990). Wilson et al. also found that having a
marital status other than married (such as single, divorced or separated) was
correlated with a higher risk of cervical inflammation (Wilson et al. 1990).
In addition, the use of contraception has been associated with an increased
risk of STIs and cervicitis (Paavonen et al. 1986, Wilson et al. 1990, Ghanem et al.
2005). In particular, the use of oral contraceptives (OCs) has been linked to
cervicitis. A study by Paavonen et al. found a strong association between the use of
oral contraceptives and mucopurulent cervicitis. The authors suggested that the
increased risk might be due to either the sexual practices of OC users or to the
hormonal effects of the medication. After adjusting for sexual behavior-related
variables, the authors found OC use was still associated with cervicitis. (However, it
is important to keep in mind that residual confounding by sexual behavior-related
variables is also possible). Paavonen et al. postulated that the hormones in OCs may
induce hyperplasia of the cervical epithelium, which may cause the cervix to be more
susceptible to infection (Paavonen et al. 1986). More cervical infections would
therefore lead to an increased risk of cervicitis. In addition to oral contraceptives,
other contraceptive methods may also be associated with inflammation of the cervix
5
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
(Sweet 1998). Spermicides, diaphragms and cervical caps may cause irritation to the
cervix, which may result in acute inflammation.
Other risk factors for STIs may also be important predictors of cervicitis.
Accordingly, variables that relate to sexual behavior such as age at first intercourse,
number of lifetime sexual partners and current number of sexual partners may be
related to the prevalence of cervicitis. In addition, socioeconomic status may also be
a predictor of cervicitis if women in lower SES groups are more likely to be sexually
active and less likely to use condoms than women in higher SES groups. Smoking is
another variable that should be investigated, since women who have STIs may be
more or less likely to smoke than women do not have STIs, and because smoking is a
risk factor for cervical cancer.
It is also important to understand how cervicitis is diagnosed and how it is
defined in relation to cervical inflammation. Cervical inflammation is known to
occur in various degrees of severity. It is defined as cervicitis when the degree of
inflammation is severe, and when one or more histological features that have been
associated with cervicitis can be identified. These features include: severe
inflammation of the epithelial and sub-epithelial tissues, inflammation with heavy
polymorphonuclear leukocyte and mononuclear infiltration, intraepithelial abscesses,
lymphoid germinal centers with transformed lymphocytes in the cervical stroma and
occasionally necrosis and ulceration of the epithelial tissue (Sweet 1998). Acute
cervicitis is often signaled by the presence of neutrophils, while chronic cervicitis is
indicated by the presence of lymphocytes. Diagnosis of cervicitis can be
6
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
accomplished through several methods including: cervical cytology, colposcopy and
cervical biopsy (Sweet 1998). Once cervicitis is diagnosed, patients can be tested
for common infections in order to determine the appropriate course of treatment.
Even if there are no symptoms, resolving cervicitis is of vital importance because it
may lead to long-term sequelae.
There are three major complications that are believed to be associated with
cervicitis, and each has the potential to lead to serious health outcomes. The first
complication is pelvic inflammatory disease (PID), which may result from the spread
of inflammation from the cervix to the organs of the upper genital tract, including the
uterus, endometrium, ovaries and the fallopian tubes (Sweet 1998, Keith et al. 1989).
PID is a potentially serious condition that may result in tubal factor infertility,
chronic pelvic pain and ectopic pregnancy (Sweet 1998). Tubal factor infertility is a
serious outcome that is implicated in about 25% of cases of infertility. Ectopic
pregnancy is the leading cause of maternal death in the first trimester of pregnancy
(Sweet 1998). PH D is often associated with Chlamydia and Gonorrhea infections that
remain untreated and lead to cervicitis (Sweet 1998).
The second major complication associated with cervicitis is adverse
pregnancy outcome, caused by the spread of inflammation from the cervix to the
intrauterine compartment during pregnancy. This may result in a host of outcomes
including inflammation and/or premature rupture of the fetal membranes, amniotic
fluid infection, premature delivery and neonatal infections (Paavonen et al. 1986).
7
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Thus, cervicitis during pregnancy can pose potential dangers to both the mother and
the infant.
Finally, there is some evidence that cervicitis may either promote or initiate
cervical cancer (Sweet 1998). Like all cancers, cervical carcinogenesis involves a
complex process, but a great deal has been learned in recent years in regard to its
etiology. Human Papillomavirus (HPV) has been identified as the most important
risk factor in the development of cervical neoplasia (Franco et al. 2003). In fact,
relative risk estimates for the association between HPV and cervical cancer are
among the strongest ever reported in epidemiologic research. Both retrospective and
prospective studies have found HPV to be associated with cervical intraepithelial
neoplasia (CIN) and invasive cancer (Franco et al. 2003). Research has identified
fifteen oncogenic HPV types, but HPV types 16 and 18 are estimated to account for
65% of invasive cervical cancers (Franco et al. 2003).
Although there is now a strong consensus that HPV causes cervical cancer,
questions remain about the mechanisms of cervical carcinogenesis. In particular,
HPV infection is common among women in the general population, but very few
women infected with HPV go on to develop cervical cancer (Hawes and Kiviat
2002). Therefore, an important goal of cervical cancer research is to identify co
factors in HPV-induced carcinogenesis. Inflammation is one such co-factor that has
been under investigation.
Over the last few years, evidence has been accumulating that suggests there is
an association between infection, chronic inflammation and cancer. There are now
8
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
many established examples of infectious agents that are major risk factors for
cancers of various types. Helicobacter pylori infection has been linked to stomach
cancer, Epstein-Barr virus is implicated as a cause of Burkitt’s lymphoma, and
Hepatitis B and C viruses are associated with liver cancer (O’Byme et al. 2001).
Inflammation has also been implicated in many of these associations; for example,
Hepatitis B and C have been shown to produce liver cancer only in the presence of
chronic inflammation (O’Byme et al. 2001). Also, chronic inflammation has been
associated with several epithelial cancers including cancers of the esophagus,
pancreas, colon, vulva and bladder (Hawes and Kiviat 2002). Based on information
from other cancers, learning more about the association between cervical cancer and
inflammation is of great interest.
A biological model has been postulated whereby chronic inflammation may
induce cervical cancer. In general, inflammation is a non-specific immune response
that results from injury to cellular tissue caused by infectious or non-infectious
agents (Castle et al. 2001). As a part of the inflammatory response, reactive
oxidative metabolites are produced (Hawes et al. 2002). Therefore, a possible
biological mechanism in which inflammation may induce cervical cancer would
involve reactive oxidative metabolites, which are produced as a result of chronic
cervical inflammation, causing genotoxic damage to the cervical DNA (Castle et al.
2003). In concert with oncogenic HPV infection, these metabolites may initiate or
cause the progression of cervical carcinogenesis. In addition to these metabolites,
other substances produced by inflammatory cells - including cytokines, chemokines,
9
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
growth factors and angiogenic factors - may also promote carcinogenesis (Hawes
and Kiviat 2002).
Some recent studies provide support for the association between cervical
inflammation and cervical cancer. A study conducted by Schwebke et al. on the
effect of lower genital tract infections on Pap smear results found that cervical
inflammation was significantly associated with squamous intraepithelial lesions
(SIL) (Schwebke et al. 1997). A second study by Castle et al. found that the
presence of cervicitis was associated with 1.9-fold increase in the risk of high-grade
cervical lesions among a group of women from Costa Rica (Castle et al. 2001).
Results from other studies suggest that women who are infected with HPV and fail to
clear the virus develop persistent inflammation and are at an increased risk of
developing cancer (reviewed by O’Byme et al. 2001). Since HPV itself is not
known to produce chronic inflammation, it is thought that co-infection with another
agent that is commonly associated with cervicitis may augment this process. C.
trachomatis and HSV-2 are agents that have been studied in association with HPV
infection and cervical cancer. However, results of epidemiologic studies have been
inconsistent, and there is limited evidence to support either infection as an important
co-factor in cervical carcinogenesis (Hawes and Kiviat 2002, Molano et al. 2005). It
may be that cervical inflammation produced in association with these and other
infections is the underlying co-factor in HPV carcinogenesis. However, further
research is needed to more fully understand the association between inflammation
and cervical cancer.
10
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
A related issue requiring further study is the possibility that cervicitis may
reduce the accuracy of cervical cancer screening. Currently, there are several
different methods used to screen for cervical cancer, with Papanicolaou (Pap) smear
being the most commonly used method in the United States (Schwebke et al 1997).
In countries with less access to screening resources, visual inspection of the cervix
has been used as an alternative to Pap smear for primary screening. Colposcopy is
regarded as the gold-standard method for cervical cancer screening and diagnosis,
and is often used for secondary screening after abnormal results are found by Pap
smear or visual inspection.
There is some evidence that inflammation may hinder the interpretation of
cervical cancer screening tests. A study by Lawley et al. found that 9 out of 78
patients with inflammation that had a normal result from Pap smear were found to
have dysplasia by colposcopy. This suggests that the presence of inflammation may
reduce the sensitivity of the Pap smear (Lawley et al. 1990). However, little research
has been conducted to investigate the effect of cervical inflammation or cervicitis on
the accuracy of colposcopy and visual inspection. In particular, the accuracy of
colposcopy is difficult to study since it is the gold-standard method to which most
other screening methods are compared.
One way to begin studying the connection between inflammation and
cervical cancer is to investigate this association in a population of women with a
high incidence of cervical cancer and a high prevalence of cervicitis (Castle et al.
2003). Latin America is one area of the world where this investigation may be
11
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
possible because both the rates of cervical cancer and the prevalence of cervicitis are
high. The average age-adjusted incidence rate for cervical cancer in Latin America
is about 40 cases per 100,000 women per year, as compared to the United States,
which has fewer than 10 cases per 100,000 women per year (Franco et al. 2003).
However, there are wide variations in incidence rates among the countries in Latin
America, and the differences in rates may be attributed to the presence of screening
activities in some countries and the lack of screening in other countries (Arrossi et al.
2003). Latin American countries with the highest incidence of cervical cancer
include Haiti, Nicaragua, Bolivia and El Salvador (Arrossi et al. 2003). However,
the true burden of the disease may even be underestimated in these countries since
many have poor systems for cancer surveillance (Arrossi et al. 2003).
Mortality from cervical cancer is also a significant concern in Latin America.
It has been estimated that less than 50% of Latin American women who develop
cervical cancer will survive longer than 5 years (Franco et al. 2003). Since cervical
cancer often occurs among relatively young women (ages 40-50), and the age
distribution in Latin America is currently skewed toward even younger age groups, it
is likely that cervical cancer morbidity and mortality will become an epidemic
problem for Latin America in the near future (Arrossi et al. 2003).
In accordance with incidence rates of cervical cancer, a high prevalence of
cervicitis has been found in Latin America. A study conducted in Costa Rica by
Castle et al. found the prevalence of cervicitis was about 45% in the population of
women studied (Castle et al. 2001). A study in Nicaragua by Claeys et al.
12
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
determined the prevalence of cervicitis was 33%; however, this number includes
only patients with clinically apparent cervicitis (Claeys et al. 2005). A study by
Giuliano et al. was able to make a comparison between cervicitis in Latin America
and the United States; they found that the prevalence of inflammatory cells was
statistically significantly higher in women from Mexico compared to women from
the United States (Giuliano et al. 2004). The information from these studies and
others show that cervicitis is a common condition among women in Latin America.
Given the high incidence of cervical cancer, Latin America may be an ideal
population in which to study the association between cervical inflammation and
cervical carcinoma.
Pilot Study in El Salvador
To begin investigating the association between cervicitis and cervical cancer
in Latin America, it is imperative that studies be carefully designed and implemented
in this population. Pilot studies are often used to collect preliminary data, which can
then be used to inform the design of more extensive and effective research. One
such pilot study was conducted in January 2004 in El Salvador, a Latin American
country that has received little attention in prior research. The study was undertaken
by group of U.S.-based physicians as a part of a public health project called
BasicHealth: El Salvador (Cremer et al. in preparation). The purpose of the study
was to conduct cervical cancer screening on women from rural El Salvador using
four separate screening methods. First, a new screening method, which involves
13
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
taking an image of the cervix with a digital camera, was tested. In addition, three
established screening methods were used: Pap smear, colposcopy and visual
inspection. The goal of the study was to compare the sensitivity and specificity of
the new digital screening method with the three more established screening
techniques.
Women were recruited for this study from the rural community of San
Sebastian, El Salvador through an outreach/education project. Women in the
community were encouraged to visit a health center that was set up to provide
cervical screening and other basic healthcare services. The women who attended the
health center were invited to participate in the study. During the three weeks in
which the study was conducted, 526 subjects were enrolled for participation. All
women who agreed to participate completed a process of informed consent. This
research was approved by the Institutional Review Board for the University of
Southern California and by the Minister of Health in El Salvador.
Participants in this study were women between the ages of 18 and 75 with an
intact uterus and cervix, no prior history of cervical, vaginal, ovarian, endometrial or
vulvar cancer and who were able to provide written informed consent. Individuals
were excluded if they had a Pap smear in the previous six months or were pregnant.
After all exclusions, 510 subjects were included in the study population.
A case report form (i.e. questionnaire) that collected information on
demographic and reproductive characteristics was administered to all study
participants by in-person interview. The form queried basic information such as
14
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
name, address and age. It also queried information about pregnancy history, history
of hormone use, previous Pap smears, gynecological history, smoking history and
sexual history. The case report form was originally designed to collect information
on variables that may be related to the risk of cervical cancer.
As a part of the study protocol, three physicians examined each subject. The
first physician collected a Pap smear and then swabbed the cervix with 5% acetic
acid solution to perform visual inspection. The result of visual inspection was scored
as positive or negative for abnormal cervical changes based upon established criteria
(Blumenthal et al. 2005). The presence of an aceto-white lesion was indicated as a
positive result. Pap smears were sent to PathNet Laboratory (San Antonio, Texas)
for cytological examination. A second physician performed digital inspection of the
cervix using a Nikkon Cool Pix 5400 digital camera. A digital photograph was taken
of the cervix and then magnified on a television screen to which the camera was
connected. The test was scored as positive or negative based on criteria similar to
those used for visual inspection (Blumenthal et al. 2005). Finally, a third physician
performed a colposcopy-directed biopsy. The colposcopic impression was scored as
positive or negative and all visible lesions identified by colposcopy were biopsied. If
no lesion was found, a biopsy was taken at 12 o’ clock. All physicians were blinded
to the results of other examiners. The staff of U.S.-based physicians who performed
these procedures included five obstetrician/gynecologists, two obstetrics/gynecology
residents, one gynecologic pathologist and one gynecologic oncologist.
15
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Cervicitis was assessed by a gynecologic pathologist by grading the amount
of inflammation on each subject’s biopsy. Each biopsy was first evaluated for the
presence of cervical cancer or its precursor lesions. Following this evaluation, the
amount of inflammation was determined; the presence of a moderate degree of
inflammation would trigger a diagnosis of cervicitis, while a marked amount of
inflammation would trigger a diagnosis of severe cervicitis. Among virtually all
biopsies, the predominant inflammatory response was chronic, composed of
lymphocytes and plasma cells. These were classified as chronic cervicitis and
divided into moderate or severe depending on the degree of inflammation.
Inflammation was also assessed separately through cytological examination of the
Pap smear results. The Pap smear was categorized for each subject as normal,
inflamed or unknown.
Preliminary analyses of the pilot data revealed a high rate of cervicitis in this
population; specifically, 74% of the study participants were classified as having
cervicitis (369 women). Since the pilot study was originally conducted to investigate
the validity and reliability of cervical cancer screening tests, the case report form
used was not designed to collect information on variables related to cervicitis. In
order to achieve this aim, the objectives of this project were (1) to conduct an
analysis of the data collected in the pilot study in order to identify factors associated
with cervicitis in this population, and (2) using the results of this analysis, to design a
new case report form. The new case report form was designed with two goals in
mind. First, it will be used to determine the major risk factors for cervicitis among
16
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
women in rural El Salvador. Second, it will be used to investigate the relationship
between cervicitis and cervical cancer, including how cervicitis may modify the
accuracy of cervical cancer screening tests. Using the original form as a guide,
known risk factors for cervicitis were considered for inclusion on the new form. It is
anticipated that the new form will be used in future studies to determine the main
factors associated with the presence of both cervicitis and cervical cancer. Using the
data on cervicitis, the goal of future research will be to investigate the association
between cervicitis and cervical cancer and cervical intraepithelial neoplasia (CIN) in
El Salvador.
17
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
CHAPTER TWO: METHODS
As described above, the data used for this project were collected from a pilot
study conducted by Cremer et al. on a population of women from rural El Salvador.
In order to begin designing a new case report form, the original form was reviewed
to enumerate all variables for which data were collected. The data taken from the
case report form were compiled into a Microsoft Access database. In addition, the
data collected on each subject from the results of the four screening tests were also
entered into the database.
As stated previously, cervicitis is defined by the presence of severe cervical
inflammation. In this study, cervical inflammation was assessed based on cytology
results from the Pap smear; while cervicitis was assessed separately using the biopsy
results from colposcopy.
Cervical biopsy results recorded by the study pathologist were later used to
determine the status of cervicitis for each subject. Subjects were categorized as
having no cervicitis, chronic cervicitis, or severe cervicitis (from lowest to highest
degree of cervical inflammation). Up to four samples of cervical tissue were taken
from each subject. When samples of tissue from a subject were scored into different
categories for cervicitis, the individual was assigned the highest designation for
cervicitis based upon all the available biopsy tissue.
After cervicitis status was scored for all participants, the data taken from the
case report form were analyzed to investigate the association between cervicitis and
demographic and reproductive factors in the study population. Chi square tests, t-
18
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
tests and Rruskal-Wallis tests were used to compare demographic and reproductive
characteristics for subjects over the three levels of cervicitis. Scheffe’s test for
multiple comparisons was used to determine which of the three groups statistically
significantly differed by age. Unconditional logistic regression was used to
investigate the univariate associations between demographic and reproductive
variables and cervicitis. The primary model used to study associations between these
variables and cervicitis compared women with cervicitis to women without
cervicitis. A secondary model compared women with severe cervicitis to women
with chronic cervicitis, while excluding women without cervicitis. After computing
odds ratios and 95% confidence intervals for the univariate models, all associations
were re-analyzed with adjustment for age.
The results of the four screening tests were also analyzed by Chi square tests
to determine whether there was an association between cervicitis, as measured by
biopsy, and the screening test results. In addition, inflammation, as determined by
Pap smear results, was analyzed for its association with the screening tests. All
statistical tests used in the analyses were two-sided and a p value of 0.05 was
considered to be statistically significant. SAS software was used for preliminary
statistical analyses (SAS version 9.0; SAS Institute, Inc. Cary, NC) and Stata
software was used for all logistic regression analyses (Intercooled Stata version 8.0;
Stata Corporation, College Station, TX).
In order to choose variables that would be included on the new case report
form, each variable in the data set was considered for its potential relationship with
19
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
cervicitis, and the medical and epidemiologic literature was reviewed to identify the
factors reported to be associated with cervicitis in different populations. The results
of the statistical analysis were used to determine which variables appeared to be
associated with cervicitis in the study population. The results were also used to
consider whether more detailed information would be needed on certain variables in
order to conduct a more accurate analysis of the data in future studies. The
information collected about each variable on the original form was assessed to
determine if more specific data should be collected on the new form. In addition, the
relevance of each variable collected was considered, and variables not related to
cervicitis in the pilot study or the literature, were eliminated from the form.
Moreover, potentially important variables that were not collected on the original case
report form were considered for inclusion on the new form.
Other considerations made in designing the case report form relate to the
feasibility and usefulness of the instrument. Potential differences in knowledge and
attitudes of Latin American women compared to U.S. women were considered.
Characteristics of the population in El Salvador were also taken into account in
determining variables that may be potential confounders. Physicians with prior
experience working in El Salvador, including Drs. Miriam Cremer and Juan Felix
were consulted in developing the form. Also, a physician from El Salvador assisted
with translating the form into Spanish. Finally, the form was designed to maximize
its usefulness in a clinical setting by limiting the length and number of items queried
and by providing a clear and concise format.
20
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
CHAPTER THREE: RESULTS
Statistical Analyses
The distribution of demographic and reproductive variables between women
with cervicitis and women without cervicitis, as measured by biopsy, is presented in
Table 1. The prevalence of cervicitis in the study population was 74%. Age was
found to be strongly associated with the presence of cervicitis (p = 0.003) (Table 1).
Use of hormone injections was also statistically significantly associated with
cervicitis (p = 0.022). Number of pregnancies and history of abnormal periods were
marginally associated with the presence of cervicitis in this population (p = 0.052
and p = 0.070, respectively).
Table 1: Self-reported demographic and reproductive characteristics of women
with and without cervicitis as measured by biopsy
No Cervicitis
n = 131
Cervicitis
n = 369
P|
Age, years 47.5 ± 13.6J 43.3 ± 13.9 0.003
No. o f pregnancies 5.8 ± 4 .0 5.1 ±3.6 0.052
Age at first intercourse, years 19.5 ± 4.2 19.4 ±4.5 0.72
No. o f sex partners, lifetime 1.8 ± 1.2 1.8 ± 1.3 0.77
% oral contraceptive use, current 25 (19.0)* 84 (22.8) 0.38
% hormone injections use, current 23 (17.6) 102 (27.6) 0.022
% previous Pap smear 115(87.8) 334 (90.8) 0.33
% ever smoker 8 (6.2) 14 (3.8) 0.27
% gynecological surgery, past 60 (45.8) 160(43.4) 0.63
% abnormal period, ever 10 (7.7) 50(13.7) 0.070
% STD, ever 3 (2.3) 12(3.3) 0.56
f P values calculated from t-test or Chi square test
% Mean ± Standard deviation
* Number (Percent)
Table 2 displays the distribution of demographic and reproductive variables
among women without cervicitis, women with chronic cervicitis and women with
21
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
severe cervicitis as measured by biopsy. Among women with cervicitis, 82% had
chronic cervicitis and 18% had severe cervicitis (Table 2). There was a statistically
significant difference in median age between the three groups (p = 0.006). In
particular, multiple comparisons tests showed that women without cervicitis were
statistically significantly older than women with chronic cervicitis and women with
severe cervicitis. There was no significant difference in age between women with
chronic cervicitis and women with severe cervicitis.
Table 2: Self-reported demographic and reproductive characteristics of women
over severity of cervicitis as measured by biopsy___________________________
No Cervicitis
n = 131
Chronic
Cervicitis
n = 301
Severe
Cervicitis
n = 68
Pt
Age, years 47 (36-58)J 43 (31-55) 40 (31.5-48.5) 0.006
Age at first intercourse, years 19 (17-22) 19 (17-22) 18(16.5-21) 0.81
No. o f pregnancies
None 3 (2.3)* 9 (3.0) 2 (2.9)
1 - 3 47 (35.9) 114(37.9) 30(44.1)
4 - 6 34 (26.0) 83 (27.6) 22 (32.4)
7 or more 47 (35.9) 95 (31.6) 14 (20.6) 0.54
No. o f failed pregnancies
None 78 (59.5) 201 (66.8) 49 (72.1)
1 29 (22.1) 55 (18.3) 13(19.1)
2 11 (8.4) 25 (8.3) 3 (4.4)
3 or more 13 (9.9) 20 (6.6) 3 (4.4) 0.51
No. o f sex partners, lifetime
0 - 1 68 (51.9) 181 (60.1) 33 (48.5)
2 - 3 54(41.2) 104 (34.6) 25 (36.8)
4 or more 9 (6.9) 16(5.3) 10(14.7) 0.039
% oral contraceptive use, current 25 (19.0) 66 (21.9) 18 (26.5) 0.49
% hormone injections use, current 23 (17.6) 83 (27.6) 19(27.9) 0.072
% previous Pap smear 115(87.8) 271 (90.3) 63 (92.6) 0.53
% ever smoker 8 (6.2) 10 (3.4) 4 (6.0) 0.35
% gynecological surgery, past 60 (45.8) 136 (45.2) 24(35.3) 0.30
% abnormal period, ever 10 (7.7) 38 (12.8) 12 (18.2) 0.092
% STD, ever 3 (2.3) 9 (3.0) 3 (4.5) 0.69
t P values calculated from Kruskal-Wallis test (for age and age at first intercourse) or Chi square test
(for all other variables)
} Median (Q1 - Q3)
* Number (Percent)
22
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
There were also significant differences between women without cervicitis
and women with chronic and severe cervicitis in regard to the number of lifetime sex
partners (p = 0.039), and marginally significant differences in use of hormone
injections (p = 0.072) and history of abnormal periods (p = 0.092) (Table 2).
Table 3 displays the univariate associations between cervicitis and the major
demographic and reproductive variables. There was a statistically significant trend
in decreasing risk of cervicitis with increasing age (p trend = 0.006) (Table 3).
Women over age 55 were about half as likely to have cervicitis as women under age
34 (OR = 0.42, 95% Cl: 0.24-0.77). In addition, there was a marginally significant
trend in decreasing risk of cervicitis with increasing number of failed pregnancies
(p trend = 0.079). Ever use of hormone pills or hormone injections was associated
with cervicitis. There was an increasing trend in the risk of cervicitis with increasing
months of use of hormone pills (p trend = 0.018) and a marginal trend of increasing
risk with increasing months of use of hormone injections (p trend = 0.051). The
variable, ‘Other treatments after Pap smear’, which identified women with a history
of vaginal infection, was significantly associated with cervicitis (OR = 0.61, 95% Cl:
0.38-0.97).
After adjusting for age, there was no longer a statistically significant trend in
risk of cervicitis for number of failed pregnancies, months of use of hormone pills or
months of use of hormone injections (Table 3). Only ‘Other treatments after Pap
smear’ remained statistically significantly associated with cervicitis after adjustment
for age (p = 0.056).
23
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Table 3: Associations between cervicitis (cervicitis vs. normal) and selected
covariates
No Cervicitis
(n = 131)
Cervicitis
(n = 369)
Crude
OR (95% Cl)
Age Adjusted
OR (95% Cl)
Age (continuous) 131 369 0.98 (0.96 - 0.99)
Age (categorized)
Under 32 22 99 1.0
3 3 - 4 0 25 73 0.65 (0 .3 4 - 1.2)
4 1 - 5 5 41 115 0.62 (0 .3 5 - 1.1)
Over 55 43 82 0.42 (0.24 - 0.77)
p trend = 0.006
Gravida (continuous) 131 369 0.95 (0 .9 0 - 1.0) 0.99 (0 .9 2 - 1.1)
Gravida (categorized)
0 pregnancies 3 11 1.0 1.0
1-3 pregnancies 47 144 0.84 (0 .2 2 -3 .1 ) 0.97 (0.25 - 3.7)
4-6 pregnancies 34 105 0.84 (0 .2 2 -3 .2 ) 1.2 (0 .2 9 -4 .6 )
7+ pregnancies 47 109 0.63 (0 .1 7 -2 .4 ) 1.1 (0 .2 7 -4 .4 )
p trend = 0.23 p trend = 0.70
Failed Pregnancies (cont.) 131 369 0.93 (0 .8 1 -1 .1 ) 0.97 (0 .8 4 - 1.1)
Failed Pregnancies (categorized)
None 78 250 1.0 1.0
1 failed 29 68 0.73 (0 .4 4 - 1.2) 0.82 (0 .4 9 - 1.4)
2 failed 11 28 0.79 (0 .3 8 - 1.7) 0.88 (0 .4 1 -1 .9 )
3+ failed 13 23 0.55 (0 .2 7 - 1.1) 0.66 (0 .3 2 - 1.4)
p trend = 0.079 p trend = 0.26
No. sex partners, lifetime 131 369 0.98 (0 .8 3 - 1.1) 0.97 (0 .8 2 - 1.1)
No. sex partners, lifetime
0 - 1 68 214 1.0 1.0
2 - 3 54 129 0.76 (0 .5 0 - 1.2) 0.71 (0 .4 7 -1 .1 )
4 + 9 26 0.92 (0 .4 1 -2 .1 ) 0.90 (0 .4 0 -2 .1 )
p trend = 0.35 p trend = 0.27
Hormone pills 131 369 1.2 (0 .7 6 -2 .1 ) 1.2 (0 .7 2 -2 .0 )
Months of use, hormone pills
0 mos 109 266 1.0 1.0
1-24 mos 17 76 1.8 (1 .0 -3 .2 ) 1 .5 (0 .7 7 -2 .8 )
25 or more mos 5 26 2.1 (0 .8 0 -5 .7 ) 1.7 (0 .6 2 -4 .8 )
p trend = 0.018 p trend = 0.17
Hormone injections 131 369 1.8 (1 .1 -3 .0 ) 1.4 (0 .7 9 -2 .5 )
Months o f use, hormone injections
0 mos 109 266 1.0 1.0
1-18 mos 11 63 2.3 (1 .2 -4 .6 ) 1.9 (0 .9 0 -3 .9 )
19 or more mos 11 39 1.5 (0 .7 2 -2 .9 ) 1.1 (0 .5 4 -2 .5 )
p trend = 0.051 p trend = 0.22
Other hormones 131 369 0.59 (0 .1 4 -2 .5 ) 0.46 (0 .1 1 -2 .1 )
Previous Pap 131 368** 1 .4 (0 .7 3 -2 .6 ) 1.4 (0 .7 2 -2 .7 )
Other treatments for Pap* 130 368 0.61 (0 .3 8 -0 .9 7 ) 0 .6 4 (0 .4 0 -1 .0 )
24
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Table 3: Continued
Smoking, ever 130 365 0.61 (0 .2 5 -1 .5 ) 0.86 (0 .5 7 - 1.3)
Gyn surgery, past 131 369 0.91 (0.61 - 1.4) 0.67 (0 .3 4 - 1.3)
Abnormal period, ever 130 364 1.9 (0 .9 4 -3 .9 ) 1.7 (0 .8 0 -3 .4 )
Hx of STDs, ever 131 362 1.5 (0 .4 1 -5 .3 ) 1.3 (0 .3 7 -4 .8 )
Age at first intercourse 131 369 0.99 (0 .9 5 - 1.0) 1.0 (0 .9 6 -1 .1 )
Age at first intercourse
15 and under 18 60 1.0 1.0
16-17 23 72 0.94 (0 .4 6 - 1.9) 0.99 (0.49 - 2.0)
18-20 44 116 0.79 (0 .4 2 - 1.5) 0.86 (0 .4 5 - 1.6)
21-24 32 73 0.68 (0 .3 5 - 1.3) 0.80 (0 .4 0 - 1.6)
25+ 14 48 1 .0 3 (0 .4 6 -2 .3 ) 1 .3 (0 .5 6 -2 .8 )
p trend = 0.58 p trend = 0.98
*Only variable significantly associated with cervicitis after adjustment for age by Likelihood ratio test
(p = 0.056)
**Some variables included in the analyses have missing data
In the analyses comparing women with severe cervicitis to women with
chronic cervicitis, only number of lifetime sex partners was statistically significantly
associated with severe cervicitis. There was a statistically significant trend in
increasing risk of severe cervicitis with increasing number of sex partners (p trend =
0.015) (data not shown). Women with four or more lifetime sex partners were 3.4
times as likely to have severe cervicitis as women with one or fewer sex partners
(OR = 3.4; 95% Cl: 1.4-8.2) (data not shown). Number of lifetime sex partners
remained statistically significantly associated with severe cervicitis after adjustment
for age (p = 0.045).
Data relating the presence of cervicitis with the results of the four screening
tests is presented in Table 4. Both the result of colposcopy and the quality of
colposcopy were highly associated with the presence of cervicitis (p = 0.020 and p <
0.001, respectively). The result of visual inspection was significantly associated
with cervicitis (p = 0.006). In addition, the result of digital inspection was
25
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
significantly associated with cervicitis (p = 0.002). Cervicitis and inflammation were
not found to be statistically significantly associated.
Table 4: Association between cervicitis as measured by biopsy and the results
of four screening procedures
Normal
n = 131
Cervicitis
n = 369
Pt
Inflammation (based on Pap Smear)*
Yes 39 134
No 81 207
Unknown 11 24 0.33
Quality of Pap Smear
Satisfactory 122 341
Unsatisfactory 9 24 0.91
Result of Pap Smear
Normal 119 338
Abnormal 3 4 0.45
Result of Digital Inspection
Negative 124 309
Suspect Dysplasia 7 59 0.002
Result of Visual Inspection
Negative 117 289
Suspect Dysplasia 14 80 0.006
Quality o f Colposcopy
Satisfactory 83 290
Unsatisfactory 47 76 <0.001
Result of Colposcopy
Normal 118 293
Low Grade Dysplasia 10 64
High Grade Dysplasia 2 9 0.020
f P values calculated by Chi square test
* Some variables have missing data
Table 5 presents the association between cervical inflammation, as
determined from Pap smear results, and the four screening tests. Quality of Pap
smear and result of Pap smear were both statistically significantly associated with the
presence of inflammation (p < 0.001 for both). Result of digital inspection was also
associated with the presence of inflammation (p = 0.003).
26
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Table 5: Association between inflammation as measured by Pap smear and the
results of four screening procedures______________________________________
Inflammation Yes
n = 175
No
n = 294
Unknown
n = 35
Pt
Quality of Pap Smear*
Satisfactory 172 294 5
Unsatisfactory 3 0 30 <0.001
Result o f Pap Smear
Normal 172 287 4
Abnormal 1 7 1 <0.001
Result o f Digital Inspection
Negative 146 262 24
Suspect Dysplasia 28 32 11 0.003
Result o f Visual Inspection
Negative 135 239 31
Suspect Dysplasia 40 55 4 0.25
Quality of Colposcopy
Satisfactory 130 219 25
Unsatisfactory 43 73 10 0.89
Result o f Colposcopy
Normal 140 244 25
Low Grade Dysplasia 28 40 9
High Grade Dysplasia 5 8 1 0.46
f P values calculated by Chi square test
* Some variables have missing data
Case Report Form
In light of the results of the statistical analyses and a review of the relevant
scientific literature, new a case report form (i.e. questionnaire) was developed based
upon the original form from the pilot study. The new case report form is presented
as Figure 1. Variables considered for inclusion on the form are included in one of
the following categories: demographic and anthropometric factors, reproductive and
pregnancy history, history of hormone use, medical history, sexual history and
infection history. The reasons for including each variable on the form will be
described in detail by category.
27
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Figure 1: Case report form
Study I D :__________
Date of Interview : _
Interview er Initials:
Demographics
N am e:______________________________________
Address:________________________________________________________________
Date of B irth:___ /____ /____ A ge:___ years old
H eight:_____ meters or____ ft._____in. W eight:____ kg. or_____lbs.
Marital Status: Single (1 )____ Married (2 )___ Separated (3 ) Divorced (4 )___
Widowed (5) Living with Partner (6 )___
Years of education:_____years
O ccupation:______________________ Occupation of head of household:________________
Number of persons living in hom e:__________ Number of rooms in hom e:__
Household Utilities (please check if yes): Electricity Potable Water____
Telephone G as___ Wood Burning Stove_____
Reproductive and Pregnancy History
Gravidity:___ Full-Term:____ Pre-Term: Spontaneous Abortions:_____
Stillbirths:_____ Ectopic Pregnancies:_____ Living Children:____
Age at m enarche:______years old Date of last menstrual period: / /
History of Hormone Use
History of birth control pill: Yes (1 )____ No (2 )____Unknown (0 )___
If yes: Current user: (1 ) Past user (2 )___
Age started:____Age ended:_____
Total tim e:____ years_____ months
History of hormone injections (for contraception): Yes (1 )___ No (2 )____ Unknown (0)
If yes: Current user (1 )___ Past user (2 )___
Age started:____ Age ended:_____
Total tim e:_____years_____months
28
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Figure 1: Continued
History of hormone therapy (for menopause): Yes (1 )___ No (2 )___ Unknown (0) _
If yes: Current user (1 ) Past user (2 )___
Age started:____ Age ended:_____
Total tim e:____ years months
Medical History
Previous Pap Smear: Yes (1 )___ No (2 )____ Unknown (0 )
Date Most R ecent:___ / /___
Result of Previous Pap Smear: Positive (1 )____ Negative (2 )____ Unknown (0 )___
Treatment Following Previous Pap Smear:
Nothing (0 ) LEEP (1 )____ Cryotherapy (2 )____ Injection (3 )___ Radiation (4 )___
Other (5) (Specify)__________
History of gynecological surgery: Yes (1 ) No (2 )___ Unknown (0 )___
History of Abnormal Period: Yes (1 )___ No (2 )___ Unknown (0 )___
Smoking History: Current (1 )___ Past (2 )___ Never (0 )___
If a Smoker: Age at Start:______years old
Number of Years of Sm oking:______years
Number of Cigarettes per D ay:_____
If Former Smoker:
Age at Stop: years old
Number of Years Since Quit Smoking:_____ years
Smokers in household (excluding self): Yes (1 )___ No (2 )___ Unknown (3 )___
If yes, number of smokers in household (excluding self):____
Sexual History
Age at First Intercourse:___ years If never had sexual intercourse, check here:
Number of Lifetime Sexual Partners:___ Number of Sexual Partners in Last Year:
29
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Figure 1: Continued
Currently Sexually Active: Yes (1 )___ No (2 )____ Unknown (0 )___
If yes, Current Contraception: None (0 )_____ Birth Control Pill (1 )_
Hormone Injections (2 )____ Condom (3 )___ Implant (4 )___
Tubal Ligation (5 )___ Male Sterilization (6) ___
Diaphragm (7 )____ Cervical Cap (8 ) IUD (9 )____
Spermicide (10) Sponge (1 1 )____ Rhythm Method (12)___
Withdrawal (1 3 )___ Abstinence (1 4 )___
Age started using current contraception:____
Past Use of Contraception: Yes (1 )___ No (2 )____ Unknown (0 )___
If yes, types used (from above list):______________ , _____________, ____________
Age started using past contraception: Age ended: ____
Infection History
History of vaginal infection (including burning, itching, pain, or discharge):
Yes (1 )___ No (2 )___ Unknown (0 )___
If yes: What was the cause: STD (1 ) Other (2 ) Unknown (0 )___
Was treatment used: Yes (1 )___ No (2 )___ Unknown (0 )___
What type of treatment: Antibiotic Pill (1 ) Cream (2 )___ None (3)
Other (4 )____________ Unknown _(0 )___
History of a Sexually Transmitted Disease: Yes (1 )____ No (2 ) Unknown (0 )_______
If yes: What STD Diagnosed: Chlamydia (1 ) Gonorrhea (2 )___
Herpes Simplex Vims (3 )___ HPV (4 ) HIV (5 )__
Other (6 )________________ Unknown (0 )____
Year(s) when Diagnosed:_________ , __________ , __________
What type of treatment: Antibiotic Pill (1 ) Cream (2 )___ None (3)
Other (4 )_____________
Current Symptoms of Infection: Yes (1 )___ No _(2 )____ Unknown (0 )___
If yes: Provide Current Symptoms: Burning (1 )_____ Itching (2 )____
Discharge (3 )___ Odor (4 )____ Vaginal Pain (5 )___
Pain Dining Sex (6 )___ Bleeding After Sex (7 )___
Other (8 )_______________
30
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Demographic and Anthropometric Factors
The original case report form collected information on several demographic
variables. These include age at time of interview, date of birth, name and address.
Age is often one of the most important variables that can be collected in assessing
the risk factors for a disease. As discussed previously, age was found to be
statistically significantly associated with cervicitis in this population (p = 0.003)
(Table 1). In addition, a statistically significant trend was found in reduced risk of
cervicitis with increasing age (p trend = 0.006) (Table 3). Moreover, when analyses
were conducted in which associations between other variables and cervicitis were
adjusted for age, very few variables remained significantly associated with cervicitis.
Based upon this information, it was concluded that age is an important predictor of
cervicitis, and as such, age is included on the new case report form. Date of birth
provides the same information as age, but is also often collected because it is a more
flexible type of measurement. For example, date of birth could be used by an
investigator conducting a follow-up study to determine a subject’s age based on the
current year instead of age at the time of interview. For this reason, date of birth is
also included on the new form. Name and address are not necessarily important for
studying cervicitis, but they are included on the new form because they are vital for
clinical purposes. While designing this form, it was important to consider variables
to suit both the research and clinical goals since it will be used for both purposes.
Other demographic variables that were not included on the original form have
been considered for inclusion on the new form. Marital status is a variable that may
31
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
have important bearing on cervicitis because it may be a proxy for sexual activity or
for number of sexual partners. For example, a woman who is married is likely to
have only one sexual partner whereas unmarried women may have more than one
partner. This may relate to the risk of cervicitis because women with multiple sexual
partners may have a higher risk of acquiring a sexually transmitted disease and thus a
higher risk of developing cervicitis. In addition, as stated previously, a study by
Wilson et al. found that having a marital status other than married was correlated
with the risk of cervical inflammation (Wilson et al. 1990). Therefore, it appears that
there is reasonable evidence that marital status may be associated with cervicitis, so
this variable will be added to the case report form. Specifically, the form will assess
whether a women is single (never married), currently married and living with spouse,
separated, divorced, widowed or unmarried and living with a partner.
Socio-economic status (SES) is a characteristic that is often assessed in
epidemiologic research because it is a proxy for numerous exposures, and can be an
important confounding or modifying factor in many associations. SES was not
assessed on the original form, so it is unclear how much variation in SES there may
be in this population. Since El Salvador is a developing country and the study
population consisted of rural residents, it could be assumed that the subjects were on
the lower end of the SES spectrum and would not be accurately measured by the
conventional SES indicators used in industrialized countries. To control for potential
confounding by SES, information on several variables will be collected and a new
indicator of SES will be developed for this population. The occupation of the study
32
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
participants will be asked as an open-ended question and will be categorized for
analysis. In addition, the occupation of the head of the household will be queried,
which will be useful additional information if most women in the study are of similar
occupation (i.e. homemaker). Years of education will be included as opposed to
asking about level of school completed because the educational system in El
Salvador may differ with that of the U.S. and it may be harder to interpret the
answers. Several questions relating to the home in which the study participant lives
will be asked including: the number of persons living in the home, the number of
rooms in the home, and whether there is access to electricity, potable water, a
telephone and gas for cooking. Income level will not be included on the form
because it is not clear whether income level could be translated into a scale that
would be useful for determining SES. The variables will be combined at a later time
into an index for SES using a method similar to that of Romieu et al. (Romieu et al.
2004).
The original form did not collect any information about anthropometric
variables such as height and weight. These variables are often associated with
disease status. For example, high body mass index (a function of height and weight)
has been associated with breast cancer risk (Hankinson and Hunter 2002). Although
there is no evidence that height or weight is associated with cervicitis, these variables
will be included on the form so that potential associations can be investigated.
33
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Reproductive and Pregnancy History
The original case report form asked several questions about reproductive and
pregnancy history. The basic information that was assessed included: number of
pregnancies (gravita), number of full-term births, number of pre-term births, number
of spontaneous abortions and the number of living children. Although all of these
variables are useful and will be included on the new form, there are some other
variables related to pregnancy history that should also be considered. In the
univariate analyses that were conducted, number of failed pregnancies was
marginally associated with the risk of cervicitis (p trend = 0.079) (Table 3). This
variable was created by summing the number of pre-term births and the number of
spontaneous abortions. However, there are other pregnancy outcomes that were not
queried that may be important in considering the number of failed pregnancies.
Stillbirths and ectopic pregnancies are other potentially important adverse pregnancy
outcomes that will be included on the new form. As mentioned previously, ectopic
pregnancies are an outcome associated with pelvic inflammatory disease and
cervicitis (Sweet 1998). Thus, collecting information on these variables may have
particular importance in studying the risk factors for cervicitis.
History of menstruation may also be important in predicting a woman’s risk
of cervicitis. The original case report form collected only information on the date of
last menstrual period. This variable can be useful in determining whether a woman
is pre- or post-menopausal, so it will be included as a part of the new form. For
example, if the last menstrual period was one or more years ago a woman could be
34
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
considered post-menopausal. Another important variable related to menstruation
history is age at menarche. This variable may be in some ways related to age at first
intercourse and sexual history since girls with an early age of menarche may become
sexually active at a younger age. In order to determine the association between age
at menarche and cervicitis, this variable will also be included on the new form.
History o f Hormone Use
The original case report form collected information on ever use and number
of years of use of hormone injections, hormone pills and other hormones. It also
queried the reason for hormone use, including birth control, hormone replacement
therapy or some other reason. As discussed previously, past studies have found an
association between the use of hormone pills (i.e. birth control pills) and cervicitis
(Paavonen et al. 1986, Wilson et al. 1990, Ghanem et al. 2005). In the pilot study,
univariate associations determined that the risk of cervicitis increased with
increasing months of use of hormone pills, and this trend was statistically significant
(p trend = 0.018) (Table 3). Also, women who used hormone injections for 1 to 18
months had 2.3 times (95% Cl: 1.2-4.6) the risk of cervicitis as compared to women
who never used hormone injections (Table 3). However, the trend of increasing risk
with increasing months of injection use was marginally significant (p trend = 0.051).
In order to more fully assess the associations between use of hormones and
cervicitis, the questions on hormone use will be re-designed on the new form. The
form will ask specifically whether the birth control pill, hormone injections (such as
35
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Depo-Provera), or hormone therapy (for menopause) were ever used. In addition, if
a woman ever used one of those types of hormones, the form will query whether the
woman is a current or a past user, and for how many years or months the hormone
has been used. The form will also query the ages when each type of hormone was
used. If there is an association between cervicitis and a form of hormone use, it is
anticipated that current users should show a stronger association with risk of
cervicitis than past users. The redesigned questions on hormone use should allow
more detailed information to be collected and will hopefully allow for a more
accurate assessment of any potential associations.
Medical History
The medical history information collected on the original form mostly relates
to gynecological medical history. In particular, past history of Pap smears was
queried, including whether a woman had a previous Pap smear, and if so, the result
of the Pap smear and information about any treatment received after a previous Pap
smear. As previous history of Pap smear relates to cervical cancer risk, it makes
questions regarding this topic important for understanding cervicitis. In addition,
previous history of Pap smear may also be important to record because it may be a
surrogate for a woman’s health care-seeking behavior. For example, a woman who
has had previous Pap smears may be more likely to seek care for other medical
reasons, such as sexually transmitted infections, than a woman who has no prior Pap
36
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
smears. Therefore, questions relating to Pap smear history have been retained on the
new form.
Other information relating to gynecological history was also obtained on the
original form. Gynecological surgical history was queried including history of the
following procedures and conditions: Caesarian section, tubal ligation, hysterectomy,
tubal pregnancy, dilatation and curettage, and elective abortion. In the analysis of
these variables, none were found to be significantly associated with cervicitis. In
addition, there is no evidence from prior studies that these conditions are risk factors
for or outcomes of cervicitis. Therefore, this section will be condensed in the new
form to only ask about a general history of gynecological surgery.
History of gynecologic problems was also queried on the original form with
questions about history of abnormal periods and history of sexually transmitted
diseases. History of abnormal periods may be an important variable to collect in
regard to cervicitis, since there could be a connection between the presence of
cervical inflammation and abnormal menstrual periods. As stated previously,
cervicitis was found to be marginally associated with history of abnormal periods (p
= 0.070) (Table 1). In addition, women who reported a history of abnormal periods
had about twice the risk of cervicitis compared to women who did not report
abnormal periods (OR = 1.9 95% Cl: 0.94-3.9) (Table 3). Due to the potential
importance of history of abnormal periods, this variable has been included on the
new form. History of sexually transmitted diseases is of key importance and will be
discussed in further detail in the section on infection history.
37
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Smoking history was also collected as a part of the original form. History of
smoking was assessed as ever or never, and if ever a smoker, the number of years of
smoking and number of cigarettes smoked per day was assessed. Among the women
in the pilot study, only about five percent were smokers. This may be due to a bias
in the sample of women included in the study, who self-selected attendance at the
screening clinic. However, it is also possible that the low prevalence of smoking is
accurate because it is known that smoking is not common among women in El
Salvador (Cremer, personal communication). The small number of smokers in the
study limited the power to detect an association between smoking history and
cervicitis. Although smoking has not been implicated as a risk factor for cervicitis, it
is an important risk factor for other diseases including cervical cancer. Hence, it is
important to include smoking history on the new case report form. The new form
will ask whether subjects are ever or never smokers, and among ever smokers,
whether they are current or former smokers. Among all smokers, the age at start of
smoking, the duration and dose of smoking will be assessed. Among former
smokers, the age at quitting and the number of years since quitting will be asked.
More specific questions will be asked about smoking in order to enhance the
accuracy of analyses to be conducted on smoking in future studies.
In order to gather more information on exposure to smoking, questions will
also be added to the form about smoking in the study participant’s household. The
form will query whether there are any smokers in the household (other than the study
participant), and if so how many smokers. If there is a low prevalence of smoking
38
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
among women in El Salvador, assessing whether there are other smokers in the home
will provide more accurate information on the overall exposure of the study subjects
to tobacco smoke.
Sexual History
Based on the results of the analyses and the review of the literature, it is clear
that sexual history is of key importance in determining the factors that are associated
with cervicitis. The original case report form collected a limited amount of
information on sexual history. Age at first intercourse and number of lifetime sexual
partners were the two main variables used to collect this information. As previously
described, preliminary data analyses found that age at first intercourse and number of
lifetime sexual partners were not significantly associated with the presence of
cervicitis (p = 0.72 and p = 0.77, respectively) (Table 1). However, in the univariate
analysis comparing severe cervicitis with chronic cervicitis, women with four or
more sexual partners were more than three times more likely to have severe cervicitis
as women with one or no lifetime sexual partners (OR = 3.4, 95% Cl: 1.4-8.2). After
adjustment for age, this association remained statistically significant (OR = 3.3, 95%
Cl: 1.3-7.9). Based on the risk factors for cervicitis and the intriguing results from
the data analysis, it appears that the number of lifetime sexual partners may have
bearing on the risk of cervicitis. It was therefore decided both age at first intercourse
and number of lifetime sexual partners will be included on the new form.
39
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
To gather a more complete assessment of sexual history, other variables will
also be added to the new form. A field will be added which will be checked off by
the interviewer if a woman has never had sexual intercourse, so that the group of
women with no history of sexual intercourse can be definitively identified. Number
of sexual partners in the last year will also be queried because it is thought that this
may give a more accurate assessment of current sexual patterns, which may relate to
the risk of cervicitis. Women will also be asked if they are currently sexually active
because this may be associated with the risk of current cervical inflammation.
Including these additional variables on the form should help provide a fuller picture
of the sexual history of the women in the study population.
Current contraception is another variable that may be important in assessing
risk of cervicitis. The original form did not include a question about this topic;
however, current contraception was recorded for most women as an add-on field.
Information was missing on this variable for 35 women in the pilot study. In
addition, since the information was queried as an open-ended question, it was
difficult to accurately categorize all types of contraception. Current contraception
may be a risk factor for cervicitis, particularly in association with contraceptive
methods such as the diaphragm, cervical cap, IUD and spermicide. Due to the
irritation these contraception methods and others may cause to the cervix, method of
contraception may be associated with the presence of cervical inflammation. A more
structured question on current contraception will be included on the new form and
will provide the following list of contraceptive methods: birth control pill, hormone
40
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
injections, condom, implant, tubal ligation, male sterilization, diaphragm, cervical
cap, IUD, spermicide, sponge, rhythm method, withdrawal, abstinence and none. It
is anticipated that providing a list of methods will make categorizing this variable
easier in the analysis. Both past use of contraception and the types of contraception
used in the past will also be included on the form. In addition, ages of use of current
contraception method and ages of use of past contraception method will be added.
These variables will be included in order to make a more complete assessment of the
patterns of contraceptive use. For example, it may be interesting to investigate
whether women use different contraceptive methods at different times in their lives.
Infection History
Collecting information on infection history is of paramount importance in
determining the factors that are predictors of cervicitis. As discussed previously,
infectious agents, such as bacteria and viruses, are the main factors already shown to
be associated with the diagnosis of cervicitis (Sweet 1998). The original case report
form did not collect detailed information on the history of infections. The only
variable included in the original form relating to this topic was history of sexually
transmitted diseases (STD). Only about 3% of the women in the pilot study reported
a history of sexually transmitted diseases, and this variable was not associated with
cervicitis (p = 0.56); (Table 1). The small number of women reporting a past history
of STDs may be a reflection of the obstacles that assessing this variable faces. First,
women who had an STD in the past may have not received medical care for
41
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
diagnosis or treatment, and thus, they would not be able to report it. Second, even if
they were diagnosed with an STD, women may not be able to recall what the agent
was or what treatment was prescribed. Third, women may be embarrassed or
ashamed to reveal that they had an STD since diagnosis of these types of infections
is often associated with some degree of shame or anxiety. Therefore, history of
infections will need to be queried in such a way that information about past STDs
can be gathered indirectly.
To begin, more detailed questions about history of sexually transmitted
diseases will be included on the new form. Women will first be asked about whether
they recall having an STD in the past. If they report a past history of an STD, they
will be asked what STD was diagnosed and the following list of STDs will be
presented: Chlamydia, Gonorrhea, Herpes Simplex Virus, Human Papillomavirus
(HPV), Human Immunodeficiency Virus (HIV) and other. If an STD not on this list
was diagnosed, women will be asked to provide the name of the agent. Other
questions will query the year of diagnosis and the treatment that was provided,
including antibiotic pills, creams or any other treatment. Although few women
reported a history of STDs in the pilot study, it was decided that collecting this
information is of vital importance for understanding cervicitis.
In order to collect information on past sexually transmitted diseases in a way
in which more women may be able to recall, several questions will be added to the
new form about prior history of vaginal infections. The first question will ask if
women had a history of burning, itching, pain or discharge in the vaginal area. If a
42
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
woman reports a past vaginal infection she will also be asked a series of other
questions including, the cause of the infection (if known), whether treatment was
used (if any), and if so, the type of treatment (including antibiotic pills, creams, or
other treatments). The reasoning behind adding questions on history of vaginal
infection to the new form resulted from some intriguing information that was
collected in the original pilot study.
Women were asked on the original form whether they had any “other” type
of treatment after a previous Pap smear, aside from procedures used to treat cervical
dysplasia or carcinoma (including LEEP, cryotherapy, injection and radiation). This
variable was recorded as an open-ended response. The information recorded was
categorized by using the written notes of the interviewer who completed the case
report form. Based on the pilot study data, 116 women (roughly 23%) reported
“other” treatments after a previous Pap smear. Of those who had “other” treatments,
105 women (90.5%) reported having a prior treatment with pills or cream for a
vaginal infection. In the univariate analyses, women who reported other treatment
after Pap smear were 0.61 times as likely to have cervicitis as women who did not
report other treatment (95% Cl: 0.38-0.97) (Table 3). Based on the information
garnered from this variable and the results of the analysis, it was thought that
collecting information on prior history of vaginal infection and prior treatments for
infection may serve as a proxy for history of STDs and would help in determining
factors that are important in predicting cervicitis.
43
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
The final piece of information that will be collected on infection history is
current symptoms of infection. The new form will include questions that ask about
current vaginal infections. If women report a current infection, they will be asked to
provide some symptoms of their infection including any of the following: vaginal
burning, itching, discharge, odor, vaginal pain, pain during sex, bleeding after sex, or
any other symptoms. Information of this type was not collected on the original form,
but may be vitally important in determining which women have an infection. Based
on the etiology of cervicitis, it is anticipated that women who report current
symptoms of infection should be much more likely to have cervicitis than women
who do not report these symptoms.
44
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
CHAPTER FOUR: DISCUSSION
The aim of designing the new case report form was to develop an instrument
that can be used to determine the risk factors for cervicitis among women in rural El
Salvador, and to study the association between the presence of cervicitis and the risk
of cervical cancer. Although the original pilot study was not designed to study
cervicitis, much of the information collected as a part of the study was useful in
determining how to approach this question. The original case report form asked
many questions that are directly relevant to cervicitis, and thus analyses of the data
that was collected provided insights into associations between many variables and
cervicitis in this population. It is anticipated that the new case report form will be
administered to women in El Salvador as a part of future research in the country.
The new form will first be tested as a part of clinical research that is on going
in El Salvador. The investigators who conducted the pilot study have a great interest
in working with this population of women and plan to conduct more research relating
to cervical cancer screening and diagnosis. Future clinical research will be able to
examine more fully the effect of cervicitis on the results of Pap smear, visual
inspection, digital inspection and colposcopy.
Epidemiologic studies will also be developed that will use this form as the
main study instrument. It is anticipated that epidemiologic studies will be able to
more accurately determine the prevalence of cervicitis in El Salvador than clinical
studies, because study participants will be enrolled who are more representative of
the general population of women in the country. A cross-sectional study would be
45
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
an ideal type of study design for investigating the prevalence of cervicitis in El
Salvador. In this type of study, a defined group of women would be identified and
those who choose to participate would be tested for the presence of cervicitis (the
outcome) and interviewed using the case report form to determine important risk
factors. In order to implement a cross-sectional study, sample size and power
calculations will be conducted to determine how many women are needed to detect
statistically significant associations between cervicitis and potential risk factors.
Designing this new case report form has provided a first step in planning
future research in El Salvador. The form has several strengths that will make it a
useful part of the research to be conducted. First, the form has been designed to
study cervicitis as a main outcome. This will make the data taken from the form
much more useful in determining risk factors for cervicitis than the original form,
since that form was not designed with cervicitis in mind. Second, information
collected on current sexually transmitted infections and current vaginal infections
may provide key information on the causes of cervicitis. In analyzing the
associations between cervicitis and other lifestyle variables, adjustment can be made
for the presence of current infections. This will aid in controlling for confounding by
current infections in studying the associations between cervicitis and other potential
risk factors. Third, the new form has been translated into Spanish by a Salvadorian
general medicine physician, which will make administering it to a Spanish-speaking
population much more efficient.
46
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
There are, however, several limitations that must be considered in using this
case report form in future studies. First of all, the form has been designed so that it
can quickly and easily be administered in a clinical setting. Although this is essential
for feasibility, it restricts the amount of detailed information that can be collected
about each variable, which may somewhat limit the accuracy of the associations that
are measured. Also, the form has been designed to be administered by persons with
a strong medical background (such a physicians or nurses). People not familiar with
medical terminology may find it difficult to administer, which may limit its
usefulness. Finally, this case report form has been designed specifically for use in
El Salvador. Additional questions or modifications may be required if it is used in
other populations. For example, race and ethnicity were not collected on this form
since women in El Salvador are largely all of Hispanic ethnicity. If this form were
used in the United States or in numerous other parts of the world, race and ethnicity
may be important variables to collect. This is because certain races may be more or
less likely to have cervicitis, which may lead to confounding in the associations
between certain exposure variables and cervicitis.
Future studies of cervicitis must also consider including other types of
exposure and outcome measurements besides those included in this case report form
to assess the risk factors for cervicitis. Conducting sexually transmitted disease
testing is one method that can be used in future studies to determine which factors
are important causative agents for cervicitis in El Salvador. As described previously,
Chlamydia trachomatis, Neisseria gonorrhoeae and Herpes Simplex virus II are all
47
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
known to be important causes of cervicitis. In addition, Human papillomavirus is the
key risk factor for cervical cancer. Therefore, it would be vital for future studies to
collect biological samples from study participants that can be tested for these
infections. With the results of STD testing on the study participants, data analyses
can be conducted that control for the presence of these infections in determining the
factors associated with cervicitis. This would allow better control of confounding by
STDs than could be accomplished by using only the data collected from the case
report form. It would also allow for an investigation of the associations between
reported symptoms and treatment and more objective biomarker measures of STDs.
Along with testing, future studies on cervicitis should also collect clinical
information on the presence of vaginal infections. Although the case report form
asks women if they have a current vaginal infection, it is also necessary for clinicians
who examine women to record whether a symptomatic infection is apparent. This
will provide another key piece of information about infection, aside from the
information that is determined from the lab test results. Since every type of
infectious agent that causes cervicitis cannot feasibly be tested, the report of
infection from clinicians will provide very important additional information.
Another element that may be added to future studies would be to ask study
participants whether they believe that they have some type of cervical condition
(such as cervical inflammation, cervical cancer or other conditions). It would be
interesting to investigate the association between subjects’ responses to this question
and the diagnosis of cervical cancer or cervicitis. If there is an association between
48
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
what a woman believes her health condition to be and her actual diagnosis, this could
be useful for targeting behavioral interventions, such as those that would encourage
women to seek cervical cancer screening.
Finally, deciding the manner in which study participants should be
categorized for cervicitis is of vital importance in determining valid associations
between cervicitis and other variables. As mentioned previously, cervicitis can be
diagnosed by cervical cytology, biopsy and colposcopy. In the pilot study, cervicitis
was assessed based on the results of the cervical biopsy that was taken during
colposcopy. The biopsy tissue was viewed under the microscope by a pathologist
who recorded comments about the condition of the tissue. These comments were
later used to categorize women for cervicitis. As stated previously, 74% of the
women with a biopsy result had either chronic or severe cervicitis.
Cervical inflammation was also assessed separately based on cytological
examination of the Pap smear results. Interestingly, only 35% (175 women) had
cervical inflammation based on the cytology results. Although common sense
dictates that cervicitis and inflammation measure the same biological process,
cervicitis and inflammation were not found to be associated in this study population.
About 76% of women who were identified by Pap smear as having inflammation
were found to have cervicitis based on biopsy. Conversely, 72% of women who
showed no evidence of inflammation based on Pap smear were found to have
cervicitis.
49
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Based on these results, it appears that biopsy and cytology do not measure
cervical inflammation in the same way. Biologically, there may be reasoning behind
this difference. During biopsy, samples of tissue are extracted directly from the
cervix; during a Pap smear, a brush is used to scrape cells from the surface of the
cervix. It is likely that some cells taken during a Pap smear may be vaginal cells.
Thus the cytology results may be more likely to reflect vaginal inflammation, while
biopsy results are likely to be displaying inflammation of the cervix. Based on these
differences, it has been decided that using biopsy tissue is a more accurate way to
measure cervicitis than using Pap smear results. In future studies, colposcopies will
be conducted on all study participants so that the results of biopsy may be used to
assess cervicitis.
In conclusion, the results of this pilot study revealed that cervicitis is a very
common condition among women in this rural El Salvadorian population. Age was
an important indicator of cervicitis, but many other variables including reproductive
history, sexual history and infection history may also play a significant role. In
addition, the results showed that the presence of cervicitis might modify the accuracy
of visual inspection, digital inspection and colposcopy.
Overall, the results of this study highlight the reasons why designing this new
case report form and studying cervicitis are important. First, cervicitis is a common
condition among women, but one that may lead to significant consequences. These
consequences, including pelvic inflammatory disease, adverse pregnancy outcomes
and cervical cancer, are major health concerns for women worldwide.
50
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Understanding the risk factors associated with cervicitis is vital for preventing and
treating this condition among the women affected by it. Second, it is anticipated that
this research on cervicitis will provide a stepping-stone for future research on the
associations between cervical inflammation and cervical intraepithelial neoplasia and
invasive cervical cancer. In order to investigate these associations among women in
Latin America, it is first important to be able to identify the women who are at the
highest risk of developing cervicitis and to determine why they are at such a high
risk.
Having this information will be vital for the development of future
epidemiologic studies on cervicitis and cervical cancer. In fact, recent epidemiologic
research has been involved with the development of biomarkers, such as the family
of interleukins, which could be used to measure levels of cervical inflammation.
Development of these biomarkers is a promising step toward being able to accurately
assess the associations between inflammation and cancer. Although there is much
that remains unknown about the process of cervical carcinogenesis, it is hoped that
this preliminary research may help to further understand these complex processes
and eventually lead to a reduction in the morbidity and mortality associated with this
disease.
51
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
REFERENCES
Arrossi S, Sankaranarayanan R, Parkin DM. 2003. Incidence and mortality of
cervical cancer in Latin America. Salud Publica Mexico 45(Suppl 3):
S306-14.
Blumenthal P, Lauterbach M, Sellors JW, Sankaranarayanan R. 2005. Training for
cervical cancer prevention programs in low-resource settings: Focus on visual
inspection with acetic acid and cryotherapy. Int J Gynaecol Obstet 89
(Suppl 2):S30-7.
Brunham RC, Paavonen J, Stevens CE. 1984. Mucopurulent cervicitis— the ignored
counterpart in women of urethritis in men. N Engl J Med 311(1): 1-6.
Castle PE, Giuliano AR. 2003. Chapter 4: Genital tract infections, cervical
inflammation, and antioxidant nutrients— assessing their roles as human
papillomavirus cofactors. JN atl Cancer Inst Monogr (31): 29-34.
Castle PE, Hillier SL, Rabe LK, Hildesheim A, Herrero R, Bratti MC, Sherman ME,
Burk RD, Rodriguez AC, Alfaro M, et al. 2001. An association of cervical
inflammation with high-grade cervical neoplasia in women infected with
oncogenic human papillomavirus (HPV). Cancer Epidemiol Biomarkers
Prev 10(10): 1021-27.
Claeys P, Gonzalez C, Gonzalez M, Van Renterghem L, Temmerman M. 2002.
Prevalence and risk factors of sexually transmitted infections and cervical
neoplasia in women’s health clinics in Nicaragua. Sex Transm Infect 78(3):
204-07.
Cremer M, Jamshidi R, Muderspach L, Tsao-Wei D, Groshen S, Felix J, Blumenthal
P. Digital camera assessment for detection of cervical intraepithelial
neoplasia in rural El Salvador. In preparation.
Critchlow CW, Wolner-Hanssen P, Eschenbach DA, Kiviat NB, Koutsky LA,
Stevens CE, Holmes KK. 1995. Determinants of cervical ectopia and of
cervicitis: age, oral contraception, specific cervical infection, smoking and
douching. Am J Obstet Gynecol 173(2): 534-43.
Franco EL, Schlecht NF, Saslow D. 2003. The epidemiology of cervical cancer.
Cancer J 9(5): 348-59.
52
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Ghanem KG, Shah N, Klein RS, Mayer KH, Sobel JD, Warren DL, Jamieson DJ,
Duerr AC, Rompalo AM, HIVERS Group. 2005. Influence of sex hormones,
HIV status, and concomitant sexually transmitted infection on cervicovaginal
inflammation. J Infect Dis 191(3): 358-66.
Giuliano AR, Papenfuss M, De Galaz EM, Feng J, Abrahamsen M, Denman C,
Guemesey De Zapien J, Navarro Henze J et al. 2004. Risk factors for
squamous intraepithelial lesions (SIL) of the cervix among women residing at
the U.S.-Mexico border. In tJ Cancer 109(1): 112-8.
Hankinson S, Hunter D. 2002. Breast Cancer. In: Adami HO, Hunter D,
Trichopoulos, D, editors. Textbook of Cancer Epidemiology. Eds. New York:
Oxford University Press p 301-39.
Hawes SE, Kiviat NB. 2002. Are genital infections and inflammation cofactors in the
pathogenesis of invasive cervical cancer? JN atl Cancer Ins tit 94(21):
1592-3.
Keith LG, Creinin M, Method MW, Hidvegi D. 1989 Cervicitis as an antecedent to
pelvic infection: a review. Int JFertil 34(2): 109-19.
Keshavarz H, Duffy SW, Sadeghi-Hassanabadi A, Zolghadr Z, Oboodi B. 2001. Risk
factors for and relationship between bacterial vaginosis and cervicitis in a
high risk population for cervicitis in Southern Iran. Eur J Epidemiol 17(1):
89-95.
La Montagne DS, Patrick LE, Fine DN, Marrazzo JM, Behalf of the Region XIPP.
2004. Re-evaluating selective screening criteria for chlamydial infection
among women in the U.S. Pacific Northwest. Sex Transm Dis 31(5): 283-9.
Lawley TB, Lee RB, Kapela R. 1990. The significance of moderate and severe
inflammation on class I Papanicolaou smear. Obstet Gynecol 76: 997-9.
Marrazzo JM, Handsfield HH, Whittington WL. 2002. Predicting chlamydial and
gonococcal cervical infection: implications for management of cervicitis.
Obstet Gynecol 100(3): 579-84.
Molano M, Weiderpass E, Posso H, Morre SA, Ronderos M, Franceshi S, Arslan A,
Meijer CJ, Munoz N, van den Brule AJ, et al. 2003 Prevalence and
determinants Chlamydia trachomatis infections in women from Bogota,
Colombia. Sex Transm Infect 79(6): 474-8.
O’Byrne KJ, Dalgleish AG. 2001. Chronic immune activation and inflammation as
the cause of malignancy. Br J Cancer 85(4): 473-83.
53
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Paavonen J, Critchlow CW, DeRouen T, Stevens CE, Kiviat N, Brunham, RC,
Stamm WE, Kou CC, Hyde KE, Corey L et al. 1986. Etiology of cervical
inflammation. Am J Obstet Gynecol 154(3): 556-64.
Romieu I, Lazcano-Ponce, E, Sanchez-Zamorano LM, Willett W, Hemandez-Avila
M. 2004. Carbohydrates and the risk of breast cancer among Mexican
women. Cancer Epidemiol Biomarkers Prev 13(8): 1283-9.
Schwebke JR, Zajackowski ME. 1997. Effect of concurrent lower genital tract
infections on cervical cancer screening. Genitourin Med 73(5): 383-6.
Sweet RL. 1998. The enigmatic cervix. Dermatol Clin 16(4): 739-45.
Wilson JD, Robinson AJ, Kinghom SA, Hicks DA. 1990. Implications of
inflammatory changes on cervical cytology. BMJ 300(6725): 638-40.
54
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
APPENDIX
Formulario para Estudio de Caso Clinico
Numero de estudio:
Fecha de la entrevista: /
Iniciales del entrevistador:
Datos Demograficos
Nom bre:______________________________________
Dom icilio:________________________________________________________________
Fecha de nacimiento:___ /____ /____ E dad:_______ anos
Estatura: mt. o _____p ies._____pulg. P eso:______ kg. o ____ lbs.
Estado Civil: Soltera(l)___ Casada(2)___ Separada(3)___ Divorciada (4 )___ Viuda
(5) Acompanada(6)___
Anos de estudio:____anos
Ocupacion:______________________ Ocupacion de cabeza de fam ilia:_____________________
Numero de personas en el hogar:____ Numero de habitaciones:____
Cuenta con servicios de: Electricidad:_____ Agua potable:____ Telefono:______
G as:____ Cocina con lena:____
Historia Clinica Obstetrica
G ravidez:___ Term ino:___ Prematuro:____________ Aborto espontaneo:_____
Nacidos vivos:____
Embarazo ectopico:____ Ninos vivos:____
Edad de la menarquia:______anos Fecha de ultima regia: / /
Historia de uso de hormonas:
o Antecedentes de uso de anticonceptivos orales: Si (1 )____ No (2 ) No sabe (0 )____
Si la respuesta es afirmativa: Esta usando: (1 )____ Uso en el pasado_(2 )____
Edad en que empezo: Edad en que termino:____
Duration: anos meses
55
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Formulario para Estudio de Caso Clinico (continuado)
o Antecedentes de uso de hormonas inyectadas: Si (1 )___ No (2 )___ No sabe (0 )__
Si la respuesta es afirmativa: Esta usando actualmente (1 )___ Uso en el pasado (2 )___
Edad en que empezo: Edad en que termino:____
Duracion: anos m eses____
o Antecedentes de uso de hormonoterapia (para la menopausia):
SI (1 )___ No (2 )___ No sabe (0 )_ _
Si la repuesta es afirmativa: Usa actualmente (1 ) _Uso en el pasado (2 )______
Edad en que empezo:_____Edad en que termino:____
Duracion: anos m eses____
Historia Medica
Exudado de PAP (papanicolau): Si (1 )___ No (2 )__ No sabe (0 )___
Fecha del ultimo PAP: / /
Resultado del ultimo PAP: Positiva (1 )_ _ N egativa(2)____ No sabe (0 )___
Tratamiento recibido tras la ultima prueba de PAP:
N ada(0) L E E P (l) Crioterapia (2 )___ Inyeccion(3) Radiacion(4)___
Otro (5) (Especificar)__________
Antecedentes de cirugia ginecologica: Si (1 )___ No (2 )___ No sabe (0 )___
Antecedentes de ciclo menstrual anormal: Si (1 )____ No (2 )____ No sabe (0 )___
Historia de tabaquismo: Actual (1 )___ Pasado (2 )____ Nunca(O)___
Si es tabaquista: Edad de inicio:______anos
Total de anos fum ando:______anos
Numero de carretillas al d ia:_____
Si ha sido fumadora anteriormente:
Edad en que lo d ejo:______anos
Numero de anos transcurridos desde que dejo de fum ar:______anos
Fumadores en el hogar (sin incluirse): Si (1 )____No (2 )_____ No sabe (0 )____
Se respondio que si, cuantos (sin incluirse):_____
Historia Sexual
Edad de la primera relacion sexual:____anos
Si nunca ha tenido relaciones sexuales, marque aqui:___
56
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
Formulario para Estudio de Caso Clinico (continuado)
Numero de parejas sexuales durante la vida:___
Numero de parejas sexuales en el ultimo an o:___
Se encuentra sexualmente activa: Si (1 )___ No _(2 )____ No sabe (0 )___
Si respondio que si, usa anticonceptivos: Ninguna (0 )____ Anticonceptivos orales (1 )_
Hormonales inyectables (2 ) Condon (3 )_Implante (4 )___ Ligadura de trompas (5)
Esterilizacion masculina (6) ____ Diafragma (7 )____ Capuchon cervical (8 ) DIU (9) _
Espermicida (1 0 )____ Esponj a (1 1 ) _ Metodo del ritmo (12)___
Eyaculacion externa (1 3 )____ Abstinencia (14)__
Edad en que comenzo el metodo en curso:____ Hasta:____
£Uso anteriormente metodos de anticonceptivos?: Si (1 )___ No (2 )___ No sabe (0 )_ _
Si respondio que si, especifique el metodo utilizado (de la lista anterior):___________,
Edad en que empezo y suspendio metodo anterior:_____ Hasta:___
Antecedentes de Infeccion
Ha tenido infecciones vaginales (incluyendo ardor, picazon, dolor o flujo):
Si (1 )____ No (2 )___ No sabe (0 )___
Si respondio que si: ^Cual fue la causa?:
Infeccion de Transmision Sexual (1 )____ Otro (2 ) No sabe (0 )______
Tuvo tratamiento: Si (1 )___ No (2 )___ No sabe (0 )___
iQ ue tipo de tratamiento?: Antibiotico oral (1 )___ Crema vaginal(2)___
Ninguno (3 ) Otro (4 ) No sabe(0)___
Ha padecido infecciones transmitidas sexualmente: Si (1 )___ No (2 )___ No sabe (0 )__
Si respondio que si: ^Cual fue el diagnostico?: Clamidia (1 ) Gonorrea (2 )___
Virus del Herpes Simple (3 )___
HPV (4 )____ VIH (5 )___
Otro (6) _______________ No sabe (0) _
En que ano(s) fue(ron) diagnosticado(s):_________ , __________ , __________
Que tratamiento tuvo: Antibiotico oral (1 )_ __ Crema vaginal (2 )___ Ninguno (3)
Otro(4)_____________
^Tiene ahora sintomas de infeccion?: Si (1 )___ No (2 ) No sabe (0 )___
Si respondio que si: ^Cuales son los sintomas que siente?: Ardor (1 )____
Picazon (2 )___ Flujo (3 ) Mai olor (4 )____ Dolor vaginal (5 )___
Dolor durante el coito (6 ) Sangrado posterior a la relation sexual (7 )__
Otro (8 )_______________
57
R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.

Asset Metadata

Creator Davis, Carol Ann (author)

Core Title Developing a case report form for assessing the risk factors for cervicitis among women in rural El Salvador

Contributor Digitized by ProQuest (provenance)

School Graduate School

Degree Master of Science

Degree Program Applied Biostatistics

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

Tag health sciences, obstetrics and gynecology,health sciences, public health,OAI-PMH Harvest,sociology, ethnic and racial studies,women's studies

Language English

Advisor Cortessis, Victoria (committee chair), Felix, Juan (committee member), Groshen, Susan (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-45444

Unique identifier UC11337854

Identifier 1430387.pdf (filename),usctheses-c16-45444 (legacy record id)

Legacy Identifier 1430387.pdf

Dmrecord 45444

Document Type Thesis

Rights Davis, Carol Ann

Type texts

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

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

Repository Name University of Southern California Digital Library

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