Direct Medical Cost of Pelvic Inflammatory
Disease and Its Sequelae: Decreasing, but Still
DAVID B. REIN, MPA, WILLIAM J. KASSLER, MD, MPH,
KATHLEEN L. IRWIN, MD, MPH, AND LARA RABIEE
Objective: To estimate direct medical costs and average
lifetime cost per case of pelvic inflammatory disease (PID).
Methods: We estimated the direct medical expenditures
for PID and its three major sequelae (chronic pelvic pain,
ectopic pregnancy, and infertility) and determined the aver-
age lifetime cost of a case of PID and its sequelae. We
analyzed 3 years of claims data of privately insured individ-
uals to determine costs, and 3 years of national survey data
to determine number of cases of PID, chronic pelvic pain,
and ectopic pregnancy. We developed a probability model to
determine the average lifetime cost of a case of PID.
Results: Direct medical expenditures for PID and its
sequelae were estimated at $1.88 billion in 1998: $1.06 billion
for PID, $166 million for chronic pelvic pain, $295 million for
ectopic pregnancy, and $360 million for infertility associated
with PID. The expected lifetime cost of a case of PID was
$1167 in 1998 dollars. The majority of those costs ($843 per
case) represent care for acute PID rather than diagnosis and
treatment of sequelae. Approximately 73% of cases will not
accrue costs beyond the treatment of acute PID.
Conclusion: The direct medical cost of PID is still substan-
tial. The majority of PID related costs are incurred in the
treatment of acute PID. Because most PID-related costs arise
in the first year from treatment of acute PID infection,
strategies that prevent PID are likely to be cost-effective
within a single year. (Obstet Gynecol 2000;95:397–402.
© 2000 by The American College of Obstetricians and
Pelvic inflammatory disease (PID) afflicts as many as 1.5
million women in the United States annually. Although
most common in young women, its consequences,
including chronic pelvic pain, ectopic pregnancy, and
infertility, can affect a woman’s health and well-being
throughout her lifetime.1The direct medical costs of
PID and two of its adverse outcomes, ectopic pregnancy
and infertility, were estimated in 1990 at approximately
$2.7 billion.2That analysis predicted that direct and
indirect costs of PID would grow to as much as $10
billion by the year 2000. The Institute of Medicine
calculated the 1994 direct costs of PID as $3.1 billion.3
Over the last decade, changes in diagnosis and treat-
ment of sexually transmitted diseases (STDs) and in the
healthcare system influenced national estimates of the
cost of PID.4Several factors might have led to a de-
crease in overall PID incidence, such as availability of
more sensitive and specific nucleic acid amplification
tests that hasten diagnosis, availability of directly ob-
served, single-dose therapies that increase adherence to
treatment for lower genital tract infections, and imple-
mentation of a national, federally funded chlamydia
screening program to detect and treat asymptomatically
infected women.5,6Significant changes in the organiza-
tion, structure, and financing of health care have re-
sulted in greater use of less expensive outpatient man-
agement of PID, and in shorter stays for hospitalized
women.7To guide clinicians and decision makers who
diagnose and treat PID or allocate healthcare resources,
we estimated the direct medical cost of PID and its three
major sequelae in 1998 dollars, using the most recent
From the United States Centers for Disease Control and Prevention,
National Center for HIV, Sexually Transmitted Diseases, and Tubercu-
losis Prevention, Division of STD Prevention, Health Services Research
and Evaluation Branch, Atlanta, Georgia; the Joint PhD Program in
Public Policy, Georgia State University/Georgia Institute of Technology,
Atlanta, Georgia; and the New Hampshire Department of Health and
Human Services, Concord, New Hampshire.
The authors thank Anne Haddix, PhD, and Guoyu Tao, PhD, for their
Supported in part by an appointment to the Research Participation
Program at the Centers for Disease Control and Prevention (CDC),
National Center for HIV, Sexually Transmitted Diseases, and Tubercu-
losis Prevention, Division of Sexually Transmitted Diseases Prevention,
administered by the Oak Ridge Institute for Science and Education
through an interagency agreement between the US Department of
Energy and the CDC.
VOL. 95, NO. 3, MARCH 20000029-7844/00/$20.00
We estimated the number of episodes of PID and its
sequelae in the United States and the direct medical cost
per episode of illness (unit costs), then multiplied the
unit costs by the number of episodes. Once direct cost
analysis was completed, we created a probability-based
model, based on decision analysis techniques, to esti-
mate an average lifetime cost per case of PID.
To estimate the number of visits for PID and its three
sequelae in the United States, we estimated the number
of hospital inpatient visits for 1993, 1994, and 1995 from
the National Hospital Discharge Survey.8Estimates
represented the annual average over the 3-year period.
Cases for each disease were selected if a disease-related
International Classification of Diseases, 9th Revision,
Clinical Modification (ICD-9-CM) code appeared as one
of seven discharge diagnoses. We established case def-
initions for PID, chronic pelvic pain, and ectopic preg-
nancy on the basis of ICD-9 diagnosis codes,9which
defined a case of PID as acute, chronic, unspecified
salpingitis (614.0–614.2); acute, chronic, or unspecified
parametritis (614.3–614.4); acute, unspecified, or other
pelvic peritonitis (614.5, 614.7); other specified or un-
specified inflammatory disease of female pelvic organs
and tissues (614.8–614.9); acute, chronic and unspeci-
fied disease of the uterus (615.0–615.1, 615.9); acute or
chronic gonococcal infection of upper genitourinary
tract, gonococcal endometritis, or salpingitus, or other
upper genitourinary tract gonococcal infection (098.10,
098.16–098.17, 098.19, 098.30, 098.36, 098.37, 098.39);
and gonococcal peritonitis (098.86). Chronic pelvic pain
was defined as any unspecified symptom associated
with the female genital organs (625.9). (Dr. Robert Rolfs,
State of Utah Department of Public Health. Dr. Rolfs
observed that a large percentage of cases of chronic
pelvic pain from a Seattle cohort study that were coded
625.9 represented a past case of acute PID with persis-
tent pain.) Ectopic pregnancy was defined as abdominal
pregnancy (633.0), tubal pregnancy (633.1), ovarian
pregnancy (633.2), and other or unspecified ectopic
pregnancy (633.8, 633.9).10–12
Outpatient cases of PID and its three sequelae diag-
nosed among privately insured women were estimated
from three surveys of outpatient services: the National
Ambulatory Medical Care Survey, the National Hospi-
tal Ambulatory Medical Care Survey—Outpatient De-
partment File, and the National Hospital Ambulatory
Medical Care Survey—Emergency Department
File.13–15Case estimates from the three files were aggre-
gated by disease to calculate one estimate of outpatient
cases. Estimates represented the annual average from 2
years of data, 1994 and 1995. Variances of outpatient
estimates were calculated with special variance files
from the National Center for Health Statistics.16–18
We assumed that 20% of chronic pelvic pain visits
and 50% of ectopic pregnancies were related to PID.19
The estimated number of outpatient cases of PID,
chronic pelvic pain, and ectopic pregnancy from the
two surveys used varied widely as a result of few
survey observations for those diagnoses. It also was
possible that the same case of PID could be counted in
more than one outpatient survey. Therefore, we vali-
dated our estimate of outpatient visits by calculating the
ratio of outpatient to inpatient cases in MarketScan data
(The Medstat Group, Ann Arbor, MI) for 1993–1995. We
applied that ratio to the National Hospital Discharge
survey estimate of inpatient PID cases. The estimate of
outpatient PID cases from the three surveys was within
10% of the MarketScan–National Hospital Discharge
We estimated the cases of outpatient PID diagnosed
at publicly financed sexually transmitted disease (STD)
clinics by first calculating the proportion of PID cases in
the Seattle, Washington, publically funded outpatient
STD clinics to the total cases of all STDs in those clinics
(Whittington W. Unpublished records from Seattle,
Washington, STD clinic system, 1998). We applied that
proportion to the estimated total number of cases at all
STD clinics estimated in a 1995 survey of STD clinics
that was weighted to the national population.20
Our case estimates of treated infertility were devised
by applying a range of probabilities of infertility related
to PID (11–30%) and infertility treatment related to PID
(25%) in the literature to the total number of PID cases
in this study.19To calculate current PID-related cases of
infertility, we assumed that numbers of PID cases in the
past (that would result in cases of infertility today) were
at least as high as those presented in this study.
Costs per episode of PID were based on actual patient
and insurance payments (as distinguished from physi-
cian charges) using the MarketScan data base, a com-
mercial data base that provided payment data for 1993,
1994, and 1995.21The MarketScan data base represents
approximately 4 million privately insured persons per
year over 3 years, approximately 12 million individual
years of health coverage.
To estimate the unit costs of PID, we added all
PID-related claims for a given episode of disease, de-
fined as all continuous claims for a case. New PID
claims filed more than 60 days after a previous claim
were defined as new episodes. Inpatient drug costs
were available in inpatient claims records. Outpatient
pharmacy costs were derived from a subset of women
for whom information about prescription drug cover-
age was available. Average outpatient drug cost was
applied to all cases of outpatient PID.22MarketScan did
398Rein et al
Pelvic Inflammatory DiseaseObstetrics & Gynecology
not provide estimates of the cost of treating a case of
PID at a publicly funded STD clinic, so we assumed that
cost was the same as treating PID in a private outpatient
To estimate the unit costs of chronic pelvic pain, we
summed patients’ outpatient claims for the year and
calculated the average yearly cost per patient. For
inpatient cases of chronic pelvic pain, we summed
patients’ claims by inpatient admission and found the
average cost per hospital admission. We included inpa-
tient pharmacy costs and costs of drugs supplied intra-
venously or by direct observation, but not outpatient
prescription and nonprescription pharmacy costs.
Unit costs for ectopic pregnancy were calculated in
the same manner. Outpatient costs were estimated on a
yearly basis, and inpatient costs were estimated per
hospital admission. Inpatient pharmacy costs, the cost
of intravenous and directly observed medications, were
included in cost estimates, but not outpatient prescrip-
tion or nonprescription pharmacy costs.
We calculated expected lifetime costs by adding costs
of adverse sequelae to costs of initial treatments. The
expected costs of sequelae were calculated by multiply-
ing their probability and cost (present value calcula-
tions were performed using a 5% discount rate). We
added those costs to the cost of treating a case of PID.23
Probabilities of each sequela were drawn from the
literature19and multiplied by the cost of each sequela to
determine which stage of disease accounted for most
PID-related costs. We assumed that half of ectopic
pregnancies were PID-related.23,24We further assumed
that ectopic pregnancies that resulted from PID did not
differ clinically or in cost from those of other etiologies.
We assumed that 20% of chronic pelvic pain was related
to previous cases of PID, that infertility developed in
11–30% of women who suffered from PID, and that of
those women, only 25% received infertility treatment.19
Treatment was stratified as inpatient or outpatient
with the use of percentages of inpatient and outpatient
cases in the national hospital and outpatient survey
data. The unit costs of chronic pelvic pain, ectopic
pregnancy, and infertility were multiplied by their
probabilities. We assumed that chronic pelvic pain
occurred on average 2 years after an episode of PID;
ectopic pregnancy 5 years after; and infertility 10 years
after.19The cost of infertility treatment was based on the
1990 estimate and adjusted to 1998 dollars, using the
consumer price index.2
We conducted a sensitivity analysis by evaluating all
unit cost values over a range of $1 to twice the unit costs
in this study. Probabilities of sequelae and inpatient or
outpatient care were evaluated as a range from 0–1. We
calculated probabilities of the most common outcomes
of a case of PID from that model to determine if acute
care of PID or care for PID sequelae accounted for the
largest proportion of costs.
We estimated approximately 1.2 million visits for PID,
1.5 million for chronic pelvic pain, and 290,000 for
ectopic pregnancy in 1995. A total of 300,000 visits for
chronic pelvic pain, 145,000 for ectopic pregnancy, and
as many as 78,000 cases of treated infertility were
estimated as related to cases of PID. Thus, we estimated
1.76 million inpatient, outpatient, and STD clinic visits
for acute PID and its three main sequelae annually from
1993–1995 (Table 1).
Total direct medical costs for PID and sequelae, with
the use of a high estimation of probability of infertility
(30%), was $1.88 billion in 1998, and with the use of a
low estimation of probability of infertility (11%), was
$1.62 billion. Of those costs, PID accounted for $1.07
billion, chronic pelvic pain for $166 million, ectopic
pregnancy for $295 million, and infertility from $120 to
$360 million (Table 1).
Cases of PID treated as inpatients were more costly.
The average lifetime cost for a case of PID treated
initially as an inpatient was $4673, compared with $777
for cases treated entirely as outpatient. The average
lifetime dollar cost per case of PID was estimated at
$1167 with the use of a midrange estimate of infertility
equal to 20% (the explicit derivation of the average cost
per case of PID, including a copy of the probabilistic
model and a table displaying the full results from the
sensitivity analysis are available from the authors upon
request). According to our model, 67.2% of PID cases
generated costs equal to the mean cost of treating PID
with no sequelae on an outpatient basis ($451). Of those
patients, 10.5% will suffer from infertility that they
choose not to treat. A total of 81.1% of cases of PID
generated costs less than the average cost of all cases of
PID ($1167). Applying our average cost per case esti-
mate of $1167 to our estimate of 1.2 million visits, we
estimated that visits for PID in 1998 would generate
$1.44 billion in current and future direct medical costs.
Sensitivity analysis showed that expected cost of a
case of PID was influenced most by probability that a
case of PID would be treated as an inpatient and the
probability of subsequent ectopic pregnancy. Our aver-
age cost estimate was not sensitive to wide variation in
the unit cost per case of an inpatient or outpatient case
We estimated that the direct medical expenditures for
PID and its three major sequelae in the United States
VOL. 95, NO. 3, MARCH 2000
Rein et al
Pelvic Inflammatory Disease
were $1.88 billion per year in 1998 dollars. This estimate
is less than the $3.12 billion of direct costs presented in
the most recent review of PID costs for 19943and is
substantially lower than the $10 billion of direct and
indirect costs projected for the year 2000.2That decrease
is due largely to fewer cases of PID and a shift from
inpatient to less costly outpatient management of PID.
The decline in PID incidence coincides with and might
be attributable to an expansion of national and regional
programs that screen women for asymptomatic chla-
mydial infections, one common cause of PID.6The
lower cost estimate also might be explained by new
methods and data sources including calculating PID
and sequelae case estimates from national survey data
that were validated with an independent database;
calculating costs on actual payments rather than physi-
cians’ charges, chart reviews, or expert opinions,
sources that tend to inflate cost estimates; and calculat-
ing the average cost model based on inpatient-to-
outpatient case ratios in the survey and MarketScan
data rather than on expert opinion. The latter method-
ologic approach was critical because the sensitivity
analysis illustrated that the expected cost model was
highly dependent on the proportion of PID cases
treated on an inpatient basis.
This study was limited by at least five factors that
tend to underestimate the costs of PID and its sequelae.
Therefore, our estimate was likely to be a minimum
estimate of PID-related costs. The direct medical cost
estimate relies on the validity and representativeness of
the MarketScan data set. The MarketScan data set is not
designed as a statistically representative source of per-
sons diagnosed with PID (as a probabilistic survey
would be), but it is an unusually large and regionally
distributed sample. This data set is likely to generate a
closer estimate of unit costs than data previously used
to estimate PID and sequelae costs.
Second, one unpublished study from Seattle that
compared insurance claims for 60 PID patients to their
chart records indicated that some PID costs—such as
the cost of diagnostic services before a PID diagnosis,
PID hospitalizations not recorded as such due to coding
error, and laboratory and diagnostic tests not marked
with a PID diagnosis code—might be missed when
costs are estimated from claims data only (personal
communication with Dr. David Magid, Seattle Health
Net). If missed costs are generalizable to cases of PID
nationwide, they could increase our cost per case esti-
mate by as much as 30%.
Third, we used a conservative estimate of the cost of
infertility diagnosis and treatment ($4625). Specifically,
we matched older estimates of infertility costs with
older probabilities of treatment rather than matching
new estimates of cost to previous estimates of demand.
We chose this method because no new estimates of
numbers of persons seeking treatment for infertility
Table 1. Number of Yearly Visits and Cost for PID and its Three Major Sequelae
Condition and type of treatment
Number of clinical
visits (? SE)
(? SD) ($)*
(in million $)*
Chronic pelvic pain
Using a low estimate of infertility
Using a high estimate of infertility
SE ? standard error; SD ? standard deviation; PID ? pelvic inflammatory disease; NA ? not applicable.
* In 1998 United States dollars.
†Infertility estimates draw from the literature to assume that between 11% (low estimate) and 30% (high estimate) of women with PID become
infertile. Of these women, only 25% seek infertility treatment.
‡These numbers reflect average lifetime cost per case of PID including sequelae, calculated by multiplying the probability of each disease event
and treatment by the cost of that event and treatment.
400 Rein et al
Pelvic Inflammatory DiseaseObstetrics & Gynecology
related to PID were available. According to more recent
studies, the current (1998) cost of infertility can range
from two to 20 times the estimate we used in this
analysis.25However, the increases in the costs of infer-
tility treatment might have reduced the probability that
women who had PID would seek infertility treatment.
Fourth, as in previous studies, our cost analysis did
not capture the outpatient prescription or nonprescrip-
tion pharmacy costs of treating chronic pelvic pain and
ectopic pregnancy because they are difficult to estimate
accurately. To the extent that those costs might be
significant, our estimates will be conservative.
Fifth, as others have done, we assumed that our cost
and case estimates for 1998 did not differ greatly from
those that used data from a few years earlier.2,5,10
Actual costs from 1998 might differ from those esti-
mated from 1993–1995 data.
Although the absolute number of cases of PID, direct
medical costs, and cost per case of PID including its
three major sequelae were lower than previous esti-
mates, the direct medical costs of PID are still substan-
tial. Direct medical costs underestimate the true burden
of a disease because they do not capture intangible costs
(decreased quality of life, emotional consequences of
chronic pelvic pain, ectopic pregnancy, and infertility)
or indirect costs (lost work days, lowered productivity
resulting from illness, or lost wages). The indirect costs
for other STDs have been estimated as at least three
times the direct medical costs.3
Our analysis found that the majority of PID-related
costs are related to treating acute PID rather than its
sequelae, indicating that PID prevention activities are
likely to be cost-effective within 1 year.26Although the
total United States burden of PID has fallen consider-
ably since 1990, additional savings might be achieved.
Examples of PID prevention efforts include screening
at-risk women for bacterial STDs, promoting condom
use and safe sexual practices, periodically assessing
patients’ behavioral risks for exposure, using sensitive
tests to diagnose lower genital tract infection, and using
appropriate treatment regimens.1
1. Centers for Disease Control and Prevention. 1998 Guidelines for
treatment of sexually transmitted diseases. MMWR Morb Mortal
Wkly Rep 1998;47(RR-1):79–88.
2. Washington AE, Katz P. Cost of and payment source for pelvic
inflammatory disease. Trends and projections, 1983 through 2000.
3. Siegel JE. Estimates of the economic burden of STDs: Review of the
literature with updates. In: Eng TR, Butler WT, eds. Institute of
Medicine. The hidden epidemic. Washington DC: National Acad-
emy Press, 1997:330–56.
4. American Social Health Association. STD counseling and treat-
ment guide. Research Triangle Park, NC: 1995:209–19.
5. Haddix AC, Hillis SD, Kassler WJ. The cost effectiveness of
azithromycin for chlamydia trachomatis infections in women. Sex
Transm Dis 1995;22:274–80.
6. Hillis S, Black C, Newhall J, Walsh C, Groseclose SL. New
opportunities for chlamydia prevention: Applications of science to
public health practice. Sex Transm Dis 1995;22:197–202.
7. Centers for Disease Control and Prevention. Prevention and man-
aged care: Opportunities for managed care organizations, purchas-
ers of health care and public health agencies. MMWR Morb Mortal
Wkly Rep 1995;44(RR-14):1–12.
8. National Center for Health Statistics. 1993, 1994, 1995 National
hospital discharge survey: Public use data tape documentation.
Hyattsville, MD: 1993–1995.
9. American Medical Association. International classification of dis-
eases, ninth revision, clinical modification (ICD-9-CM). Salt Lake
City, UT: Medicode Publications, 1996.
10. Washington AE, Arno PS, Brooks MA. The economic cost of pelvic
inflammatory disease. JAMA 1986;255:1735–8.
11. Washington AE, Cates W Jr, Zaidi AA. Hospitalizations for pelvic
inflammatory disease: Epidemiology and trends in the United
States, 1975 to 1981. JAMA 1984;251:2529–33.
12. Rolfs RT, Galaid EI, Zaidi AA. Pelvic inflammatory disease: Trends
in hospitalizations and office visits, 1979 through 1988. Am J Obstet
13. National Center for Health Statistics. 1994, 1995 National ambula-
tory medical care survey. Public-use data tape documentation.
Hyattsville, MD: 1994–1995.
14. National Center for Health Statistics. 1994, 1995 National hospital
ambulatory medical care survey—outpatient department file. Pub-
lic-use data tape documentation. Hyattsville, MD: 1994–1995.
15. National Center for Health Statistics. 1994, 1995 National hospital
ambulatory medical care survey—Emergency department file.
Public-use tape documentation. Hyattsville, MD: 1994–1995.
16. National Center for Health Statistics. 1994, 1995 NAMCS and
NHAMCS OD/ER Variance Files. Public-use tape. Hyattsville,
17. Software for Survey Data Analysis (SUDAAN), Version 7.50.
Research Triangle Park, NC: Research Triangle Institute, 1997.
18. Shah BV. SESUDAAN: Standard errors program for computing
standardized rates from sample survey data. Research Triangle
Park, North Carolina: Research Triangle Institute, 1981.
19. Westro ¨m L, Mårdh P. Acute pelvic inflammatory disease (PID). In:
Holmes KK, Mårdh P, Sparling PF, Wiesner PJ, Cates W Jr, Lemon
SM, et al, eds. Sexually transmitted diseases. New York: McGraw-
Hill Information Services Company, 1990:593–613.
20. Landry DJ, Forrest JD. Public health departments providing sexu-
ally transmitted disease services. Fam Plann Perspect 1996;28:
21. MarketScan®. MarketScan®database user’s guide. Ann Arbor, MI:
The MEDSTAT Group Inc., 1997.
22. Physicians’ desk reference. 51st ed. Montvale, NJ: Medical Eco-
nomics Company, 1997.
23. Haddix AC, Teutsch SM, Schaffer PA, Dun ˜et DO. Prevention
effectiveness: A guide to decision analysis and economic evalua-
tion. New York: Oxford University Press, 1996.
24. Curran JW. Economic consequences of pelvic inflammatory
disease in the United States. Am J Obstet Gynecol 1980;138:848–
25. Van Voorhis BJ. Cost of therapy, infertility treatment: What is a
cost-effective approach? Drug Benefit Trends July 1998:22–9.
26. Scholes D, Stergachis A, Heidrich FE, Andrilla H, Holmes KK,
Stamm WE. Prevention of pelvic inflammatory disease by screen-
ing for cervical chlamydial infection. N Engl J Med 1996;334:
VOL. 95, NO. 3, MARCH 2000
Rein et al
Pelvic Inflammatory Disease
Address reprint requests to: Download full-text
David B. Rein, MPA
Division of STD Prevention
Centers for Disease Control and Prevention
1600 Clifton Road, Mail Stop E-44
Atlanta, GA 30333
Received May 21, 1999.
Received in revised form August 17, 1999.
Accepted August 27, 1999.
Copyright © 2000 by The American College of Obstetricians and
Gynecologists. Published by Elsevier Science Inc.
402 Rein et al
Pelvic Inflammatory DiseaseObstetrics & Gynecology