Secondary Logo

Share this article on:

Trends in Pelvic Inflammatory Disease Hospital Discharges and Ambulatory Visits, United States, 1985–2001

Sutton, Madeline Y. MD, MPH; Sternberg, Maya PhD; Zaidi, Akbar PhD; St. Louis, Michael E. MD; Markowitz, Lauri E. MD

doi: 10.1097/01.olq.0000175375.60973.cb
Article

Objective: The objective of this study was to describe the estimated trends in incidence of pelvic inflammatory disease (PID) among reproductive-aged women in hospital and ambulatory settings.

Study: Analyses of PID estimates were performed. Three nationally representative surveys conducted by the National Center for Health Statistics (NCHS): National Hospital Discharge Survey (NHDS), National Hospital Ambulatory Medical Care Survey (NHAMCS), and National Ambulatory Medical Care Survey (NAMCS), were used to obtain the estimates of PID (defined by International Classification of Diseases, 9th Revision codes). National Disease and Therapeutic Index (NDTI) estimates were reviewed for comparison.

Results: Rates of hospitalized PID declined 68% overall from 1985 through 2001 (P <0.0001). Ambulatory data support a decrease in PID from 1985 to 2001. From 1995 to 2001, approximately 769,859 cases of acute and unspecified PID were diagnosed annually, 91% in ambulatory settings.

Conclusions: PID has decreased in hospital and ambulatory settings. The expanded national surveys in outpatient and emergency departments provide more complete estimates for PID. Optimal management of PID should target ambulatory settings, where the majority of cases are diagnosed and treated.

Although pelvic inflammatory disease (PID) cases have decreased in the United States in recent years, over 760,000 reproductive-aged women continue to be diagnosed with acute PID annually.

From the Division of Sexually Transmitted Disease Prevention, National Center for HIV, STD, and TB Prevention, Centers for Disease Control and Prevention, Atlanta, Georgia

The authors thank the clinicians and their staff who voluntarily participated in the NCHS surveys and the field representatives from the U.S. Census Bureau who collected the data.

Correspondence: Madeline Y. Sutton, MD, MPH, Centers for Disease Control and Prevention, DSTDP 1600 Clifton Road, Mailstop E-02, Atlanta, GA 30333. E-mail: zxa3@cdc.gov.

Received for publication February 4, 2005, and accepted April 15, 2005.

PELVIC INFLAMMATORY DISEASE (PID), an upper genital tract inflammation often caused by sexually transmitted infections, is the most common gynecologic reason for hospitalization1 and emergency department visits2 in the United States. Serious sequelae that can result from the fallopian tube damage and tissue scarring associated with PID include infertility, ectopic pregnancy, chronic pelvic pain, and recurrent PID.3–6 Recent annual estimates of the direct costs of care for acute PID and its sequelae are at $2 billion.7 When indirect costs of PID such as productivity losses and losses from premature death were considered, total annual cost estimates were as high as $10 billion using year 2000 dollars.8,9 The high financial and societal costs of this common gynecologic disease process make PID incidence and trend data important components of understanding the impact of this disease.

Many cases of PID are caused by sexually transmitted organisms that migrate from the vagina and/or cervix to the uterus and fallopian tubes.10 From 16% to 75% of women with acute PID have 2 common sexually transmitted organisms, Neisseria gonorrhoeae and/or Chlamydia trachomatis, isolated from the lower genital tract.11–15 The gonococcus often produces a more intense inflammatory reaction in the tubal lumen than the reaction caused by chlamydial infection.16 Hence, the patient with gonococcal PID may present more acutely, often requiring hospitalization. It is estimated that approximately 15% of women with cervical infection by N. gonorrhoeae subsequently show clinical signs of acute pelvic inflammatory disease, compared with 10% of cervical chlamydial infections.16,17

The clinical criteria for diagnosis of PID are often uncertain and lack sensitivity. However, it is a diagnosis often made in reproductive-aged women with symptoms, because undiagnosed and untreated, it can cause significant long-term morbidity. Although PID is not a nationally notifiable disease, national survey datasets can provide estimates of diagnosed cases. Previous estimates of PID hospitalizations and office visits showed an annual average of 676,000 total cases (including 181,700 hospital discharges for acute and unspecified PID) diagnosed in the 1980s,18 and 267,200 hospitalized cases of PID in the late 1970s.19 Data from outpatient and emergency department settings were not available at that time. In this article, we analyzed National Center for Health Statistics datasets to estimate trends in PID hospitalizations and ambulatory visits from 1985 to 2001, including new datasets from outpatient department and emergency department settings.

Back to Top | Article Outline

Methods

Data Sources

Data were analyzed from 3 U.S.-based annual national probability surveys conducted by the National Center for Health Statistics (NCHS): National Hospital Discharge Survey (NHDS), National Hospital Ambulatory Medical Care Survey (NHAMCS), and National Ambulatory Medical Care Survey (NAMCS) for 1985 to 2001. Patient diagnoses were coded using International Classification of Diseases, 9th Revision (ICD-9) during this time period. Data analyses were inclusive of several ICD-9 codes for pelvic inflammatory disease (Table 1). Data were available for all years of the NHDS. However, the NHAMCS-emergency department (ED) and NHAMCS-outpatient department (OPD) datasets were available from 1992 to 2001 only. The methods of data collection are summarized subsequently.

TABLE 1

TABLE 1

The NHDS uses inpatient discharges as its sampling unit and is an ongoing national probability survey of inpatients discharged from nonfederal short-stay hospitals with 6 or more beds. The NHDS collects data from a sample of approximately 270,000 inpatient records acquired from a national sample of approximately 500 hospitals representing all 50 states and the District of Columbia. The response rate has been greater than 90% in recent years. The final data are weighted to represent approximately 30 million hospital discharges per year.20

The basic sampling unit for the NHAMCS is the patient visit or encounter. NHAMCS includes a national 4-stage probability sample of visits to nonfederal emergency and outpatient departments of U.S. noninstitutional general and short-stay hospitals.22 A fixed panel of 600 hospitals is used for the NHAMCS sample. Approximately 425 eligible hospitals are sampled annually, and the response rate has been 95% in recent years.

The basic sampling unit for the NAMCS is also the physician-patient encounter or visit. Only visits to the offices of nonfederally employed physicians classified by the American Medical Association (AMA) or the American Osteopathic Association (AOA) as “office-based, patient care” are included in NAMCS.21 The NAMCS uses a multistage probability design. For PID estimates, only first office visits are counted, because the management of PID often involves multiple follow-up visits. Approximately 1500 physicians are sampled each year. The response rate is above 60%.

Because the NAMCS data have not been collected consistently each year during the entire period from 1985 to 2001, the National Disease and Therapeutic Index (NDTI) annual data of total physician office first visits for PID were reviewed for comparison. NDTI is a commercially conducted (IMS America), annual, 2-stage, stratified, random cluster sample survey of patient encounters with physicians in office-based, private practices in the United States. Approximately 3000 physicians are surveyed quarterly.23 Physicians who are selected report on all patient encounters during one 48-hour period each quarter, but only office visit encounters were included in this analysis. Weighted national estimates are then projected from this sample. Approximately 80% to 90% of the NDTI physician participants report and receive an incentive for their participation.

Back to Top | Article Outline

Definition of Pelvic Inflammatory Disease

The ICD codes were used to classify cases of PID for this analysis (Table 1). Discharge diagnoses data from the NCHS surveys were coded using International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM). Patient discharges and encounters were included as PID diagnoses if one of several ICD-9 codes was listed as one of the final diagnoses. PID was subclassified as acute if at least one acute PID code was listed, as chronic if at least one chronic but no acute code was listed and as unspecified if neither acute nor chronic codes were listed. The NDTI diagnoses of PID were coded with the International Classification of Diseases, 7th Revision (ICD-7) through 1986 and with ICD-9 from 1987 to 2001 (Table 1). Although chronic PID was rarely reported, the NDTI estimates include acute, unspecified, as well as chronic disease classifications.

Back to Top | Article Outline

Data Analysis and Statistical Methods

This article focuses on acute and unspecified PID as distinct infectious processes occurring in women. Previous studies support the fact that chronic PID is an epidemiologically different diagnosis with distinctly different outcomes compared with acute and unspecified PID.16,18 Chronic PID, which is characterized by long-term sequelae of acute infection such as adhesions and hydrosalpinx, is often bacteriologically sterile24 and as such represents a distinctly different clinical syndrome that occurs in an often older subset of women and more often leads to surgical intervention. In our data, women with acute and unspecified PID were demographically similar with regard to age and region of country for diagnosis. Women in the reproductive-aged group from 15 to 44 years represented over 90% of those diagnosed with PID of acute and unspecified duration. For these reasons, this article focuses mainly on analyses of acute and unspecified-duration PID of women in the 15- to 44-year age group. For overall hospital discharge PID trends and total PID estimates, chronic PID estimates are included in an effort to remain consistent with and comparable to previously published hospital discharge PID estimates. The data for acute and unspecified PID are presented as: 1) annual PID estimates from NHDS, NAMCS, and NDTI for the years for which the data were available between 1985 and 2001; 2) average age- and region-specific hospitalization rates (NHDS), emergency room and outpatient department visit rates (NHAMCS), and physician visit rates (NAMCS) between 1995 and 2001; and 3) average annual estimates for each NCHS data source between 1995 and 2001. The NHAMCS-ED and -OPD surveys were started in 1992 and therefore, long-term and meaningful trend data are limited for these newer data sources and NAMCS were not conducted from 1986 to 1988.

Specialized software for complex survey designs was used to obtain point estimates and variances based on the weights and design variables provided by NCHS for each survey. Before calendar year 2002, NAMCS and NHAMCS public use files did not contain the sample design variables to compute standard errors while taking the sampling design into account. In 2002, a 5-year research project culminated in releasing the data from 1993 to 2001 for NAMCS and 1995 to 2001 for NHAMCS to mask sample design variables so that they could be released without compromising the confidentiality of the participants. Therefore, the years 1995 to 2001 are used for the average annual estimates, because these are both the most recent years of available data and years for which standard errors can be estimated using the publicly released data along with complex survey design software. The denominator of the age- and region-specific hospitalization rates (NHDS), emergency room and outpatient department visit rates (NHAMCS), and physician visit rates (NAMCS) were calculated using an average number of females during the 7-year period (1995–2001) based on the U.S. Census population estimates (NCHS, CD-ROM series 13, no. 21A).

Data management was performed in SAS V8 and all statistical analyses were performed using SAS-callable SUDAAN. A 2-sample t test was used to test the change in the number of hospitalizations (NHDS) for all diagnosed PID cases between 1985 and 2001. Differences between proportions were tested using Wald chi-squared tests, and t tests were used to test differences between means. A 0.05 significance level was used to declare significance for all hypothesis tests.

Back to Top | Article Outline

Results

Trends (1985–2001)

Hospitalizations for Pelvic Inflammatory Disease.

Between 1985 and 2001, the estimated rate of hospitalizations for all diagnosed PID declined by 68%, from 4.4 to 1.4 per 1000 women ages 15 to 44 (Fig. 1). This trend was significant at a P value of <0.0001 using a 2-sample t test. Hospitalization rates for acute and unspecified PID also decreased overall and for each age group (Fig. 2). The 15- to 19-year age group experienced a 53% decrease; a 77% decrease occurred for the 20- to 24-year age group; 66% decrease for the 25- to 29-year group; and 36%, 71%, and 15% declines, respectively, for the 30–34-, 35–39-, and 40–44-year age groups.

Fig. 1

Fig. 1

Fig. 2

Fig. 2

Back to Top | Article Outline

Ambulatory Visits for Pelvic Inflammatory Disease Office Visits (NDTI and NAMCS, 1985–2001).

Among women ages 15 to 44, NDTI data show some variation in number of initial visits to physicians' offices for acute and unspecified PID between 1985 and 2001, but overall there was a 47% decrease in the estimated rates of PID cases diagnosed in this setting (Fig. 3). NAMCS shows a 68% decrease in estimated rates of first visit PID between 1985 and 2001 (Fig. 3). NAMCS data have standard errors greater than 30% for years 1985 to 2001, and these data should be interpreted with caution. The NAMCS and NDTI estimates, although initially widely divergent in the late 1980s, show a narrower gap by 2001. Between 1985 and 2001, based on NAMCS data, the decrease in office visits was consistent among all age groups and regions (data not shown).

Fig. 3

Fig. 3

Back to Top | Article Outline

Annual Estimates (1995–2001)

Hospitalizations for Pelvic Inflammatory Disease.

For the period of 1995 to 2001, an annual average of 88,743 hospitalizations for PID occurred. Of these, a mean of 20,363 (23%) were classified as being for acute PID and 45,172 (51%) for PID of unspecified duration. The mean rate of hospitalization for acute and unspecified PID from 1995 to 2001 was highest for women ages 20 to 24 years, with the 15- to 19-year age group ranking second, followed by the 30- to 34-year age group (Table 2). Rates of hospitalization for acute and unspecified PID were highest in the south and lowest in the west (P<0.0001) (Table 2). Average rates of acute and unspecified hospitalized PID were 0.6 per 1000 for women classified as white and 2.1 per 1000 for women classified as black, reflecting a 3.5-fold disparity that was consistent for each year from 1995 to 2001.

TABLE 2

TABLE 2

Back to Top | Article Outline

Ambulatory Visits for Pelvic Inflammatory Disease (NAMCS, NHAMCS-OPD, NHAMCS-ED).

Overall, between 1995 and 2001, an estimated 735,316 women aged 15 to 44 were diagnosed with PID in ambulatory settings in the United States annually; 96% or 704,325 had acute and unspecified PID. This combined figure includes women diagnosed in physician offices, hospital outpatient departments, and hospital emergency departments. The age group with the highest rate of PID diagnoses in these combined ambulatory settings was the 25- to 29-year age group (Table 2). The region with the highest rate of PID ambulatory diagnoses was the south (Table 2). In the combined ambulatory settings, black women were diagnosed with PID at a rate of 22.2 per 1000 population, 2.3 times greater than the rate of diagnosis among white women (9.8 per 1000 population).

Back to Top | Article Outline

Office Visits (NAMCS).

An annual average of 429,096 initial PID visits were made to private physicians' offices between 1995 and 2001 (Table 3), for a mean annual rate of 7.3 per 1000 women ages 15 to 44. Ninety-four percent (403,554) of those initial PID visits were for acute and unspecified PID. Like the overall ambulatory PID rates by 5-year age groups (Table 2), the mean annual physician office visit rates for acute and unspecified pelvic inflammatory disease from 1995 to 2001 were highest among the 25- to 29-year age group (13.6 per 1000). In contrast to overall ambulatory rates, mean annual physician office visit rates for acute and unspecified PID were highest in the west (8.3 per 1000), followed by the south (7.6 per 1000), midwest (5.3 per 1000), and northeast (4.3 per 1000), respectively. Women classified as black were diagnosed at a rate of 9.6 per 1000, whereas women classified as white were diagnosed at a rate of 6.1 per 1000.

TABLE 3

TABLE 3

Based on NAMCS data, most PID diagnoses between 1995 and 2001 were made by obstetrician-gynecologists (47%), general or family practitioners (35%), and internists (10%). This pattern did not change over the study period. No data were available for midlevel providers such as nurse practitioners and physicians' assistants.

Back to Top | Article Outline

Outpatient Department Visits (NHAMCS-OPD).

The annual average of acute and unspecified PID outpatient department visits for NHAMCS between 1995 and 2001 was 45,106 (94% of all PID diagnosed in this setting), giving a rate of 0.8 per 1000 women ages 15 to 44 (Table 3). Annual visits in this setting were highest among those 15 to 19 years, with 1.3 visits per 1000 (Table 2). Although overall ambulatory visits were most common in the south (Table 2), the outpatient visit portion was highest in the midwestern region of the country (1.2 per 1000 women) followed by the northeast (0.7 per 1000), west (0.69 per 1000), and south (0.5 per 1000). In outpatient department settings, black women were almost 3 times more likely to be diagnosed with PID than white women, with rates of 1.7 and 0.6 per 1000, respectively, from 1995 to 2001.

Back to Top | Article Outline

Emergency Department Visits (NHAMCS-ED).

The annual average of emergency department visits for PID between 1995 and 2001 was 258,235 (4.2 per 1000 women ages 15–44), 99% of which was for acute and unspecified PID (Table 3). The most commonly affected age group was 20 to 24 year olds (7.9 per 1000). Consistent with the overall data, emergency department visit rates were highest in the southern region area of the country (5.7 per 1000). The midwest had the next highest rate of diagnosed cases (4.8 per 1000), followed by the northeast (3.5 per 1000) and west (1.9 per 1000), respectively. In emergency department settings, black women were diagnosed with acute and unspecified PID at rate of 11.0 per 1000, which was 3.5 times the rate of diagnosis among white women (3.1 per 1000) between 1995 and 2001.

Back to Top | Article Outline

Total Hospitalized and Ambulatory Pelvic Inflammatory Disease.

The annual average estimate of all diagnosed acute and unspecified ambulatory and hospitalized PID, based on recent NCHS data sets from 1995 to 2001, was 769,859 (Table 3). When chronic cases are included, the annual average estimate is 824,059 (Table 3). This estimate represents hospital discharge and office visits estimates, and, for the first time, includes estimates from outpatient and emergency departments. Hospital discharges for PID were only 11% of the total estimated cases of PID, and 89% of cases were diagnosed in ambulatory settings (Table 3).

None of our current data sources provides information regarding which hospitalized women may have been previously evaluated in ambulatory settings. However, even if all of the hospitalized women had had a previous ambulatory evaluation, this would account for an overlap of approximately 11% of the annually diagnosed cases of PID in the United States (Table 3).

Back to Top | Article Outline

Discussion

Two principal findings frame these analyses: both hospital discharges and ambulatory estimated cases of PID decreased significantly during the study time period. Nevertheless, our current annual estimate of acute and unspecified PID diagnosed in the United States (769,859) is greater than previously published estimates of PID that had fewer data sources available.18,19 Previous average annual estimates of all hospitalized PID (267,200 from 1975–1981),19 or hospitalizations and office visits (676,000 all PID, including 581,600 for acute PID, from 1979–1988),18 used available NCHS hospital discharge and physician office visit data surveys as available. Our data show continuing, significant decreases in hospitalizations for acute and unspecified PID (65,534 from 1995–2001), a trend that is similar to earlier reports of hospitalized acute PID cases.18 However, with over 824,000 reproductive-aged women diagnosed with various stages of PID in this country each year, clinically evident PID remains an important public health concern for women in both hospital and ambulatory settings.

The highest rate of hospitalizations reported for PID in 1987 to 1988 was in 15 to 19 year olds.18 Our analyses found hospitalization rates highest in 20 to 24 year olds (Table 2). Although 15 to 19 year olds had fewer hospitalizations than older age groups for this analysis, our denominator was U.S. Census data for all 15- to 19-year-old women, which may not be as accurate as a denominator that includes only sexually active adolescents. The rate among sexually active 15 to 19 year olds may be more than double our estimate if less than 50% of the 15- to 19-year-old population is sexually active as reported.25

The previously unreported NHAMCS estimates of patients seen in emergency and outpatient departments show that these settings account for as much as 39% of acute and unspecified PID cases diagnosed and treated on an ambulatory basis. This new information underscores the importance of providers in ambulatory settings as caregivers of women who present for clinical care in these settings.

These data have several important limitations. First, some of the surveys have limited precision because of their small sample sizes and often large sampling errors. Second, these national probability sampling surveys provide only estimates and not an actual count of patients encountered that a national surveillance system may allow. A problem for both surveys, and for a potential national surveillance system for PID, is the lack of sensitivity and specificity of the clinical diagnosis itself. The clinical diagnosis has a sensitivity of 65% when compared with the diagnostic gold standard of laparoscopy.26 However, laparoscopy is an invasive surgical procedure requiring anesthesia and is not a frequently used method of diagnosis for PID. The current estimates of PID, which are clinically based, would likely be higher if the gold standard of laparoscopy were used for diagnosis. Third, these data are probably not representative of all types of PID occurring in the United States. Approximately two thirds of women may have no or mild symptoms of PID, which go undiagnosed. In addition, the ICD coding system has limitations and may be causing under- or overreporting, depending on the additional influences of available clinical management options and reimbursement issues. Some PID clinical diagnostic codes used to conduct surveillance surveys have a poor positive predictive value (as low as 18%) when compared with the Centers for Disease Control and Prevention clinical case definition.27

The trend of racial disparities in diagnosed PID is of concern; in ambulatory and hospitalized settings, black women had rates of diagnosis of disease that were 2 to 3 times the rate in white women. Data were not available for women of Hispanic ethnicity for these analyses. The subjective method by which PID is diagnosed may allow racial and socioeconomic biases to influence this diagnosis at many levels.28,29 Therefore, the lack of information regarding the impact and meaning of race and ethnicity of patient and/or provider on a diagnosis of PID requires that the racial disparity data be interpreted with caution. Additional studies are warranted to further understand the potential role of race/ethnicity in making or having a diagnosis of PID.

The overall decrease in diagnoses of PID in the United States through the year 2001 appears to correlate with national declines in gonorrhea and chlamydial infection rates in areas that have had long-term programs in place for routine screening and treatment.30 The rate of gonorrhea declined 73% between 1975 and 2001,30 paralleling the trend in hospitalized and overall PID rates during that same time period. Trends in reported chlamydial infections are not as well defined, largely because national surveillance of chlamydial infection is still evolving. The northwestern part of the United States, an area with one of the first and most comprehensive chlamydia screening and prevalence monitoring programs, has demonstrated a 55.4% decline in chlamydia positivity among 15- to 44-year-old women since 1988, and this parallels the overall downward trend in reported cases of gonorrhea and estimates of PID in the northwestern part of the country (NCHS, data not shown). As national efforts to screen and report cases of chlamydial infection continue to improve, so will our ability to compare trends with PID sequelae.

A large part of the economic and psychosocial burden of PID is associated with adverse sequelae, which occur in approximately 25% of untreated women and account for 80% of PID-associated costs.31 Histologic studies estimate that approximately 50% of ectopic pregnancies occur in oviducts damaged by previous PID.16 In certain European countries, changes in PID trends have been associated with similar changes in sequelae such as ectopic pregnancy.32 Similar to hospitalizations and management trends for PID, hospitalizations for ectopic pregnancies have declined overall in the United States from the late 1980s through 2001, and use of outpatient management regimens for ectopic pregnancies have increased.33

Prevention of PID and its sequelae are important reasons for public health efforts to maintain and institute programs that seek to reduce the incidence of etiologic genital infections caused by N. gonorrhea and C. trachomatis. One randomized trial showed that chlamydia screening programs can lead to a 60% decrease in the incidence of PID.34 New efforts are warranted at routine surveillance for PID using administrative datasets, which also link diagnosed sexually transmitted diseases to PID occurrences. In the absence of a national surveillance system for PID, these NCHS survey results should continue to be used as our best estimates for the burden of disease related to PID in the United States. The expanded national hospital ambulatory care surveys for emergency and outpatient departments are an important advance in providing these estimates. Based on these new data findings, efforts to promote optimal management of PID should target ambulatory settings, where an estimated 91% of acute and unspecified cases of PID are diagnosed. Furthermore, results from the Pelvic inflammatory disease Evaluation And Clinical Health (PEACH) randomized trial found no measurable difference in effectiveness between inpatient and outpatient treatment regimens, and suggest that opting for outpatient management, when appropriate, has the added benefit of $500 million cost savings based on 1998 dollars.35

Important questions remain unanswered regarding PID, including the proportions that are linked directly to sexually transmitted pathogens and the true numbers of cases that may remain unrecognized as a result of lack of symptoms. If as much as two thirds of PID is “subclinical” or undiagnosed,16 the true burden of disease for PID may be nearly 2.5 million affected women each year. Future attempts to better understand the true impact of PID should attempt to link information regarding etiologic organisms and further develop methods of understanding silent, undiagnosed cases of PID.

Back to Top | Article Outline

References

1. Velebil P, Wingo PA, Xia Z, Wilcox LS, Peterson HB. Rate of hospitalization for gynecologic disorders among reproductive-age women in the United States. Obstet Gynecol 1995; 86:764–769.
2. Curtis KM, Hillis SD, Kieke BA, Brett KM, Marchbanks PA, Peterson HB. Visits to emergency departments for gynecologic disorders in the United States, 1992–1994. Obstet Gynecol 1998; 91:1007–1012.
3. Lepine LA, Hillis SD, Marchbanks PA, Joesoef MR, Peterson HB, Westrom L. Severity of pelvic inflammatory disease as a predictor of the probability of live birth. Am J Obstet Gynecol 1998; 178:977–981.
4. Westrom L, Joesoef R, Reynolds G, Hagdu A, Thompson SE. Pelvic inflammatory disease and fertility. A cohort study of 1,844 women with laparoscopically verified disease and 657 control women with normal laparoscopic results. Sex Transm Dis 1992; 19:185–92.
5. Joesoef MR, Westrom L, Reynolds G, Marchbanks P, Cates W. Recurrence of ectopic pregnancy: The role of salpingitis. Am J Obstet Gynecol 1991; 165:46–50.
6. Mardh PA, Lind I, Svensson L, Westrom L, Moller BR. Antibodies to Chlamydia trachomatis, Mycoplasma hominis, and Neisseria gonorrhoeae in sera from patients with acute salpingitis. Br J Vener Dis 1981; 57:125–129.
7. Rein DB, Kassler WJ, Irwin KL, Rabiee L. Direct medical cost of pelvic inflammatory disease and its sequelae: Decreasing, but still substantial. Obstet Gynecol 2000; 95:397–402.
8. Washington AE, Katz P. Cost of and payment source for pelvic inflammatory disease. Trends and projections, 1983 through 2000. JAMA 1991; 266:2565–2569.
9. Curran JW. Economic consequences of pelvic inflammatory disease in the United States. Am J Obstet Gynecol 1980; 138:848–851.
10. Mardh PA. An overview of infectious agents of salpingitis, their biology, and recent advances in methods of detection. Am J Obstet Gynecol 1980; 138:933–938.
11. Kristensen GB, Bollerup AC, Lind K, et al. Infections with Neisseria gonorrhoeae and Chlamydia trachomatis in women with acute salpingitis. Am 1985; 61:179–184.
12. Moller BR, Mardh PA, Ahrons S, Nussler E. Infection with Chlamydia trachomatis, Mycoplasma hominis and Neisseria gonorrhoeae in patients with acute pelvic inflammatory disease Sex Transm Dis 1981; 8:198–202.
13. Eschenbach DA, Buchanan TM, Pollock HM, et al. Polymicrobial etiology of acute pelvic inflammatory disease. N Engl J Med 1975; 293:166–171.
14. Sweet RL, Draper DL, Schachter J, James J, Hadley WK, Brooks GF. Microbiology and pathogenesis of acute salpingitis as determined by laparoscopy: What is the appropriate site to sample? Am J Obstet Gynecol 1980; 138:985–989.
15. Parker CA, Topinka MA. The incidence of positive cultures in women suspected of having PID/salpingitis. Acad Emerg Med 2000; 7:1170–1174.
16. Westrom L, Eschenbach D. Pelvic inflammatory disease. In: Holmes KK, ed. Sexually Transmitted Diseases. New York: McGraw-Hill, 1999:784–785.
17. Eschenbach DA. Acute pelvic inflammatory disease: etiology, risk factors and pathogenesis. Clin Obstet Gynecol 1976; 19:147–169.
18. Rolfs RT, Galaid EI, Zaidi AA. Pelvic inflammatory disease: Trends in hospitalizations and office visits, 1979 through 1988. Am J Obstet Gynecol 1992; 166:983–990.
19. Washington AE, Cates W Jr, Zaidi AA. Hospitalizations for pelvic inflammatory disease. Epidemiology and trends in the United States, 1975 to 1981. JAMA 1984; 251:2529–2533.
20. Dennison C, Pokras R. Design and operation of the National Hospital Discharge Survey: 1988 redesign. Vital Health Stat 2000; 1.
21. Cherry DK, Burt CW, Woodwell DA. National Ambulatory Medical Care Survey: 1999 summary. Advance Data From Vital and Health Statistics, no 322. Hyattsville, MD: National Center for Health Statistics, 2001.
22. Ly N, McCaig LF, Burt CW. National Hospital Ambulatory Medical Care Survey: 1999 Outpatient Department Summary. Advance Data From Vital and Health Statistics, no 321. Hyattsville, MD: National Center for Health Statistics, 2001.
23. IMS America Information Service Manual: National Disease and Therapeutic Index, 1996.
24. Mishell DR, Stenchever MA, Droegemueller W, Herbst AL. Infections of the upper genital tract. In: Comprehensive Gynecology, 3rd ed. St. Louis: Mosby-Year Book, 1997:661–690.
25. Brener N, Lowry R, Kann L, et al. Trends in sexual risk behaviors among high school students—United States, 1991–2001. MMWR Morb Mortal Wkly Rep 2002; 51:856–859.
26. Jacobson L, Westrom L. Objectivized diagnosis of acute pelvic inflammatory disease. Diagnostic and prognostic value of routine laparoscopy. Am J Obstet Gynecol 1969; 105:1088–1098.
27. Ratelle S, Yokoe D, Blejan C, et al. Predictive value of clinical diagnostic codes for the CDC case definition of pelvic inflammatory disease (PID). Sex Transm Dis 2003; 30:866–870.
28. Lawson EJ, Rodgers-Rose LF, Rajaram S. The psychosocial context of black women's health. Health Care for Women International 1999; 20:279–289.
29. Grodstein F, Rothman K. Epidemiology of pelvic inflammatory disease. Epidemiology 1994; 5:234–241.
30. Centers for Disease Control and Prevention. Sexually Transmitted Disease Surveillance, 2002. Atlanta: US Department of Health and Human Services, September 2003.
31. Yeh J, Hook EW, Goldie SJ. A refined estimate of the average lifetime cost of pelvic inflammatory disease. Sex Transm Dis 2003; 30:369–378.
32. Westrom L. Decrease in incidence of women treated in hospital for acute salpingitis in Sweden. Genitourin Med 1998; 64:59–63.
33. Zane SB, Kieke BA Jr, Kendrick JS, Bruce C. Surveillance in a time of changing health care practices: Estimating ectopic pregnancy incidence in the United States. Matern Child Health J 2002; 6:227–36.
34. Scholes D, Stergachis A, Heidrich FE, Andrilla H, Holmes KK, Stamm WE. Prevention of pelvic inflammatory disease by screening for cervical chlamydial infection. N Engl J Med 1996; 34:1362–1366.
35. Ness RB, Soper DE, Holley RL, et. al. Effectiveness of inpatient and outpatient treatment strategies for women with pelvic inflammatory disease: Results from the Pelvic Inflammatory Disease Evaluation and Clinical Health (PEACH) Randomized Trial. Am J Obstet Gynecol 2002; 186:929–937.
© Copyright 2005 American Sexually Transmitted Diseases Association