Pelvic inflammatory disease (PID) leads to more hospitalizations among reproductive-aged women than other benign gynecological conditions such as ovarian cysts, menstrual disorders, and uterine leiomyomata.1 From 1975 to 1981, PID accounted for more than 350,000 hospital admissions and 150,000 surgical procedures per year in the United States.2 While tuboovarian abscesses are generally reported as complicating 10–15% of hospitalized cases of acute PID, rates up to 34% have been reported.3 In 1990, PID and its major health sequelae, such as infertility and ectopic pregnancy, cost 4.2 billion dollars.4 In 1998, direct medical expenditures for PID and its sequelae were estimated at 1.88 billion dollars.5
The epidemiology of hospitalizations and surgical procedures performed for PID and tuboovarian abscess in the United States in the past decade is not well described. A recent Scandinavian study showed a 26% reduction in hospitalized cases of PID in Norway over a 10-year period.6 Population-based research is important for identifying medical and hospital practice patterns that can inform health policy. Understanding both the distribution of and changes in the inpatient management of PID across large and diverse populations can help revise resource distribution to best serve the needs of society as practice patterns change.
Using the California Patient Discharge Database from the Office of Statewide Health Planning and Development and the California Census, we estimated hospitalization rates for PID and tuboovarian abscess from 1991 through 2001. Our objective was to show trends in the inpatient management of PID and tuboovarian abscess, with a special focus on age and racial/ethnic differences.
MATERIALS AND METHODS
We used the California Patient Discharge Database of the California Office of Statewide Health Planning and Development, which tracks all hospital discharges in the state of California. The database includes diagnostic and procedural codes according to the International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM). In addition, the database lists a primary diagnosis, up to 24 secondary discharge diagnoses, and 20 procedures as well as patient demographic information and outcome variables, including length of hospital stay, readmissions, and inpatient mortality. The institutional review board of the University of California, Davis, approved this study.
We used ICD-9-CM codes 614.0, 614.1, 614.3, and 614.4 to define PID and 614.2 to define tuboovarian abscess. We also used ICD-9-CM codes to categorize the surgical procedures performed for these diagnoses: total hysterectomy, 68.4; subtotal hysterectomy, 68.3; vaginal hysterectomy, 68.5; unilateral or bilateral oophorectomy, 65.3, 65.5; unilateral or bilateral salpingo-oophorectomy, 65.4, 65.6; laparoscopic-assisted vaginal hysterectomy, 68.51; percutaneous drainage of tuboovarian abscess, 66.91; and percutaneous drainage of pelvic abscess, 54.91. Because our objective was to describe hospitalizations for PID and tuboovarian abscess, we included only patients who had PID or tuboovarian abscess listed as the primary discharge diagnosis.
We analyzed the data for PID or tuboovarian abscess discharges for females aged 10 years or older and by these racial or ethnic categories: white (non-Hispanic), black (non-Hispanic), Hispanic, and Asian. To calculate population rates of hospitalizations for a given age category, we used the number of discharges with the primary diagnosis of PID or tuboovarian abscess within the specified age group as the numerator. The denominator was the number of women in the state of California in that age group based on yearly United States Census Bureau population estimates provided by the California Department of Finance. Population rates for each racial or ethnic group were calculated as the number of PID or tuboovarian abscess discharges for a specific racial or ethnic category divided by the number of women in California in that racial or ethnic group between specified age categories. The annual incidence of hospitalizations was defined as the number of PID or tuboovarian abscess discharges per 10,000 women in the state of California for a given year from 1991 to 2001. We also estimated hospitalization rates for PID and tuboovarian abscess by both age and racial/ethnic category. To calculate the proportion of hospitalizations that included procedures for a specific age or racial/ethnic group in a particular year, we combined the number of procedures for PID and tuboovarian abscess in that age or racial/ethnic category performed in a particular year and divided by the total number of age- or race/ethnic-specific women hospitalized for PID and tuboovarian abscess in that year.
To estimate inpatient mortality associated with PID or tuboovarian abscess, we used the disposition code for “died.” We also examined readmissions within 90 days with the primary diagnosis of PID or tuboovarian abscess and compared the mean length of stay for hospitalizations with subsequent readmissions with the average length of stay for hospitalizations without readmissions.
In 1991, 3,348 women were hospitalized with a primary diagnosis of PID at a rate of 2.6 per 10,000 women; in 2001, 1,463 women were hospitalized with PID at a rate of 1.0 per 10,000 (Table 1). Therefore, from 1991 to 2001, the hospitalization rate for PID in California decreased by 61.5%. Tuboovarian abscess hospitalization rates declined by only 33.3%, overall, during the same time period (from 0.6 to 0.4 per 10,000 women). Over the 10 years, the percentage of women hospitalized for PID or tuboovarian abscess who underwent hysterectomy remained stable (range 16.9–17.1%), while the proportion of hospitalizations associated with oophorectomy and drainage procedures increased steadily from 1991 to 2001 (oophorectomy, 27.3% in 1991 to 35.7% in 2001; drainage procedures, 2.6% in 1991 to 7.6% in 2001).
We observed the highest PID hospitalization rates and the steepest decline in rates over the 10 years in women 20–39 years of age. In 1991, 2,287 women in this age group were hospitalized with PID in California, whereas only 639 women were hospitalized in 2001. In contrast, rates in teens and in women over age 40 were stable (Fig. 1). During this same period, hospitalization rates for tuboovarian abscess decreased in all age groups except in the group of women 40 years of age or older who had a 25.0% increase in rates (from 0.3 per 10,000 in 1991 to 0.4 per 10,000 in 2001; 10- to 19-year age group, 0.3 to 0.1 from 1991 to 2001; 20- to 29-year age group, 0.9 to 0.5; 30- to 39-year age group, 1.0 to 0.6).
Among women hospitalized with PID or tuboovarian abscess between 1991 and 2001, a higher proportion of women over 40 years of age underwent hysterectomy and oophorectomy (32.1% had a hysterectomy and 56.8% had an oophorectomy). For women 30–39 years of age, 16.3% had a hysterectomy and 27.3% had an oophorectomy. For those 20–29 years of age, less than 5% of hospitalizations included hysterectomy and 10% included oophorectomy. Although very few teens hospitalized for PID or tuboovarian abscess underwent hysterectomy (5 total in the 10 years), 3.7% had an oophorectomy. We found that, for all age groups, the highest proportion of hospitalizations that included drainage procedures were in 2001 (7.8% among 10–19 year olds; 4.9% among 20–29 year olds; 6.8% among 30–39 year olds; and 9.1% among women 40 years or older).
Black women between 20 and 39 years of age had the highest PID hospitalization rates and the steepest decline in rates from 1991 to 2001 compared with other racial/ethnic groups. The number of black women 20–39 years of age hospitalized for PID was 439 in 1991 and declined to 136 in 2001. (Fig. 2). The proportion of hospitalizations that included hysterectomy or oophorectomy was lowest among black women compared with white, Hispanic, or Asian women (Fig. 3). This finding persisted when we compared black women with white, Hispanic, and Asian women in the same age categories. The proportion of hospitalizations that included drainage procedures was similar among all 4 racial/ethnic groups (Fig. 3).
From 1991 to 2001, the average length of hospital stay for PID and tuboovarian abscess did not differ (4.9 ± 1.7 days for PID; and 4.9 ± 1.6 days for tuboovarian abscess), and 4.8% of women required readmission within 90 days. No difference was observed in average length of stay for hospitalizations associated with subsequent readmission compared to those without subsequent re-hospitalizations. Over the ten years, a total of 18 deaths occurred among women hospitalized for PID or tuboovarian abscess (mortality: 0.006%), and two deaths occurred during readmission hospitalizations for PID.
We found that hospitalization rates in California for PID and tuboovarian abscess have declined from 1991 to 2001. The steepest decline was in women 20–39 years of age. In contrast, PID hospitalization rates for women 40 years of age or older remained stable. Although black women between 20 and 39 years of age had the highest PID hospitalization rates, they had the lowest proportion of hospitalizations that included hysterectomy or oophorectomy compared with other racial/ethnic groups.
The decline in hospitalizations for PID was twice that for tuboovarian abscess. One possible explanation for this finding is that PID cases treated in the ambulatory setting have increased over the years, whereas tuboovarian abscess remains one of the Centers for Disease Control (CDC) criteria for hospitalization.5 The 2003 CDC Surveillance Report comments that, in the United States, hospitalizations for PID have remained relatively constant between 1995 and 2002.7 Similarly, an earlier study conducted by the CDC using data from the Hospital Discharge Survey and the Commission on Professional and Hospital Activities found that the trend in PID hospitalizations appeared stable between 1970 and 1975.8 From 1979 to 1988, the most stable hospitalization rate for PID was found in the 15- to 19-year age group.9 In contrast, our findings show a clear decline in PID hospitalizations, especially in women 20–39 years of age in California. One reason for the difference between our data and that of the CDC could be the differences in the databases analyzed. The CDC used the National Hospital Discharge Survey, which contains a sample of hospital discharges across the whole United States, whereas the California Discharge Database contains all of the hospital discharges in the state of California. Why there might be regional differences in PID hospitalization rates is unclear, but one explanation for the decline in PID hospitalization rates in California could be the influence of managed care organizations and the effect of economic pressures by insurance companies in California. However, we find no evidence in the literature to support this speculation. Furthermore, not all patients in the California Discharge Database are managed care patients, and therefore, we do not believe that the managed care environment in California significantly affected the results of our study.
In our study, the most stable PID hospitalization rates were noted in women 40 years of age or older. Indeed, we found that the tuboovarian abscess hospitalization rate actually increased by 25% in this older group of women, while tuboovarian abscess hospitalization rates declined in all other age categories. Tuboovarian abscesses do occur most commonly in the third and fourth decades of life.10 The increasing hospitalization rates in older women may be explained by recurrent infections resulting from increasing outpatient management in the younger age groups and a higher incidence of other diagnoses that may confuse the clinical picture, such as diverticulitis and ovarian neoplasms.11 A recent Scandinavian study posits that the unchanged number of tuboovarian abscess cases over a 10-year period may represent a different etiology for tuboovarian abscess than for acute PID.6 A study from Israel theorizes that there is probably a new trend in the epidemiology of tuboovarian abscess occurring in older women.12 Our findings represent important epidemiologic trends in tuboovarian abscess incidence that require further investigation.
Tuboovarian abscesses can complicate PID in up to 34% of women requiring hospitalization for acute PID.13 Therapy for these abscesses initially consists of antibiotic treatment, which results in clinical response rates between 35% and 87%, depending on the regimen used.14 If the patient does not respond to medical therapy within 48–72 hours, surgical procedures are recommended, including hysterectomy, oophorectomy, and percutaneous drainage procedures.15 Appropriately, we found that women 40 years of age or older, who are less likely to desire fertility preservation, were more likely to have a hysterectomy. Several investigators have reported good results from oophorectomy alone in patients with unilateral tuboovarian abscesses.3,16 Although there is a risk that a hysterectomy will be required, oophorectomy does offer the patient with a tuboovarian abscess an alternative to the complete removal of all reproductive organs. Furthermore, there has been increasing interest in interventional radiological approaches to tuboovarian abscess management in an attempt to decrease morbidity, length of hospital stay, and cost. Currently, ultrasound-guided percutaneous drainage of tuboovarian abscesses can be successful in 95% of patients.17 Our study shows that the proportion of hospitalizations that included oophorectomy and drainage procedures has increased steadily from 1991 to 2001. Our results suggest that these newer recommendations are making their way into clinical practice.
Interestingly, although black women had higher hospitalization rates than other racial/ethnic groups, they had the lowest proportion of hospitalizations that included hysterectomy. Our initial concern was that this finding reflected higher rates of hospitalization in young black women, but when we compared black women with white, Hispanic, or Asian women of similar ages, a significantly lower proportion of black women received hysterectomies. Our finding that black women were the least likely to have a hysterectomy or oophorectomy during hospitalization for PID or tuboovarian abscess could have several explanations. Perhaps, black women were not as likely to consent to these procedures because of differing attitudes toward hysterectomy or oophorectomy.18 Another possible reason is that these procedures were not indicated because black women hospitalized for PID or tuboovarian abscess either responded better to medical therapy or were not sick enough to require surgical intervention, perhaps because of a lower threshold for hospitalization.
Our study has several limitations. The first is that, because we use the California Patient Discharge Database, the accuracy of our estimates depends on the accuracy and completeness of ICD-9-CM coding for diagnoses and procedures, the collection of demographic characteristics listed on discharge summaries, and accuracy of medical record documentation. Although there are no studies validating the coding accuracy for gynecological conditions, a recent study assessed the validity of obstetric complications in the California Patient Discharge Database and found that third- and fourth-degree perineal lacerations were reported accurately, with estimated sensitivities and positive predictive values exceeding 90%. Most postpartum complications were reported with 70% sensitivity but at least 80% positive predictive value.19 In addition, there was a validation study within the California Hospital Outcomes Project, which was an initiative mandated by the State of California and conducted by the Office of Statewide Health Planning and Development, to develop public reports comparing hospital outcomes for selected medical conditions and surgical procedures for patients treated in hospitals throughout California. This study found that, when community-acquired pneumonia was the coded discharge diagnosis in the California Patient Discharge Database, 90.5% of cases met definite or possible clinical and radiographic criteria for community-acquired pneumonia.20 Because PID is most often a clinical diagnosis without definitive laboratory or radiological findings, the coding accuracy may be reduced. On the other hand, similar to community-acquired pneumonia, tuboovarian abscesses are most often diagnosed radiographically, likely increasing coding accuracy.
In our study, only hospitalizations and procedures linked with a primary diagnosis of PID or a primary diagnosis of tuboovarian abscess were included in our analysis because we did not wish to include hospitalizations in which these diagnoses were not firmly established. This decision can be considered a strength of our study because it is unlikely that the hospitalizations and procedures were done for any condition other than PID or tuboovarian abscess. However, for this same reason, hospitalization rates may be underestimated. In addition, because the database does not include patients with PID or tuboovarian abscess treated in the ambulatory setting, our study does not estimate the true incidence of these conditions.
The findings for PID could be explained by increasing outpatient management of this infectious disease. The proportion of hospitalizations that included oophorectomy and drainage procedures has increased from 1991 to 2001, which reflects the increasing use of conservative surgical approaches, especially in younger women. Furthermore, in the past, patients with tuboovarian abscess were either hospitalized for a prolonged period on intravenous antibiotics or were surgically extirpated, but our study suggests that the management of tuboovarian abscess in California reflects recent evidence-based practice guidelines. Our finding that black women had the lowest percentage of hospitalizations for PID or tuboovarian abscess that included hysterectomy or oophorectomy needs further investigation.
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;86:764–9.
2. 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.
3. Landers DV, Sweet RL. Tubo-ovarian abscess: contemporary approach to management. Rev Infect Dis 1983;5:876–84.
4. Washington AE, Katz P. Cost of and payment source for pelvic inflammatory disease: trends and projections, 1983 through 2000. JAMA 1991;266:2565–9.
5. 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.
6. Sorbye IK, Jerve F, Staff AC. Reduction in hospitalized women with pelvic inflammatory disease in Oslo over the past decade. Acta Obstet Gynecol Scand 2005;84:290–6.
7. Centers for Disease Control and Prevention. STD Surveillance 2003. Available at: http://www.cdc.gov/std/stats
/. Retrieved October 31, 2005.
8. St John RK, Jones OG, Blount JH, Zaidi AA. Pelvic inflammatory disease in the United States: epidemiology and trends among hospitalized women. Sex Transm Dis 1981;8:62–6.
9. Rolfs RT, Galaid EI, Zaidi AA. Pelvic inflammatory disease: trends in hospitalizations and office visits, 1979 through 1988. Am J Obstet Gynecol 1992;166:983–90.
10. Franklin EW 3rd, Hevron JE Jr, Thompson JD. Management of the pelvic abscess. Clin Obstet Gynecol 1973;16:66–79.
11. Gorbach SL, Bartlett JG. Anaerobic infections. 1. N Engl J Med 1974;290:1177–84.
12. Halperin R, Levinson O, Yaron M, Bukovsky I, Schneider D. Tubo-ovarian abscess in older women: is the woman’s age a risk factor for failed response to conservative treatment? Gynecol Obstet Invest 2003;55:211–5.
13. Ledger WJ. A historical review of pelvic infections. Am J Obstet Gynecol. 1988;158 (3 Pt 2):687–93.
14. McNeeley SG, Hendrix SL, Mazzoni MM, Kmak DC, Ransom SB. Medically sound, cost-effective treatment for pelvic inflammatory disease and tuboovarian abscess. Am J Obstet Gynecol 1998;178:1272–8.
15. Fabiszewski NL, Sumkin JH, Johns CM. Contemporary radiologic percutaneous abscess drainage in the pelvis. Clin Obstet Gynecol 1993;36:445–56.
16. Golde SH, Israel R, Ledger WJ. Unilateral tuboovarian abscess: a distinct entity. Am J Obstet Gynecol 1977;127:807–10.
17. Worthen NJ, Gunning JE. Percutaneous drainage of pelvic abscesses: management of the tubo-ovarian abscess. J Ultrasound Med 1986;5:551–6.
18. Shelton AJ, Lees E, Groff JY. Hysterectomy: beliefs and attitudes expressed by African-American women. Ethn Dis 2001;11:732–40.
© 2006 The American College of Obstetricians and Gynecologists
19. Romano PS, Yasmeen S, Schembri ME, Keyzer JM, Gilbert WM. Coding of perineal lacerations and other complications of obstetric care in hospital discharge data. Obstet Gynecol 2005;106:717–25.