Volume 6 - Number 2 1997
Special Issue: STDs and Sexual/Reproductive Health

Published by SIECCAN
The Sex Information & Education Council of Canada

The Effects of Undetected and Untreated Sexually Transmitted Diseases: Pelvic Inflammatory Disease and Ectopic Pregnancy in Canada

Noni E. MacDonald
Professor, Paediatrics & Microbiology
University of Ottawa
Children's Hospital of Eastern Ontario
Ottawa, Ontario

Robert Brunham
Professor and Head
Department of Medical Microbiology
J.C. Wilt Medical Microbiology Laboratories
University of Manitoba
Winnipeg, Manitoba

Abstract:

Pelvic Inflammatory Disease (PID) resulting from undetected and untreated sexually transmitted disease (STD) is a major cause of infertility and ectopic pregnancy. This paper examines the relationships between STDs and PID, and between PID and reproductive sequelae, such as ectopic pregnancy and tubal infertility. The diagnosis, polymicrobial, pathogenesis, and risk factors of PID are described. The incidence/prevalence of PID in Canada is discussed and control strategies are proposed.

Key words:

• Pelvic Inflammatory  Disease, Sexually transmitted disease, Ectopic pregnancy, Tubal infertility, Chlamydia trachomatis

Correspondence concerning this paper should be addressed to Dr. Noni MacDonald, Professor, Microbiology and Paediatrics, 401 Smyth Rd., Ottawa, Ontario, K1H 8L1. Tel: (613) 737-2651; Fax: (613) 738-4832.

Introduction

The Report of the Royal Commission on New Reproductive Technologies (1993) emphasized the relationship between infertility, pelvic inflammatory disease (PID) and sexually transmitted diseases (STDs), and placed a high priority on prevention and treatment of STDs and PID to minimize the prevalence of infertility. The documentation in support of this view indicated that one third to one half of women who acquire an STD will develop PID, that this group of women represents about 80% of all cases of PID (Goeree & Gully, 1993), and that current or prior PID is associated with a sizeable percentage of cases of tubal infertility and ectopic pregnancy (Rootman, 1992; Westrom, 1980). This paper examines the relationships between STDs and PID, and between PID and reproductive sequelae such as ectopic pregnancy and tubal infertility. We identify goals for reducing the prevalence of these sequelae through interrelated strategies for prevention, diagnosis, and treatment of STDs and PID.

Pelvic Inflammatory Disease

Definition

Pelvic inflammatory disease (PID) is a serious syndrome of the female reproductive system which results from the spread of infections (most often sexually transmitted infections such as Chlamydia trachomatis and Nisseria gonnorrhoea), from the vagina and endocervix to the uterus, fallopian tubes and ovaries. PID commonly manifests as endometritis (infection of the lining of the uterus) or salpingitis (infection of the fallopian tubes) and less commonly as pelvic peritonitis and/or inflammation of contiguous structures (MacDonald & Bowie, 1995; Westrom & Mardh, 1990).

Diagnosis

PID is best understood as a syndrome; it has a variety of symptoms that can vary in occurrence and severity including: lower abdominal pain; dysuria; urinary frequency; purulent endocervical discharge; dyspareunia; adnexal tenderness (pressure sensitivity of tissues adjacent to the uterus); cervical motion tenderness; and fever (Table 1). Clinical diagnosis of PID is difficult because of the wide variation in symptoms and signs, and the high rate of asymptomatic infection, particularly among adolescents (Paradise & Grant, 1992). In the classic study by Jacobson and Westrom (1969), only 65% of 814 women with clinically diagnosed PID had infection confirmed by laparoscopy, 23% had no observable abnormality, and 12% had other findings such as ectopic pregnancy, endometriosis, and appendicitis. Unfortunately, no single or combined set of symptoms, signs, or laboratory test results is highly sensitive and specific for PID (Kahn, Walker, Washington, Landers, & Sweet, 1991).

Given the present limitations of clinical evaluation, laparoscopy is the only tool that can provide a definitive diagnosis. Unfortunately, this is a surgical procedure and is therefore not a useful tool for routine diagnosis. Despite the low sensitivity and specificity of clinical examination for detection of PID, the importance of detecting even mild cases has led to the identification of accepted minimum criteria for clinical diagnosis (MacDonald & Bowie, 1995; MMWR, 1991) (see Table 2). These minimum criteria detect only about 60% of laparoscopically proven cases (Westrom & Mardh, 1990). On examination, many patients with mild PID do not have a fever nor do they show evidence of mucopurulent cervicitis. While additional criteria such as fever, cervicitis, elevated erythrocyte sedimentation rate, and laboratory documentation of cervical infection with N. gonorrhoea or C. trachomatis increase the specificity of the diagnosis (see Table 2), they also decrease the sensitivity. Only 16% of women with PID manifest both minimum and routine criteria (Westrom & Mardh, 1990). Unfortunately, the consequences of both symptomatic and asymptomatic "silent" PID can be long-term and severe. At least 1 in 4 women with PID experience one or more serious sequelae including ectopic pregnancy, infertility, recurrent infection, chronic abdominal pain or tuboovarian abscess (Paradise & Grant, 1992; Buchan, Vessey, Goldacre & Fairweather, 1992).

Microbial Etiology and PID

Recent studies and reviews of PID have emphasized the polymicrobial etiology of this syndrome (Cates, Wasserheit & Marchbanks, 1994; Rice & Schachter, 1991). Laparoscopic findings indicate that there are two major groups of pathogens: (1) the STD organisms N. gonorrhoea and C. trachomatis; and (2) the varied aerobic and anaerobic organisms that are endogenous to the lower genital tract. Recent studies have also shown that the organisms associated with bacterial vaginosis are similar to the nongonococcal, nonchlamydial bacteria frequently isolated from the upper genital tract in women with PID (Sweet, 1995). In North America, chlamydia has been isolated in up to 51% of PID cases (Bowie & Jones, 1981). N. gonorrhoea, while a prominent pathogen in some populations with PID (e.g., urban American blacks), is much less common in others (e.g., Scandinavian women) (Mardh, Moller & Paavonen, 1981). Although the growing list of possible causative agents has implications for the treatment regimes used with both inpatients and outpatients, Canada does not yet have broadly based multi-centred studies on the etiology of PID in our population.

Pathogenesis

The precise mechanism by which the infectious organisms reach the uterus and fallopian tubes is not known, although ascending infection is thought to be an important component (Westrom & Wolner-Hansen, 1993). In populations that have been carefully studied, 30-40% of women with untreated gonorrhoea or chlamydial cervicitis developed clinical symptoms of acute PID (Cates, Wasserheit & Marchbanks, 1994). A unifying hypothesis suggested by Cates (1994) proposes that acute polymicrobial PID begins with a cervical infection with C. trachomatis or N. gonorrhoea. Rapid growth of these cervical pathogens changes the microenvironment of the lower genital tract (e.g., altered pH and oxygen availability) which, in turn, permits overgrowth of the organisms that are normally present there. If immune defences are low, these organisms, along with the original infectious agent, can ascend along contiguous mucosal surfaces into the uterus and fallopian tubes.

Risk Factors

Many of the studies that have examined risk factors associated with PID have failed to distinguish the intervening risk factors for lower tract STD (Cates et al., 1994). This is an important consideration, since risk factors for STD also influence risk for PID (Cates, Rolfs & Aral, 1990). The major risk factors associated with PID are shown in Table 3. Sexually active adolescents are three times more likely to be diagnosed with PID than are 25-29 year old women (Bell & Holmes, 1984). Use of barrier methods of contraception decreases this risk, while IUD use is associated with increased risk, especially in the first month after insertion of the IUD (Cates et al., 1990; Farley, Rosenberg, Rowe, Chen, & Meirik, 1992). On average, women with a past history of PID — and particularly of gonorrhoea-associated PID — have a higher rate of infection than those without a history of PID. Studies in the United States have suggested that race may be a risk factor in that a higher proportion of African-American women have PID; however, in this assessment, race may be confounded by such factors as the higher gonorrhoea rates in this population and/or lesser access to appropriate health care. Although comparable Canadian ethnocultural or sociocultural data on the prevalence of PID are not available, it seems likely that populations with known high rates of STD (e.g., street youth and some First Nations communities) might also have above average rates of PID.

Douching has also been associated with acute PID (Wolner-Hansen et al., 1990), perhaps because it alters the growth milieu of the vaginal environment and/or flushes the infective agent up into the uterus. Cigarette smoking has also been implicated as a risk factor (Marchbanks, Lee & Peterson, 1990), but this risk factor could be an indirect effect of some associated aspect of lifestyle or of alteration of the immune system or other such influence (Cates et al., 1994).

Magnitude of PID in Canada

Since PID is not a notifiable disease in Canada and is inherently difficult to diagnose, it is hard to determine the incidence or prevalence of PID. A study in Calgary found that 6.5% of women aged 14-50 presenting to physicians for contraceptive advice had a history of PID. Based on a self-report study in the U.S. (Aral, Mosher & Cates, 1991), 1 in 10 women said they had had PID during their reproductive years. Hospital separation data are currently the most reliable sources regarding the incidence and prevalence of PID, although these data are skewed towards more serious, symptomatic PID (i.e., those who are sufficiently affected to go to hospital). However, only 10-15% of all PID cases in Canada are thought to be treated in hospital (Todd, 1986), and even this estimate may be "generous" given the recent sharp decrease in hospital beds and the attendant pressure to move care to the outpatient setting. In the U.S., for example, Rolfs, Galaid and Zaidi (1992) reported that hospitalization for PID decreased between 1979 and 1988. Possible reasons for the decrease included: changes in practice patterns with respect to hospitalization; a relative decrease in the prevalence of gonorrhea-associated PID; and a relative increase in chlamydia-associated PID which is clinically less severe.

The Laboratory Centre for Disease Control STD Division has recently reviewed the Statistics Canada hospital morbidity data for the fiscal years 1983/84 through 1993/94 to determine the number of hospitalizations for PID by age grouping (Table 4) and by province (Table 5). The Canadian Diagnostic Short Code 137 was used to obtain the data. This corresponds to ICD-9 Code 614 which records "inflammatory diseases of ovary, fallopian tube, and pelvic peritoneum". Since the number of hospitalizations for PID reflects the number of hospital discharges, the database may include several admissions in one year for some patients. Incidence rates per 100,000 females were calculated by age group (Table 4) and province (Table 5) using by selected age groups and by province using population data supplied by Statistics Canada.

In the 11 years from 1983/84 to 1993/94, records indicate that about 2.3% of Canadian women aged 15-44 experienced PID of sufficient severity to warrant hospitalization. Overall, both the number of hospitalizations for PID and the rate of PID declined appreciably from 1983/84 to 1993/94 with the most notable age-specific declines occurring through the late 1980s and early 1990s, and primarily in the younger age groups. For example, the age-specific rate for 15-19 year olds and 20-24 year olds dropped by 50% from 1988/89 to 1993/94 (Table 4). The decline in hospitalizations for PID in Canada since the mid- to late 1980s is similar to changes seen in Holland (Coutinho, Rijsdijk, Van den Hoek & Leenrvaar-Kuijpers, 1992), Sweden (Westrom, 1988) and the United States (Rolfs et al., 1992). This recent trend differs from the pattern of change reported by Jessamine and Todd (1984) for the period from 1971 to 1980 during which PID-related hospitalizations increased in all age groups and most heavily among 15-19 year olds (51% increase) and 20-24 year olds (35% increase).

A study done in Manitoba over the period from 1981-1990 (Orr et al., 1994) considered outpatient PID cases as well as hospitalizations, and showed that outpatient visits to physicians over the 10-year period decreased 33% (from 695 to 463) while hospitalizations decreased by 48% (from 139 to 71). This encouraging decline in PID in Manitoba may be a reflection of improved detection and treatment of chlamydia infections subsequent to a province-wide control program in 1987 (Orr et al., 1994). A 33% decline in PID in Wisconsin has been similarly attributed to a comprehensive chlamydia prevention program (Hillis et al., 1995). Indeed the impact of the chlamydia control programs in Manitoba and Wisconsin may have had a greater impact on PID than was measured, since neither study would have been able to assess the incidence or reduction of "silent" PID (which might have been considerable given that chlamydia PID is usually less severely symptomatic than that associated with gonococcal PID). The cost-effectiveness of such STD control programs warrants attention in light of the estimated $34 million annual direct and indirect cost of hospitalizations for gonococcal ($13 million) and chlamydia PID ($21 million) (personal communication, Laboratory Centre for Disease Control).

Sequelae of PID

Ectopic Pregnancy

Implantation of a fertilized ovum in the fallopian tube, or less commonly in the abdominal cavity, results in ectopic pregnancy. Ciliary dysfunction within the fallopian tube is the proximate cause of ectopic pregnancy, and prior tubal infection with N. gonorrhoea and/or C. trachomatis can result in selective loss of ciliated epithelial cells from the fallopian tubes. It is estimated that prior tubal infection with STD agents causes about 50% of the cases of ectopic pregnancy, although this rate will be affected by the incidence rates of chlamydia and gonococcal infection in a community. As well, because ectopic pregnancy is a delayed sequela of tubal influence, the proportion of ectopic pregnancy attributable to STDs likely reflects STD incidence rates 5-10 years prior to case recognition.

Currently in Canada, as in many other developed countries, ectopic pregnancy rates are increasing (Table 6). This trend is surprising given the declining rates of gonorrhea and, in some jurisdictions, chlamydia. The explanation for the widening gap between ectopic pregnancy rates and STD rates is likely multifactorial. Since ectopic pregnancy has multiple causes, non-STD-related ectopic pregnancy rates may be increasing, while those cases due to STDs may be declining. Older maternal age, tubal ligation, prior tubal surgery, and IUD use may all play a causal role in non-STD-related ectopic pregnancy, and important changes in these risk factors are occurring among pregnancy-seeking/vulnerable women.

As well, since ectopic pregnancy is a late sequela of STD, a lag period between STD control and declining ectopic pregnancy rates may be occurring. In Finland, ectopic pregnancy rates only began to decline 5 to 10 years after the introduction of a chlamydia control program. In Manitoba, ectopic pregnancy rates have not declined despite 3 years of intensive chlamydia control (Table 7). Further research into the epidemiology of "cause-specific" ectopic pregnancy and its relationship to control of chlamydia and gonococcal infection is required.

Tubal Infertility

Scarring of the fallopian tubes associated with STD-related PID frequently leads to tubal infertility. The longer the delay between the onset of PID and treatment, the greater the likelihood of tubal infertility, a problem compounded by the relatively asymptomatic nature of many PID cases. Currently, we do not have enough data to accurately specify the exact proportion of female infertility that is the result of PID. However, a number of serologic studies suggest that at least 64% of cases of tubal infertility are linked to chlamydia infections (e.g., Gump, Gibson, & Ashikaga, 1983; Kosseim & Brunham, 1986). In addition, the Royal Commission on New Reproductive Technologies (1993) estimated that at least 20% of all infertility among Canadian couples is attributable to damage to the female partner's fallopian tubes which is a direct result of PID caused by STD infection.

Control of PID

PID can be viewed as an intermediary between acute STD and chronic problems such as tubal infertility, chronic abdominal pain and ectopic pregnancy. This relationship is influenced by an interaction of three environments: (1) the genital microbial environment; (2) the individual behavioural environment; and (3) the socio-geographic environment. Opportunities for control of PID occur in each. A nationwide aggressive chlamydia control program aimed at early diagnosis and treatment, especially in high risk target groups, may reap large benefits by decreasing chlamydia associated PID and its sequelae. Behavioural intervention programs that increase barrier contraceptive use, and frequent testing for silent chlamydia infection in high risk populations such as street youth, have the potential to decrease PID. To be most effective, PID control programs will need to include targeted programs that change the social and geographic factors which influence spread of STD within the population (e.g., survival sex among street youth and the urban homeless). Control will also require improved treatment of PID, both inpatient and outpatient, in order to minimize inflammatory damage and to facilitate early detection of a recurrence. Further research to determine the optimal therapeutic regimen of antibiotics and anti-inflammatories is required.

In order to determine the breadth and extent of this STD problem, surveillance data on gonococcal and chlamydia infection by age, gender, region, and risk group are needed, as are PID rates and ectopic pregnancy rates by age and region. To improve the quality of the estimates for PID, both hospital separation and outpatient billable diagnosis codes need to be used. The changes in hospital practice for admission with PID suggest that further health economic data are needed to clarify the economic burden of PID when outpatient treatment costs are included.

Multi-level interventions will be needed to disrupt the pattern that links STDs, PID and reproductive sequelae such as ectopic pregnancy and tubal infertility. Table 8 identifies achievable goals and suggests strategies for achieving them.

Table 1 - Symptoms and Signs in 623 Patients with Laparoscopically Proven Pelvic Inflammatory Disease*

Table 2 - Criteria for Clinical Diagnosis of Pelvic Inflammatory Disease*

Table 3 - Risk factors for Pelvic Inflammatory Disease

Table 4 - Changes in Number and Rate of Hospital Treated PID Cases in Canada, 1983/84-1993/94

Table 5 - PID in Canada, 1983/84 to 1993/94: Number of Cases and Rates of Hospital Separations by Province

Table 6 - Ectopic Pregnancy Trends, Canada 1988-1993

Table 7 - Annual Rates of Hospitalization and Outpatient Physician Visits for Pelvic Inflammatory Disease and Ectopic Pregnancy in Manitoba, Canada (1981-1990)

Table 8 - Suggested Goals for Reduction of PID and Ectopic Pregnancy in Canada

Figure 1

Etiological relationships among STD, PID and Tubal Infertility ^*

* Source: Cates, W. Jr. et al., 1994.

References

• Aral, S.O., Mosher, W.D., & Cates, W., Jr. (1991). Self-reported pelvic inflammatory disease in the United States, 1988. Journal of the American Medical Association, 266, 2570-2573.

• Bell, T.A., Holmes, K.K. (1984) Age specific risks of syphilis, gonorrhea, and hospitalized pelvic inflammatory disease in sexually experienced US women. Sexually Transmitted Diseases, 11, 291-295.

• Bowie, W.R., Jones, H. (1981). Acute pelvic inflammatory disease in outpatients: Association with C. trachomatis and N. gonorrhoea. Annals of Internal Medicine, 95, 685-688.

• Buchan, H., Vessey, M., Goldacre, M., & Fairweather, J. (1992). Morbidity following pelvic inflammatory disease. British Journal of Obstetrics & Gynaecology, 100, 558-562,

• Cates, W.Jr., Wasserheit, J.N., & Marchbanks, P.A. (1994). Pelvic inflammatory disease and tubal infertility: The preventable condition. Annals of the New York Academy of Sciences, 709, 179-195.

• Cates, W.Jr., Rolfs. R.T., & Aral, S.O. (1990). Sexually transmitted diseases, pelvic inflammatory disease, and infertility: An epidemiology update. Epidemiological Review, 12, 199-200.

• Coutinho, R.A., Rijsdijk, A.J, Van den Hoek, J.A.R., & Leenrvaar-Kuijpers, A. (1992). Decreasing incidence of pelvic inflammatory disease in Amsterdam. Genitourinary Medicine, 68, 253-255.

• Farley, T.M.M., Rosenberg, M.J., Rowe, P.J., Chen, J-H. & Meirik, O. (1992). Intrauterine devices and pelvic inflammatory disease: An international perspective. Lancet, 339, 785-788.

• Goeree, R. & Gully, P. (1993). The burden of chlamydial and gonococcal infection in Canada. In Research Volumes of the Royal Commission on New Reproductive Technologies 1993. Ottawa: Minister of Government Services.

• Gump, D., Gibson, M., & Ashikaga, T. (1983). Evidence of prior pelvic inflammatory disease and its relationship to Chlamydia trachomatis antibody and intrauterine contraceptive device use in infertile women. American Journal of Obstetrics and Gynecology, 153, 153-159.

• Hillis, S.D., Nakashima, A., Amsterdam, L., Pfister, J., Vaughn, M., Addiss, D., Marchbanks, P.A., Owens, L.M., & Davis, J.P. (1995). The impact of a comprehensive chlamydia prevention program in Wisconsin. Family Planning Perspectives, 27, 108-111.

• Jacobson, L. & Westrom, L. (1969) Objectivised diagnosis of acute PID. American Journal of Obstetrics and Gynecology, 105, 1088-1098.

• Jessamine, G. & Todd, M. (1984). Trends in hospitalization for pelvic inflammatory disease, Canada 1971-1980. International Conjoint STD Meeting, Montreal, 1984.

• Kahn, J.G., Walker, C.K., Washington. E., Landers, D.V. & Sweet, R.L. (1991). Diagnosed pelvic inflammatory disease: a comprehensive analysis and consideration for developing a new model. Journal of the American Medical Association, 266, 2594-2604.

• Kosseim, M., & Brunham, R. (1986). Fallopian tube obstruction as a sequela to Chlamydia trachomatis. European Journal of Clinical Microbiology, 5, 584-590.

• MacDonald, N.E., & Bowie, W.R. (1995). 1995 Update Canadian Guidelines for the Prevention, Diagnosis, Management, and Treatment of Sexually Transmitted Disease in Neonates, Children, Adolescents, and Adults. Canadian Communicable Disease Report, 21S4, 25-33.

• Marchbanks, P.A., Lee, N.C., & Peterson, H.B. (1990). Cigarette smoking as a risk factor for pelvic inflammatory disease. American Journal of Obstetrics & Gynecology, 162, 639-644.

• Mardh, P-A., Moller, B.R., & Paavonen, J. (1981) Chlamydial infection of the female genital tract with emphasis on pelvic inflammatory disease: A review of Scandinavian studies. Sexually Transmitted Diseases, 8, 140-155.

• Mortality, Morbidity Weekly Report. (1991) Pelvic inflammatory disease: Guidelines for prevention and management. MMWR, 40(RR-5), 1-25.

• Orr, P., Sherman, E., Blanchard, J., Fast, M., Hammond, G. & Brunham, R. (1994). Epidemiology of infection due to Chlamydia trachomatis in Manitoba, Canada. Clinical Infectious Diseases, 19, 876-883.

• Paradise, J.E., & Grant, L. (1992) Pelvic inflammatory disease in adolescents. Pediatric Review, 13, 216-223.

• Rice, P.A., & Schachter, J. (1991). Pathogenesis of pelvic inflammatory disease: What are the questions? Journal of the American Medical Association, 266, 2587-2593.

• Rolfs, R.T., Galaid, E.I., & Zaidi, A.A. (9912). Pelvic inflammatory disease: Trends in hospitalization and office visits, 1979 through 1988. American Journal of Obstetrics & Gynecology, 166, 983-990.

• Rootman, I. (1992). Prevention of infertility: Literature of applicable theoretical models. Report produced for the Royal Commission on New Reproductive technologies, January, 1992.

• Royal Commission on New Reproductive Technologies. (1993). Proceed with care: Final report of the Royal Commission on New Reproductive Technologies. Ottawa: Minister of Government Services.

• Shafer, M., & Sweet, R.L. (1990). Pelvic inflammatory disease in adolescent females. Adolescent Medicine. State of the Art Reviews, 1, 545-564.

• Sweet, R.L. (1995). Role of bacterial vaginosis in pelvic inflammatory disease. Clinical Infectious Diseases, 20(S2), S271-275.

• Todd, M.J. (1986). Trends in hospitalization for pelvic inflammatory disease in Canada 1971-81. Canadian Medical Association Journal, 134, 1157-1158.

• Westrom, L. (1980) Incidence, prevalence, and trends of acute pelvic inflammatory disease and its consequences in industrialized countries. American Journal of Obstetrics and gynecology, 138, 880-892.

• Westrom, L. & Wolner-Hansen, P. (1993). Pathogenesis of pelvic inflammatory disease. Genitourinary Medicine, 69, 9-17.

• Westrom, L., & Mardh, P-A. (1990). Acute pelvic inflammatory disease (PID). In Holmes, K.K., Mardh, P-A., & Sparling, P.F, (Eds.), Sexually Transmitted Diseases, 2nd edition (pp.593-613), New York: McGraw-Hill.

• Westrom, L. (1988). Decrease in incidence of women treated in hospital for acute salpingitis in Sweden. Genitourinary Medicine, 64, 59-63.

• Wolner-Hanssen, P., Eschenback, D.A., Paavonen, J., Kiviat, N., Stevens, C.E., Critchlow, C., DeRouen, T., & Holmes, K.K. (1990). Decreased risk of symptomatic chlamydial pelvic inflammatory disease associated with oral contraceptive use. Journal of the American Medical Association, 263, 54-59.

[Previous] [Table of Contents]

Last Updated: November 26, 1997