eMedicine Specialties > Obstetrics and Gynecology > Infections

Pelvic Inflammatory Disease

Author: Suzanne Moore Shepherd, MD, MS, DTM&H, FACEP, FAAEM, Associate Professor, Department of Emergency Medicine, Hospital of the University of Pennsylvania; Director of Education and Research, PENN Travel Medicine
Contributor Information and Disclosures

Updated: Feb 4, 2010

Introduction

Background

Pelvic inflammatory disease (PID) is an infectious and inflammatory disorder of the upper female reproductive tract, including the uterus, fallopian tubes, and adjacent pelvic structures. It is a common and serious disorder, initiated by infection, that ascends from the vagina and cervix. PID may produce tubo-ovarian abscess and extend to produce pelvic peritonitis and Fitz-Hugh-Curtis syndrome (perihepatitis) (see image below).

Pathophysiology

Most cases of PID occur in 2 stages. The first stage involves acquisition of a vaginal or cervical infection. The original sexually transmitted infection (STI) may be asymptomatic. An estimated 10-20% of untreated chlamydial or gonorrheal infections progress to PID.

In PID, the second stage of upper female genital tract infection occurs by direct ascent of micro-organisms from the vagina and cervix. Although the exact mechanism of ascent is unknown, studies have suggested that a number of factors may be involved. Cervical mucus provides a functional barrier against upward spread; however, the efficacy of this mechanism may be decreased by hormonal changes that occur during ovulation and menstruation. Alterations in the cervicovaginal microenvironment may result from antibiotic treatment and STIs that can disrupt the balance of endogenous flora, causing normally nonpathogenic organisms to overgrow and ascend. Opening of the cervix during menstruation with retrograde menstrual flow may facilitate ascent of micro-organisms. Intercourse may contribute to the ascent of infection due to rhythmic mechanical uterine contractions. Bacteria may be carried along with sperm into the uterus and tubes.

Risk factors for PID include multiple sexual partners, a history of prior STIs, and a history of sexual abuse.³ Frequent vaginal douching has also been implicated.^4,5 Younger age has been found to be associated with increased risk, suggested to be due to some combination of increased cervical mucosal permeability, a larger zone of cervical ectopy, a lower prevalence of protective chlamydial antibodies, and increased risk-taking behaviors. Surgical procedures, such as endometrial biopsy, curettage, and hysteroscopies break the cervical barrier, predisposing women to ascending infections.

The organisms most commonly isolated in many, if not most, cases of acute PID are Neisseria gonorrhoeae and Chlamydia trachomatis. As such, therapy has been directed primarily against these organisms. However, unlike infections in other areas of the body, newer studies obtaining more sensitive and specific laparoscopic cultures have found acute PID to be polymicrobial in up to 30-40% of cases. Organisms involved include Gardnerella vaginalis, Mycoplasma hominis, Ureaplasma urealyticum, herpes simplex virus-2 (HSV-2), Trichomonas vaginalis, cytomegalovirus, Haemophilus influenzae, Streptococcus agalactiae, enteric gram-negative rods, and anaerobes. N gonorrhoeae and C trachomatis may be instrumental in the initial infection of the upper tract, with anaerobes, facultative anaerobes, and other bacteria increasingly isolated as inflammation increases and abscesses form.

The microbiology of PID has also been found to reflect both the predominant STIs prevalent within a specific population and also less common organisms seen in specific populations. Bacterial vaginosis (BV) is suggested to play a role in the initiation of ascending infection in a subset of women with heavy growth of BV-associated organisms, such as G vaginalis, more than 2 recent sexual partners, and especially after recent abortion or gynecologic surgery.^6,7

A 2006 cross-sectional study examined the role of HSV-2 and T vaginalis in PID. In this study of 736 women, those with T vaginalis demonstrated a 4-fold increase in the histologic evidence of acute endometritis. Co-infection of HSV-2 with N gonorrhoeae, C trachomatis, and BV was also associated with histologic evidence of acute endometritis. HSV-2 was demonstrated to be associated with fallopian tube inflammation and lower tract ulcerations that may contribute to disruption of the endocervical canal mucus barrier.⁸

HIV infection has been found to be associated with an increased incidence of C trachomatis infection, Candida, and human papillomavirus. Those with HIV infection also have an increased risk of progression to PID and tubo-ovarian abscess.⁹

Actinomycete species have been identified almost exclusively in those patients with IUDs.²

PID may result from Mycobacterium tuberculosis in endemic areas.¹⁰

Different forms of contraception may affect PID incidence and severity. Appropriately used barrier contraception has clearly been shown to decrease the acquisition of most STIs. The CDC recommends that spermicides and condoms containing nonoxynol-9 should be avoided, as a number of African studies have demonstrated that nonoxynol-9 can cause vaginal lesions and may increase the risk of HIV transmission. While the level of nonoxynol-9 in condoms is lower than the level associated with vaginal lesions, these are also not recommended because they are more expensive, have a shorter shelf life, and have been associated with urinary tract infections.

Oral contraceptive pills (OCPs) have been found to have differing effects on PID risks. Oral contraceptives are thought to increase the risk of endocervical infection, probably by increasing the zone of cervical ectopy. OCPs have been found to decrease the risk of symptomatic PID, possibly by increasing cervical mucous viscosity, decreasing menstrual anterograde and retrograde flow, and modifying local immune responses. More recent data suggest that OCPs may not have any effect on PID incidence.⁴

Bilateral tubal ligation (BTL) has not been found to provide protection against PID; however, patients with BTL may have delayed or milder forms of PID.¹¹

Intrauterine device (IUD) use has been associated with a 2- to 9-fold increased risk for PID, but recent data suggest that the risk with current IUDs may be significantly less.¹² Kelly et al found a rate of 9.6 cases of PID per 1000 IUD insertions, with the most significant risk in the first 20 days.¹³ Meirik et al validated early risk of PID within the first month after insertion and also found that the risk appears to be modified by the number of patient sexual partners, the age of the user, and the community prevalence of STIs.¹⁴

Pregnancy decreases the risk of PID once the cervical os is protected by the mucous plug. However, PID can occur during the early first trimester until the mucous plug is solidified, and it may produce fetal loss.

In the upper tract, a number of microbial and host factors appear to play a role in the degree of host inflammation and resultant scarring. Uterine infection is usually limited to the endometrium, but may be more invasive in a gravid or postpartum uterus. Tubal infection initially affects the mucosa, but acute, complement-mediated transmural inflammation may develop rapidly and increase in intensity with subsequent infections. Inflammation may extend to uninfected parametrial structures, including the bowel. Infection may extend by spillage of purulent materials from the fallopian tubes or via lymphatic spread beyond the pelvis to produce acute peritonitis and acute perihepatitis (Fitz-Hugh-Curtis syndrome).

Bjartling et al have found less symptomatic urethral infection and decreased lower abdominal findings produced by a less virulent variant strain of Chlamydia trachomatis.¹⁵ Den Hartog examined the role of 5 single nucleoside polymorphisms (SNPs) in 4 genes encoding pattern recognition receptors in local tubal cells and circulating immune cells (eg, macrophages). The presence of 2 or more SNPs in patients appeared to correlate with increased laparoscopically identifiable tubal pathology.¹⁶ Genetic polymorphisms of PID pathogens has been suggested to affect the likelihood that a lower tract infection will progress to frank PID. CHSP60 antigen expression in C trachomatis¹⁷ and P9Opa(b) protein expression in N gonorrhoeae¹⁸ are examples of specific bacterial genes implicated in the pathology of PID.

Frequency

United States

According to the CDC, more than 1 million women are estimated to experience an episode of PID every year. The disease produces approximately 2.5 million office visits and 125,000-150,000 hospitalizations yearly. The true incidence is probably much higher due to incomplete and untimely conventional nonelectronic reporting methods and because many cases of silent and smoldering PID occur and are discovered only when the patient develops chronic complications of this process.

From 1995-2001, 769,859 cases of PID were reported in the United States, 91% of which were diagnosed in the ambulatory setting.¹⁹ Long-term sequelae, including chronic pelvic pain, ectopic pregnancy, tubal factor infertility (TFI), and implantation failure in in vitro fertilization attempts may occur in up to 25% of patients, ultimately affecting approximately 11% of reproductive-aged women. More than 100,000 women are estimated to become infertile each year as a result of PID.²⁰

International

While no specific data are available for PID incidence, in 1999, the World Health Organization (WHO) estimates that approximately 340 million new cases of curable STIs occur annually worldwide in individuals aged 15-49 years.²¹ A number of factors contribute to the difficulty in determining the actual worldwide incidence and prevalence of PID, including lack of patient recognition of disease, difficulties in access to care, the often subjective method of disease diagnosis, lack of diagnostics and laboratory facilities in many developing countries, and under-funded, and over-stretched public health systems.²² Worldwide, WHO has determined that STIs rank in the top 5 disease categories for which adults seek care. Women in resource-poor countries, especially those in sub-Saharan Africa and Southeast Asia, experience an increased rate of complications and sequelae.

The annual rate of PID in high-GNP countries has been reported to be as high as 10-20 per 1000 women of reproductive age. Public health efforts implemented in Scandinavia to decrease the prevalence of STIs have been quite effective.²³

Mortality/Morbidity

A delay in diagnosis or treatment can result in long-term reproductive sequelae, such as tubal infertility. Each repeat episode of PID doubles the risk for tubal factor infertility. Women with a history of PID have a 7- to 10-fold increased risk for ectopic pregnancy (tubal pregnancy) compared with women with no history of PID. Chronic pelvic pain can also follow PID and occurs in 25-75% of women. Direct mortality, which does not include those women dying from ruptured ectopic pregnancies, is estimated to occur in 0.29 patients per 100,000 cases in those women aged 15-44 years. The most common cause of death is rupture of a tubo-ovarian abscess, with the mortality of rupture remaining approximately 5-10%, despite improved diagnostic methodology and current treatment options.

Race

An increased rate of PID, approximately 2.3 times the rate seen in white women, is reported in lower socioeconomic groups in the United States. This is felt not to be a result of race or socioeconomic class per se, but rather is attributed to early coitarche, multiple sexual partners, and delayed medical care seeking. Racial and socioeconomic biases may influence both the likelihood of determining this often subjective diagnosis and which potential subjects are included in large, urban emergency departments and STI clinics often used as data collection centers for major PID studies.

Sex

PID is an infection of the female genital tract.

Age

PID may occur more frequently in adolescents (ie, 15-19 y), but it can occur in any patients who are sexually active. Age distributions vary with geographic location and etiology. Young age at first intercourse increases risk for PID.

Short et al found that age younger than 25 years and smoking were independently associated with Mycoplasma genitalium infection, which is associated with pelvic inflammatory disease.²⁴

Clinical

History

Patients can present with a variety of symptoms, and significant variation may be seen in described symptoms, ranging from lower abdominal pain to dysuria. Women may be asymptomatic or seriously ill. Symptoms occur most commonly early in the menstrual cycle or at the end of menses, which has been attributed to low progesterone levels at this time, with consequent thinning of the cervical mucosal barrier.

• Lower abdominal pain is present in more than 90% of documented cases and is by far the most common presenting symptom.
  • Usually pain is described as dull, aching or crampy, bilateral, and constant; it begins a few days after the onset of the last menstrual period and tends to be accentuated by motion, exercise, or coitus.
  • Pain may be exacerbated by movement or sexual activity (dyspareunia).
  • Pain from PID usually lasts less than 7 days; if pain lasts longer than 3 weeks, the likelihood that PID is the correct diagnosis declines substantially.
• Abnormal vaginal discharge is present in approximately 75% of cases.²⁵
• Unanticipated vaginal bleeding, often postcoital, coexists in about 40% of cases.
• Temperature higher than 38°C (30%), nausea, and vomiting manifest late in the clinical course of the disease.

Physical

The diagnosis of acute PID is primarily based on historical and clinical findings. This is imprecise, with no single piece of historical, physical, or laboratory information found to be specific or sensitive for the disease. A large number of patients may present atypically and exhibit no or few symptoms; however, more than one-fourth of these patients meet objective criteria for upper tract infection on laparoscopic examination. The sensitivity of the pelvic examination is only 60%. The physical examination usually reveals lower abdominal tenderness, cervical motion tenderness, and uterine or adnexal tenderness. Rebound tenderness and involuntary guarding may be noted and suggest associated peritonitis. The positive predictive value (PPV) of these findings will vary depending on the prevalence of PID in a given population.

One large multicenter trial found adnexal tenderness to be the most sensitive physical examination finding (95% sensitive, p <0.001).²⁶ Mucopurulent cervicitis is common, and, if absent, it provides a significant negative predictive value (NPV). Adnexal fullness or disproportionate unilateral adnexal tenderness may indicate the development of a tubo-ovarian abscess. Molander et al found 3 variables to be significant predictors of the diagnosis: adnexal tenderness (p <0.001), fever (p <0.001), and an elevated sedimentation rate (ESR) (p <0.001). In this study, these variables correctly classified 65% of patients with laparoscopically documented PID (95% CI, 61-99%).²⁷

Right upper-quadrant tenderness, especially if associated with jaundice, may indicate associated Fitz-Hugh-Curtis syndrome. A prospective cohort study in 117 incarcerated adolescents documented a 4% incidence of Fitz-Hugh-Curtis syndrome in those with mild-to-moderate PID.²⁸

Signs of associated STI, such as HSV, syphilis, and HPV, should be noted and evaluated.

For the diagnosis of PIV in sexually active young women, the Centers for Disease Control and Prevention (CDC) recommend uterine/adnexal tenderness or cervical motion tenderness as the minimal clinical criteria for the diagnosis of PID.

Additional criteria may be used to enhance the specificity of the minimum criteria:

• Temperature higher than 38.3°C (101°F)
• Abnormal cervical or vaginal mucopurulent discharge
• Presence of white blood cells (WBCs) on saline microscopy of vaginal secretions
• Elevated erythrocyte sedimentation rate
• Elevated C-reactive protein level
• Laboratory documentation of cervical infection with Neisseria gonorrhoeae or Chlamydia trachomatis

A more recent study examined the addition of chlamydial IgG antibody titers to use of the clinical picture. Laparoscopic analysis of 100 women with lower abdominal pain, in whom C trachomatis IgG antibody titers were obtained, suggested that women with a high titer of chlamydial antibodies, acute pelvic pain, and a clinical picture suggestive of PID, were more likely to have salpingitis than adhesions alone. Those patients with high titers and chronic pelvic pain, but with a clinical picture that did not suggest PID, were more likely to have adhesions alone. The investigators concluded that their limited data suggested that serologic testing might help to formulate the diagnosis; however, at this time they felt that the only way to prevent upper tract damage was to continue to treat early based on suspicion.²⁹

Causes

The classic high-risk patient is a menstruating woman younger than 25 years who has multiple sex partners, does not use contraception, and lives in an area with a high prevalence of STD. PID is also more prevalent among individuals who are young at first intercourse. The IUD confers a relative risk of 2.0-3.0 for the first 4 months following insertion, but then it decreases to baseline thereafter. Women who are not sexually active have a very low incidence of upper genital tract infection, as do women who have undergone tubal sterilization.

• Chlamydia trachomatis³⁰ : C trachomatis, an intracellular bacterial pathogen, is the predominant STD organism causing PID. Clinically, infection with this obligate intracellular parasite may manifest as mucopurulent cervicitis.
• Cytomegalovirus (CMV): CMV has been found in the upper genital tracts of women with PID, suggesting a potential role of CMV in PID.
• Endogenous microflora: In iatrogenically induced infections, the endogenous microflora of the vagina predominate.
• Gardnerella vaginalis
• Haemophilus influenzae
• Enteric gram-negative organisms (Escherichia coli)
• Peptococcus species
• Streptococcus agalactiae
• Bacteroides fragilis: This can cause tubal and epithelial destruction.
• Neisseria gonorrhoeae: In the United States, the role of N gonorrhoeae as the primary cause of PID has decreased.
• Mycoplasma genitalium: M genitalium has been isolated in the endometrium and fallopian tubes of women who have PID.³¹

• Search for CME/CE on This Topic »

• Pelvic Inflammatory Disease (Emergency Medicine)
• Pelvic Inflammatory Disease/Tubo-ovarian Abscess (Radiology)
• Salpingitis (Obstetrics and Gynecology)
• Oophoritis (Obstetrics and Gynecology)

• Female Sexual Problems Overview
• Ectopic Pregnancy Overview
• Birth Control FAQs
• Birth Control Overview
• Pregnancy and Reproduction Center
• Pelvic Inflammatory Disease Overview
• Pelvic Inflammatory Disease Treatment
• Pelvic Inflammatory Disease Causes
• Sexually Transmitted Diseases Center
• Pelvic Inflammatory Disease Symptoms

• Chlamydia Vaginal Infection Linked to Pelvic Inflammatory Disease
• Video Intervention May Improve Treatment for Partners of Teen Girls With Pelvic Inflammatory Disease
• Laparoscopic Uterosacral Nerve Ablation May Not Be Helpful for Women With Chronic Pelvic Pain

• Adolescents With Initial Diagnosis of Pelvic Inflammatory Disease at Significant Risk for Repeat STIs
• Failure of Cefoxitin and Doxycycline to Eradicate Endometrial Mycoplasma Genitalium and the Consequence for Clinical Cure of Pelvic Inflammatory Disease
• Effectiveness of Chlamydia Screening: Systematic Review