Pelvic Inflammatory Disease

IMPORTANT POINTS

1. Pelvic inflammatory disease (PID) is a polymicrobial infection, usually initiated by endocervical infection with Neisseria gonorrhoeae or Chlamydia trachomatis.

2. More than 25% of women affected with PID suffer one of several serious sequelae, including tubal infertility, tubo-ovarian abscess, chronic pelvic pain, or ectopic pregnancy.

3. PID is diagnosed primarily through clinical assessment, with laboratory and radiographic studies supporting or confirming it.

4. Treatment of PID includes broad-spectrum antibiotic coverage for anaerobic bacteria as well as for the causative organisms.

5. Primary and secondary prevention is paramount in reducing the social, emotional, and economic burden of PID and its sequelae.

Definition

Pelvic inflammatory disease (PID) is an ascending polymicrobial genital tract infection that occurs in sexually active females. It includes an array of inflammatory disorders, including endometritis, parametritis, salpingitis, oophoritis, tubo-ovarian abscess (TOA), peritonitis, and perihepatitis. Neisseria gonorrhoeae and Chlamydia trachomatis are usually the causative agents of PID, but microflora from the vagina and the bowel also may contribute to its pathogenesis.

The serious medical and social consequences of PID, including chronic pain, ectopic pregnancy, and infertility, demand improved efforts at prevention, recognition, and management. Given the declining age of sexual initiation in the United States, strategies designed to curtail PID and its sequelae begin in the pediatrician’s office.

Epidemiology

PID affects 1 million American women annually at an estimated annual cost of $4.2 billion. Sexually active adolescents account for 20% to 30% of cases and face a tenfold increase in risk compared with adults. PID results in more than 600,000 hospital admissions yearly and leads to serious long-term complications in 25% of women who are infected. The likelihood of sequelae is particularly high among adolescents because of late presentation, delayed diagnosis, and inadequate treatment.

The increased risk of PID during adolescence reflects both the biologic susceptibility of the immature cervix and the high prevalence of sexual behaviors that are risky. The transitional zone between the columnar and squamous epithelium of the vagina is located more distally in the child and adolescent than in the adult. As the zone recedes into the endocervix, the likelihood of pathogenic organisms reaching it declines. In addition to the endogenous biological risk, American teenagers are initiating sexual intercourse earlier than ever. In 1997, 17% of 7th-and 8th-grade females and 49% of 9th-through 12th-grade females reported a history of sexual intercourse. Sexually active adolescents are three times more likely than adults to be infected with C. trachomatis and account for one third of all cases of gonorrhea reported to the Centers for Disease Control and Prevention (CDC).

The risk of sexually transmitted diseases (STDs) and PID increases with failure to use condoms, the number of lifetime sexual partners, new partners within 3 months, past history of STD or PID, and the presence of an intrauterine device. Use of oral contraception decreases the risk of PID by as much as sevenfold. The protective effect of oral contraceptives appears to be related to thickening of the cervical mucus, less dilatation of the endocervical canal, milder uterine contractions with menses, and diminished menstrual flow. However, these agents do not decrease the risk of vaginal or cervical infection with sexually transmitted organisms.

The protection or risk associated with other factors is controversial. For example, use of a diaphragm may provide some protection by blocking the entry of microorganisms into the cervical canal. Vaginal douching may predispose to PID by propelling microorganisms from the vagina into the uterus. Cigarette smoking has been associated with PID, but it probably is a marker of general risk behaviors rather than a specific cause of PID.

Pathogenesis

PID usually begins with cervical infection that is caused by C trachomatis, N gonorrhoeae, or both. More than 50% of women who have PID have evidence of gonococcal or chlamydial infection. Among women who have documented gonococcal or chlamydial infection of the cervix, 30% to 50% have evidence of uterine and/or tubal involvement. Following lower tract infection with N gonorrhoeae or C trachomatis, the normal vaginal lactobacilli are supplanted by facultative organisms such as Escherichia coli, Bacteroides sp, anaerobic cocci, Mycoplasma hominis, and Ureaplasma urealyticum. The polymicrobial nature of PID has been confirmed by cultures of the upper tract obtained laparoscopically. Consequently, although N gonorrhoeae and C trachomatis are the likely inciting organisms, treatment of PID must include broad-spectrum antibiotic coverage.

The normal host defense against infection of the endometrium consists of an acid vaginal pH, a thick layer of vaginal epithelial cells, and the cervical mucus plug. Pathogenic microorganisms gain access to the endocervical canal by adhering to columnar epithelium. As described earlier, the immature cervix has more ectopy, or exposed columnar epithelium, than the adult cervix, thus predisposing the adolescent to infection. Inflammatory disruption of the vaginal environment and cervical barrier facilitates ascension of the inciting sexually transmitted pathogens and other microorganisms from the vagina into the normally sterile environment of the endometrium, fallopian tubes, and peritoneum. The spread of organisms from the vagina to the uterus is aided further during menstruation when the cervical os is open and the mucus plug is absent. Blood and tissue in the endometrial cavity may support bacterial growth, and myometrial contraction during menses may propel microorganisms upward into the tubes.

Laparoscopy, endometrial biopsy, and ultrasonography have helped clarify the pathogenesis of PID. Although laparoscopy is considered the gold standard for diagnosis, it seldom is indicated clinically or is practical. Furthermore, laparoscopy may miss endometritis or mild salpingitis. Specimens obtained through endometrial biopsy suggest that plasma cell endometritis is present in 89% of women who have PID and may be a hallmark of the disease. Ultrasonographic studies are nonspecific, but they often reveal fluid in the uterus or cul-de-sac, indistinct tissue planes, increased adnexal volume, and hydrosalpinx.

As the time between the onset of infection and diagnosis (and subsequent treatment) increases, the risk of TOA formation increases. The presence of microorganisms in the upper genital tract, in addition to the decreased tubal motility associated with inflammation, ultimately leads to organization of an abscess in up to 19% of adolescents who have PID. Other complications of PID, such as peritonitis, result from microorganisms spilling from the fimbriated ends of the fallopian tubes into the peritoneal cavity. This material may track along the paracolic gutter and cause inflammation of the hepatic capsule and diaphragm, resulting in the right upper quadrant and referred scapular pain of perihepatitis (Fitz-Hugh—Curtis syndrome).

The final stage in the pathogenesis of PID is resolution of inflammation, with its consequent risk of tubal scarring and pelvic adhesions. PID, therefore, is a major cause of chronic pelvic pain, ectopic pregnancy, and tubal infertility.

Clinical Characteristics

The most common presenting complaint of women who have PID is lower abdominal pain. Associated symptoms include vaginal discharge, irregular bleeding, dysmenorrhea, dyspareunia, dysuria, nausea, vomiting, and fever. Women who have gonococcal infection may evidence more acute inflammation (peritoneal signs, fever, leukocytosis) than those who have nongonococcal infection because of the endotoxin produced by N gonorrhoeae. Most women who have acute PID present during the first half of the menstrual cycle. Presentation later in the cycle indicates an infection of longer duration and increases the likelihood that a TOA has organized.

The severity of PID can vary widely, depending on the stage of progression, host factors, and etiologic organisms. Cases of silent PID have been well-described in which infertile women who have tubal scarring and serum antibodies to C trachomatis or N gonorrhoeae recall no symptoms suggestive of PID. Patients who are chronically ill or immunocompromised may present with either severe systemic illness from disseminated disease or with milder pain from an impaired inflammatory response. Decreased severity of symptoms does not translate into decreased likelihood or severity of sequelae, particularly because a milder presentation may delay diagnosis and initiation of therapy. PID caused by C trachomatis tends to be associated with less fever and pain and fewer systemic symptoms than gonococcal PID, but it also is associated with high rates of infertility. Atypical presentations of PID are common and complicate the differential diagnosis. For example, the symptoms of Fitz-Hugh—Curtis syndrome may mimic hepatitis or cholecystitis. The diagnosis of PID can be confirmed by abnormal results of pelvic examination and normal results of liver function studies.

The range of signs and symptoms can make the clinical diagnosis of PID difficult. The positive predictive value of a clinical diagnosis when compared with laparoscopy ranges from 65% to 90%, depending on the criteria used and the epidemiologic characteristics of the patient population. No single finding is sufficiently sensitive or specific to yield a reliable diagnosis. Rather, the diagnosis rests on combinations of findings obtained through a careful history, targeted physical examination, and selected laboratory studies. The positive predictive values of these criteria tend to be high among sexually active adolescents because of their biological and behavioral risks for STD.

The CDC recommends the use of three minimum criteria and eight optional, additional criteria for the diagnosis of PID (Table 1⇓ ) . Empiric treatment should be undertaken if lower abdominal tenderness, adnexal tenderness, and cervical motion tenderness are present on physical examination. Additional criteria supporting the diagnosis include an oral temperature of more than 38.3°C (100.9°F), abnormal cervical or vaginal discharge, elevated erythrocyte sedimentation rate or C-reactive protein, and laboratory evidence of cervical infection with N gonorrhoeae or C trachomatis. Definitive diagnosis occasionally is warranted and can be made by demonstrating endometritis on endometrial biopsy; TOA or thickened, fluid-filled tubes on radio-graphic studies; and laparoscopic findings suggesting PID. The CDC urges clinicians to maintain a low threshold of suspicion for the diagnosis and empiric treatment of PID. Experts generally agree that beginning antibiotic treatment is unlikely to compromise the course of other causes of lower abdominal pain, while delaying such treatment increases the risk of sequelae if the pain is due to PID.

Diagnosis

The assessment of an adolescent female who has lower abdominal pain begins with a sensitive, private interview by a clinician who can assure the patient of confidentiality. If PID is suspected, a pelvic examination is paramount for diagnosis. In addition to the laboratory studies noted in the CDC recommendations, a urine pregnancy test is essential to exclude the possibility of uterine or ectopic pregnancy. If an abnormal vaginal discharge is present, a sample should be examined microscopically for trichomoniasis, candidiasis, and bacterial vaginosis. Most patients who have suspected PID also should be evaluated with urinalysis and urine culture for urinary tract infection and serologic test for syphilis. All patients who have STDs warrant discussion about the risk of human immunodeficiency virus (HIV) infection and should be offered counseling and testing.

Laboratory documentation of C trachomatis or N gonorrhoeae on endocervical samples supports but does not confirm the diagnosis of PID because cervicitis may exist without endometritis or salpingitis. Furthermore, the diagnosis of PID should not await the results of either culture or rapid identification techniques. Endocervical samples should be collected and sent to the laboratory prior to initiating antibiotic therapy. These specimens should be examined for C trachomatis and N gonorrhoeae only. Cultures for anaerobes, mycoplasmas, or enteric organisms should not be performed because these organisms may normally colonize the vagina.

Four types of tests are used to detect C trachomatis. Cell culture, traditionally considered the gold standard, detects fewer than 80% of chlamydial infections. Antigen detection tests, which include direct-smear fluorescent antibody (DFA) and enzyme-linked immunosorbent assay (ELISA), yield quicker results than culture but are less sensitive and specific. Nucleic acid hybridization or genetic probe tests detect 50% to 70% of infections. Genetic amplification, which includes polymerase chain reaction (PCR) and ligase chain reaction (LCR), is the newest and most promising technology for the detection of C trachomatis. In addition to having detection rates approaching 95%, the tests can be performed on vaginal and urine samples as well as on cervical samples.

Although endocervical culture has been used traditionally for the detection of N gonorrhoeae, LCR now offers a readily available alternative. Its sensitivity and specificity for N gonorrhoeae exceeds 97%, and the same specimen can be used to test for both N gonorrhoeae and C trachomatis. The disadvantage of LCR is its high cost. As its cost decreases, the ease of use and high performance characteristics of LCR should make it the preferred test for screening and diagnosing C trachomatis and N gonorrhoeae infections.

Pelvic ultrasonography can be very helpful in narrowing the differential diagnosis of PID and identifying complications of the disease. It always should be performed in the adolescent who has suspected PID and a positive pregnancy test. Beyond 6 weeks’ gestation, the absence of an intrauterine sac on ultrasonography raises suspicion of an ectopic pregnancy and warrants immediate gynecologic consultation. Ultrasonography, particularly if performed both transabdominally and transvaginally, can help in the diagnosis of appendicitis, appendiceal abscess, adnexal torsion, ovarian cyst, and TOA. Any adolescent who has suspected PID and does not respond promptly to antibiotic therapy should be evaluated by ultrasonography for the presence of TOA.

Laparoscopy rarely is necessary to establish the diagnosis of PID. It is indicated only when the diagnosis is in doubt or the patient has failed to respond to appropriate antibiotic therapy. Because the vast majority of adolescents who have PID respond promptly to antibiotic therapy, surgical intervention rarely is necessary.

Management

Although broad-spectrum antibiotic coverage results in symptomatic improvement in most patients, the risk of long-term sequelae remains high. Adolescents are at particularly high risk for future reproductive complications because of their tendency not adolescents who have PID. Hospitalization is especially recommended if compliance is unpredictable, the diagnosis is uncertain, or TOA is suspected. Hospitalization is essential if the patient is pregnant, has HIV infection, or is too ill to tolerate or has failed to respond to outpatient therapy. All patients who are managed as outpatients must be re-evaluated within 72 hours of the initiation of antibiotics.

Treatment regimens for PID must provide antimicrobial coverage for N gonorrhoeae, C trachomatis, anaerobes, streptococci, and gram-negative facultative bacteria. The CDC recommends several regimens for inpatient treatment and two regimens for outpatient treatment (Tables 2⇓ and 3⇓ ) . Parenteral Regimen A should be continued for at least 48 hours after clinical improvement and followed by continuation of doxycycline 100 mg orally twice daily to complete a 14-day course. Parenteral Regimen B also should be continued for at least 24 hours after clinical improvement and followed by either doxycycline 100 mg orally twice daily or clindamycin 450 mg orally four times daily to complete a 14-day course. Data that support the use of the alternative regimens are limited. In the setting of TOA, clindamycin or metronidazole with doxycycline is used as continued therapy for expanded anaerobic coverage. Oral Regimens A and B both provide good coverage against the likely pathogens of PID, but little evidence is available regarding long-term outcomes. Oral Regimen A provides better anaerobic coverage but is more costly than Regimen B. Of note is the absence of single-dose azithromycin in all regimens. Although shown to be effective in the treatment of chlamydial cervicitis, its use in the treatment of PID remains controversial.

Patients should demonstrate clinical improvement (ie, defervescence and decreased pain) within 72 hours of the initiation of treatment. Outpatients who do not improve within 72 hours require hospitalization, and inpatients who do not improve within 3 to 5 days require further diagnostic evaluation and/or surgical intervention. The subset of adolescents in whom PID is complicated by TOA may require prolonged hospitalization (mean of 7 days) for continued broad-spectrum intravenous antibiotic therapy. Progress in these patients is documented by decreasing abdominal tenderness on physical examination and abscess size on ultrasonography.

Patients who have baseline laboratory documentation of N gonorrhoeae or C trachomatis should be screened again for these organisms following completion of therapy. Sexual partners of women who have PID should be treated for N gonorrhoeae and C trachomatis, even if neither patient nor partner has laboratory evidence of infection by these organisms. The likehood of both asymptomatic infection in males and reinfection in treated females is sufficiently high to warrant empiric eatment of all male sexual contacts of females who have PID.

Prognosis

The primary sequelae of acute PID are chronic pain, ectopic pregnancy, and infertility; all are due to the tubal and intra-abdominal scarring that may occur as the inflammation resolves. Up to 18% of women treated for PID report chronic pelvic pain. In one large cohort study, women who had a history of PID were ten times more likely than those who had no history to require subsequent hospitalization for abdominal pain and eight times more likely to have a hysterectomy.

PID is the single most common predisposing factor for ectopic pregnancy, with one episode of PID increasing the risk six- to tenfold. Adolescents who have PID should be counseled about this risk and should understand the importance of effective contraception and early evaluation for suspected pregnancy. The clinician’s index of suspicion for an ectopic pregnancy should be high for any adolescent who has a history of PID and is in the early weeks following conception, even if she is asymptomatic.

The risk of infertility increases with the number of episodes of PID. One episode is associated with a 13% to 21% risk, two episodes with a 35% risk, and three or more episodes with a 55% to 75% risk. Care must be taken in counseling an adolescent who has had PID about the risk of infertility. Although it is important that she understand the complications associated with infection, she neither should assume that she is unable to conceive nor have unprotected intercourse in an attempt to prove fertility. Instead, the importance of contraception until a pregnancy is both planned and desired should be emphasized.

For many young women, PID is the first experience with severe pain, illness, and hospitalization. For others, it leads to prolonged suffering and disappointment about reproductive expectations. The emphasis for all adolescents should be on primary and secondary prevention. Primary prevention involves early education about STD prevention and aggressive screening of all sexually active adolescents. Secondary prevention focuses on those adolescents who have or have had PID.

Counseling should begin with the initiation of antibiotic therapy. If the adolescent is hospitalized, the inpatient stay provides an important opportunity for intense education and discussion. The process of prevention continues after treatment, with ongoing encouragement for abstinence from sexual intercourse or safer sexual behaviors. All adolescents who have PID must understand their high risk of recurrence. With this understanding may come the motivation both to protect themselves from infection and to seek care for regular STD screening.