Pelvic Inflammatory Disease
Mark G. Martens
DEFINITIONS
Bacterial vaginosis (BV)—A vaginal condition demonstrated by a replacement of the normal lactobacillus-predominant vaginal flora with a mixture of facultative and obligate anaerobes, mycoplasma, and other species, resulting in a malodorous discharge. The presence of Gardnerella vaginalis without symptoms is not synonymous with BV.
Colpotomy—Incision into and through the vaginal epithelium into a cavity (often abscess or pelvic/abdominal cavity).
Febrile morbidity—Temperature equal to or greater than 100.4°F or 38°C on two separate occasions (>4 hours apart), starting 24 hours after the initiating event.
Fever—Temperature equal to or greater than 99.6°F or 37.5°C.
Lower genital tract infection (LGTI)—An alternative term used to describe infection limited to the cervix and/or vulva and vagina.
Mollicutes—Class of organisms that includes Chlamydia, Mycoplasma, Ureaplasma, and other species.
Outpatient therapy—Previously synonymous with oral therapy; however, the term now includes oral, ambulatory intravenous, or intermittent injectable therapies.
Parenteral therapy—Usually denotes intravenous (antibiotic) therapy, but not necessarily given while hospitalized.
Pelvic inflammatory disease (PID)—A general term used to refer to infection and inflammation of the upper genital tract in women.
Phlegmon—A massive infiltration of inflammatory cells into infected soft tissue resulting not in an abscess, but a “woody induration” sensation upon examination/palpation.
Prevotella—Newly named genus that now includes several species formerly called Bacteroides (e.g., Bacteroides bivia).
Salpingitis—Infection/inflammation of one or both of the fallopian tubes (often incorrectly used as an equivalent to all PID).
Tuboovarian abscess (TOA)—A term used to describe an abscess incorporating the fallopian tube and ovary.
Upper genital tract infection (UGTI)—A more descriptive term used alternatively to PID.
 FIGURE 30.1 Chlamydia rates by sex, United States, 1991-2011. (Data from the Centers for Disease Control.) |
Pelvic inflammatory disease (PID) is one of the most serious infections faced by women today. Also, it is one of the most common gynecologic reasons for hospitalization and emergency department visits in the United States each year.
Untreated or unsuccessfully treated women may suffer lifethreatening consequences, and even adequately treated women are at much higher risk for potentially serious sequelae. PID is a spectrum of diseases initially involving the cervix, uterus, and fallopian tubes. Acute PID, the acute clinical syndrome, is most often attributed to an ascending spread of microorganisms from the vagina and endocervix to the endometrium, fallopian tubes, and contiguous structures. The terms acute PID and acute salpingitis are often used interchangeably, but PID is not limited to tubal infection only. A more descriptive term to differentiate the severity and extent of various forms of PID was introduced by Hemsell and colleagues: upper genital tract infection (UGTI). This is differentiated from lower genital tract infection (LGTI) because response to treatment appears to be different in these two entities. PID has also been categorized into inpatient or outpatient treatment groups. However, this may not accurately reflect the severity of illness, but rather the predictive accuracy of the caregiver for hospitalization.
EPIDEMIOLOGY
Sexually transmitted infections (STIs) have been reported at epidemic proportions in the United States, with the Centers for Disease Control and Prevention (CDC) estimating almost 20 million new infections each year at a cost of $6 billion each year. The incidence of Chlamydia trachomatis infections has increased 8% since 2010, to a rate of 458 per 100,000 people, with a total of 1,412,791 cases reported in 2011 (Fig. 30.1).
Neisseria gonorrhoeae incidence has also increased to 4% since 2010 with a rate of 104 per 100,000 people, with more than 320,000 cases reported in 2011 in the United States. For women, acute PID is the most common and important complication of STIs. Bell and Holmes in the 1980s estimated that 1 million women a year were treated for acute salpingitis in the United States, but recent estimates by Sutton and colleagues estimated a decrease in cases of PID to approximately 770,000 per year and a 68% decrease in hospitalized PID from 1995 to 2001.
The CDC estimates that of the 750,000 to 1 million women per year who experience an episode of PID, up to 15% will become infertile as a result of the infection. Hospitalization rates have been declining over the past decade. However, it is uncertain if this is due to decreasing incidence of the disease or a change in clinical management to treat more patients as outpatients. A consequence of the latter scenario is that despite
the long-term trend of decreased hospitalization, as confirmed by Sørbye and colleagues (35% reduction in hospitalized cases of PID), they also found that there was a 65% increase (P = 0.013) of tuboovarian abscesses in these patients.
the long-term trend of decreased hospitalization, as confirmed by Sørbye and colleagues (35% reduction in hospitalized cases of PID), they also found that there was a 65% increase (P = 0.013) of tuboovarian abscesses in these patients.
 FIGURE 30.2 PID—Hospitalizations of women aged 15 to 44 years, United States, 2001-2010. (Data from the Centers for Disease Control.) |
However, after hitting an all-time low in hospitalizations for PID in 2009, there was a marked increase in 2009 to 2010, and we are anxious to see if this will become a long-term trend (Fig. 30.2).
In addition to the 200,000 women who are hospitalized each year with a diagnosis of salpingitis or PID, the disease also generates nearly 2.5 million visits to physicians and an estimated 150,000 surgical procedures for complications every year. In terms of overall incidence, acute PID occurs in 10% to 15% of women during their lifetime, with a diagnosis in 1% to 2% of young, sexually active women each year. Therefore, PID is the most common serious bacterial infection in women aged 16 to 25 years, and the resultant morbidity exceeds that produced by all other infections combined for this age group.
ETIOLOGY
Within certain geographic areas or populations, N. gonorrhoeae is a common cause of PID. However, most cases of acute PID are the result of a polymicrobial infection caused by organisms ascending from the vagina and cervix to infect the lining of the endometrium and fallopian tubes. Approximately 85% of cases are spontaneous, noniatrogenic infections that occur in sexually active women of reproductive age. The remaining 15% of infections occur after procedures that break the cervical mucous barrier, such as with the placement of an intrauterine device (IUD), endometrial biopsy, or uterine curettage, which allow vaginal flora to infect the upper genital tract.
In the United States, nontuberculous PID was traditionally separated into gonococcal and nongonococcal disease, depending on the isolation of N. gonorrhoeae from the endocervix. However, a variety of organisms can be isolated from the endocervix. Therefore, it is difficult to determine which of these organisms cause PID and which are coexistent cervicovaginal flora representing vaginal colonizers in the upper genital tract at time of diagnosis. While upper genital tract organisms are probably more indicative of the causative organisms, they are often difficult to obtain without suspecting endocervical contamination from the diagnostic procedure. However, studies by Martens and colleagues of transcervical cultures of the infected uterine cavities did not find significant contamination with this approach with either protected catheters or suction curettes. Bacterial organisms cultured directly from tubal fluid may commonly include N. gonorrhoeae, C. trachomatis, endogenous aerobic and anaerobic bacteria, and genital Mycoplasma species. Laparoscopic studies have demonstrated a correlation of no more than 50% between endocervical and tubal cultures, but the presence of N. gonorrhoeae is almost always considered an important causative factor. Even in the presence of N. gonorrhoeae, direct fallopian tube cultures have demonstrated that tubal infections are often polymicrobial. The type and number of species vary depending on the stage of the disease. Gonorrhea, for example, is often cultured from the cervix during the first 24 to 48 hours of the disease but is often absent later. Similarly, fewer organisms are cultured late in the disease, and anaerobic bacteria such as Prevotella, Bacteroides, Peptococcus, and Peptostreptococcus species tend to predominate. Whether these anaerobes play a causative role or increase in number and frequency as a result of the inflammatory response is uncertain. Sweet has summarized the literature by stating that in approximately one third of women with PID, N. gonorrhoeae is the only organism recovered by direct tubal or cul-de-sac culture, one third have a culture positive for N. gonorrhoeae plus a mixture of endogenous aerobic and anaerobic flora, and the remaining third have only aerobic and anaerobic organisms. Chow and colleagues and Monif and colleagues have postulated that the intense inflammatory nature of gonococcus may initiate acute PID and produce tissue damage. This damage changes the local environment, which in turn allows anaerobic and aerobic organisms from the vaginal and cervical flora to invade the upper genital tract. Both Eschenbach and Sweet have suggested that not all PID follows gonococcal infection and that acute PID initially may also have a polymicrobial etiology.
According to Sweet and Gibbs, approximately 20% of all women with salpingitis have tubal cultures positive for C. trachomatis. Also, both N. gonorrhoeae and C. trachomatis are found in the same individual 25% to 40% of the time. Scandinavian studies by Eilard and coworkers have reported the recovery of C. trachomatis from the cervix in 22% to 47% of women with acute PID. C. trachomatis by itself produces a mild form of salpingitis with an insidious onset. In contrast to gonorrhea, Chlamydia can remain in the fallopian tubes for months or years after initial colonization of the upper genital tract. Svensson and colleagues found that women with C. trachomatis infection at laparoscopy had the most severe fallopian tube involvement, probably because of its clinically silent or minimally symptomatic nature, which results in difficult or delayed diagnosis and therefore delayed or absent treatment. The two major sequelae of acute PID are tubal infertility and ectopic pregnancy. These have been strongly associated with prior chlamydial infection as a consequence of intratubal and peritubal adhesions.
Although C. trachomatis is generally believed to be one of the most common causes of PID, its etiologic role is very different when compared to N. gonorrhoeae. N. gonorrhoeae is a Gram-negative diplococcus with rapid growth that is due to a short reproductive cycle of about 20 to 40 minutes to divide. This results in a logarithmic increase in the number of organisms once N. gonorrhoeae reaches a relatively sterile area such as the endometrium or fallopian tube, where growth is relatively unimpeded. This rapid increase in the number of Gram-negative bacteria usually results in a rapid and intense inflammatory response by the woman’s host defenses. The response to this rapid bacterial growth is proliferation and aggregation of white blood cells (WBCs) and their inflammatory products. Migration of this bacterial and leukocytic mixture in two directions, both through the fallopian tube to the ovary and peritoneal cavity and back to the cervix and vagina, causes the symptoms that are pathognomonic of acute PID— abdominal/pelvic pain and cervicovaginal discharge.
For years, N. gonorrhoeae was thought to be the primary PID pathogen. Its often severe immunologic response, mostly due to its release of lipopolysaccharide, resulted in acute and severe pain, high fevers, and a marked WBC response. This triad resulted in an easily recognized clinical course, a prompt diagnosis and often early hospitalization, and initiation of antibiotic treatment.
Chlamydia trachomatis, however, is a slow-growing intracellular organism. Its lack of mitochondria results in its obligatory intracellular existence and also causes its growth cycle to be extremely slow compared with N. gonorrhoeae and other nonintracellular microorganisms. The growth cycle of Chlamydia is 48 to 72 hours; therefore, several weeks to months are required for the growth to reach numbers sufficient to cause clinical symptoms, if at all. Its slow growth does not induce a rapid or violent inflammatory response. This explains the slow and insidious nature of the symptoms of acute C. trachomatis infections. However, because of its intracellular growth cycle, the release of elementary bodies (its infectious vehicle) occurs by rupture of the cell that it has invaded. In addition, Linhares and Witkin have eloquently documented the serious immunopathologic consequences of the 60-kDa heat shock protein (hsp60) from Chlamydia on the fallopian tube.
Thus, the repeated occurrence of elementary body infection of susceptible cells—and their subsequent destruction by rupture—along with a chronic inflammatory response are the major mechanisms by which C. trachomatis causes disease in acute and chronic pelvic infections. Also, because the slow growth and the chronic inflammatory response often result in subtle clinical symptoms, treatment is often delayed or not started at all, adding to the extended tissue destruction and PID sequelae.
The lack of acute symptoms does not lessen the importance of Chlamydia as a PID pathogen. Not only does the tissue destruction result in severe complications such as ectopic pregnancy and infertility, but also the tissue damage provides fertile ground for the growth of secondarily infecting aerobic and anaerobic bacteria. This necrotic tissue is an excellent growth medium, and the epithelial damage enhances the breakdown of the surface defense mechanisms. The importance of Chlamydia was documented during the 1980s when treatment of acute PID during new antibiotic research trials was initially believed to be successfully accomplished with regimens not active against C. trachomatis. However, although success was evident at short-term follow-up, long-term follow-up demonstrated that treatment of C. trachomatis was necessary. PID regimens without Chlamydia coverage resulted in an increased incidence of recurrent episodes and long-term complications such as abscesses and chronic pelvic pain, with resultant increased surgical intervention. Therefore, current treatment of PID includes C. trachomatis coverage, even though this organism may not be the cause of the acute symptoms. Therefore, despite the lack of immediate and acute symptoms, Chlamydia remains an important pathogen in PID, as subtle or absent symptoms are more difficult to diagnose and treat early, resulting in much of the serious long-term sequelae found with PID. This is supported by the high incidence of Chlamydia antibodies in patients with acute PID, ectopic pregnancy, and infertility in several studies.
Despite the focus on these two STIs, nongonococcal and nonchlamydial pathogens cause more than 60% of PID infections. Besides N. gonorrhoeae and C. trachomatis, aerobic and anaerobic bacteria and other microorganisms, particularly other mollicutes, have been implicated as etiologic agents in acute salpingitis.
Haggerty et al., Cohen et al., Short et al., and others have all demonstrated the clinical significance of Mycoplasma genitalium in PID. Other novel pathogens are certain to be identified as the Human Microbiome Project identifies new, difficult-to-culture organisms found to be associated with PID.
However, M. genitalium is the current pathogen generating much attention after recent studies by Clausen et al. and Svenstrup and colleagues have found an association with M. genitalium and tubal factor infertility. More importantly, current PID treatment regimens may not adequately treat M. genitalium. Since it is a mollicute similar to Chlamydia, it lacks a cell wall and thus is resistant to cell wall-inhibiting antibiotics such as the penicillins and cephalosporins.
Also, studies by Björnelius and Falk and their colleagues in men and women with M. genitalium have found persistence of the organism despite treatment with levofloxacin and tetracyclines. However, these same researchers have found azithromycin to have better activity against M. genitalium. Moxifloxacin, a newer quinolone, also appears to have good in vitro activity.
In the large National Institutes of Health (NIH) PID study called the PID Evaluation and Clinical Health trial, Haggerty reported that 41% of M. genitalium-positive women experience microbiologic failure to eradicate the organism with standard Centers for Disease Control (CDC) PID treatment regimens, and more importantly, 44% experienced clinical failure. These results will most certainly be addressed at the next CDC review of PID treatment guidelines scheduled for 2015.
Bacterial vaginosis (BV)-related organisms such as Gardnerella, Mycoplasma hominis, and Ureaplasma urealyticum have also been suggested as causal agents in acute salpingitis by Ness and colleagues. However, their role remains controversial, as other studies have not found an association with PID. Cervical cultures positive for both M. hominis and U. urealyticum have been recovered from women with PID. However, the rate of isolation can be as high as 75%, which is not statistically different from that of women who are sexually active but without PID (baseline rate of about 50%), as found by Lemeke and Lsonka. Also, Clarke et al. have published evidence that Cytomegalovirus (CMV) may be associated with PID, but additional confirmatory investigations are limited.
Group B streptococcus, Escherichia coli, and other facultative anaerobic agents are also associated with PID, but fortunately are generally adequately covered by current CDC guidelines. In light of all the new data, the main point to emphasize is that although N. gonorrhoeae and C. trachomatis are significant pathogens, they are still only associated with 30% to 40% of all PID episodes. Thus, while current standard of care requires testing for both these organisms, negative results do not mean the patient does not have PID, and treatment should be initiated or completed based on the patient clinical course or additional microbiologic testing if possible.
RISK FACTORS
Several factors that predispose to the development of acute PID have been identified. Risk factors are important considerations in both the clinical management and prevention of UGTIs. First and foremost, there is a strong correlation between exposure to STIs and PID. In the United States, recent studies have confirmed this association with the recovery of N. gonorrhoeae or C. trachomatis in 40% to 50% of patients hospitalized with acute PID. Age at first intercourse, frequency of intercourse, number of sexual partners, and marital status are all associated with the frequency of exposure to STIs and thus are associated with PID. Women with multiple partners have an increased risk (four to six times normal) for development of PID, compared with women who have monogamous sexual relations. Additional studies by Ness and colleagues have found significant factors including age at first sex, cervicitis, history of PID, family income, smoking, medroxyprogesterone use, and sex with menses.
The incidence of acute PID decreases with advancing age. Adolescent girls are at significant risk for development of acute salpingitis. Westrom reported that nearly 70% of women with PID were younger than 25 years of age, 33% experienced their first infection before the age of 19, and 75% were nulliparous. The risk for development of acute PID in a sexually active adolescent female patient was 1:8, whereas the risk was 1:80 for a sexually active woman 24 years of age or older. Several reasons have been suggested for this increased risk. The two microorganisms most commonly considered to be the inciting agents in cases of PID, N. gonorrhoeae and C. trachomatis, have a predilection for columnar epithelium. As suggested by Schaefer and by Sweet and colleagues, cervical columnar epithelium is exposed to a greater extent in younger individuals and recedes into the cervical canal with increasing age.
Clinical and laboratory studies have documented that the use of contraceptives change the relative risk for development of PID. Multiple case-control studies have shown an increased risk of acute PID in women who use an IUD. It has been estimated that IUD users have a threefold to fivefold increased risk for development of acute PID, with the greatest risk the first 20 days after insertion. The incidence of infection has recently been described by Sufrin and colleagues in over 57,000 IUD insertions in California from 2005 to 2009. They reported an overall rate of approximately 1 in 200 patients.
Barrier methods of contraception (condoms, diaphragms, and spermicidal preparations) are effective both as mechanical obstructive devices and as chemical barriers. A nearly 60% decrease in the risk of PID has been demonstrated among women using a barrier method of contraception. Ness and colleagues found a 30% to 60% decrease in recurrent PID in women using condoms consistently.
Oral contraceptives have also been shown to reduce the risk and severity of acute PID. The mechanism for such protection remains speculative. The thicker cervical mucus produced by the progestin component of oral contraceptives is believed to inhibit sperm and accompanying bacteria penetration into the upper genital tract. The decrease in duration of menstrual flow accompanying oral contraceptive use theoretically creates a shorter interval for bacterial colonization. Svensson and coworkers reported that in addition to protecting against PID, the use of oral contraceptive pills was associated with a better prognosis for future fertility than was seen in women with acute PID using other contraceptive methods or no contraceptive methods. The pill’s ability to inhibit ovulation also helps prevent an open nidus for tuboovarian abscess formation.
Surgical procedures of the female genital tract also place the patient at risk for PID. About 15% of pelvic infections occur after procedures that break the cervical mucous barrier, allowing for colonization of the upper genital tract. Eschenbach and Holmes reported that these procedures include endometrial biopsy, curettage, IUD insertion, hysteroscopy, and hysterosalpingography. The incidence of UGTI associated with first-trimester abortions is approximately 1 in 200 cases. Recent practice has emphasized the use of prophylactic antibiotics in high-risk cases to attempt to decrease the incidence of iatrogenic acute PID. A randomized trial by Jackson and a randomized trial by the Luton and Dunstable Hospital Study Group have indicated that the treatment of BV with metronidazole substantially reduced postabortion PID.
Acute salpingitis occurring in a woman with a previous tubal ligation was once believed to be rare. However, Phillips and D’Abling reported that acute PID developed in the proximal stump of previously ligated fallopian tubes in 1 of 450 women hospitalized for acute salpingitis. In addition, it is suspected that many cases may be undiagnosed because of the absence of peritoneal signs from the prevention of retrograde spillage of bacteria and inflammatory exudate in the pelvic cavity.
Previous acute PID is also a risk factor for future episodes of the disease. Another acute tubal infection develops in approximately 25% of women who have had acute PID. The exact mechanism for this increased susceptibility has not been determined, but it may be loss of the natural protective mechanisms of the fallopian tube lining against microorganisms. This increased risk may also be related to the sexual habits of the woman involved, such as reinfection from an untreated male partner. Eschenbach has documented that more than 80% of male contacts are not treated.
Genetic factors for PID have been further delineated recently. Paavonen and Taylor et al. have demonstrated that variants in the genes that regulate toll-like receptors can interfere with the innate immune system and are associated with an increased progression of infection, especially with C. trachomatis.
DIAGNOSIS
Acute PID presents with a broad spectrum of clinical symptoms. The differential diagnosis of acute PID includes acute appendicitis, endometriosis, torsion or rupture of an adnexal mass, ectopic pregnancy, and cervicitis.
Common clinical manifestations include lower abdominal pain, cervical motion tenderness, and adnexal tenderness and may include fever, cervical discharge, and leukocytosis. Historically, the diagnosis of acute PID was not established unless the patient had the triad of lower abdominal and pelvic pain, fever, and leukocytosis.
Jacobson and Westrom have shown that all three criteria are present in only 15% to 30% of documented PID cases. In fact, 50% of patients initially present with a normal temperature and WBC count. Pain in the lower abdomen and pelvis is by far the most common symptom of acute PID. It occurs in more than 90% of patients at initial presentation. The pain is usually described as constant and dull and is accentuated by motion and sexual activity. Generally, the pain is of recent onset, usually less than 7 days. About 75% of patients with PID have an associated endocervical infection and coexistent purulent vaginal discharge. Nausea and vomiting are relatively late symptoms in the course of the disease. Abnormal vaginal bleeding, especially menorrhagia or spotting, is noted in about 40% of patients. The CDC has established the criteria for making the diagnosis of salpingitis based on clinical grounds. They specify that only signs of pelvic and abdominal pain are required for a potential diagnosis of PID and eliminate leukocytosis and fever as essential criteria (Table 30.1).
TABLE 30.1 Criteria for the Diagnosis of Acute PID | |||||||||
|---|---|---|---|---|---|---|---|---|---|
|
Abdominal pain can also be associated with perihepatic inflammation and adhesions, more commonly known as the Fitz-Hugh-Curtis syndrome. It is believed to develop in 1% to 10% of patients with acute PID. Signs and symptoms include right upper quadrant pain, pleuritic pain, and tenderness in the right upper quadrant when the liver is palpated (Fig. 30.3). Usually, the symptoms and signs of this syndrome are preceded by the clinical onset of acute PID. Due to the upper abdominal pain, this condition is often mistakenly diagnosed as either acute cholecystitis or pneumonia. Fitz-Hugh-Curtis syndrome is believed to develop from vascular or transperitoneal dissemination of either N. gonorrhoeae or C. trachomatis to produce the perihepatic inflammation. Other organisms may be involved, but limited data exist on their causality.
Jacobson and Westrom attempted to correlate the clinical diagnosis of acute salpingitis with laparoscopic pelvic findings. Of 814 women in whom laparoscopy was performed for presumed acute PID, 512 (65%) had visual evidence of salpingitis, 184 (23%) had normal visual findings, and 98 (12%) had other pelvic pathology. Because of the positive clinical findings consistent with PID, many of the patients with normal findings were suspected to have early PID with endometritis and endosalpingitis, without the visual evidence of tubal or pelvic damage. Thus, laparoscopy is limited as a method of diagnosing the early stages of PID, but it is crucial to rule out the 12% “other” pathology, especially the non-PID surgical emergencies, such as appendicitis, and noninfectious entities requiring different treatment modalities, such as endometriosis or hemorrhagic cyst.
 FIGURE 30.3 Violin-string adhesions of chronic Fitz-Hugh-Curtis syndrome. (Image reprinted from Medscape Reference, http://emedicine. medscape.com/, 2013, with permission, available at: http://emedicine. medscape.com/article/256448-overview) |
Despite these shortcomings of early diagnosis, laparoscopic visualization of the pelvis is still important as the most
accurate method of confirming the diagnosis and assessing the severity of acute PID. However, it is logistically and economically impractical for all patients suspected of having acute PID to undergo diagnostic laparoscopy in the United States. Therefore, the diagnosis of most episodes of acute PID is often made on the basis of clinical history and physical examination. Although it is suggested that laparoscopy be offered to all patients with an uncertain diagnosis, it is strongly indicated for patients who are not responding to therapy, to confirm the diagnosis, to obtain cultures from the cul-de-sac or fallopian tubes, or to drain pus if necessary. In summary, laparoscopic studies have shown the following:
accurate method of confirming the diagnosis and assessing the severity of acute PID. However, it is logistically and economically impractical for all patients suspected of having acute PID to undergo diagnostic laparoscopy in the United States. Therefore, the diagnosis of most episodes of acute PID is often made on the basis of clinical history and physical examination. Although it is suggested that laparoscopy be offered to all patients with an uncertain diagnosis, it is strongly indicated for patients who are not responding to therapy, to confirm the diagnosis, to obtain cultures from the cul-de-sac or fallopian tubes, or to drain pus if necessary. In summary, laparoscopic studies have shown the following:
The clinical diagnosis of acute PID may be inaccurate.
Acute PID is sometimes found in patients undergoing laparoscopy for other causes of pelvic pain.
Laparoscopy is a relatively safe method in most cases for making the visual diagnosis of the latter stages of PID and thus assessing future fertility prognosis and planning.
Laparoscopy is an excellent means of obtaining cultures directly from the tube and peritoneal cavity.
The appearance of the pelvic organs can vary from erythematous, indurated, edematous oviducts, to pockets of purulent material, to a large pyosalpinx or tuboovarian abscess. However, although no disease may be evident in early stages, it is imperative to render treatment to all stages to avoid long-term sequelae.
Other less invasive methods of diagnosis have been suggested for verifying a clinical diagnosis of acute PID.
Because most cases of UGTI are associated with and preceded by LGTI, examination of the endocervix for inflammation, Gram stain, and culture for both N. gonorrhoeae and C. trachomatis are important for proper evaluation. A negative Gram-stained smear of the endocervix does not absolutely rule out upper tract infection. However, most studies have found that acute PID is rare without a concomitant increase in inflammatory cells in the vagina and the cervix.
Endometrial biopsy is one alternative to laparoscopy. Paavonen and associates reported a 90% correlation between histologic endometritis and laparoscopically confirmed salpingitis. However, results may be delayed up to 2 to 3 days, making its clinical applicability limited.
Ultrasonography is of limited value for patients with mild or moderate pelvic PID. Thus, the routine use of sonography in patients with early PID is often not helpful. However, ultrasound is indicated in distinguishing an adnexal mass, especially in patients who demonstrate a lack of response to antimicrobial therapy in the initial 48 to 72 hours of therapy. Sonohysterography, an ultrasound examination using the instillation of saline to better define pelvic structures, is not indicated at this time for patients suspected of having PID, because no studies have been performed to demonstrate its safety in the event that pathogens are dispersed into the upper genital tract in the process of instilling the saline.
Culdocentesis, with evidence of purulent peritoneal fluid, is a helpful, but painful, method to help diagnose acute PID. With acute PID, the WBC count of peritoneal fluid is greater than 30,000 cells/mL, compared with a WBC count of 1,000 cells/mL in women without peritoneal inflammation. However, other infections, such as appendicitis and diverticulitis, among others, can also cause purulent pelvic fluid and a false diagnosis of PID.
Laboratory tests can be obtained, but their results lack sufficient sensitivity and specificity to make them an important factor in establishing the diagnosis. Leukocytosis is not a reliable indicator of acute PID, nor does it accurately correlate with the severity of tubal inflammation or need for hospitalization. Less than 50% of women with acute PID have a WBC count greater than 10,000 cells/mL. Similarly, the erythrocyte sedimentation rate (ESR), which for years was a routine laboratory test for women with acute PID, is nonspecific and is a crude indicator of severity of disease. The ESR is elevated higher than 15 mm/hour in about 75% of women with laparoscopically confirmed acute salpingitis. However, 53% of women with pelvic pain and normal-appearing pelvic organs also have an elevated ESR. Plasma proteins, such as C-reactive protein and antichymotrypsin, have been studied to determine whether they help in the diagnosis of acute PID. They have been found to be more sensitive than the ESR. Other investigators have found that decreased or absent isoamylase in peritoneal fluid in cases of acute PID is the best nonculture laboratory test for the disease. The major disadvantages of this test are that it requires several hours to complete and that peritoneal fluid must be obtained.
Other evaluations have revealed various inflammatory cytokines to be associated with pelvic infections; however, these tests are not yet commercially available to a useful extent.
SEQUELAE
Infertility
One fourth of all women who have had acute salpingitis experience one or more long-term sequelae. The most common is involuntary infertility, which occurs in about 20% of patients. Thus, PID ranks as one of the major causes of infertility. Before antibiotic therapy, 50% to 70% of women who had experienced UGTIs became sterile. The sequelae of infections vary from a patent oviduct, to peritubular and periovarian adhesions that may interfere with ovum pickup, and/or to complete tubal obstruction. The infertility rate increases directly with the number of episodes of acute pelvic infection. Also, women with mild disease are seven times less likely to suffer tubal obstruction than women with severe PID.
Ectopic Pregnancy
The chance of ectopic pregnancy is increased 6- to 10-fold in patients with a previous episode of acute salpingitis. Pathologic studies estimate that at least 50% of ectopic pregnancies occur in fallopian tubes damaged by previous salpingitis. The mechanism for the increased rate is believed to be interference of ovum transport through the tube or entrapment of the ovum secondary to intraluminal tubal damage.
Chronic Pelvic Pain
The chance that chronic pelvic pain will develop in a woman after acute salpingitis is four times that of control subjects without pelvic infection (20% vs. 5%). Chronic pelvic pain can be caused by a hydrosalpinx. A hydrosalpinx is presumably the end-stage development of a pyosalpinx. The pain can also be related to adhesions surrounding the tube and ovary. All patients with chronic pelvic pain believed to be caused by acute PID should undergo laparoscopy or laparotomy to establish the cause of the chronic pain and rule out other diseases such as endometriosis, which require different treatment, or the presence of pelvic adhesions, which can often be directly resolved.
A tuboovarian complex is a collection of pus within an anatomic space created by adherence to adjacent organs. The incidence of true adnexal abscess is about 10% in women with acute PID. Landers and Sweet noted a 20% rate of early treatment failure (after 48 to 72 hours) of antibiotic therapy as a result of persistent pain or enlargement of a tuboovarian
abscess or complex. In addition, according to Landers and Sweet, 31% required an operation several weeks to months after their acute infections for persistent disease or pain.
abscess or complex. In addition, according to Landers and Sweet, 31% required an operation several weeks to months after their acute infections for persistent disease or pain.
MORTALITY
Before antibiotic therapy, the mortality rate associated with acute PID was 1%. Most of these deaths resulted from rupture of tuboovarian abscesses. Today, death associated with PID is rare, as effective antibiotic treatment is frequently initiated. However, the mortality rate can still be as high as 5% to 10% for ruptured tuboovarian abscesses, if treatment is delayed or inadequate. Death is mostly the result of subsequent development of adult respiratory distress syndrome (ARDS), a condition often associated with serious infection or delayed treatment after intra-abdominal spillage of pus and bacteria.
TREATMENT
The therapeutic goals in the management of acute PID include both elimination of the acute infection and symptoms and prevention of long-term sequelae such as infertility, ectopic pregnancy, and chronic pelvic pain. Antibiotic treatment should be started as soon as cultures have been obtained and diagnosis is confirmed or strongly suspected. Treatment is based on the consensus that PID is polymicrobial in cause. Empirical antibiotic protocols should cover a wide range of bacteria, including N. gonorrhoeae, C. trachomatis, anaerobic rods and cocci, Gram-negative aerobic rods, Gram-positive aerobes, and Mycoplasma species. Despite general agreement that broadspectrum therapy is appropriate, questions persist regarding optimal therapeutic regimens.
Controversy has arisen over the issue of outpatient treatment with oral antibiotics versus inpatient treatment with parenteral antibiotics. There is limited data available to evaluate the efficacy of hospital versus ambulatory management of acute PID, due to the ethics of randomizing seriously ill PID patients to outpatient therapy. However, in the United States, three of four women with acute pelvic infection are treated as outpatients for their disease.
In Scandinavia, which has a different health care system, most women are treated as inpatients. In 2010, the CDC published recommended treatment guidelines for outpatient management of acute PID (Table 30.2). Some of the treatment regimens are based on the controversial premise that it may be adequate to cover just a few of the major etiologic agents (N. gonorrhoeae and C. trachomatis) involved in acute PID. As a result, studies have documented a 10% to 20% treatment failure rate for women receiving oral antibiotics as outpatients compared with a 5% to 10% failure rate for women without an abscess receiving intravenous antibiotics as inpatients, where broader coverage is used.
It is important to reevaluate patients within 48 to 72 hours of initiating outpatient therapy to determine the response of the disease. If a poor response has been obtained, the patient should be hospitalized with parenteral antibiotics and possible laparoscopic evaluation in the hope of preventing or limiting the sequelae of PID or discovering a different diagnosis, such as appendicitis or diverticulitis.
Ideally, every woman with acute PID should be hospitalized for the first few days for parenteral antibiotic treatment. Because this may not be practical because of limited economic or physical facility resources, the clinician who diagnoses acute PID in the office or emergency department is faced with the question of which patient to hospitalize. Indications for the hospitalization of patients with acute PID are also defined by the CDC in the 2010 guidelines (Table 30.3). In the past, the CDC has suggested that all adolescents with PID be hospitalized because of their high noncompliance rate and to optimize treatment to prevent damage to the reproductive tract, which could affect future fertility and/or result in chronic pelvic pain. Recently, Kelly and colleagues found a high incidence of recurrence in adolescents, perhaps related to poor compliance. However, the CDC’s new policy is to use the same criteria for hospitalization as for older women. This recommendation is not agreed upon by all infectious disease groups because of the seriousness of inadequately treated PID in a younger, often noncompliant population.
Another indication for hospitalization is the presence of an adnexal or pelvic abscess. Outpatient therapy may not provide antibiotic levels high enough to penetrate an abscess, and rupture of the abscess may have serious consequences.
Also, women in whom the definitive diagnosis of acute PID is in question should also be hospitalized, and additional diagnostic measures should be instituted. As previously stated, at least 10% of all patients have other serious diagnoses, such as acute appendicitis, ectopic pregnancy, or adnexal torsion, and these should be ruled out. Patients with serious illness, patients with nausea and vomiting, patients who are unable to follow or tolerate outpatient therapy, and patients with a previously failed outpatient oral regimen also should be hospitalized and given parenteral antibiotics. A trial by Ness and colleagues investigating the outcomes in patients randomized to inpatient or outpatient did not find differences in short-term outcomes, but difficulties in patient selection and randomization may not permit these results to be applicable to all PID patients.
The 2010 CDC guidelines for inpatient treatment of acute PID describe two regimens (Table 30.4). Regimen A is a combination of oral or parenteral doxycycline plus intravenous cefoxitin or cefotetan. Other third-generation cephalosporins can be substituted, such as ceftizoxime (Cefizox) or cefotaxime (Claforan). All of these agents are effective against penicillinase-producing N. gonorrhoeae, Peptostreptococcus, and other anaerobic species, as well as E. coli and other aerobic (facultative) species. Ceftriaxone is recommended by the CDC; however, its poor anaerobic activity and lack of controlled trials in PID patients do not make it an acceptable alternative for several investigators. Doxycycline can be given intravenously if the patient is unable to tolerate oral therapy, but it must be infused very slowly to prevent pain and sclerosis of the vein. Oral doxycycline has been demonstrated to be equally effective as parenteral doxycycline because of the slow growth cycle of Chlamydia and the requirement of prolonged treatment.
A recent study by Viberga and colleagues investigating the microbiology of IUD-related infection found an increased recovery of Fusobacterium and Peptostreptococcus species, which generally are covered by both cephalosporin and clindamycin. A possible disadvantage of the cephalosporin-doxycycline combination is that these two antibiotics may be less than ideal for anaerobic infections such as in a pelvic abscess.
Regimen B is a combination of clindamycin and an aminoglycoside (gentamicin). This combination provides excellent activity against anaerobes, Gram-negative aerobes, and Grampositive aerobes. Historically, it has been the preferred regimen for patients with an abscess, IUD-related infections, or pelvic infections after a diagnostic or operative procedure. However, there are few data to prove that it is significantly more effective than the cephalosporin regimens. A possible disadvantage of regimen B is that it may not provide optimal activity against C. trachomatis and N. gonorrhoeae. Clindamycin in high doses (900 mg in 8 hours) has good activity against Chlamydia, and in vitro studies by Martens and colleagues have demonstrated
effectiveness against 90% of C. trachomatis strains. Doxycycline is the most effective chlamydial agent, according to in vitro testing, but needs to be used for at least 7 days to complete treatment when the patient is switched from parenteral to posthospitalization therapy. Also, the CDC recommendation of once-daily dosing for gentamicin is not based on any data on PID patients and should be used only if indicated for renal considerations until more studies are completed.
effectiveness against 90% of C. trachomatis strains. Doxycycline is the most effective chlamydial agent, according to in vitro testing, but needs to be used for at least 7 days to complete treatment when the patient is switched from parenteral to posthospitalization therapy. Also, the CDC recommendation of once-daily dosing for gentamicin is not based on any data on PID patients and should be used only if indicated for renal considerations until more studies are completed.
TABLE 30.2 CDC-Recommended Treatment Regimen for Oral Therapy of Acute PID | ||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||
Each regimen stresses two concepts: the polymicrobial etiology of acute pelvic infection and the necessity of protecting against C. trachomatis and N. gonorrhoeae. With both regimens, the CDC recommends a minimum of at least 24 hours of intravenous treatment after clinical improvement. Both protocols also require completion of a 14-day course of oral antibiotics (doxycycline or clindamycin) to eradicate slow-growing organisms such as C. trachomatis and other mollicutes.
TABLE 30.3 Criteria for Hospitalization of Patients with Acute PID | |||
|---|---|---|---|
|
TABLE 30.4 CDC-Recommended Treatment Regimens for Parenteral Therapy of Acute PID | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
