sign in sign up
<<
Home , Bacteroides, Cervicitis, Culdocentesis, Oophoritis, Parametritis, Salpingitis

J Am Board Fam Pract: first published as 10.3122/jabfm.7.2.110 on 1 March 1994. Downloaded from 1 Pelvic Inflammatory Disease: Diagnosis i And Management

Martin Quan, MD

Bacllground: Acute pelvic inflammatory disease (PID) is a major gynecologic health problem in the United States, aftlicting more than 1 million women each year and generating annual direct and indirect costs estimated at $4.2 billion. Family physicians aut play an important role in the prevendon, as well as diagnosis and treatment, of PID. Methods: A MEDLINE search for articles published from 1985 to the present was made using the key words "pelvic inflammatory disease," "," "salpingids," ''mho-ovarian abscess," "adnexids," "pelvic abscess," "parametrids," and "oophorids." The bibliographies of these articles and the author's personal files were also sources of information. Results and Conclusions: A number of risk factors have been linked to PID, including young age, age at first intercourse, muldple sex partners, the presence of , vaginal douching, the use of an intrauterine contracepdve device, and a history of a sexually transmitted disease. The diagnosis of PID represents a major clinical challenge that requires a careful history and physical examination coupled with selective and knowledgeable use of the diagnostic tests and procedures currently available. Broad-spectrum antibiodcs, which represent the cornerstone of therapy, must adequately cover the polymicrobial spectrum of pathogens implicated in this infecdon, which includes gonorrhoeae, traehomatis, and specific cervicovaginal anaerobic and aerobic . The numerous sequelae associated with PID, which include inferti1ity, , and chronic syndromes, underscore the need for effective measures for preventing pelvic inflammatory disease. (J Am Board Fam Pract 1994; 7:110-23.)

Acute pelvic inflammatory disease (PID) is an "pelvic inflammatory disease," "endometritis," ascending of the female genital tract "," "tubo-ovarian abscess," "adnexitis," involving the , fallopian tubes, and adjacent "pelvic abscess," "," and "." http://www.jabfm.org/ pelvic structures. It is the most frequent serious The bibliographies of these articles and the author's infection encountered by United States womenl personal files were also sources of information. and is responsible for more than 250,000 hospital admissions and 2.5 million outpatient visits each Incidence year. Direct and indirect costs of PID were esti­ Acute PID is not a reportable disease, and thus mated at greate.r than $4.2 billion in 1990.2 precise incidence rates are unavailable. The inci­

In light of the medical and socioeconomic conse­ dence appears to be increasing, however, as evi­ on 23 September 2021 by guest. Protected copyright. quences associated with this disorder, it is impera­ denced by estimates of 850,000 cases per year3 in tive that the primary care physician be proficient in the mid-1970s rising to more than 1.4 million its recognition and treatment. This report reviews cases annually by the late-1980s.2 From a survey the clinical aspects of acute PID with a special focus conducted in 1988, Aral, et a1.4 found that nearly on its epidemiology, diagnosis, and management. 11 percent of United States women of reproduc­ tive age report having had PID in the past. Methods A MEDLINE search for articles published from Epidemiology 1985 to the present was made using the key words A number of epidemiologic risk factors are asso­ ciated with PID. Awareness of these factors is of Submitted, revised, 12 November 1993. value not only in assessing a patient's risk for From the Division of Family Medicine, School of Medicine, having PID but also when counseling female pa­ University of California, Los Angeles. Address reprint requests to Martin Quan, MD, UCLA Family Health Center, 220 UCLA tients regarding contraceptive methods and sex­ Medical Plaza, Room #220, Los Angeles, CA 90024-1683. ual practices.

110 JABFP March-April 1994 Vol. 7 No.2 J Am Board Fam Pract: first published as 10.3122/jabfm.7.2.110 on 1 March 1994. Downloaded from

Age contraceptives were capable of inhibiting the Young women are at greatest risk for acute PIO, spread of into the upper genital tract and nearly 70 percent of cases occur in women in patients with chlamydial of the . younger than 25 years. Scandinavian studies have determined the risk of PIO developing in sexually Vaginal Douching active adolescent girls to be 1 in 8 compared with In a multivariate analysis that controlled for a 1 in 80 in women older than 24 years of age.s number of potentially confounding variables (in­ cluding age, age at first intercourse, history of Age at First Intercourse or in­ A sexual debut before the age of 16 years was fection or PIO, current birth control method, and reported by Lidegaard and Helm6 to be associated presence of bacterial vaginosis), WoIner-Hanssen with a twofold increased risk of PIO compared and associates13 reported that women who with women whose age at first intercourse was 18 douched three or more times a month were 3.4- years or older. Similarly, Aral, et al.4 found that fold more likely to have acute PIO than women white women who were sexually active before the who douched less than once a month. This find­ age of 15 years were three times more likely to ing was corroborated by Scholes, et al. 14 in a have been treated for PIO than those who initi­ case-control study, which found that women who ated intercourse after 19 years of age. douched once or more a week had a 3. 9-fold increase in risk. Possible mechanisms for this in­ Number ofSex Partners creased risk include flushing of cervicovaginal The risk of acute PIO is increased two- to three­ bacteria into the uterus by the act of douching, as fold for those women who have multiple sex well as a douching-induced alteration of the vagi­ partners.4,7 nal environment to one less protective against vaginal pathogens. 13,14 History ofPID Four to 23 percent of patients who have had one Bacterial Vaglnosls episode of PIO have a subsequent episode.8 A positive Gram stain for bacterial vaginosis was found by Eschenbach, et a1. 1S to be significantly History ofa Sexually transmitted DIsease associated with the clinical diagnosis of acute

A self-reported history of gonococcal, chlamydial, PIO. Corroboration of this finding was provided http://www.jabfm.org/ or herpesvirus infection was reported by Aral and by Paavonen, et al. 16 who reported a strong asso­ associates4 to increase a woman's likelihood of ciation between bacterial vaginosis and patients PIO by 2.5-fold. with histopathologic evidence ofPIO.

Method ofContraception Other It is well established that the use of contraceptives Nonwhite groups are generally regarded as being influences the risk of development of PIO. The at increased risk for PIO, as are women of a low on 23 September 2021 by guest. Protected copyright. presence of an intrauterine contraceptive device socioeconomic class. An increased risk has also (IUD) is associated with a 1.5- to 2.6-fold in­ been observed in women never married, divorced, creased risk of developing acute PIO.1,9 Barrier or separated. 17,18 Marchbanks, et al. 19 recently re­ methods of contraception afford protection ported a twofold increased risk of PIO among against PIO, and the Women's Health Study current and former cigarette smokers. found a 40 percent reduction of risk in users of diaphragms or .9 The use of combined Microbiology oral contraceptives has generally been shown to Studies relying on advanced microbiological have a protective effect, reducing the risk of PIO techniques and upper genital tract cultures have by up to 50 to 80 percent.9,10 Although concerns clearly established the polymicrobial causation of have been expressed regarding the possible en­ acute PIO.16,17,20-23 Moreover, these studies have hancement of Chlamydia trachomatis infections also demonstrated the lack of concordance be­ among oral contraceptive users,l1 a study by tween findings on cervical cultures and findings Wolner-Hanssen, et al. 12 suggested that oral on the microbiologic investigation of upper geni-

Pelvic Inflammatory Disease 111 J Am Board Fam Pract: first published as 10.3122/jabfm.7.2.110 on 1 March 1994. Downloaded from

tal tract infections. Organisms most frequently species, bivius, Bacteroides disiens, and 1 implicated in PID, individually as well as some­ . Facultative aerobes are iso­ times in combination, include Neisseria gonorrhoeae, lated less frequently than anaerobes and include Chlamydia trachoma tis, endogenous cervicovaginal gram-negative bacilli (including Escherichia colt)' bacteria, and the genital . group B streptococcus, Gardnerella vaginalis, and Haemophilus injluenzae. 20,21,23,25 Neisseria gonorrboeae I N gonorrhoeae is a strictly aerobic, gram-negative Genital MycoplllSmllS diplococcus that has long been recognized as an hominis and Ureapiasma urealyticum important pathogen in acute PID. Although the are the two genital mycoplasmas linked to PID. number of cases reported to the Centers for Dis­ Although commonly recovered from the lower ease Control (CDC) in 1991 was 700,000, an genital tract of patients with PID, they have been additional one to two cases are thought to occur infrequently isolated from the upper genital for every reported case. 24 Ten to 17 percent of tract. 17,23 \Vhether they represent true pathogens women with untreated gonococcal infections of or are simply commensal or~isms (whose pres­ the cervix develop PID, with 66 to 77 percent of ence is simply a marker of the sexual activity of the these women becoming symptomatic during patient) remains an area of controversy that awaits or within 7 days of the onset of menses.25 further elucidation. N gonorrhoeae has been isolated from the lower genital tract in 27 to 80 percent of cases of acute Pathogenesis PID and from the upper genital tract in 13 to 33 N gonorrhoeae and C. trachomatis have tradition­ percent of cases. 16 ally been regarded as primary pathogens - sexu­ ally transmitted agents deemed capable of initiat­ Chlamydia trllCbotnlltis ing infection in a normal upper genital tract. C. trachomatis is an obligate, intracellular patho­ Cervicovaginal bacteria have generally been cast gen that currently ranks as the most common in the role of secondary pathogens - opportunis­ sexually transmitted disease in the United States tic organisms capable of infecting only an upper with more than 3 million cases each year.26 Simi­ genital tract that has previously been im­ lar to gonococcal PID, chlamydial PID begins as munologically compromised. Causes of such a

a cervical infection with an ascending infection compromised state would include (1) an ante­ http://www.jabfm.org/ developing in 10 to 30 percent of untreated pa­ cedent infection caused by a primary pathogen tients.25 C. trachoma tis has been implicated on the (which subsequently can be replaced by secondary basis of lower genital tract specimens in 5 to 51 invaders), (2) instrumentation of the uterus (e.g., percent of cases of acute PID 17 and on the basis of cervical dilatation, IUD insertion, endometrial upper genital tract specimens in 0 to 33 percent biopsy, , abortion proce­ of cases.17,27 dure), or (3) the overgrowth of endogenous on 23 September 2021 by guest. Protected copyright. anaerobic and aerobic organisms at the cervical­ Endogenous CenJlcovaginal BlICteria vaginal interface as seen in patients with bacterial A variety of anaerobic and facultative aerobic bac­ vaginosis.8,15,17 teria have been implicated in the causation of In the early 1980s, Sweet and co-workers21 pre­ acute PID. These organisms, which are derived sented data documenting the presence of cervico­ from the endogenous cervicovaginal flora, have vaginal bacteria in the upper genital tract during been isolated from the upper genital tract in 38 to the early phases ofPID, thereby lending indirect 84 percent of patients with PID.20,21,23,28 Anaerobic support to the hypothesis that these organisms bacteria are the major cause of tubo-ovarian might also be capable of independently initiating abscesses and have been isolated in 63 to 100 per­ infection. In this laparoscopic study, anaerobic cent of cases. Conversely, chlamydia has never been bacteria were found in the upper genital tract in isolated from a tubo-ovarian abscess, and gonor­ 55 percent of patients experiencing their first epi­ rhea is isolated in less than 4 percent of cases. 29 sode of PID and were isolated from the fallopian Anaerobic bacteria regarded as pathogens in tubes in 37 percent of patients within 24 hours PID include Peptococcus species, Peptostreptococcus of symptom onset. Moreover, in 19 percent of

112 JABFP March-April 1994 Vol. 7 No.2 J Am Board Fam Pract: first published as 10.3122/jabfm.7.2.110 on 1 March 1994. Downloaded from patients, anaerobes represented the sole isolates. tients with acute PID (fable 1),34-37 with the ad­ Although it could be argued that the isolated nexal teJ?derness being unilateral in 8 to 2 0 per­ cervicovaginal bacteria represented secondary in­ cent.37.38 Rebound tenderness can be elicited in 61 vaders in patients who had had a previous (albeit to 76 percent of patients, and the presence of an unrecognized) subclinical infection attributable was reported in 16 to 49 percent. to or chlamydia, such data clearly dem­ is a variable finding present in 24 to 60 onstrated the polymicrobial nature ofPID even in percent of patients. the early stages of clinical disease. Although considerable overlap exists, the clini­ In an effort to explain the ascension of cervicovagi­ cal appearance of the patient can offer clues to the nal bacteria, as well as gonorrhea and chlamydia, into responsible pathogen(s). For example, gonococcal the upper genital tract, a number oftheories have been PID is classically associated with an acute, abrupt proposed.30 The vector theory suggests that patho­ illness, with a greater occurrence of fever and a gens present in the lower genital tract are transported shorter duration of symptoms prior to the seeking in a piggyback fashion by organisms possessing greater of medical attention compared with patients with powers of locomotion. Both Trichunwnas vaginalis and nongonococcal PID. Conversely, chlamydial PID spennatozoa have been shown in vitro to be capable of tends to be more subtle and has a lesser degree of transporting potential pathogens that adhere to their fever and milder clinical signs and symptoms.39 surfaces and have been nominated as possible vectors. Nongonococcal-nonchlamydial PID is more likely Arguing against T. /is acting as a vector for in the patient with a tubo-ovarian abscess, as well gonorrhea, however, is a study by Street, et aJ.31 that as the patient who has recently undergone a pro­ demonstrated a reduced viability of the latter organ­ cedure requiring uterine instrumentation.17.23 ism when incubated with T. vaginalis. The "in-suck" theory hypothesizes that pressure gradients between Laboratory Evaluation the lower and upper genital tracts are generated as the Blood Studies result of uterine contractions occurring during coitus Blood studies traditionally ordered to evaluate or menstruation and that such gradients might facili­ suspected acute PID lack both sensitivity and speci­ tate the ascent of organisms into the upper tract.30 ficity. Leukocytosis is associated with numerous The existence of such a gradient was established by inflammatory conditions other than PID and can Fox and associates32 who recorded a fall in intrauterine be absent in 33 to 50 percent ofcases.38.40 Both the

pressure to a - 25 an H20 following female orgasm. erythrocyte sedimentation rate (ESR) and the http://www.jabfm.org/ Furthermore, it has been shown by Egli and Newton33 serum C-reactive protein (CRP) determination that uterine contractions induced by oxytocin adminis­ are also nonspecific indicators of inflammation. tration can transport inert carbon particles placed in the Although classically elevated in patients with vaginal cavity up into the fallopian tubes. PID, they can be normal (ESR less than 15 mmlh and CRP less than 20 mglL) in 25 percent36 and Clinical Findings 26 percent41 of cases, respectively. Acute PID causes a broad spectrum of illness on 23 September 2021 by guest. Protected copyright. ranging from mild and subclinical illness to severe CervlCflI Gram Stain symptoms with obvious signs of . The cervical Gram stain is a potentially useful Lower abdominal pain is the most common though oftentimes overlooked test for investigat­ symptom cited by patients with acute PID but can ing the patient with suspected acute PID. The be absent in 6 percent of patients (Table 1).34-37 finding of 10 or more white cells per oil immer­ The pain is less than 14 days in duration in 83 sion field is diagnostic for mucopurulent cervici­ percent and typically is aggravated by coitus, tis42 and is believed to corroborate the diagnosis movement, and the Valsalva maneuver.38 Other of PID.40·43 In addition, in the setting of acute symptoms include an abnormal pelvic pain, the finding of gram-negative intracel­ in 38 to 73 percent, in 16 to 50 per­ lular diplococci identified within three or more cent, gastrointestinal upset in 10 to 56 percent, and neutrophils on Gram stain has a reported 68 per­ urinary symptoms in approximately 20 percent. cent sensitivity and a 98 percent specificity for Adnexal and cervical motion tenderness are the gonococcal cervicitis38,44 and is similarly support­ physical findings most frequently elicited in pa- ive of the diagnosis of PID.

Pelvic Inflammatory Disease 113 J Am Board Fam Pract: first published as 10.3122/jabfm.7.2.110 on 1 March 1994. Downloaded from

Table 1. Historical AspeC1s of Acute Pelvic Inflammatory Disease: Percentage ofSymptonas and Signs Encountered.

Morcos, et al. 34 Wolner-Hanssen, et al. 35 Jacobson & Westrom36 Bongard, et al. 37 Symptoms and signs (n = 134) (n = 76) . (n =623) (n = 45) Abdominal pain 100 100 94 Vaginal discharge 73 68 55 38 Vaginal bleeding 16 50 36 31 Nausea or vomiting 28 31 10 56 Urinary symptoms 22 21 19 History of PID or STD 25 21 44 Adnexal tenderness 90 100 92 78 Unilateral 20 Cervical motion tenderness 80 82 Abdominal rebound-guarding 61 76 Adnexal mass 19 24 49 16 Fever 30 24 41 60

Examination of tbe Male Partner the quantitative serum pregnancy tests, which can If available, the male sex partner of the patient detect hCG levels as low as 5 mIU/mL. These should be examined for the presence of a , tests have a reported 98.8 to 100 percent sensitiv­ because such a finding is thought to corroborate the ity for ectopic pregnancy, and a negative result clinical diagnosis of acute PID.38 Because more than virtually excludes this diagnosis. 52 one-half of men with gonococcal urethritis45 and one-third of men with chlamydial urethritis46 can be Cervical Studies to Detect N. gonorrhoeae and asymptomatic, a urethral Gram stain and cultures C. trachomatis for gonorrhea and chlamydia (or an appropriate Laboratory confirmation of an endocervical in­ nonculture test) should be obtained even in the fection with gonorrhea or chlamydia is regarded absence of symptoms. In a study by Wasserheit, et as supporting evidence for the diagnosis of acute

al.,47 examination of the male partner would have PID. Cultures for gonorrhea and chlamydia re­ http://www.jabfm.org/ provided confinnatory evidence for the diagnosis of main the reference standard for diagnosis despite PID in more than 50 percent of cases. a reported sensitivity of 85 to 95 percent53 and 80 to 95 percent, 54 respectively. Nonculture methods Pregnancy Testing have been developed and are now available. In Excluding ectopic pregnancy is a critical diagnos­ addition to being less technically demanding, tic step when examining the patient with sus­ these methods offer several other advantages over pected PID. In light of the inadequacy of the culture including a more rapid turnaround, a on 23 September 2021 by guest. Protected copyright. clinical history and physical examination in differ­ wider clinical availability, as well as less rigorous entiating these two entities,36,48 modem manage­ specimen transport and storage requirements be­ ment mandates that a sensitive pregnancy test be cause organism viability is not required.43 ,53,54 routinely obtained. Urine monoclonal antibody, N onculture tests for the laboratory detection of enzyme-linked immunosorbent pregnancy tests N gonorrhoeae include the enzyme immunoassay become positive at human chorionic gonadotro­ and the DNA probe test. Studies evaluating the pin levels (hCG) as low as 50 mIU/mL in the performance of the enzyme immunoassay in female urine and detect 90 to 96 percent of ectopic preg­ populations with gonorrhea have reported sensitivi­ nancies.49,50 Qualitative serum pregnancy tests, ties ranging from 74 to 100 percent and specificities of which include serum monoclonal antibody tests 86 to 100 percent. 55 A DNA probe test that binds with and qualitative radioimmunoassays, become posi­ gonococcal ribosomal RNAis now commercially avail­ tive at serum hCG levels as low as 25 mIU/mL able and in two recent studies demonstrated a sensi­ and can detect 96 percent of patients with extra­ tivity of 86 percent and aspecificity of 99 to 100 per­ uterine pregnancies.51 Even more sensitive are cent in women with a high disease prevalence.53 ,56

114 JABFP March-April 1994 Vol. 7 No.2 J Am Board Fam Pract: first published as 10.3122/jabfm.7.2.110 on 1 March 1994. Downloaded from

N onculture systems for detecting chlamydia mass with cystic and solid elements incorporating include 'the direct smear immunofluorescent anti­ the uterus. Even with the improved image resolu­ body test, the enzyme immunoassay, and a DNA tion offered by endovaginal sonography, careful probe test. The direct smear immunofluorescent clinical correlation is required because the sono­ antibody test and the enzyme immunoassay are graphic findings associated with PID lack speci­ both antigen detection tests; the former uses fluo­ ficity and can be seen in other intrapelvic dis­ rescein-tagged, species-specific monoclonal anti­ orders.60.61 In addition to being of value in bodies to C. trachomatis to identify elementary examining the patient in whom pain, obesity, or bodies in infected secretions, whereas the latter is uncooperativeness precludes an adequate pelvic designed for immunochemical detection of solu­ examination, a pelvic ultrasound scan is also a bilized chlamydial antigens. When studied in fe­ highly accurate method for detecting pelvic male populations with chlamydial infections, the abscesses (93 percent sensitivity and 99 percent direct smear immunofluorescent antibody and specificity) and can be used to monitor their re­ the enzyme immunoassay tests have sensitivities sponse to medical therapy. 62 ranging from 61 to 99 percent and 60 to 98 per­ cent, respectively, and specificities of 89 to 99 percent and 86 to 98 percent, respectively.43 The Transcervical endometrial biopsy is an office pro­ clinical use of a commercially available DNA cedure that has been proposed as a possible alter­ probe for C. trachomatis was recently studied by native to laparoscopy in the diagnosis of acute Iwen, et al.,54 and a sensitivity of 93 percent and a PID. The histopathologic diagnosis of endome­ specificity of 98 percent were reported. tritis is based on the finding of plasma cells in the endometrial stroma, with the severity proportion­ Culdocentesis ate to the degree of stromal infiltration.16.63 In a Culdocentesis is a rapid and relatively simple di­ study assessing the diagnostic value of endome­ agnostic aid that detects and samples fluid present trial biopsy in 27 patients with suspected PID (18 in the cul-de-sac. Purulent fluid found on culdo­ of whom were subsequently determined to have centesis corroborates the diagnosis of acute PID57 acute salpingitis on laparoscopy), Paavonen, et a1.63 but can also be seen in other causes of peritonitis, reported a sensitivity of 89 percent, a specificity of such as acute appendicitis or a ruptured diverticu­ 67 percent, and a positive predictive value of 84 lar abscess. In a series of 204 patients with acute percent. A major drawback of this technique is that http://www.jabfm.org/ PID (with moderate to severe adnexal tenderness its results might not be available for several days. and involuntary guarding on examination), culdocentesis was productive of purulent fluid in Dtagnostk LIlparoscopy 82 percent of cases.20 Aerobic, anaerobic, and Diagnostic laparoscopy is a procedure that en­ gonococcal cultures should be obtained on the ables the clinician to visualize the fallopian tubes fluid aspirated. Because involved pathogens are and other pelvic structures. Laparoscopic criteria likely to be present on both Gram stain and cul­ required for the diagnosis of PID include abnor­ on 23 September 2021 by guest. Protected copyright. ture, a Gram stain of the fluid can help differenti­ mal erythema and edema of the fallopian tubes ate such pathogens from the 30 percent of isolates and spontaneous or expressible tubal inflamma­ likely to represent vaginal contaminants of the tory exudate.16.36 Although regarded by many in­ specimen.58 Contraindications to culdocentesis vestigators as the reference standard for the diag­ include a mass in the cul-de-sac (absolute) and a nosis of PID, questions regarding the diagnostic markedly retroflexed uterus (relative).59 accuracy of laparoscopy were recently raised by Sellors, et al. 64 In their study they compared PID Pelvic Ultraso1UJgrapby diagnosed visually by laparoscopy with PID diag­ Pelvic sonography plays primarily an adjunct role nosed histopathologically from a fimbrial mini­ in the diagnosis of acute PID. Sonographic find­ biopsy specimen obtained at laparoscopy, and ings consistent with PID include distention and only 50 percent of patients with histopathologi­ dilation of the fallopian tubes; enlargement of the cally diagnosed PID had tubal erythema and edema , tubes, and ligaments; fluid in the cul-de-sac; on laparoscopic examination, and less than 40 and the appearance of a complex, multiloculated percent had a spontaneous or expressible exudate.

Pelvic Inflammatory Disease 115 J Am Board Fam Pract: first published as 10.3122/jabfm.7.2.110 on 1 March 1994. Downloaded from

Although some authorities have advocated that 7. A finding of mucopurulent on the laparoscopy be routinely performed in all patients basis of finding more than 5 to 10 white cells in whom acute PID is suspected, the risk-benefit . per oil immersion field on Gram stain of the ratio and cost effectiveness of such an approach endocervical discharge remain uncertain. Although generally regarded as a safe procedure when performed by an experi­ Recognizing that insistence on stringent diagnos­ enced laparoscopist, it is not without risk. There tic criteria would result in an unacceptable num­ are an estimated five deaths for every 100,000 ber of cases of mild PID going undiagnosed and laparoscopies performed and five major morbid untreated, the CDC published diagnostic guide­ events for every 1000 performed, including blood lines in 1991 that call for a lowering of the clinical vessel injury, penetration of a hollow viscus, threshold for making a presumptive diagnosis of bleeding, and gas embolism.65 Nevertheless, diag­ PID in patients with mild disease. 57 Provided nostic laparoscopy should be strongly considered competing diagnoses can be adequately excluded, for patients with suspected PID unresponsive to the CDC recommends that a provisional diagno­ medical therapy, as well as patients for whom the sis of PID be made and a therapeutic trial of diagnosis remains unclear despite a comprehen­ initiated in patients who simply have sive evaluation and for whom the need exists to lower abdominal, adnexal, and cervical motion exclude a surgical emergency, such as ectopic tenderness on examination. pregnancy or acute appendicitis. Regardless of how the diagnosis of PID is made, the importance of close clinical follow-up cannot be Establishing the Diagnosis overstated. Patients treated for PID require peri­ The clinical differentiation of acute PID from odic evaluations, and, if no clinical improvement is other causes of acute pelvic pain is a notoriously seen in 2 to 3 days, other possible diagnoses (i.e., difficult task. Jacobson and Westrom36 examined appendicitis, , ruptured , by laparoscope 814 women thought clinically to or adnexal torsion) warrant serious reconsidera­ have acute PID, but they confirmed the diagnosis tion.57 Diagnostic laparoscopy (or possibly endome­ in only 65 percent. In 23 percent of patients a trial biopsy) should be strongly considered for such normal pelvis was found, and in 12 percent an­ patients, as well as the patient whose clinical symp­ other pathologic disorder was found (e.g., acute toms are severe enough to require that a definitive appendicitis, ectopic pregnancy, endometriosis, diagnosis be made rapidly. http://www.jabfm.org/ and complications of ovarian cysts). Similar expe­ riences were reported by Morcos, et al. 34 and Management Allen and Schoon48 who were able to confirm their After making the diagnosis of PID, the physician prelaparoscopic diagnoses of PID in only 76 and must next decide whether to hospitalize the patient. 61 percent of patients, respectively. Despite hospitalization being standard practice in In light of the unreliability of the clinical diagno­ many European countries, only 15 to 30 percent of on 23 September 2021 by guest. Protected copyright. sis ofPID, many authorities have argued that a diag­ patients with acute PID are hospitalized in the nosis ofPID cannot be confidently or reliably made United States.23,66 The superiority of hospitaliza­ unless specific criteria derived from laparoscopic tion and inpatient treatment to ambulatory man­ studies are fulfilled.66 Such criteria generally re­ agement has never been proved,67 and studies are quire lower abdominal, cervical motion and ad­ urgently needed comparing the efficacy ofthese two nexal tenderness, plus at least one of the following: options with respect to both short-term and long­ term outcomes. The CDC favors hospitalization in 1. Temperature exceeding 3 SoC the following situations:57 2. Leukocytosis exceeding 10,500 per cubic mm 3. Purulent material aspirated on culdocentesis 1. The diagnosis is uncertain, or a surgical emer­ 4. Inflammatory mass on pelvic ultrasonography gency, such as ectopic pregnancy or acute 5. Erythrocyte sedimentation rate exceeding appendicitis, cannot be excluded 15 mm/h 2. A pelvic abscess is suspected 6. Culture or nonculture evidence of gonococcal 3. The patient is pregnant or chlamydial infection of the endo cervix 4. The patient is an adolescent

116 JABFP March-April 1994 Vol. 7 No.2 J Am Board Fam Pract: first published as 10.3122/jabfm.7.2.110 on 1 March 1994. Downloaded from

5. Severe illness precludes outpatient management Table 2. 1993 Centers for Disease Control GuideUnes* 6. The patient is unable to follow or tolerate an for the Inpatient Treatment for Acute Pelvic outpatient regimen Inflammatory Di~. 7. The patient fails to respond to outpatient Regimen A 2 g intravenously every 6 h or cefotetan therapy 2 g intravenously every 12 h plus doxycycline 8. Clinical follow-up within 72 hours of starting 100 mg orally or intravenously every 12 hours. Continue for at least 48 h after clinical improve­ treatment cannot be arranged ment and follow with doxycycline 100 mg orally twice a day for a total of 14 d Anttblolk Therapy Regimen B 900 mg intravenously every 8 h plus Antibiotics remain the cornerstone of treatment a gentamicin loading dose intravenously or intra­ for acute PID. Although the efficacy of current muscularly of 2 mglkg followed by a maintenance dose of 1.5 mglkg every 8 h. Continue for at least antibiotic therapy in preventing PID sequelae re­ 48 h after clinical improvement and follow mains unclear, a number of studies attest to anti­ with either doxycycline 100 mg orally twice a day biotics providing high rates of clinical cure.67 In or clindamycin 450 mg orally 4 times a day to complete 14 d of total therapy addition, in a recent study reported by Hillis, et al.,68 starting treatment early (within 3 days of the ·Source: Centers for Disease Control. 69 onset of symptoms) in patients with PID was associated with a threefold reduced risk of subse­ to be efficacious as a single agent in quent or ectopic pregnancy. patients with uncomplicated PID. Concerns When prescribing an antibiotic regimen, it is about the lack of activity by ofloxacin against important that physicians recognize the poly­ anaerobic bacteria (a limitation shared by all microbial nature of PID and that they prescribe quinolone antibiotics), however, are responsible adequate coverage. CDC treatment guidelines pub­ for the recommendation that it be used in combi­ lished in 1993 provide recommendations for inpa­ nation with either clindamycin or . tient and outpatient treatment (fables 2 and 3).69 The other option, which calls for an intramuscu­ Inpatient regimen A (cefoxitin or cefotetan plus lar injection of a cephalosporin followed by oral doxycycline) (Table 2) offers excellent activity doxycycline, is active against chlamydia and gon­ against gonorrhea and chlamydia and good activ­ orrhea but provides less than optimal coverage ity against gram-negative bacilli and anaerobic against anaerobes and gram-negative enteric bacilli.

bacteria (including Bacteroides fragilis), although A number of antibiotic regimens other than http://www.jabfm.org/ the emergence of cefoxitin-resistant anaerobes those currently recommended by the CDC have has been described.7° Inpatient regimen B also been proposed and studied.23 ,67 When con­ (clindamycin plus gentamicin) provides optimal sidering alternative regimens, physicians should activity against anaerobes and gram-negative or­ be aware of the spectrum of antimicrobial activity ganisms and is particularly favored in patients provided and recognize that the wide distribution who have suspected tubo-ovarian and pelvic of antimicrobial-resistant N gonO'lThoeae has pre- abscesses. The efficacy of these two regimens was on 23 September 2021 by guest. Protected copyright. recently confirmed in a study by Landers, et al.,40 Table 3. 1993 Centers for Disease Control Recom­ which also supported the adequacy of high-dose mendations* for Ambulatory Management of Acute Pelvic Inflammatory Disease. clindamycin in the treatment of C. trachomatis. The inclusion of the combination of ofloxacin Regimen A Cefoxitin 2g intramuscularly plus probenecid plus either clindamycin or metronidazole as an (1 g) or ceftriaxone 250 mg intramuscularly (or an equivalent third-generation cephalosporin, outpatient antibiotic option represents the major such as ceftixoxime or cefotaxime) followed by change seen in the 1993 CDC guidelines per­ doxycycline 100 mg orally twice a day for 14 d. taining to the management of PID (fable 3). For patients unable to tolerate doxycycline, 500 mg orally four times a day Ofloxacin provides excellent coverage against can be substituted gonorrhea and gram-negative enteric bacilli. Al­ Regimen B Ofloxacin 400 mg orally twice a day for 14 d plus though a lack of efficacy against chlamydia has clindamycin 450 mg orally four times a day for been reported for ciprofloxacin,71 ofloxacin has 14 d or ofloxacin 400 mg orally twice a day plus been highly active against chlamydia,67,72 and metronidazole 500 mg orally twice a day for 14 d both Soper, et al.7 3 and Wendel, et aP4 found ·Source: Centers for Disease Control.69

Pelvic Inflammatory Disease 117 J Am Board Fam Pract: first published as 10.3122/jabfm.7.2.110 on 1 March 1994. Downloaded from cluded the use of and penicillin as Surgical Therapy first-line therapy. 75 Surgical interventions have a limited role in the Monotherapy with a second- or third-genera­ management of acute PID. Possible indications tion cephalosporin provides good coverage for surgery include (1) absence of clinical re­ against gonorrhea and gram-negative rods. Al­ sponse despite optimal antibiotic coverage for though these agents (particularly cefoxitin and 48 to 72 hours, (2) a pelvic mass that enlarges cefotetan) also possess activity against anaerobes, or persists despite therapy, (3) intraperitoneal more effective antianaerobic agents are avail­ bleeding caused by erosion of a major vessel, able including metronidazole, chloramphenicol, (4) suspected rupture or leakage of a tubo-ovarian clindamycin, and 13-lactam-I3-lactamase inhibitor abscess, and (5) a pointing abscess that can be combinations,?6 Furthermore, all cephalosporins drained extraperitoneally.62 are ineffective against chlamydia, a weakness un­ Available surgical procedures include diagnos­ derscored in a study by Sweet, et al. 28 that isolated tic laparoscopy or laparotomy, laparoscopic drain­ chlamydia in post-treatment upper genital tract age and lysis of adhesions, computed tomogra­ cultures in 12 of 13 patients with chlamydial PID phy-guided drainage of abscesses, posterior despite apparendy adequate clinical cure follow­ colpotomy, and the surgical excision of involved ing monotherapy with a 13-lactam antibiotic. In­ structures (e.g., unilateral adnexectomy, total ab­ adequate activity against chlamydia is also the dominal -bilateral salpingo-oopho­ major shortcoming for regimens calling for rectomy). Current surgical practice calls for metronidazole in combination with an aminogly­ conservative management whenever possible, coside as well as mono therapy with a 13-lactam- limiting extirpation to those structures that have 13-lactamase inhibitor combinations (i.e., ampi­ to be resected.62 ,79 cillin-sulbactam, ticarcillin-clavulanate, amoxicillin­ clavulanate).67 Sequelae Combination therapy with doxycycline and In addition to its immediate impact, acute PID is metronidazole provides excellent activity against responsible for a number of sequelae as well. The chlamydia and anaerobes but unreliable coverage development of a tubo-ovarian abscess is a well­ against gonorrhea and gram-negative bacilli,?7 recognized complication seen in 3 to 16 percent Doxycycline combined with either a third-genera­ of patients hospitalized with PID,?9 Recurrent tion cephalosporin or aztreonam provides excel­ PID is another important complication and is http://www.jabfm.org/ lent coverage against gonorrhea, chlamydia, and reported in 4 to 23 percent of cases. 8 PID is also a gram-negative aerobic rods but less than optimal common cause of chronic pelvic pain, and 17 to coverage against anaerobic bacteria.76 An outpa­ 18 percent of patients develop chronic pain syn­ tient combination antibiotic regimen offering dromes that include and painful fairly comprehensive antimicrobial coverage is adhesions.5 amoxicillin-clavulanate plus doxycycline. 78 A Pelvic inflammatory disease can also have a on 23 September 2021 by guest. Protected copyright. major drawback of this regimen, however, is a dramatic impact on the reproductive potential of reported 20 percent discontinuation rate because the woman afflicted. Acute PID is a major cause of associated nausea and diarrhea. of infertility, rendering more than 11 percent of those afflicted infertile. The risk of infertility re­ AdlittiotUll Measures sulting from tubal occlusion rises with each suc­ General supportive measures, such as bed rest, ceeding episode, for an incidence of 8 percent sexual abstinence until cure is achieved, hydra­ after the first episode, 20 percent following the tion, and the provision of antipyretics and analge­ second episode, and 40 percent following the sia, are recommended in the management of third.80 Other factors associated with a greater acute PID. Placing the patient in a semi-Fowler likelihood of infertility following an episode of position has been suggested in the past for com­ PID include an age older than 25 years at the time fort and to facilitate drainage.23 If the patient has of diagnosis and severe inflammatory changes an , removal of the device is found at laparoscopy. Conversely, the risk of tubal generally recommended following the institution factor infertility arising from PID appears to be of antibiotic therapy.23,25 reduced in patients who are taking oral contra-

118 JABFP March-April 1994 Vol. 7 No.2 J Am Board Fam Pract: first published as 10.3122/jabfm.7.2.110 on 1 March 1994. Downloaded from ceptives, as well as those who have nonchlamydial Treatment ofSex Partners PID.17,68 All sex partners of women with PID should be Pelvic inflammatory disease has also been a examined for gonococcal and chlamydial infec­ major contributing factor to the dramatic increase tions, as well as other STDs. The importance of in ectopic pregnancies seen in the United States contact tracing is underscored by estimates that during the past two decades, numbering more male partners of women with PID have infection than 88,000 cases in 1989.81 A history or patho­ rates ranging up to 53 percent for chlamydial logic evidence of PID has been found in 20 to 56 infections and 41 percent for gonorrhea. 57 CDC percent of cases of ectopic pregnancy. It has been guidelines call for epidemiologic treatment (i.e., estimated that an episode of PID increases the presumptive treatment at the time of initial exam­ risk of having an ectopic gestation by 2.0- to ination before culture results are available) of 7.5-fold.82 gonorrhea and chlamydia for all sex partners of patients with PID.69 Prevention In light of the enormous medical and socioeco­ Sexual eounsellng nomic costs associated with PID, it is evident that The advisability of abstinence and the practice of effective measures directed toward its prevention should be routinely discussed with all are needed. The importance of these measures is patients at risk for contracting a sexually transmit­ further underscored by the uncertainty surround­ ted disease. Suggested areas of discussion include ing the efficacy of current therapy in preventing delaying the age at first intercourse, limiting the the sequelae ofPID,23,67,83 as well as the growing number of sex partners, and avoiding sexual rela­ appreciation of the existence of subclinical or so­ tions with potential partners at risk for having called silent PIDY Although research is still re­ an STD.83 In addition, sexually active patients quired to determine the magnitude of this condi­ should· be strongly encouraged to use condoms tion, as well as to develop criteria for its clinical as a measure to reduce the risk of transmission diagnosis, evidence for its occurrence can be de­ ofSTDs. rived from studies that report the absence of a history of PID in more than half of patients with Contraceptive Counseling despite having tubal scar­ Barrier methods of contraception, such as me­ ring and peritubal adhesions indistinguishable chanical barriers or locally applied spermicidal http://www.jabfm.org/ from that seen in patients with a history of PID agents, are thought to reduce the risk of develop­ and despite serologic evidence of past chlamydial ing PID and are favored for women at risk for and gonococcal infection.84-86 Preventive efforts exposure to STDs.83,87,88 Although further study include measures directed at minimizing the risk is required,83 recent investigations suggest that for recurrent disease, as well as epidemiologic combined oral contraceptives can also confer pro­ measures designed to prevent the transmission of tection against PID .10,89 Conversely, the use of the on 23 September 2021 by guest. Protected copyright. sexually transmitted diseases (STDs). intrauterine contraceptive device increases the risk of PID9,89 and thus should be discouraged in Treatment anil Follow-up women at risk for PID. Clinicians need to imptess upon patients the ne­ cessity of completing the full course of therapy Recognltlonanil Mallllgement ofCervlcttls and the importance of close clinical follow-up. Because it is generally believed that most cases of Reevaluation of the patient 48 to 72 hours after PID begin as a cervical infection,17,90 early recog­ initiating treatment, as well as following comple­ nition and treatment of cervicitis represent im­ tion of therapy, is crucial to determine the ade­ portant areas of prevention. Although universal quacy of therapy.23 Patients with PID also require screening for gonorrhea and chlamydia has been examination for other STDs (including , proposed as a potentially effective public health , and human papillomavirus) and, measure, selective screening (determined on the at a minimum, should be provided counseling basis of disease prevalence, as well as the availabil­ regarding the need for HIV testing and safer ity, costs, and accuracy of the screening test) is sexual practices. considered to be a more realistic, cost-effective

Pelvic Inflammatory Disease 119 J Am Board Fam Pract: first published as 10.3122/jabfm.7.2.110 on 1 March 1994. Downloaded from alternative. In a cost-effective analysis by Phillips, each year and generating an estimated $4.2 billion et al.,91 routine screening with gonorrhea cultures in direct and indirect costs in 1990. A number of was deemed justifiable when the prevalence of important risk factors for PID have been identi­ infection exceeded 1.5 to 2.1 percent. A similar fied, including young age, multiple sex partners, analysis for chlamydial cervicitis justified screen­ use of an intrauterine contraceptive device, vagi­ ing with the rapid antigen test when the disease nal douching, bacterial vaginosis, and a history of prevalence exceeded 7 percent and screening with a sexually transmitted disease. Acute PID has culture in groups in whom the prevalence ex­ been clearly established as a polymicrobial infec­ ceeded 14 percent.92 High-risk groups of women tion with N. gonorrhoeae, C. trachomatis, and en­ in whom routine screening has been recom­ dogenous cervicovaginal aerobic and anaerobic mended have included prostitutes; illicit drug bacteria playing important pathogenic roles. Al­ users; and female patients examined in jails, vene­ though diagnostic laparoscopy represents the ref­ real disease clinics, and emergency departments. 83 erence standard for the diagnosis of PID, eco­ Screening studies should also be considered in nomic and logistical considerations often dictate sexually active women who are adolescents, who that a presumptive clinical diagnosis of PID be have abnormal friability of the endocervix when made. The unreliability of a clinical diagnosis of swabbed during examination, who report multi­ PID mandates that, in addition to a careful history ple sex partners or a new partner within the pre­ and physical examination, adjunctive diagnostic vious 2 months, who report recent contact with a tests and procedures be selectively and knowl­ person with an STD, or who have a partner who edgeably used to help corroborate the diagnosis, might have had other partners during the preced­ as well as to help exclude competing diagnoses. ing 3 months.83 ,91-94 Broad-spectrum antibiotic therapy remains the cornerstone of therapy for acute PID, with surgi­ Prophylactic Antibiotics cal management generally reserved for patients In an effort to reduce the incidence of iatrogenic who fail to respond to medical therapy. The numer­ PID, the use of prophylactic antibiotics before ous sequelae of PID, which include infertility, procedures requiring uterine instrumentation chronic pelvic pain, and ectopic pregnancy, serve to has been proposed. A role for prophylactic anti­ underscore the importance of preventive measures. biotics before IUD insertion is supported by stud­ ies that suggest that the greatest risk of PID asso­ http://www.jabfm.org/ ciated with IUD use occurs shortly following References 1. Grimes DA. Intrauterine devices and pelvic inflam­ insertion.9 In a study by Sinei and associates,95 the matory disease: recent developments. Contraception routine administration of oral doxycycline at the 1987; 36:97-109. time of IUD insertion reduced the likelihood of 2. Washington AE, Katz P. Cost of and payment source PID by 31 percent. On the basis of a literature for pelvic inflammatory disease. Trends and projec­ tions, 1983 through 2000.JAMA 1991; 266:2565-9.

review by Grimes, et al.,96 antibiotic prophylaxis on 23 September 2021 by guest. Protected copyright. given before curettage abortion reduced associ­ 3. Curran Jw, Economic consequences of pelvic in­ flammatory disease in the United States. Am J Ob­ ated infectious morbidity by one-half. Finally, stet Gyneco11980; 138:848-51. recognizing the growing body of evidence im­ 4. Aral SO, Mosher WD, Cates W Jr. Self-reported plicating bacterial vaginosis in the pathogenesis pelvic inflammatory disease in the United States, of PID, Thomason, et al. 97 have recently advo­ 1988.JAMA 1991; 266:2570-3. cated prophylactic treatment of bacterial vagino­ 5. Westrom L. Incidence, prevalence, and trends of acute pelvic inflammatory disease and its conse­ sis in asymptomatic women scheduled to undergo quences in industrialized countries. Am J Obstet ambulatory invasive procedures, such as endo­ Gyneco11980; 138:880-92. metrial biopsy, IUD insertion, , and 6. Lidegaard 0, Helm P. Pelvic inflammatory disease: hysterosalpingography. the influence of contraceptive, sexual, and social life events. Contraception 1990; 41:475-83. Conclusion 7. Lee NC, Rubin GL, Grimes DA. Measures of sexual behavior and the risk of pelvic inflammatory disease. Acute pelvic inflammatory disease is a major pub­ Obstet Gyneco11991; 77:425-30. lic health problem in the United States, afflicting 8. Westrom L. Pelvic inflammatory disease: bacteriol­ more than 1 million women in the United States ogyand sequelae. Contraception 1987; 36:111-28.

120 JABFP March-April1994 Vol. 7 No.2 J Am Board Fam Pract: first published as 10.3122/jabfm.7.2.110 on 1 March 1994. Downloaded from

9. Lee NC, Rubin GL, Borucki R. The intrauterine de­ 25. Washington AE, Sweet RL, Shafer MA. Pelvic in­ vice and pelvic inflammatory disease revisited: new flammatory disease and its sequelae in adolescents.] results from the Women's Health Study. Obstet Adolesc Health Care 1985; 6:298-310. Gyneco11988; 72:1-6. 26. Schachter]. Chlamydial infections. West] Med 10. Panser LA, Phipps WR. Type of oral contraceptive 1990; 153:523-34. in relation to acute, initial episodes of pelvic inflam­ 27. Kiviat NB, Wolner-Hanssen P, Peterson M, matory disease. Contraception 1991; 43 :91-9. Wasserheit], Stamm WE, Eschenbach DA, et al. 11. Washington AE, Gove S, Schachter ], Sweet RL. Localization of Chlamydia trachomatis infection by Oral contraceptives, Chlamydia trachomatis infection, direct immunofluorescence and culture in pelvic in­ and pelvic inflammatory disease. A word of caution flammatory disease. Am ] Obstet Gynecol 1986; against protection.]AMA 1985; 253:2246-50. 154:865-73. 12. Wolner-Hanssen P, Eschenbach DA, Paavonen], 28. Sweet RL, Schachter], Robbie MO. Failure of beta­ Kiviat N, Stevens CE, Critchlow C, et al. Decreased lactam antibiotics to eradicate Chlamydia trachomatis risk of symptomatic chlamydial pelvic inflammatory in the despite apparent clinical cure of disease associated with oral contraceptive use.]AMA acute salpingitis.]AMA 1983; 250:2641-5. 1990; 263:54-9. 29. Landers DV, Sweet RL. Current trends in the diag­ 13. Idem. Association between vaginal douching and nosis and treatment of tuboovarian abscess. Am ] acute pelvic inflammatory disease. ]AMA 1990; ObstetGynecol1985; 151:1098-110. 263:1936-41. 30. Keith LG, Berger GS, Edelman DA, Newton W, 14. Scholes D, Daling]R, Stergachis A, Weiss NS, Wang Fullan N, Bailey R, et al. On the causation of pelvic S, Grayston]T. Vaginal douching as a risk factor for inflammatory disease. Am ] Obstet Gynecol 1984; acute pelvic inflammatory disease. Obstet Gynecol 149:215-24. 1993; 81:601-6. 31. Street DA, Wells C, Taylor-Robinson D, Ackers ]P. 15. Eschenbach DA, Hillier S, Critchlow C, Stevens C, Interaction between Trichomo1lllS vaginalis and other DeRouen T, Holmes KK. Diagnosis and clinical pathogenic micro-organisms of the human genital manifestations of bacterial vaginosis. Am] Obstet tract. Br] Vener Dis 1984; 60: 31-8. Gyneco11988; 158:819-28. 32. Fox CA, Wolff HS, Baker]A. Measurement of intra­ 16. Paavonen], Teisala K, Heinonen PK, Aine R, Laine vaginal and intrauterine pressures during human S, Lehtinen M, et al. Microbiological and histopath­ coitus by radio-telemetry. ] Reprod Fertil 1970; ological findings in acute pelvic inflammatory dis­ 22:243-51. ease. Br] Obstet Gynaecol1987; 94:454-60. 33. Egli GE, Newton M. The transport of carbon 17. Cates W]r, Rolfs RT ]r, Aral SO. Sexually transmit­ particles in the human female reproductive tract. ted diseases, pelvic inflammatory disease, and infer­ Fertil Steril1961; 12(2):151-5. tility: an epidemiologic update. Epidemiol Rev 1990; 34. Morcos R, Frost N, Hnat M, Petrunak A, Caldito G.

12:199-220. Laparoscopic versus clinical diagnosis of acute pelvic http://www.jabfm.org/ 18. Washington AE, Aral SO, Wolner-Hanssen P, inflammatory disease.] Reprod Med 1993; 38:53-6. Grimes DA, Holmes KK. Assessing risk for pelvic 35. Wolner-Hanssen P, Mardh P, Svensson L, Westrom inflammatory disease and its sequelae.]AMA 1991; L. Laparoscopy in women with chlamydial infection 266:2581-6. and pelvic pain: a comparison of patients with and 19. Marchbanks PA, Lee NC, Peterson HB. Cigarette without salpingitis. Obstet Gyneco11983; 61:299-303. smoking as a risk factor for pelvic inflammatory dis­ 36. Jacobson L, Westrom L. Objectivized diagnosis of ease. Am] Obstet Gyneco11990; 162:639-44. acute pelvic inflammatory disease. Diagnostic and 20. Eschenbach DA, Buchanan TM, Pollock HM, For­ prognostic value of routine laparoscopy. Am] Obstet on 23 September 2021 by guest. Protected copyright. syth PS, Alexander ER, Lin ]S, et al. Polymicrobial Gyneco11969; 105:1088-98. etiology of acute pelvic inflammatory disease. 37. Bongard F, Landers DV, Lewis F. Differential diag­ N EnglJ Med 1975; 293:166-71. nosis of appendicitis and pelvic inflammatory dis­ 21. Sweet RL, Draper DL, Hadley WK. Etiology of ease. A prospective analysis. Am ] Surg 1985; acute salpingitis: influence of episode number and 150:90-6. duration of symptoms. Obstet Gyneco11981; 58:62-8. 38. Eschenbach DA, Epidemiology and diagnosis of 22. Hemsell DL, Nobles B], Heard MC, Hemsell PG. acute pelvic inflammatory disease. Obstet Gynes;ol Upper and lower reproductive tract bacteria in 126 1980; 55(5 Suppl):142S-153S. women with acute pelvic inflammatory disease. Micro­ 39. Svensson L, Westrom L, Ripa KT, Mardh P. Differ­ bial susceptibility and clinical response to four thera­ ences in some clinical and laboratory parameters in peutic regimens.] Reprod Med 1988; 33:799-805. acute salpingitis related to culture and serologic 23. Bumakis RG, Hildebrandt NB. Pelvic inflammatory findings. Am] Obstet Gyneco11980; 138:1017-21. disease: a review with emphasis on antimicrobial 40. Landers DV, Wolner-Hanssen P, Paavonen ], therapy. Rev Infect Dis 1986; 8:86-116. Thorpe E, Kiviat N, Ohm-Smith M, et al. Combina­ 24. Zenilman]M. Gonorrhea: clinical and public health tion antimicrobial therapy in the treatment of acute issues. Hosp Pract 1993; 28(2A):29-35, 39-40,43-5, pelvic inflammatory disease. Am] Obstet Gynecol passim. 1991; 164:849-58.

Pelvic Inflammatory Disease 121 J Am Board Fam Pract: first published as 10.3122/jabfm.7.2.110 on 1 March 1994. Downloaded from

41. Lehtinen M, Laine S, Heinonen PK, Teisala K, 56. Granato PA, Franz MR. Use of the Gen-Probe Miettinen A, Aine R, et al. Serum C-reactive protein PACE system for the detection of Neisseria determination in acute pelvic inflammatory disease. gonorrhoeae in urogenital samples. Diagn Microbiol Am] Obstet Gyneco11986; 154: 158-9. Infect Dis 1990; 13:2l7-21. 42. Brunham RC, Paavonen], Stevens CE, Kiviat N, 57. Centers for Disease Control. Pelvic inflammatory Kuo CC, Critchlow CW, et al. Mucopurulent cervi­ disease: guidelines for prevention and management. citis - the ignored counterpart in women of ure­ MMWR 1991; 40(RR-5):1-25. thritis in men. N Engl] Med 1984; 311: 1-6. 58. Cunningham FG, Hauth ]C, Gilstrap LC, Herbert 43. Stamm WE. Diagnosis of Chlamydia trachrmultis gen­ WN, Kappus SS. The bacterial pathogenesis of itourinary infections. Ann Intern Med 1988; acute pelvic inflammatory disease. Obstet Gynecol 108:710-7. 1978; 52:161-4. 44. Wald ER. Gonorrhea: diagnosis by Gram stain in the 59. Eisinger SH. Culdocentesis. ] Fam Pract 1981; female adolescent. Am] Dis Child 1977; l31:1094-6. 13:95-101. 45. Handsfield HH, Lipman TO, Harnisch ]P, Tronca 60. Patten RM, Vincent LM, Wolner-Hanssen P, E, Holmes KK. Asymptomatic gonorrhea in men. Thorpe E]r. Pelvic inflammatory disease. Endovagi­ Diagnosis, natural course, prevalence, and signifi­ nal sonography with laparoscopic correlation.] Ul­ cance. N EnglJ Med 1974; 290: 117 -23. trasound Med 1990; 9:681-9. 46. Bell TA, Grayston]T. Centers for Disease Control 61. Cacciatore B, Leminen A, Ingman-Friberg S, guidelines for prevention and control of Chlamydia Ylostalo P, Paavonen ]. Transvaginal sonographic trachomatis infections. Ann Intern Med 1986; findings in ambulatory patients with suspected pelvic 104:524-6. inflammatory disease. Obstet Gynecol 1992; 80:912 -6. 47. Wasserheit ]N, Bell TA, Kiviat NB, Wolner­ 62. Walker CK, Landers DV. Pelvic abscesses: new Hanssen P, Zabriskie V, Kirby BD, et al. Microbial trends in management. Obstet Gynecol Surv 1991; causes of proven pelvic inflammatory disease and ef­ 46:615-24. ficacy of clindamycin and tobramycin. Ann Intern 63. Paavonen], Aine R, Teisala K, Heinonen PK, Med 1986; 104:187-93. Punnonen R. Comparison of endometrial biopsy 48. Allen LA, Schoon MG. Laparoscopic diagnosis of and peritoneal fluid cytologic testing with laparos­ acute pelvic inflammatory disease. Br ] Obstet copy in the diagnosis of acute pelvic inflammatory Gynaecol1983; 90:966-8. disease. Am] Obstet Gyneco11985; 151:645-50. 49. Barnes RB, Roy S, Yee B, Duda M], Mishell DR]r. 64. Sellors], Mahony], Goldsmith C, Rath D, Mander Reliability of urinary pregnancy tests in the diagno­ R, Hunter B, et al. The accuracy of clinical findings sis of ectopic pregnancy. ] Reprod Med 1985; and laparoscopy in pelvic inflammatory disease. Am 30:827-31. ] ObstetGynecol1991; 164:113-20. 50. Cartwright PS, Victory DF, Moore RA, Anderson 65. Dodson MG, Faro S. The polymicrobial etiology of

]R, Brown DH. Performance of a new enzyme­ acute pelvic inflammatory disease and treatment regi­ http://www.jabfm.org/ linked immunoassay urine pregnancy test for the de­ mens. Rev Infect Dis 1985; 7(Supp 4):696S-702S. tection of ectopic gestation. Ann Emerg Med 1986; 66. Sweet RL. Sexually transmitted diseases, pelvic in­ 15:1198-9. flammatory disease, and infertility in women. Infect 51. Olson CM, Holt]A, Alenghat E, Grego S, Lumpkin Dis Clin North Am 1987; 1(1):199-215. ]R, Geanon GD. Limitations of qualitative serum 67. Peterson HB, Walker CK, Kahn ]G, Washington beta-HCG assays in the diagnosis of ectopic preg­ AE, Eschenbach DA, Faro S. Pelvic inflammatory nancy.] Reprod Med 1983; 28:838-42. disease: key treatment issues and options. ]AMA 52. Kim DS, Chung SR, ParkMI, Kim VP. Comparative 1991; 266:2605-11. on 23 September 2021 by guest. Protected copyright. review of diagnostic accuracy in tubal pregnancy: a 68. Hillis SD, ]oesoef R, Marchbanks PA, Wasserheit 14-year survey of 1040 cases. Obstet Gyneco11987; ]N, Cates W]r, Westrom L. Delayed care of pelvic 70:547-54. inflammatory disease as a risk factor for impaired 53. Panke ES, Yang LI, Leist PA, Magevney P, Fry R], fertility. Am] Obstet Gyneco11993; 168:1503-9. Lee RF. Comparison of Gen-Probe DNA probe test 69. Centers for Disease Control. 1993 sexually transmit­ and culture for the detection of Neisseria gonorrhoeae ted diseases treatment guidelines. MMWR 1993; in endocervical specimens. ] Clin Microbiol 1991; 42(RR-14):75-81. 29:883-8. 70. Jones RN, Barry AL, Aldridge KE, Gerlach EH. 54. Iwen PC, Blair TMH, Woods GL. Comparison of Comparative antimicrobial activity of aminothiazolyl the Gen-Probe PACE 2 system, direct fluorescent­ methoxyimino cephalosporins against anaerobic antibody, and cell culture for detecting Chlamydia bacteria, including 100 cefoxitin-resistant isolates. trachomatis in cervical specimens. Am] Clin Pathol Diagn Microbiol Infect Dis 1987; 8:157-63. 1991; 95:578-82. 71. Hooton TM, Rogers ME, Medina TG, Kuwamura 55. Lieberman RW, Wheelock]B. The diagnosis of gon­ L. Ewers C, Roberts PL, et al. Ciprofloxacin com­ orrhea in a low-prevalence female population: en­ pared with doxycycline for nongonococcal urethri­ zyme immunoassay versus culture. Obstet Gynecol tis. Ineffectiveness against Chlamydia trachomatis due 1987; 69:743-6. to relapsing infection.]AMA 1990; 264:1418-21.

122 JABFP March-April1994 Vol. 7 No.2 J Am Board Fam Pract: first published as 10.3122/jabfm.7.2.110 on 1 March 1994. Downloaded from

72. Faro S, Martens MG, Maccato M, Hammill HA, 85. Westrom L, Wolner-Hanssen P. Pathogenesis of pel­ Roberts S, Riddle G. Effectiveness of ofloxacin in vic inflammatory disease. Genitourin Med 1993; the treaUDent of Chlamydia trachomatis and Neisseria 69:9-17. gonorrhoeae cervical infection. Am J Obstet Gynecol 86. Patton DL, Moore DE, Spadoni LR, Soules MR, 1991; 164:1380-3. Halbert SA, Wang SP. A comparison of the fallopian 73. Soper DE, Brockwell NJ, Dalton HP. Microbial eti­ tube's response to overt and silent salpingitis. Obstet ology of urban emergency department acute salpin­ Gyneco11989; 73:622-30. gitis: treaUDent with ofloxacin. AmJ Obstet Gynecol 87. Horsburgh CR Jr, Douglas JM, LaForce FM. Pre­ 1992; 167:653-60. ventive strategies in sexually transmitted diseases for 74. Wendel GD Jr, Cox SM, Bawdon RE, Theriot SK, the primary care physician. US Preventive Services Heard MC, Nobles BJ. A randomized trial of Task Force.JAMA 1987; 258:814-21. ofloxacin versus cefoxitin and doxycycline in the out­ 88. Rosenberg MJ, Rojanapithayakorn W, Feldblum PJ, patient treaUDent of acute salpingitis. Am J Obstet Higgins JE. Effect of the contraceptive sponge Gyneco11991; 164:1390-6. on chlamydial infection, gonorrhea, and candidiasis. 75. Schwarcz SI(, Zenilman JM, Schnell D, Knapp JS, A comparative clinical trial. JAMA 1987; 257: Hook EW 3d, Thompson S, et al. National sur­ 2308-12. veillance of in Neisseria 89. Buchan H, Villard-Mackintosh L, Vessey M, Yeates gonorrhoeae. The Gonococcal Isolate Surveillance D, McPherson K Epidemiology of pelvic inflamma­ Project.JAMA 1990; 264:1413-7. tory disease in parous women with special reference 76. Styrt B, Gorbach SL. Recent developments in the to intrauterine device use. Br J Obstet Gynaecol understanding of the pathogenesis and treaUDent of 1990; 97:780-8. anaerobic infections (2). N Engl J Med 1989; 90. Rice PA, Schachter J. Pathogenesis of pelvic inflam­ 321:298-302. matory disease. What are the questions? JAMA 77. Brunham Re. Therapy for acute pelvic inflamma­ 1991; 266:2587-93. tory disease: a critique of recent treaUDent trials. Am 91. Phillips RS, Safran C, Aronson MD, Taylor We. J Obstet Gyneco11984; 148:235-40. Should women be tested for gonococcal infection of 78. WoIner-Hanssen P, Paavonen J, Kiviat N, Landers the cervix during routine gynecologic visits? An eco­ D, Sweet RL, Eschenbach DA, et al. Ambulatory nomic appraisal. AmJ Med 1989; 86:297-302. treaUDent of suspected pelvic inflammatory disease 92. Phillips RS, Aronson MD, Taylor WC, Safran C. with augmentin, with or without doxycycline. Am J Should tests for Chlamydia trachomatis cervical infec­ ObstetGynecol1988; 158:577-9. tion be done during routine gynecologic visits? An 79. Reed SD, Landers DV, Sweet RL. Antibiotic treaUDent analysis of the costs of alternative strategies. Ann In­ of tuboovarian abscess: comparison of broad-spectrum tern Med 1987; 107:188-94. beta-Iactam agents versus clindamycin- containing reg­ 93. Phillips RS, Hanff PA, Holmes MD, Wertheimer A,

imens. AmJ Obstet Gynecoll991; 164: 15 56-62. Aronson MD. Chlamydia trachomatis cervical infec­ http://www.jabfm.org/ 80. Westrom L, Joesoef R, Reynolds G, Hagdu A, tion in women seeking routine gynecologic care: cri­ Thompson SE. Pelvic inflammatory disease and fer­ teria for selective testing. Am J Med 1989; tility. A cohort study of 1844 women with 86:515-20. laparoscopically verified disease and 657 control 94. Phillips RS, Hanff PA, Wertheimer A, Aronson women with normal laparoscopic results. Sex MD. Gonorrhea in women seen for routine gyneco­ Transm Dis 1992; 19:185-92. logic care: criteria for testing. Am J Med 1988; 81. Washington AE, Katz P. Ectopic pregnancy in the 85:177-82. United States: economic consequences and payment 95. Sinei SI(, Schulz KF, Lamptey PRo Grimes DA, MatiJK, on 23 September 2021 by guest. Protected copyright. source trends. Obstet Gyneco11993; 81:287-92. Rosenthal SM, et al. Preventing ruCD-related pelvic 82. Chow WH, Daling JR, Cates W Jr, Greenberg RS. infection: the efficacy of prophylactic doxycycline at Epidemiology of ectopic pregnancy. Epidemiol Rev insertion. Br J Obstet Gynaeco11990; 97:412-9. 1987; 9:70-94. ' 96. Grimes DA, Schulz KF, Cates WJr. Prophylactic an­ 83. Washington AE, Cates W Jr, Wasserheit IN. Pre­ tibiotics for curettage abortion. Am J Obstet venting pelvic inflammatory disease. JAMA 1991; Gyneco11984; 150:689-94. 266:2574-80. 97. ThomasonJL, Gelbart SM, Scaglione NJ. Bacterial 84. Cates W Jr, Wasserheit IN. Gonorrhea, chlamydia, vaginosis: current review with indications for asymp­ and pelvic inflammatory disease. Curr Opin Infect tomatic therapy. Am J Obstet Gynecol 1991; 165: Dis 1990; 3:10-9. 1210-7.

Pelvic Inflammatory Disease 123