INTRODUCTION — Pelvic inflammatory disease (PID) refers to acute and subclinical infection of the upper genital tract in females, involving any or all of the uterus, fallopian tubes, and ovaries; this is often accompanied by involvement of the neighboring pelvic organs. It results in endometritis, salpingitis, oophoritis, peritonitis, perihepatitis, and/or tubo-ovarian abscess.
The majority of PID cases (85 percent) are caused by sexually transmitted pathogens or bacterial vaginosis-associated pathogens. Fewer than 15 percent of acute PID cases are not sexually transmitted and instead are associated with enteric (eg, Escherichia coli, Bacteroides fragilis, Group B streptococci, and Campylobacter spp) or respiratory pathogens (eg, Haemophilus influenzae, Streptococcus pneumoniae, Group A streptococci, and Staphylococcus aureus) that have colonized the lower genital tract [1]. Post-operative pelvic peritonitis and abscess, pregnancy-related pelvic infection, injury or trauma-related pelvic infection, and pelvic infection secondary to spread of another infection (eg, appendicitis, diverticulitis, tumor) can also produce a very similar clinical picture. However, the etiologic differences among these processes, principally in that they are not caused by a sexually transmitted infection (STI), have significant implications for treatment and prevention. Infectious complications of gynecologic surgery and pregnancy are discussed elsewhere. (See "Posthysterectomy pelvic abscess" and "Septic pelvic thrombophlebitis" and "Postpartum endometritis".)
PID represents a spectrum of infection and there is no single diagnostic gold standard. Clinical diagnosis remains the most important practical approach. Several expert guidelines discuss the clinical approach to the diagnosis of PID. These include the United States Centers for Disease Control and Prevention guidelines on the management of STIs, the International Union against STI European guidelines for the management of PID, and the British Association for Sexual Health and HIV guidelines on the management of PID [2-4]. The discussion in this topic is generally consistent with these guidelines.
The clinical features and diagnosis of sexually transmitted PID will be reviewed here. The pathogenesis, microbiology, risk factors for acquisition, treatment, and sequelae associated with this disorder are discussed separately. (See "Pelvic inflammatory disease: Pathogenesis, microbiology, and risk factors" and "Pelvic inflammatory disease: Treatment in adults and adolescents".)
CLINICAL FEATURES
Patients at risk — Any sexually active female is at risk for sexually transmitted infection (STI) associated pelvic inflammatory disease (PID), but those with multiple sexual partners are at the highest risk. Additionally, age younger than 25, a partner with a sexually transmitted infection, and a history of prior PID or a sexually transmitted infection are important risk factors. The use of barrier contraception is protective but depends on appropriate use. These are discussed in detail elsewhere. (See "Pelvic inflammatory disease: Pathogenesis, microbiology, and risk factors", section on 'Risk factors'.)
While it is rare to have PID during pregnancy because the mucus plug and decidua seal off the uterus from ascending bacteria, PID can occur in the first 12 weeks of gestation before this occurs.
Females who undergo instrumentation of the cervix (eg, termination of pregnancy) are at higher risk of infection ascending to cause PID. Older females less commonly present with PID, but when they do, the cause is more likely to be non-STI related. Salpingitis has been reported in premenarcheal girls and adolescents who are not sexually active, but it is very rare. In such situations, respiratory and enteric bacteria should be also considered.
Spectrum of disease — The term PID encompasses a wide spectrum of clinical presentations. The time course of presentation is typically acute over several days, but a more indolent presentation over weeks to months can also occur. Some females do not present to care with symptoms of PID but are later suspected to have had it because of tubal factor infertility. Even acute symptomatic PID represents a spectrum of clinical disease, from mild, vague pelvic symptoms to severe pain associated with tubo-ovarian abscess and, rarely, fatal intra-abdominal sepsis. In some females, the inflammatory process can extend to the liver capsule to cause perihepatitis (the Fitz-Hugh Curtis syndrome).
These varied clinical syndromes are discussed in more detail in the sections that follow.
Acute symptomatic PID — Acute symptomatic PID is characterized by the acute onset of lower abdominal or pelvic pain, pelvic organ tenderness, and evidence of inflammation of the genital tract. The findings can be subtle and nonspecific.
Symptoms — Lower abdominal pain is the cardinal presenting symptom in females with PID. The abdominal pain is usually bilateral and rarely of more than two weeks' duration [2-4]. The character of the pain is variable, and in some cases, may be quite subtle. The recent onset of pain that worsens during coitus or with jarring movement may be the only presenting symptom of PID. The onset of pain during or shortly after menses is particularly suggestive [5].
The majority of females with PID have mild to moderate disease and only a minority develop peritonitis or pelvic abscess, which are usually manifest by more severe pain, greater tenderness on examination, and systemic features such as fever.
Abnormal uterine bleeding (post-coital bleeding, inter-menstrual bleeding, menorrhagia) occurs in one-third or more of patients with PID [6,7]. Other non-specific complaints include urinary frequency and abnormal vaginal discharge.
Examination findings — On physical examination, most females with PID have abdominal tenderness on palpation, greatest in the lower quadrants, which may or may not be symmetrical. Rebound tenderness, fever, and decreased bowel sounds are usually limited to females with more severe PID.
Acute cervical motion, uterine, and adnexal tenderness on bimanual pelvic examination are the defining characteristic of acute symptomatic PID [7,8]. Purulent endocervical discharge and/or vaginal discharge is also common. However, significant lateralization of adnexal tenderness is uncommon in mild to moderate PID.
Laboratory findings — Most laboratory findings in PID are nonspecific. Although PID is usually an acute process, only a minority of PID patients with more severe disease exhibit peripheral blood leukocytosis [9]. Similarly, an elevated erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) have poor sensitivity and specificity.
Perihepatitis — Perihepatitis (Fitz-Hugh Curtis Syndrome) occurs in the setting of PID when there is inflammation of the liver capsule and peritoneal surfaces of the anterior right upper quadrant. There is generally minimal stromal hepatic involvement. It occurs in approximately 10 percent of females with acute PID and is characterized by right upper quadrant abdominal pain with a distinct pleuritic component, sometimes referred to the right shoulder. Marked tenderness in the right upper quadrant can be seen on exam. The severity of the pain in this location may mask the diagnosis of PID and lead to concerns regarding cholecystitis [10]. Aminotransferases are usually normal or only slightly elevated [11,12].
On laparoscopy or visual inspection, perihepatitis manifests as a patchy purulent and fibrinous exudate ("violin string" adhesions), most prominently affecting the anterior surfaces of the liver (not the liver parenchyma).
The syndrome was first associated with gonococcal salpingitis in 1920 [13] and subsequently with Chlamydia trachomatis [14,15] and possibly Mycoplasma genitalium [16].
Tubo-ovarian abscess — A tubo-ovarian abscess is an inflammatory mass involving the fallopian tube, ovary, and, occasionally, other adjacent pelvic organs. Females with a tubo-ovarian abscess may have a palpable adnexal mass on examination. Other associated clinical findings are discussed elsewhere. (See "Epidemiology, clinical manifestations, and diagnosis of tubo-ovarian abscess", section on 'Clinical presentation'.)
Subclinical PID — Subclinical infection of the upper reproductive tract that does not prompt a woman to present to medical care but is severe enough to produce significant sequelae appears to be relatively common [7]. Females with tubal factor infertility that appears likely to have been a result of past episodes of PID often give no history of PID [17-19]. As an example, in one study of 112 infertile females, 36 had adhesions or distal tube occlusion on laparoscopy suggestive of PID, but only 11 had a documented history of a PID diagnosis [19].
Previously undiagnosed PID has also been identified in females with a history of previous mild symptoms, but with an endometrial biopsy that demonstrates excess neutrophils and plasma cells, consistent with inflammation and PID. Lower genital tract infection with gonorrhea, chlamydia, or bacterial vaginosis is a risk factor for this finding [20,21]. As an example, in a study that included 562 females at risk for but without clinical findings suggestive of PID, 13 percent of them had endometritis on endometrial biopsy, and rates of cervical C. trachomatis isolation were similar to females with clinically evident PID [7]. Subclinical episodes of PID may occur more frequently in oral contraceptive users [22,23].
Chronic PID — An indolent presentation of PID with low-grade fever, weight loss, and abdominal pain has been reported following pelvic infection due to actinomycosis and tuberculosis. An association between an indwelling IUD and risk of actinomycosis has been suggested, although this relationship remains unclear. The diagnosis of this pathogen is discussed elsewhere. (See "Abdominal actinomycosis" and "Clinical manifestations, diagnosis, and treatment of miliary tuberculosis", section on 'Clinical manifestations'.)
EVALUATION — The evaluation of the woman with acute pelvic pain is discussed in detail elsewhere (see "Evaluation of acute pelvic pain in nonpregnant adult women"). Elements of the evaluation that are specific to female patients with suspected PID are outlined below.
Initial evaluation — The possibility of PID should be considered in any sexually active female patient who presents with lower abdominal pain and pelvic discomfort. The index of suspicion for PID should be high, especially in adolescents. The goal of the initial evaluation of females with suspected PID is to establish a presumptive clinical diagnosis of PID, assess for additional findings that increase the likelihood of that diagnosis, and evaluate for other potential causes of pelvic pain.
A presumptive clinical diagnosis of PID can be made on the basis of history and physical exam findings alone. Although laboratory testing is also done at the initial evaluation of all patients with suspected PID, empiric treatment should not be delayed while awaiting results of these supportive tests. (See 'Diagnosis' below and "Pelvic inflammatory disease: Treatment in adults and adolescents", section on 'Threshold for treatment'.)
For females who are acutely ill and may have complications of PID, who do not improve with empiric therapy for PID, or in whom the diagnosis remains uncertain, additional diagnostic tests, such as pelvic imaging, can be useful. (See 'Additional evaluation for diagnostic uncertainty' below.)
History — The history should focus on potential risk factors for PID. In particular, a sexual history should be taken, assessing for new sexual partners and consistent use of condoms.
Additionally, the history should clarify the onset (usually recent) and character of pelvic pain (usually constant and aching), with the understanding that even subtle and mild symptoms can be consistent with PID (see 'Symptoms' above). Other symptoms such as prominent urinary or gastrointestinal symptoms that might indicate alternative diagnoses should also be sought. (See 'Differential diagnosis' below.)
Physical and pelvic exam — All females, including adolescents, suspected of having PID should undergo bimanual exam to evaluate for cervical motion, uterine, or adnexal tenderness. Additionally, speculum exam should be performed to evaluate for cervical mucopurulent discharge. (See 'Examination findings' above.)
Pelvic organ tenderness is the defining characteristic of acute symptomatic PID. One study found that adnexal tenderness was the sign that correlated best with the finding of endometritis on endometrial biopsy [8]. Other diagnoses should also be considered if uterine and adnexal tenderness are not prominent.
The presence of a palpable adnexal mass may suggest a tubo-ovarian abscess complicating PID, but it could also reflect other disease processes in the differential diagnosis of PID. (See "Approach to the patient with an adnexal mass", section on 'Diagnostic evaluation' and "Differential diagnosis of the adnexal mass" and "Epidemiology, clinical manifestations, and diagnosis of tubo-ovarian abscess".)
The value of conducting a pelvic examination in an asymptomatic adolescent has been questioned and received some attention by professional associations [2,24,25]. However, in an adolescent presenting with lower abdominal pain and vaginal discharge (spotting or bleeding), a presumptive diagnosis of cervicitis and PID salpingitis cannot be made without a pelvic examination [26]. Imaging studies will not reveal acute salpingitis. (See 'Imaging techniques' below.)
Point-of-care and laboratory tests — The following tests should be performed for all females suspected of having PID:
●Pregnancy test
●Microscopy of vaginal discharge (where available)
●Nucleic acid amplification tests (NAATs) for C. trachomatis and Neisseria gonorrhoeae
●NAAT for Mycoplasma genitalium (where available)
●HIV screening
●Serologic testing for syphilis
Testing for patients suspected of PID should always begin with a pregnancy test to rule out ectopic pregnancy and complications of an intrauterine pregnancy, the main obstetric differential diagnoses of PID. Saline microscopy of vaginal discharge is to assess for increased white blood cells (WBC) in vaginal fluid which is sensitive for PID. The absence of WBC could thus suggest an alternate diagnosis. However, the finding is not very specific for PID. Microscopy can also identify coexisting bacterial vaginosis and trichomoniasis. Positive NAATs for C. trachomatis, N. gonorrhoeae, or M. genitalium support the diagnosis of PID, but negative NAATs do not rule out PID. HIV and syphilis testing are to evaluate for other sexually transmitted infections that share similar risk factors with PID.
Additional tests can be potentially useful in certain situations. A Gram stain from cervical discharge can also be a useful diagnostic tool, but is often not available in community settings. If a cervical Gram stain is positive for gram-negative intracellular diplococci (suggestive of N. gonorrhoeae) when interpreted by an experienced microscopist, the probability of PID greatly increases [2-4]. However, if the Gram stain is negative, it is of limited value because most cases of PID are not caused by gonorrhoea, and the sensitivity of microscopy is only around 60 percent.
A complete blood count, erythrocyte sedimentation rate, and C-reactive protein are often obtained in patients seen in hospital-based settings who have more severe clinical presentations, including fever, and may warrant inpatient therapy [2]. These tests have low sensitivity and specificity for the diagnosis of PID in general, but can be useful in assessing severity (including cases of suspected tubo-ovarian abscess) and monitoring the response to treatment. Urinalysis should also be checked in females with urinary symptoms. Hepatitis B virus testing may be appropriate depending on the patient's risk history and vaccination history.
Testing for M. genitalium is useful to help guide the choice of specific antimicrobial therapy [3]. (See "Mycoplasma genitalium infection in males and females".)
Additional evaluation for diagnostic uncertainty — Additional testing may be warranted for females who are acutely ill (eg, with fever, peritonitis, or a pelvic mass), whose symptoms are atypical (eg, with an abnormal site or duration of symptoms) or do not improve significantly within 72 hours after starting empiric antibiotic therapy, or who have persisting pain after completing therapy. These findings suggest the possibility of complications of PID (such as a tubo-ovarian abscess) or alternate diagnosis. Pelvic imaging can be helpful to evaluate for these. Ultrasound is generally the preferred imaging modality if an abscess or adnexal pathology is suspected clinically, as it produces high-quality images of the upper genital tract and does not expose the patient to radiation. Otherwise, computed tomography (CT) or magnetic resonance imaging (MRI) may be more helpful in identifying gastrointestinal pathology or alternate diagnoses. (See 'Imaging techniques' below.)
Laparoscopy and transcervical endometrial biopsy are uncommonly performed. (See 'Other studies' below.)
Imaging techniques — Pelvic imaging can help evaluate for alternative causes of pelvic pain or complications of PID (such as a tubo-ovarian abscess). However, the absence of radiographic findings consistent with PID does not rule out the possibility of PID and should not be a reason to forgo or delay therapy for presumptive PID. Ultrasound is the imaging technique that has been most studied, and there is limited evidence for the use of CT or MRI in females with suspected PID [27-29]; however, CT and MRI are useful to exclude alternative diagnoses in females with an atypical and severe presentation.
The interpretation of sonographic findings is operator dependent, and there are often minimal changes seen in females with uncomplicated PID. Thickened, fluid-filled fallopian tubes with or without free pelvic fluid [30] and the cogwheel sign (cogwheel appearance on a cross-section of the tube) may be present. Doppler studies may demonstrate tubal hyperemia suggestive of pelvic infection. Among females who have endometritis, ultrasound may show fluid or gas within the endometrial canal, heterogeneous thickening, or indistinctness of the endometrial stripe, but these findings are inconsistent. When a tubo-ovarian abscess is present, a complex thick-walled, multilocular cystic collection can be seen in the adnexa, typically with internal echoes or multiple fluid levels. (See "Epidemiology, clinical manifestations, and diagnosis of tubo-ovarian abscess", section on 'Imaging studies'.)
Doppler ultrasound scanning may be useful to identify areas of increased blood flow associated with inflammation but is not used routinely because of the limited evidence to support its utility [30].
Other studies — Findings on laparoscopy or transcervical endometrial biopsy can confirm the clinical diagnosis of PID, but these tests are uncommonly performed.
●Laparoscopy – Despite its value in confirming a diagnosis of PID, laparoscopy is not sensitive enough to be considered the diagnostic gold standard. The specificity of laparoscopy is high, but its sensitivity is as low as 50 percent when compared with fimbrial histopathology because it does not detect isolated endometritis or mild intra-tubal inflammation [31]. Additionally, it is an invasive procedure, particularly for a condition that does not typically warrant surgical intervention, and it is not universally available in the acute setting.
Laparoscopy can be a useful part of the diagnostic workup for PID when imaging studies have not been definitively informative in the following situations:
•In a patient who has failed outpatient treatment for PID, to look for alternative causes of the patient's symptoms
•In a patient whose symptoms are not clearly improving or worsening after approximately 72 hours of inpatient treatment for PID, which suggests that PID may not be the correct diagnosis
In addition, some surgeons may proceed directly to laparoscopy in an acutely ill patient with a high suspicion of a competing diagnosis that would be diagnosed and intervened on through laparoscopy (eg, appendicitis). Consent for laparotomy at the same procedure should be obtained in advance for these patients.
Laparoscopic abnormalities consistent with PID include tubal erythema, edema, and adhesions; purulent exudate or cul-de-sac fluid; and abnormal fimbriae.
●Transcervical endometrial biopsy – This can be used to detect endometritis, which is associated with salpingitis. However, it is not used routinely because the correlation is not 100 percent, there is a delay associated with processing the biopsy (which means that the result seldom influences the decision to treat), and there may be difficulty in interpreting the histology due to the patchy nature of the inflammation, thus limiting consistency.
DIAGNOSIS — The presumptive clinical diagnosis of PID is made in sexually active young females, especially those at high risk for sexually transmitted infections (STIs), who present with pelvic or lower abdominal pain and have evidence of cervical motion, uterine, or adnexal tenderness on exam.
The sensitivity of this clinical diagnosis is only 65 to 90 percent [6,32,33], but because of the potential for serious reproductive sequelae if PID treatment is delayed or not given, this presumptive diagnosis is sufficient to warrant empiric antimicrobial therapy for PID. Even patients with minimal or subtle findings should be treated since the potential consequences of withholding therapy are great. (See "Pelvic inflammatory disease: Treatment in adults and adolescents", section on 'Threshold for treatment'.)
In general, adding more diagnostic criteria increases the specificity, but decreases the sensitivity of the diagnosis. The following additional findings can be used to support the clinical diagnosis of PID [2]:
●Oral temperature >101°F (>38.3°C)
●Abnormal cervical or vaginal mucopurulent discharge or cervical friability
●Presence of abundant numbers of white blood cells (WBCs) on saline microscopy of vaginal secretions (eg, >15 to 20 WBCs per high power field or more WBCs than epithelial cells)
●Documentation of cervical infection with N. gonorrhoeae, C. trachomatis, or M. genitalium
The United States Centers for Disease Control and Prevention also lists an elevated C-reactive protein (CRP) or erythrocyte sedimentation rate (ESR) as findings that may increase the specificity of the diagnosis of PID. However, these tests are not particularly specific. In one study, a CRP ≥60 mg/L (or 6 mg/dL) or ESR ≥40 mm/h had a specificity of only 61 percent for severe PID [34].
Certain findings can suggest against PID, such as the combination of normal cervical discharge and absence of WBCs on microscopy of vaginal secretions. Also, prominent gastrointestinal and urinary symptoms can suggest other etiologies of pelvic pain. (See 'Differential diagnosis' below.)
For females who have undergone additional testing, certain findings on pelvic imaging (eg, thickened, fluid-filled tubes/oviducts or tubo-ovarian complex), laparoscopy (eg, tubal erythema, edema or adhesions; purulent exudate), or biopsy (eg, histologic evidence of endometritis) can help to confirm the diagnosis of PID, although their absence does not rule out the possibility of PID. These are discussed in detail elsewhere. (See 'Additional evaluation for diagnostic uncertainty' above.)
Standards for the diagnosis of subclinical PID remain to be established. It is typically diagnosed retroactively in females who are ultimately found to have tubal factor infertility. Subclinical PID can also be identified incidentally in females undergoing laparoscopy for other reasons.
DIFFERENTIAL DIAGNOSIS — The differential diagnosis of PID is broad and includes other pelvic pathology, urinary tract processes, and gastrointestinal tract disorders. The main differential diagnoses for PID and their suggestive features are summarized in the Table (table 1). Other diagnoses associated with acute pelvic pain are discussed in further detail elsewhere. (See "Evaluation of acute pelvic pain in nonpregnant adult women", section on 'Causes'.)
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Sexually transmitted infections".)
INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5th to 6th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10th to 12th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.
Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)
●Basics topics (see "Patient education: Pelvic inflammatory disease (The Basics)")
●Beyond the Basics topics (see "Patient education: Gonorrhea (Beyond the Basics)" and "Patient education: Chlamydia (Beyond the Basics)")
SUMMARY AND RECOMMENDATIONS
●Definition – Pelvic inflammatory disease (PID) refers to acute infection of the upper genital tract structures, including the uterus, fallopian tubes, and/or ovaries, and encompasses a wide spectrum of clinical presentations. (See 'Spectrum of disease' above.)
●Presenting features – Patients with acute symptomatic PID generally complain of recent onset of lower abdominal pain in association with new vaginal discharge and/or intermenstrual bleeding. Constitutional symptoms may occur and include fever and chills. Perihepatitis can also occur and present with marked tenderness in the right upper quadrant. Pelvic organ tenderness on palpation is the defining exam finding. (See 'Clinical features' above.)
●When to suspect PID – The possibility of PID should be considered in any sexually active female patient who presents with pelvic discomfort. The index of suspicion for PID should be high especially in adolescents. Individuals with multiple sexual partners are at the highest risk. (See 'Initial evaluation' above and 'Patients at risk' above.)
●Initial evaluation – The goal of the initial evaluation of females with suspected PID is to establish a presumptive clinical diagnosis of PID, assess for additional findings that increase the likelihood of that diagnosis, and evaluate for other potential causes of pelvic pain. Additionally, infections or co-morbidities that can occur in individuals at risk for PID should be tested for. (See 'Initial evaluation' above.)
In addition to a history, including sexual history, and physical and pelvic exams, the following tests should be performed:
•Pregnancy test
•Microscopy of vaginal discharge (where available)
•Nucleic acid amplification tests for C. trachomatis, N. gonorrhoeae, and M. genitalium
•HIV screening
•Serologic testing for syphilis
●Additional evaluation for selected patients – For patients who are acutely ill, who do not improve with empiric therapy for PID, or in whom the diagnosis remains uncertain, imaging is warranted to evaluate for complications of PID (eg, tubo-ovarian abscess) or alternative diagnoses (table 1). (See 'Additional evaluation for diagnostic uncertainty' above and 'Differential diagnosis' above.)
●Presumptive clinical diagnosis – The presumptive clinical diagnosis of PID is made in sexually active young females, particularly those with risk factors for sexually transmitted infections (STIs), who present with pelvic or lower abdominal pain and have evidence of cervical motion, uterine, or adnexal tenderness on exam. (See 'Diagnosis' above.)
●Low threshold for presumptive treatment – Because of the potential for serious reproductive sequelae if PID treatment is delayed or not given, this presumptive diagnosis is sufficient to warrant empiric antimicrobial therapy for PID. (See "Pelvic inflammatory disease: Treatment in adults and adolescents", section on 'Threshold for treatment'.)
●Additional supportive findings – Other findings, such as fever, mucopurulent discharge or cervical friability, abundant white blood cells on saline microscopy of vaginal secretions, and detection of genital infection with N. gonorrhoeae, C. trachomatis, or M. genitalium, support the clinical diagnosis of PID. Although not obtained in all patients, imaging studies with characteristic findings, laparoscopy abnormalities consistent with PID, and histologic evidence of endometritis on biopsy can help to confirm the diagnosis. (See 'Diagnosis' above.)
ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Charles Livengood, MD, who contributed to an earlier version of this topic review.
