Practice Easily Missed?

Pelvic inflammatory disease

BMJ 2013; 346 doi: (Published 23 May 2013) Cite this as: BMJ 2013;346:f3189
  1. Emily C Bartlett, foundation year 2 doctor in genitourinary medicine1,
  2. Wendy B Levison, general practitioner2,
  3. Pat E Munday, retired consultant genitourinary physician3
  1. 1King’s College Hospital, Denmark Hill, London SE5 9RS, UK
  2. 2Callowland Surgery, Watford, Hertfordshire, UK
  3. 3West Hertfordshire Hospitals NHS Trust, Watford General Hospital, Watford, Hertfordshire, UK
  1. Correspondence to: E C Bartlett Emily.Bartlett{at}
  • Accepted 12 April 2013

A 22 year old woman in a new sexual relationship presented to her general practitioner with slightly painful intermenstrual bleeding for 10 days. She had pelvic tenderness but no cervical excitation on bimanual examination. The doctor suspected that the bleeding was a side effect of the oral contraceptive pill, started three months earlier. However, as he was unable to rule out pelvic inflammatory disease (PID), endocervical swab samples were taken to test for the sexually transmitted infections caused by Neisseria gonorrhoeae and Chlamydia trachomatis, and broad spectrum antibiotics prescribed, after which the symptoms resolved. The swab sample confirmed the presence of C trachomatis.

What is PID?

PID is due to infection of the upper female genital tract resulting in a wide range of pelvic pathology, from mild endometritis to pelvic peritonitis. Causative organisms are either sexually transmitted (such as C trachomatis, N gonorrhoeae, Mycoplasma genitalium) or endogenous vaginal organisms (for example, Bacteroides species) that ascend into the pelvic area from the lower genital tract through the cervix.

How common is PID?

  • Because of the difficulties in diagnosis, the exact prevalence of PID is not known

  • Recent data suggest that the rate of definite PID diagnosis in primary care is about 280/100 000 person years1

  • This may underestimate the true incidence, as many cases are relatively asymptomatic

  • PID is most common in women aged 20-291

Why is PID missed?

PID may present with classic symptoms and signs or with mild and subtle symptoms that may be easy to miss, according to a review of the evidence in national guidelines written by the British Association for Sexual Health and HIV.2

The best recent data on easily missed PID come from a retrospective audit in a walk-in centre for sexual health. This showed significantly different rates for the diagnosis of PID among 23 doctors, compared with similar rates for genital warts (a control condition). Although overdiagnosis of PID could not be excluded, the authors postulated that if high diagnosing doctors were overdiagnosing PID, then some women would not display the epidemiological and clinical features of the condition. Instead, the women with a diagnosis by both high and low diagnosing doctors had a similar epidemiological and clinical profile, suggesting underdiagnosis of PID by low diagnosing doctors. It is possible that some doctors may have been looking for severe symptoms, thus having too high a threshold for diagnosis.3

Why does this matter?

The sequelae of PID include ectopic pregnancy, infertility, and chronic pelvic pain, and the economic costs of treating these are substantial.4 The best outcome data on sequelae of PID come from a longitudinal cohort study concerning 2500 women, summarised by Weström et al.5 A retrospective reanalysis of these data showed that delaying antibiotic treatment for more than three days after the onset of symptoms was associated with a nearly threefold greater risk of post-PID infertility or ectopic pregnancy than in those who sought care promptly (absolute risk 20% v 8.3%).6 A retrospective cohort study in the United States similarly showed that women who had symptoms for two or more days before treatment were more likely to experience infertility than those with symptoms for two or fewer days before treatment (absolute risk 14% v 54%).7 In a third study with a different patient population, the adverse outcomes were less strongly correlated with time to treatment.8

How is PID diagnosed?


Although not 100% sensitive, laparoscopy is the ideal investigation to diagnosis PID.9 As this procedure is not available or appropriate in primary care, a clinical diagnosis must be made based on symptoms and signs and the exclusion of competing diagnoses.

PID is most common in women aged 20-29.1 Historically, the diagnosis of PID has been based on the findings of abdominal tenderness with or without adnexal or cervical excitation tenderness on bimanual examination, with supporting results from investigations confirming an infective or inflammatory process. In one large study of women admitted to hospital with a clinical diagnosis of PID, the sensitivity of abdominal tenderness, cervical motion tenderness, and adnexal tenderness compared with diagnostic laparoscopy was 61.2%, 79.9%, and 90.3% respectively.10 In a logistic model, the presence of these three signs together was found to be 82.1% sensitive but only 28.6% specific. Other symptoms and signs include nausea or vomiting (sensitivity 27.6%, specificity 69.0%), vaginal discharge (73.1%, 35.7%), vaginal bleeding (16.4%, 73.8%), urinary symptoms (21.6%, 83.3%), proctitis (7.5%, 95.2%), fever (29.9%, 69.0%), and an adnexal mass (90.3%, 21.4%).9

Similar studies have confirmed that these symptoms and signs are neither sensitive nor specific when compared with laparoscopy.11 12 Therefore no combination of symptoms and signs has been adopted in UK national guidelines as the minimal criteria for diagnosis, and a high index of suspicion for PID is recommended.13

In practice a thorough history screening for symptoms and a sexual history should be taken. A change of partner or lack of condom use may be important, as sexually transmitted organisms are common causal agents. A history of recent obstetric events (such as vaginal delivery or miscarriage) or gynaecological procedures (such as the insertion of an intrauterine device or termination of pregnancy) is relevant, as disturbance to the cervix may introduce infection to the upper genital tract. Assessment should include an abdominal and vaginal bimanual examination and visualisation of the cervix.

Where the history and examination findings are compatible with a diagnosis of PID, the patient should be treated presumptively.


As the diagnosis of PID is made clinically, investigations to confirm an infective or inflammatory process are required to lend support to it. Given the importance of early antibiotic treatment to prevent sequelae, positive results should not be a prerequisite to justify antibiotic treatment.13

Endocervical swabs for C trachomatis and N gonorrhoeae, using nucleic acid amplification tests where available, are recommended in all patients with suspected PID.2 14 A positive result shows the presence of one of these organisms in the lower genital tract, supporting the diagnosis of PID. However, as these organisms were detected in less than 12% of 22 000 cases of PID seen in genitourinary medicine clinics in England in 2011, a negative swab result does not exclude a diagnosis of PID.15

Ultrasound scans are helpful in excluding other diagnostic possibilities, but studies showing both high sensitivity and specificity for PID are lacking.16

An increased erythrocyte sedimentation rate (>15 mm/hr) supports a clinical diagnosis. It is a sensitive (81%) but non-specific (33%) predictor of PID.17

Ectopic pregnancy, which is a major differential diagnosis, should be excluded by a pregnancy test.

How is PID managed?

Mild or moderate disease can be safely managed in the outpatient setting, whereas clinically severe disease requires hospital admission for parenteral antibiotics.2 13 In the outpatient setting, oral antibiotics should be prescribed to cover C trachomatis, N gonorrhoeae, and anaerobes.2 13

A 14 day course of antibiotics (ceftriaxone, doxycycline, and metronidazole) is recommended, as the optimum treatment duration is unknown, and few data exist to support a shorter course.18 19 Although some doctors may be concerned about prescribing unnecessary antibiotics, treatment is recommended because the consequences of missing the diagnosis are profound. In those who fail to respond to treatment, laparoscopy is essential to confirm the diagnosis or to make an alternative diagnosis. Once PID has been diagnosed all current sexual partners should be screened and offered empirical treatment, and recent partners (from the previous six months) should be screened for sexually transmitted infections.2 13

Key points

  • Consider a diagnosis of PID in young women with any combination of bilateral lower abdominal pain, abnormal vaginal or cervical discharge, abnormal vaginal bleeding, deep dyspareunia, fever, adnexal tenderness, and cervical motion tenderness

  • A history of a new sexual partner, change of contraceptive from barrier contraception, or previous PID may be crucial in aiding the diagnosis

  • If PID is suspected, treat empirically with a 14 day course of antibiotics to cover Chlamydia trachomatis, Neisseria gonorrhoeae, and anaerobes, to achieve clinical and microbiological resolution

  • Subsequent infertility may be prevented with early antibiotic treatment


Cite this as: BMJ 2013;346:f3189


  • This is one of a series of occasional articles highlighting conditions that may be more common than many doctors realise or may be missed at first presentation. The series advisers are Anthony Harnden, university lecturer in general practice, Department of Primary Health Care, University of Oxford, and Richard Lehman, general practitioner, Banbury. To suggest a topic for this series, please email us at easilymissed{at}

  • We thank J Ross for his helpful comments.

  • Contributors: ECB had the original idea for the paper and wrote the first draft. PEM and WBL revised the manuscript and contributed to the final draft. All authors approved the final version to be published. PEM is the guarantor.

  • Competing interests: All authors have completed the ICMJE uniform disclosure form at (available on request from the corresponding author) and declare: no support from any organisation for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work in the previous three years, no other relationships or activities that could appear to have influenced the submitted work.

  • Provenance and peer review: Not commissioned; externally peer reviewed.

  • Patient consent not required (patient anonymised, dead, or hypothetical).


View Abstract

Sign in

Log in through your institution

Free trial

Register for a free trial to to receive unlimited access to all content on for 14 days.
Sign up for a free trial