• A diagnosis of pelvic inflammatory disease (PID) should be made on clinical grounds.
• Negative swab results do not rule out a diagnosis of PID.
• Do not delay making a diagnosis and initiating treatment whilst waiting for the results of laboratory tests.
• Ectopic pregnancy should be ruled out.
• Suspect PID if any of the following symptoms are present:
• Pelvic or lower abdominal pain (usually bilateral).
• Deep dyspareunia particularly of recent onset.
• Abnormal vaginal bleeding (intermenstrual, postcoital, or 'breakthrough') which may be secondary to associated cervicitis and endometritis.
• Abnormal vaginal or cervical discharge as a result of associated cervicitis, endometritis, or bacterial vaginosis. This is often very slight and may be transient, especially with chlamydial infection.
• Right upper quadrant pain due to peri-hepatitis (Fitz–Hugh–Curtis syndrome).
• Peri-hepatitis occurs in 10–20% of women with PID.
• It is characterized by the development of adhesions between the liver and the peritoneum, causing right upper quadrant pain.
• On examination look for:
• Lower abdominal tenderness — usually bilateral.
• Adnexal tenderness (with or without a palpable mass), cervical motion tenderness, uterine tenderness (on bimanual vaginal examination).
• Abnormal cervical or vaginal mucopurulent discharge (on speculum examination).
• A fever of greater than 38°C, although the temperature is often normal.
• Take endocervical swabs for gonorrhoea and chlamydia and a high vaginal swab. Consider taking blood for a white cell count, erythrocyte sedimentation rate (ESR), and C-reactive protein.
• For more information see Investigations.

Making a diagnosis of PID on history and examination alone

• These recommendation are based on expert opinion in guidelines on the management of acute pelvic inflammatory disease (PID) from the Royal College of Obstetricians and Gynaecologists [RCOG, 2009], the British Association for Sexual Health and HIV [BASHH, 2005a], and guidelines from the Department of Health and Human Services Centres for Disease Control and Prevention [CDC, 2006].
• Making an accurate clinical diagnosis of PID from the symptoms and signs has been described as 'little better than tossing a coin' [Ross, 2002] and clinicians should have a low threshold for initiating treatment.
• Symptoms and signs for suspected PID lack sensitivity and specificity [RCOG, 2009]. The positive predictive value of making a clinical diagnosis of PID compared with a laparoscopic diagnosis (using laparoscopic diagnosis as the gold standard) is 65–90% [RCOG, 2009]. However, although used as the gold standard, laparoscopy may have a low sensitivity, as 15–30% of women with suspected PID have no signs of acute infection on laparoscopy even when organisms have been found in their fallopian tubes [RCOG, 2009].
• The positive predictive value of a clinical diagnosis of PID also depends on epidemiological factors, including [CDC, 2006]:
• The age of the woman (PID is more common in adolescents).
• Whether the woman is attending a genito-urinary medicine clinic.
• Whether the woman is in a setting where the rates of chlamydia, gonorrhoea, and bacterial vaginosis are high [BASHH, 2005a].
• There is evidence from a large cross-sectional analysis that:
• Adnexal tenderness has a high sensitivity for PID.
• The finding most strongly associated with endometritis was a positive test result for Chlamydia trachomatis or Neisseria gonorrhoeae.

Endocervical and high vaginal swabs

• The recommendation that all women with suspected PID should be tested for C. trachomatis and N. gonorrhoeae (in general by taking endocervical swabs) is based on expert opinion in guidelines from the Royal College of Obstetricians and Gynaecologists [RCOG, 2009], the British Association for Sexual Health and HIV [BASHH, 2005a], the European guideline for the management of pelvic inflammatory disease [Ross et al, 2008], and guidelines from the Department of Health and Human Services Centres for Disease Control and Prevention [CDC, 2006]. These are the most common sexually transmitted organisms detected in PID.
• CKS recommends taking a high vaginal swab to look for other vaginal infections such as bacterial vaginosis and candidiasis.

Erythrocyte sedimentation rate (ESR), C-reactive protein, and leucocyte count

• Increased ESR, C-reactive protein, and leucocyte count supports the diagnosis of PID and can provide useful measures of disease severity [RCOG, 2009].
• However, the ESR or C-reactive protein and white cell count may be normal in mild or moderate PID [Ross et al, 2008].

• Risk factors for developing pelvic inflammatory disease include:
• Factors related to sexual behaviour:
• Young age (less than 25 years).
• Early age of first coitus.
• Multiple sexual partners.
• Recent new partner (within the previous 3 months).
• History of sexually transmitted infection in the woman or her partner.
• Recent instrumentation of the uterus or interruption of the cervical barrier:
• Termination of pregnancy.
• Insertion of an intrauterine device (within the past 6 weeks).
• Hysterosalpingography.
• In vitro fertilization and intrauterine insemination.

These recommendations are based on expert opinion in guidelines from the Royal College of Obstetricians and Gynaecologists [RCOG, 2009], the British Association for Sexual Health and HIV [BASHH, 2005a], the European guideline for the management of pelvic inflammatory disease [Ross et al, 2008], and a non-systematic review [Barrett and Taylor, 2005].

• Lower abdominal pain in a young woman may also be due to:
• An ectopic pregnancy.
• Threatened abortion.
• Ruptured corpus luteal cyst.
• Acute appendicitis:
• Nausea and vomiting occur in most women with appendicitis, but in only 50% of women with pelvic inflammatory disease.
• Cervical motion pain occurs in about 25% of women with acute appendicitis.
• Endometriosis.
• Gastrointestinal disorders (most commonly irritable bowel syndrome).
• Complications of an ovarian cyst, such as rupture, torsion, or haemorrhage.
• Urinary tract infection.
• Functional pain (that is of unknown physical origin). There may be longstanding symptoms.
• Mittelschmerz pain.

These recommendations are based on expert opinion in guidelines from the Royal College of Obstetricians and Gynaecologists [RCOG, 2009], the British Association for Sexual Health and HIV [BASHH, 2005a], the European guideline for the management of pelvic inflammatory disease [Ross et al, 2008], and a guideline that was adapted for use in the Department of Pediatrics at the University of Texas — Houston Health Science Center [Eissa and Cromwell, 2003].

• Do a pregnancy test if pregnancy is a possibility.
• Take endocervical swabs for Chlamydia trachomatis and Neisseria gonorrhoeae.
• If uncertain, confirm with the local laboratory which testing methods are available, the samples required, and how soon these should reach the laboratory.
• In general, chlamydia should be tested for by taking endocervical samples, using nucleic acid amplification tests (NAATs). First-catch urine samples or vulvovaginal swabs may be accepted for NAAT testing by some laboratories.
• Endocervical swabs for N. gonorrhoeae should be sent in transport medium to arrive at the laboratory for culture within 24 hours. NAATs may be used, but a positive NAAT should be confirmed using a second NAAT test with a different primer sequence, or with culture.
• Adequate sample collection is important. When taking an endocervical swab, the swab should be inserted inside the cervical os and firmly rotated against the endocervix. Swabbing a collection of discharge will result in an inadequate specimen, so it is generally recommended that excess cervical secretions are cleaned away prior to taking the swab.
• There is no need to take a urethral swab unless local guidelines suggest this.
• If possible, look for endocervical or vaginal pus cells under a microscope on a wet-mount vaginal smear:
• If absent, a diagnosis of pelvic inflammatory disease (PID) is unlikely.
• Excess leucocytes are associated with PID but they are also found in women with lower genital tract infection.
• Consider performing the following tests. If elevated they support the diagnosis of PID but are non-specific:
• Erythrocyte sedimentation rate (ESR).
• C-reactive protein.
• Leucocyte count.

Taking swabs for Chlamydia trachomatis and Neisseria gonorrhoeae

• The recommendation that all women with suspected PID should be tested for C. trachomatis and N. gonorrhoeae is based on expert opinion in guidelines from the Royal College of Obstetricians and Gynaecologists [RCOG, 2009], the British Association for Sexual Health and HIV [BASHH, 2005a], the European guideline for the management of pelvic inflammatory disease [Ross et al, 2008], and guidelines from the Department of Health and Human Services — Centres for Disease Control and Prevention [CDC, 2006].
• A positive result supports a diagnosis of PID and reinforces the need to treat sexual partners.
• However, a negative result does not exclude PID [BASHH, 2005a].

Interpreting the results of a wet-mount smear

• The absence of endocervical or vaginal pus cells suggests that the diagnosis is not PID. The negative predictive value for a diagnosis of PID is 95% (that is, 19/20 people who have a negative test result will not have the disease) [BASHH, 2005a; RCOG, 2009].
• The presence of pus cells is a non-specific finding. The positive predictive value is low (that is, the proportion of people who have a positive test result and who have PID is only 17%) [BASHH, 2005a; RCOG, 2009].

Erythrocyte sedimentation rate (ESR), C-reactive protein, and leucocyte count

• Increased ESR or C-reactive protein, and leucocytosis, all support the diagnosis of PID and can provide a useful measure of disease severity [RCOG, 2009].
• However, the ESR, C-reactive protein, or white cell count may be normal in mild or moderate PID [Ross et al, 2008].