Diagnosis

The diagnosis of PID remains a challenge due to the wide vari­ation in symptoms and signs which may overlap with other condi­tions. There is no single symptom, physical finding, or laboratory test that is sensitive or specific enough to definitively diagnose PID.

Clinical signs

As previously discussed, PID is usually a clinical diagnosis based on pelvic tenderness on bimanual examination indicated by cer­vical motion tenderness, adnexal tenderness, and uterine tender­ness. The positive predictive value of a clinical diagnosis of PID is around 65-90% when compared with laparoscopy. Unsurprisingly, the accuracy of the clinical diagnosis and the experience of the clin­ician have been shown to be positively correlated (49). Specificity is increased by finding signs of lower genital tract inflammation such as increased friability of the cervix with contact bleeding and a mucopurulent endocervical discharge. Systemic features such as fever and vomiting are rare in mild to moderate disease, but more common in severe infection, particularly with N. gonorrhoeae.

Laboratory testing

All patients with suspected PID should have a pregnancy test to rule out possible ectopic pregnancy. Vulvovaginal NAATs for N. gonorrhoeae and C. trachomatis should be taken, as well as tests for HIV and syphilis. Licensed NAATs for M. genitalium are now becoming available and testing should also be performed in women with PID where possible. Microscopy of a saline preparation of va­ginal fluid (‘wet mount’) should be performed looking for increased numbers of white cells (more than one leucocyte per epithelial cell). The absence of endocervical or vaginal pus cells has a good negative predictive value for a diagnosis of PID (94.5%) but their presence is non- specific with a poor positive predictive value (17%) (41).

Table 43.1 Criteria for diagnosis

(>10 mg/L) are also useful and have been found to have good sensi­tivity (93%) and specificity (83%) in some studies (51).

Imaging

Transvaginal ultrasound (TVUS) scanning is widely available, rela­tively non-invasive, and useful in ruling out other diagnoses such as a ruptured ovarian cyst or ectopic pregnancy.

Power Doppler technology detects increased vascularization consistent with hyperaemia and inflammation. Switching to this modality during the examination may increase diagnostic sensi­tivity. Power Doppler may also allow detection of PID at an earlier stage, before the typical sonographic signs of PID have developed. However, this technique requires a high level of expertise to differ­entiate significant pathology from ‘flash artefacts' generated by iliac and other pelvic vessels adjacent to the adnexae (53).

Magnetic resonance imaging (MRI) is more sensitive (95% vs 81%) and specific (89% vs 78%) than TVUS in women with laparoscopy-proven PID (54), but is more expensive and not typic­ally available for mild to moderate cases. MRI is useful in complex or equivocal cases and may reduce the need for diagnostic laparotomy.

Computed tomography (CT) is not commonly used as it is rela­tively insensitive, expensive, inaccessible, and leads to unacceptable radiation exposure in women of reproductive age. In hospitalized patients, and where there is diagnostic uncertainty, CT can be useful in confirming alternative diagnoses such as acute appendicitis or urinary tract pathology.

Laparoscopy

Laparoscopy has historically been used as the gold standard diag­nostic procedure against which other investigations have been compared; however, as the management of PID has shifted to the outpatient setting, its use has diminished.

Indications for laparoscopy include when pelvic or tubo-ovarian abscess is suspected, or where a different diagnosis cannot be ex­cluded. Also, laparoscopy should be considered when there is no clinical improvement after 3 days of adequate antibiotic therapy. As well as confirming (or refuting) the diagnosis, it also offers the chance to perform therapies such as abscess drainage, rinsing and suctioning, lysis of adhesions, or salpingectomy if indicated. However, laparoscopy has the disadvantage of being invasive, re­quiring a general anaesthetic, and being relatively expensive. Due to these factors and concerns around standardization, it is rarely used for routine diagnosis in mild or moderate disease.

Histology

Endometrial sampling is safe, relatively inexpensive, and less inva­sive than laparoscopy. Endometrial biopsy showing endometritis on histology has also been used as a gold standard in terms of PID diag­nosis. The classic finding of plasma cells in the endometrial stroma gives a sensitivity of 89-92% for predicting laparoscopically con­firmed salpingitis (57, 58). In the PEACH study, however, the pres­ence of histological endometritis was not correlated with a greater risk of long-term adverse sequelae (ectopic pregnancy and infer­tility) when compared with participants without endometritis (5), suggesting not all women with endometritis had tubal inflammation. It is also recognized that ‘uncomplicated' endocervical infection can cause endometritis, which would explain the lower specificities re­ported with this diagnostic test (63-87%). Other drawbacks include the risk of introducing infection, delayed diagnostic reports, and variable histopathological definitions and interpretation. Samples can, however, undergo traditional microbiological analysis or molecular testing for novel organisms. In future, newer methodolo­gies could be used to evaluate the interaction between pathogens and the host response in different individuals.