There is evidence from a large cross-sectional analysis that adnexal tenderness has a high sensitivity for pelvic inflammatory disease (PID) and that the finding most strongly associated with endometritis is a positive test result for Chlamydia trachomatis or Neisseria gonorrhoeae.

• A cross-sectional analysis of 651 women enrolled in the Pelvic Inflammatory Disease Evaluation and Clinical Health (PEACH) study considered the clinical predictors of endometritis in women with PID [Peipert et al, 2001]:
• Adnexal tenderness was found to have a high sensitivity (95%) but a low specificity (4%). This means that if this symptom was used alone then many women would be over-treated although it is unlikely many women with PID would be missed.
• Combining symptoms (lower abdominal tenderness, adnexal tenderness, and cervical motion tenderness) lowers the sensitivity to 83% but improves the specificity (22%).
• The finding most strongly associated with endometritis was a positive test result for Chlamydia trachomatis or Neisseria gonorrhoeae (adjusted odds ratio 4.3, 95% CI 2.89 to 2.63).
• Women who had a high temperature as well as a high white cell count were at a significantly increased risk of endometritis (p < 0.001) but women with an elevated temperature alone at presentation were at significantly less risk of having endometritis, possibly because the elevated temperature was due to another cause (such as viral gastroenteritis) that did not increase the white cell count [Ross, 2002].