Factors predicting STI-positive cervicitis and pelvic inflammatory disease, and implications on antibiotic use: a cross-sectional analysis.

Background: Clinical diagnoses of cervicitis and pelvic inflammatory disease (PID) are often subjective, and not all are associated with sexually transmitted infections (STI). Presumptive treatment for these syndromes may lead to unnecessary antibiotics in those subsequently found to be STI-negative. We investigated patient and clinician characteristics associated with accurately diagnosing bacterial STI-positive cervicitis and PID. We quantified the amount of unnecessary antibiotics prescribed in STI-negative cases. Methods: A cross-sectional study of 243 cervicitis and 179 PID patients presenting to Sydney Sexual Health Centre between January 2020 and 2023 was conducted. Patient demographics, sexual behaviours, symptoms, examination, results and clinician type were extracted from electronic medical records. Results: A total of 70 cervicitis (28.8%) and 28 PID (15.6%) patients were bacterial-STI positive. Age <25 years (aOR 2.78, 95% CI 1.25–6.25, P < 0.001) was associated with bacterial STI-positive cervicitis; presence of endocervical mucopurulent discharge on examination (aOR 6.05, 95% CI 2.20–16.68, P < 0.001) and current use of contraception (aOR 7.69, 95% CI 2.13–25.00, P = 0.002) were associated with bacterial STI-positive PID. For cervicitis, there were 139 doxycycline, 11 azithromycin and 5 ceftriaxone courses that were given to STI-negative patients. For PID, there were 148 doxycycline, 7 azithromycin and 152 ceftriaxone courses that were given to STI-negative patients. Conclusions: This study identified several characteristics that predicted STI-positive cervicitis and PID. However, none predicted STI-positivity with high degrees of confidence. Presumptive treatment leads to significant unnecessary antibiotic use, but ceftriaxone may have additional coverage beyond gonorrhoea. Rapid molecular point-of-care testing should be trialled to reduce unnecessary antibiotics for these syndromes – starting with cervicitis. This study uses a cross-sectional dataset from a sexual health clinic to determine which patient factors are associated with accurately diagnosing cases of cervicitis and pelvic inflammatory disease caused by STIs, and to quantify the amount of antibiotics prescribed for cases found to be STI-negative. Hence, it points out potential ways the current management paradigm leads to antibiotic overuse. Further research and investment into rapid point-of-care testing is needed to improve precision in antibiotic prescribing for these complex syndromes.