Background: Pelvic inflammatory disease (PID) is mainly treated in the outpatient setting. The CDC recommends several antibiotic treatment regimens, however there is no evidence of differences in PID complication outcomes between treatments while almost three-fold differences in cost exist between the least and most expensive.

Methods: We constructed a Markov decision model to estimate the incremental cost-effectiveness of CDC recommended outpatient antibiotic regimens for mild to moderate PID (i.e. when severe illness was absent), calculating incremental costs per QALY gained if PID complication rates differed between regimens. The baseline model examined 18-year-olds from the societal perspective over a 10 year time horizon using a 3% discount rate. PID outcome frequencies and utilities were obtained by primary data collection; 2004 US average wholesale prices were used for antibiotic costs. PID outcome costs were obtained from the literature. Multiple sensitivity analyses were performed.

Results: Antibiotic costs vary between $64 (cefoxitin and doxycycline [CD]) and $188 (ofloxacin and metronidazole [OM]). If the more expensive therapy decreases the relative risk of PID complications by 1% compared to the less expensive (due to greater effectiveness and/or adherence) then: 1) OM compared to CD costs $38,300/QALY gained and 2) the more expensive therapy is cost saving if the cost difference is £$11, or costs £$50,000/QALY if the cost difference is £$185. A nearly linear relationship between greater decreases in relative PID complication risk and cost difference impact is noted (e.g. if the relative risk difference is 2%, cost savings occur if cost differences are £$22).

Conclusion: Within the cost range of CDC recommended antibiotic regimens for outpatient PID treatment, use of more expensive antibiotics is economically reasonable if relatively small decreases in PID complication rates occur due to greater treatment effectiveness or adherence. Further work is needed to investigate differences in antibiotic effectiveness for modifying PID complication risk.