Are Antibiotics Necessary with Sacral Neuromodulation? A Cost-Effectiveness Analysis

Objectives: Currently, preoperative antibiotics are recommended prior to sacral neuromodulation (SNM) procedures, but there are no definitive guidelines to direct postoperative antibiotic prophylaxis. This study assessed cost-effectiveness of three antibiotic strategies for prevention of SNM infections requiring explantations.

Methods: A decision analysis model was created using TreeAge to compare three antibiotic strategies with SNM: intravenous preoperative and oral postoperative antibiotics (PrePostAbx), preoperative antibiotics without postoperative antibiotics (PreAbxOnly), and no antibiotics preoperatively or postoperatively (NoAbx). We modeled SNM success and continuation, risk of antibiotic side effects, infection requiring explantation, and re-implantation +/- repeat infection utilizing existing literature. Modeled costs (2022 $US) included antibiotics +/- side effects, explantation due to infection, and potential for re-implantation with additional infection risk. A 1-year time horizon was used given most infections develop during this time period, and effectiveness was reported as quality adjusted life years (QALY). Monte Carlo simulation (n=10,000) accounted for uncertainty in model inputs via a probabilistic sensitivity analysis and was used for primary analysis since costs and QALYs were not normally distributed as explants/reimplants were rare but had high impact on costs and QALYs. Monte Carlo Simulation calculated net monetary benefit (NMB) of the three scenarios, instead of the incremental cost-effectiveness ratio (ICER), due to very similar QALYs of the non-normally distributed data. Otherwise, minuscule changes in the QALY denominator leads to large ICER fluctuations, and NMB more easily compares >2 strategies.

Results: Monte Carlo Analysis showed PrePostAbx had the highest median costs ($28,848) which was $19 more than PreAbxOnly. NoAbx had the lowest median costs ($28,784) and highest QALY (0.95) when accounting for antibiotic side effects (PrePostAbx and PreAbxOnly QALY=0.91). In 70.2% of 10,000 iterations, NoAbx was dominant with lower costs and higher QALYs. At a $100,000 willingness-to-pay threshold, the resulting NMBs were similar (PrePostAbx=$62,158, PreAbxOnly=$62,176, and NoAbx=$66,216). Utilizing our initial infection rate of 3.23% from the literature, NoAbx resulted in the greatest NMB and therefore was the most cost-effective and preferred strategy. Importantly the baseline SNM infection rate had the highest contribution to variability in model calculations and if >5% experience infection/explantation, NoAbx would no longer be cost-effective.

Conclusions: Assuming an infection rate of < 5%, SNM performed without IV preoperative or postoperative oral antibiotics is cost-effective with the lowest costs and highest quality of life. Importantly, there is limited literature on infection rates without antibiotics for SNM and future research is needed to confirm model conclusions. Prophylactic antibiotics are given for suspected decreased infection rates, but with low baseline rates, any marginal gains should be balanced against other consequences. Our field has seen significant changes in prophylactic antibiotics recommendations for procedures such as hysteroscopy and diagnostic laparoscopy. It may be time to consider a change in prophylactic antibiotics for SNM.