Optimal antiseptic skin preparation agents for minimizing surgical site infection following surgery: cost and cost-effectiveness analysis

Abstract

Abstract Background The application of antiseptic skin agents prior to incision minimizes the rate of surgical site infection. Despite their ubiquity, the optimal skin preparation agent remains uncertain. A retrospective economic analysis was conducted to complement the results from the NEWSkin Prep trial which prospectively compared three preparation agents. Methods A cost and cost-effectiveness analysis was performed from a healthcare service perspective to compare chlorhexidine with 70% ethanol, and aqueous povidone-iodine, against povidone-iodine with 70% ethanol. Resource use estimates accounted for hospital admissions, readmissions associated with surgical site infection, outpatient and general practitioner attendances, visits from community nurses and therapeutic consumables. The measure of effectiveness comprised the net difference in number of patients with surgical site infections per 1000 patients. Costs were compared using a two-sample Welch's t-test. Deterministic and probabilistic sensitivity analyses were performed to evaluate the incremental cost-effectiveness ratio. Results The null hypothesis that the mean costs for the trial arms were significantly different was not rejected (Welch's t-test P value: 0.771 for chlorhexidine with 70% ethanol against povidone-iodine with 70% ethanol; and 0.955 for aqueous povidone-iodine against povidone-iodine with 70% ethanol). Based on bootstrap averages, the chlorhexidine with 70% ethanol intervention generated 8.0 fewer surgical site infections per 1000 patients and net cost savings of €151,698 (Euros) per 1000 patients compared with povidone-iodine with 70% ethanol, and aqueous povidone-iodine produced a net cost saving of €37,494 per 1000 patients but generated an additional 11.6 surgical site infections per 1000 patients compared with povidone-iodine with 70% ethanol. The comparison of chlorhexidine with 70% ethanol to povidone-iodine with 70% ethanol was sensitive to the inclusion of cost outliers, while the comparison of aqueous povidone-iodine to povidone-iodine with 70% ethanol was sensitive to the estimated cost per surgical site infection. Conclusion Based on the outcomes from the NEWSkin Prep study, this economic analysis found no definitive evidence in favour of any one of the study comparators. Future model-based economic analyses of alternative skin preparations should critically address the quality of evidence and integrate the results from the NEWSkin Prep study.