Cost-effectiveness of pre-operative Staphylococcus aureus screening and decolonization

Abstract

Objective: We developed a decision analytic model to evaluate the impact of a preoperative Staphylococcus aureus decolonization bundle on surgical site infections (SSIs), health-care-associated costs (HCACs), and deaths due to SSI.

Methods: Our model population comprised US adults undergoing elective surgery. We evaluated 3 self-administered preoperative strategies: (1) the standard of care (SOC) consisting of 2 disinfectant soap showers; (2) the "test-and-treat" strategy consisting of the decolonization bundle including chlorhexidine gluconate (CHG) soap, CHG mouth rinse, and mupirocin nasal ointment for 5 days) if S. aureus was found at any of 4 screened sites (nasal, throat, axillary, perianal area), otherwise the SOC; and (3) the "treat-all" strategy consisting of the decolonization bundle for all patients, without S. aureus screening. Model parameters were derived primarily from a randomized controlled trial that measured the efficacy of the decolonization bundle for eradicating S. aureus.

Results: Under base-case assumptions, the treat-all strategy yielded the fewest SSIs and the lowest HCACs, followed by the test-and-treat strategy. In contrast, the SOC yielded the most SSIs and the highest HCACs. Consequently, relative to the SOC, the average savings per operation was $217 for the treat-all strategy and $123 for the test-and-treat strategy, and the average savings per per SSI prevented was $21,929 for the treat-all strategy and $15,166 for the test-and-treat strategy. All strategies were sensitive to the probability of acquiring an SSI and the increased risk if SSI if the patient was colonized with SA.

Conclusion: We predict that the treat-all strategy would be the most effective and cost-saving strategy for preventing SSIs. However, because this strategy might select more extensively for mupirocin-resistant S. aureus and cause more medication adverse effects than the test-and-treat approach or the SOC, additional studies are needed to define its comparative benefits and harms.

Fig. 1.

Decision analytic trees. The trees were designed to evaluate the impact of a preoperative (pre-op) decolonization bundle on reducing surgical site infections (SSIs), healthcare-associated costs, and death due to SSI. Carrier and noncarrier refer to carriage of Staphylococcus aureus (SA). (A) 3 strategies. (B) Subtree of strategy efficacy.

Fig. 2.

Surgical site infections (SSIs), muprocin resistance, and side effects from medications modeled for a population of 10,000 surgical patients. Modeled patients assigned to 1 of 3 preoperative treatment strategies: “treat all,” “test and treat,” and the standard of care (SOC).