What is the likely impact on surgical site infections in Australian hospitals if smoking rates are reduced? A cost analysis

Abstract

Introduction Assisting smokers to quit before surgery reduces surgical site infection (SSI) risk. The short-term economic benefits of reducing SSIs by embedding tobacco dependence treatment in Australian hospitals are unknown. Estimated annual number of SSIs prevented, and hospital bed-days (HBD) and costs saved from reducing smoking before surgery are calculated. Methods The most recent number of surgical procedures and SSI rates for Australia were sourced. The number of smokers and non-smokers having a SSI were calculated using the UK Royal College of Physicians reported adjusted odds ratio (1.79), and the proportion of SSIs attributable to smoking calculated. The potential impact fraction was used to estimate reductions in SSIs and associated HBDs and costs from reducing the smoking rates among surgical patients from 23.9% to 10% or 5% targets. Uncertainty around the final estimates was calculated using probabilistic sensitivity analysis. Results In 2016–17, approximately 40,593 (95% UI 32,543, 50,239) people having a surgical procedure in Australia experienced a SSI leading to 101,888 extra days (95% UI 49,988, 200,822) in hospital. If the smoking rate among surgical patients was reduced to 10%, 3,580 (95% UI 2,312, 5,178) SSIs would be prevented, and 8,985 (95% UI 4,094, 19,153) HBDs and $19.1M (95% UI $7.7M, $42.5M) saved in one year. If the smoking rate was reduced to 5%, 4,867 (95% UI 3,268, 6,867) SSIs would be prevented, and 12,217 (95% UI 5,614, 25,642) HBDs and $26.0M (95% UI $10.8M, $57.0M) would be saved. Conclusions The findings suggest achieving smoking rate targets of 10% or 5% would provide substantial short-term health and economic benefits through reductions in SSIs. Embedding tobacco dependence treatment in Australian hospitals would provide value for money by reducing costs and improving clinical quality and safety. A more comprehensive, modelled economic evaluation synthesising the best available evidence is needed to confirm findings.