Seven-day vs four-day infusion set replacement interval and catheter-related infections

Abstract

Importance Catheter-related bloodstream infections (CRBSIs) are associated with longer hospital stays and increased mortality risks. The Replacement at Standard vs Prolonged Interval (RSVP) trial found that central venous access devices and peripheral arterial catheter infusion set replacement intervals can be extended from 4 to 7 days without increasing infection risk by more than 2%. However, the trial did not consider the full cost and health trade-offs of extending replacement intervals. Objective To estimate the changes in total cost and health benefits associated with a decision to adopt 7-day vs standard 4-day replacement intervals via decision-analytic modeling. Design, Setting, and Participants In this economic evaluation, a decision tree model was developed on August 26, 2025, from a health care perspective incorporating data from patients in the RSVP randomized clinical trial conducted from 2011 to 2016 across 10 Australian hospitals. Data were analyzed from December 12, 2016, to April 23, 2019. Main Outcomes and Measures A probabilistic sensitivity analysis generated a joint distribution of the expected change to cost and effectiveness outcomes. Parameter uncertainty was assessed, a cost effectiveness acceptability curve was plotted, and a value of information analysis was done. Results Of 2941 patients from the RSVP trial (median age, 59.0 years; range, 47-68 years), 62.9% were male. Of patients, 62.6% were admitted to intensive care, and common specialties were medical, hematology, and emergency surgical. A CRBSI risk of 1.78% in the 7-day group and 1.46% in the standard 4-day group was reported. The 7-day strategy was associated with annual savings of approximately 52 million Australian dollars (A$) (95% uncertainty interval [UI], −A$42 841 427 to A$181 823 300) from fewer set changes, offset by A$3.1 million (95% UI, −A$6 974 903 to A$14 099 754) in additional costs for treating CRBSI. The expected health outcomes were 395 (95% UI, −945 to 1739) additional infections, 103 (95% UI, −246 to 452) excess deaths, and 1724 (95% UI, −4199 to 7925) life-years lost at the population level. At a willingness-to-pay threshold of A$28 033, the probability that the 7-day strategy is cost-effective was 50.3%, and the probability that it is cost-saving was 82.67%. Conclusions and Relevance In this economic evaluation of the RSVP trial, a small increase in CRBSI risk was associated with large cost-savings and substantial health losses at the population level. While the RSVP trial assumed that a 2% equivalence margin was acceptable, the use of such arbitrary difference margins might not consider information that could change decision-making outcomes. The clinical and economic impacts of small increases to infections appear to be important for decision-making inference.