Should intravenous catheters be replaced routinely?

1036 www.thelancet.com Vol 380 September 22, 2012 Currently the US Centers for Disease Control and Prevention (CDC) state that peripheral catheters do not need to be replaced more frequently than every 72–96 h to reduce the risk of infection and phlebitis in adults. Although results from some observational studies have shown that the risk of phlebitis rises with increasing catheter dwell time, other studies have not confi rmed this fi nding. Catheter replacement trials are frequently limited by study design and small sample size. Therefore, the study in The Lancet by Claire Rickard and colleagues, which compares intravenous catheter replacement in adults every 3 days with replacement when clinically indicated, is a major contribution to this debate. It is a large (3283 patients), multisite, randomised trial with high quality methods, excellent enrolment (97%) and follow-up (100%), and broad inclusion criteria. The investigators postulated that occurrence of phlebitis and other complications would be equivalent when intravenous catheters were replaced when clinically indicated compared with routine changes every third day. Indeed, the occurrence of the primary outcome of phlebitis was 7% in both groups (absolute risk diff erence 0·41%, 95% CI –1·33 to 2·15). Rickard and colleagues acknowledge that the non-masking of research nurses was a limitation that could have biased the recording of phlebitis. However, the high quality of this study provides a strong basis for their Should intravenous catheters be replaced routinely? associated with death are inconsistent, because cost can be high, for example with death after many days in the ICU, or low with sudden postoperative death. How can we reconcile Pearse and colleagues’ study with that by Wunsch and colleagues, who looked at variation in critical care services across the USA, Canada, and western Europe? Wunsch identifi ed a substantial diff erence in ICU admissions, for example a ten-times diff erence between the USA and Germany, and a seventimes diff erence between the UK and Germany. The Netherlands, with one of the lowest mortality rates in Pearse and colleagues’ study, was in the lowest rank in terms of availability of ICU beds of the eight countries assessed by Wunsch and colleagues. Such data suggest that quality assurance in surgery relies on several factors, of which the availability of ICU beds is only one. In future studies, we need to learn more about the relevant issues and optimum processes to secure quality. Targets could include the type of intensive care beds needed, volume, university versus community hospitals, and surgeons’ qualifi cations. Costs for the overall postoperative course would also be key, to allow us to propose cost-eff ective and relevant corrective measures.