bundle demonstrated a significant but opposite association of the partial bundles: the HR of the maintenance bundle was 1.68 [1.19‐2.36], i.e. a 68% increased CRBSI risk when the two daily checks were complied with, but the HR of the insertion bundle was 0.50 [0.30‐0.85], i.e. a 50% reduced risk when the four insertion items were complied with. Ongoing participation, overall nor for the first four years only, and calendar time were not significant. Catheter duration Compliance to the bundle, especially the maintenance bundle, could result in shorter catheter stays, which also contributes to a smaller CRBSI risk. In our study, bundle compliance was associated with a small but significant decrease in catheter duration, when adjusted for patient and CVC characteristics: ‐0.15 days (95% CI ‐0.05 to ‐0.26). When the two partial bundles were modelled instead, the compliance to the maintenance bundle was associated with a significant decrease in duration: ‐0.29 days (‐ 0.18 to ‐0.39), whereas the insertion bundle was associated with a similarly small increase: +0.29 days (0.12 ‐ 0.47). Adding calendar time and number of participation years to the model resulted in comparable estimates (not shown). DISCUSSION During and following a national patient safety programme, introducing a CRBSI intervention bundle, Dutch hospitals successfully decreased their CRBSI rates. Overall bundle compliance increased from 60% in the first to 73% in the fourth bundle participation year. In the ICU there was no association between bundle compliance and CRBSI risk. Outside the ICU, where the incidence density of infection was approximately 2.5 times higher than in the ICU, compliance with the overall bundle was marginally significantly associated with the CRBSI risk. However, the effect was composed of a significant risk reduction of 50% when the insertion bundle was complied with and a significant risk increase of 68% when the maintenance bundle was complied with. This last finding might be the result of an increased awareness to diagnose CRBSI when a daily check on the insertion site was performed. The same analysis performed on CVCs with a duration of at least six days showed similar results overall and for CVCs with ≥ 6 days in ICU, but the effect of ongoing participation/calendar time was now insignificant. For CVCs with ≥ 6 days outside the ICU the effect of insertion bundle compliance was no longer significant, whereas compliance to the maintenance bundle remained significantly associated, as can be expected (data not shown). 9 227 The effect of a CRBSI prevention bundle in the Netherlands
RkJQdWJsaXNoZXIy MTk4NDMw