The lack of association in the ICU might be explained by the low CRBSI incidence from the start compared with other studies at the moment the bundle was introduced [7]. In a European study (PROHIBIT), the implementation of a comparable bundle in ICUs was effective but not in one study arm (of three) where baseline CRBSI incidence was as low as 1.4/1000 CVC days already[8]. Ista et al. demonstrated in their meta‐analysis that risk reduction was significantly lower in studies with a lower baseline central line‐ associated bloodstream infection (CLABSI) incidence than in studies with CLABSI incidence > 5/1000 CVC days[9]. Another similarity between this study and ours was the relatively high compliance with the insertion bundle at baseline: 82% and 85%, respectively. Nevertheless, ICU rates decreased over time, which may be related to the decrease in catheter duration in the ICU and other, undocumented, improvements in infection prevention practices. Outside the ICU the CRBSI rate was higher though, possibly in part because CVCs here were more often used for total parenteral nutrition (47.9 versus 18.4%), and nursing staff probably had less experience with CVCs. In the Netherlands, central line‐teams, when present, are usually not active outside the ICU. Improving CVC insertion procedures outside the ICU led to a significant reduction in the risk to develop CRBSI. Unexpectedly, compliance with the maintenance bundle, which was lower here than in the ICU, was associated with an increased risk of infection. As mentioned above, compliance with the daily checks presumably increased the awareness of CRBSI whereas some cases might have gone unnoticed previously. It is also possible that the compliance increased in the event of a suspected increased risk of CRBSI of a patient. Furuya et al. found a marginally significant protective association between ‘daily check’ and CLABSI rates in univariable analysis, but in this study the compliance and CLABSI rates were reported on ICU level[10]. The relatively few studies that reported on the effect of (varying) CRBSI prevention bundles in non‐ICU wards have been mostly positive too[11‐14], but did not adjust for other relevant patient and CVC characteristics. A successful implementation of the intervention bundle, especially of the insertion bundle item regarding insertion site and the two daily checks, could also have resulted in a decrease in catheter duration and an increase in insertions via subclavian veins. The median overall catheter duration decreased from six to five days and, from four to three days in the ICU. Outside the ICU catheter duration did not decrease, but the obligation to monitor hospital‐wide from 2016 onwards may have impacted this. Compliance with the overall bundle was associated with a small but significant reduction in catheter duration, as one would expect. When evaluated as partial bundles, compliance with the maintenance bundle was associated with a (larger) reduction, whereas compliance to the insertion bundle was associated with a longer duration. It is possible that precautions 228 Chapter 9
RkJQdWJsaXNoZXIy MTk4NDMw