slightly. CVCs were increasingly inserted in the jugular vein and became less frequently used for antibiotics and dialysis, and instead more often for ‘other indications’. Catheter duration decreased from a median of six days (IQR 3‐9) in 2009‐2013 to five (3‐8) thereafter. For CVC days in the ICU (of CVCs with at least one day in the ICU) this was five (IQR 3‐7) and four (2‐7) respectively, whereas for CVC days outside the ICU (of CVCs with at least one day outside the ICU) this was five (IQR 3‐9) and five (2‐9). In all 66 hospitals participating in the CRBSI surveillance, patient and CVC characteristics demonstrated comparable trends (Supplementary Table S2). CRBSI incidence The average CRBSI rate based on the hospital years with >80% bundles recorded, decreased from 4.0 to 1.6/1000 CVC days, resembling the overall trend from 3.7/1000 in 2009 to 1.3/1000 in 2019 (Figure 1). Figure 2 demonstrates the CRBSI incidence density against number of participation years. During the first three years the CRBSI rate decreased from 2.0 (median 1.6, IQR 0.0‐3.0) to 1.5/1000 CVC days (median 1.0, IQR 0.0‐ 2.0). Ongoing participation was associated with a further decrease. Bundle compliance The mean compliance to the overall bundle in the first participation year was 64% (median 75% IQR 46‐92%). For the insertion bundle this was already 90% (median 93% [85‐98%]). The use of 0.5% chlorhexidine in alcohol was practiced for the majority of patients in almost all hospitals. For the maintenance bundle the mean compliance was 69% (median 86% [56‐96%]). During the first participation years, the insertion bundle compliance increased (Figure 3) and slightly decreased thereafter. Improving the maintenance bundle and consequently the overall bundle appeared to be more challenging at first. There were, however, substantial differences between participating hospitals. Supplementary Figure A1 shows bundle compliance per calendar year. Association with the CRBSI risk Hospital wide: In multivariable analysis of hospital wide data, i.e. without distinction between ICU and non‐ICU, the parameters sex, insertion vein and catheter indication were significantly associated with the CRBSI risk (Table II). Male sex, insertion in the jugular and femoral vein and total parenteral nutrition were associated with increased risk whereas antibiotics and hemodynamic monitoring were associated with reduced risk. The overall bundle compliance was not significantly associated with CRBSI risk in the multivariable model (the hazard ratio (HR) of 1.24 (95% confidence interval (CI) 0.97‐ 9 221 The effect of a CRBSI prevention bundle in the Netherlands
RkJQdWJsaXNoZXIy MTk4NDMw