countries, SDD is seldom included in a bundle [101] or otherwise implemented, as the higher antibiotic resistance levels might lead to enhanced selection of resistant bacteria [102]. When the advocated best practices have become integrated in regular care (‘business as usual’), one should consider updating the bundle, e.g. adding glucose control to SSI bundles [103]. Consideration is also needed when practices are less well integrated. For example, compliance to the maintenance elements of the Dutch CRBSI bundle have remained rather low. The catheter‐insertion elements have, on average, been well integrated, and there are no new interventions that could become part of a “CRBSI bundle 2.0”. But the maintenance bundle elements, e.g. daily assessment of the indication for the CVC, require ongoing effort to hold HCW attention. Implementing bundles of best practices is usually effective although causal relationships may be impossible to establish due to lack of randomised controlled trials [93] or adherence measurements showing that practices were indeed better adhered to [85]. In the PROHIBIT study and the DHPSP programme, adherence was recorded and multivariable analyses demonstrated that compliance in settings with relatively high CRBSI rates was indeed associated with lowered rates. The number of confounders was limited, however, and the improvement may have come not only from bundle compliance but from other positive consequences of the interventions and programme as a whole. Recording adherence could be a motivating part of an intervention itself, especially when giving feedback to the observed HCWs. It can also increase the insight in the success of a more challenging intervention, e.g. aiming at hand hygiene (chapter 7 and 8), and this insight could lead to more tailored interventions. However, recording compliance entails quite a lot of work/organisational effort, and its benefits should be weighed against this [28]. There is the possible drawback of the Hawthorne effect: people aware of being observed perform better than they would normally do [104‐106]. The few studies on the duration of this effect have shown it to be transient [107], and as the observations in the PROHIBIT study occurred weekly over 2.5 years, we expect that HCWs soon became used to being observed. Moreover, we were interested in the change in bundle compliance and hand hygiene. In case the Hawthorne effect did affect our results, we expect that compliance in the beginning of the study may have been overestimated and the effect of the intervention on these process parameters underestimated. Compliance measurements aside, well‐received infection prevention interventions, in bundles or not, may have additional beneficial effects: best practices to decrease one type of HAI may incidentally decrease other types, as they improve the general health of 248 Chapter 10
RkJQdWJsaXNoZXIy MTk4NDMw