This thesis provides an overview of the impact of healthcare associated infections (HAI) on patients in Dutch hospitals during the last two decades, as measured through the national surveillance program. It discusses the frequency of the major device‐associated HAI and their risk factors over time, and, within the setting of an inter‐rater reliability study, the contribution of HAI to mortality. It also presents some of the advances in HAI prevention, with a focus on CRBSI and the effect of introducing so‐called “bundles” of best practices. The first part of this chapter focusses on the merits and limitations of surveillance programs and surveillance data (section 10.1), the actual HAI rates in the Netherlands (10.2), and the future of HAI surveillance (10.3). It will moreover discuss the preventability of HAI and death following a HAI, as this is connected to the contribution of HAI to mortality (10.4). The second part contains two sections: on the effectiveness of bundles (10.5) and the importance of hand hygiene, with an extension to hand hygiene in Dutch hospitals (10.6). Some concluding notes follow in 10.7. PART I 10.1 The merits and limitations of surveillance programs and their data All over the world a growing number of hospitals participate in HAI surveillance [1‐5]. HAI surveillance has proven to be an essential instrument to lower HAI rates and thus protect patients, as demonstrated in many evaluations [1, 6‐10] and this thesis. Surveillance systems, in order to be effective, need to meet a number of requirements[11]: Ongoing and systematic data collection. Protocols ensure uniform application of patient and procedure selections, variable recording, and infection criteria, through time. Monitoring HAI in a large‐scale surveillance program in which the infection and inclusion criteria are standardised additionally allows hospitals to compare their HAI rates with other hospitals, However, despite a protocol variation in local data collection can exist and limit the quality of the surveillance data. For the Dutch surveillance surgical site infection (SSI) programme, on‐site validation in the past has proven that the surveillance data are reliable [12], and the same was concluded for CRBSI. Validation visits within the European PPS demonstrated a good specificity for HAI (98‐99%). The sensitivity was 72‐83% [13, 14]. Although this validation study found inter‐rater reliability for pneumonia was very good (kappa 1.00), most other studies reveal that inter‐rater reliability for VAP is low among radiologists [15, 16]. It has furthermore been demonstrated for SSI that surveillance quality is positively associated with better detection and therefore higher SSI rates and that culturing practices, which may vary among hospitals, affect the number of CRBSI found (‘the 238 Chapter 10
RkJQdWJsaXNoZXIy MTk4NDMw