Marjon Borgert

197 Summary and future perspectives SUMMARY Actually integrating new research ndings into daily practice is challenging, especially if they require changes to behaviour, clinical practices, the organization, or the way professionals collaborate. In many cases patients come to harm because evidence- based recommendations or guidelines are not followed consistently. Multiple studies have shown that patients receive only half of the recommended care. 1,2 In health care, there is a slow uptake of new research ndings in guidelines and daily care. 3 This thesis focuses on the implementation of strategies for improving patient safety and quality of care of critically ill patients. Part I focuses on improving patient safety for critically ill patients on nursing wards by implementing a rapid response system (RRS). Part II focuses on improving patient safety for critically ill patients in the intensive care unit (ICU) by implementing evidence-based care bundles. Part I. Improving patient safety of critically ill patients on nursing wards. In part I, we focusses on the implementation of the RRS on nursing wards. RRSs are developed to improve the care for deteriorating patients in hospitals. 4,5 Previous studies have shown that most patients who su er from serious adverse events, such as an unplanned ICU admission, cardiac arrest and unexpected death, have vital sign abnormalities prior to these adverse events. 6,7 However, these signs are not always recognized in time by nurses or are not adequately and timely acted upon. 8 RRS involves the recognition of patients’ conditions prior to deterioration by using a ‘track-and- trigger system’, such as the modi ed early warning score (MEWS). 9 The rapid response team (RRT) should be called in case the patient’s condition deteriorates beyond a certain MEWS threshold. Chapter 2 describes the implementation of the MEWS on nursing wards of the Academic Medical Center in Amsterdam. In this quasi-experiment, we studied the e ects of protocolized measurement (i.e. three times daily) of the MEWS versus measurement when clinically indicated. The study was conducted between September and November 2011. All patients admitted to the hospital for at least one overnight stay were included. Nursing wards were randomized to measure the MEWS three times daily or on indication. In total, 902 patients were included in this study, and a set of 6598 vital sign measurements were registered in the patient les during the three study weeks. The results showed that in the protocolized randomization arm, the MEWS was calculated in 70%. On wards were the MEWS was measured on indication the MEWS wasmeasured in only 2%, di erence 67.9%, 95%con dence interval 66.3 to 77.0, P -value < 0.001. Furthermore, there were 90 calls to the primary physician on the ward in the protocolized arm versus nine calls on the wards randomized on indication. The results indicate that measuring the MEWS three times daily results in improved compliance to