Checklists and crisis management in surgical emergencies

Changes in healthcare culture have resulted in a number of initiatives designed to improve patient safety and quality of care. Besides doctors, governments, healthcare providers, quality improvement organizations, health insurers and the media have all shown an interest in these developments. The surgical community should, however, improve care by considering the facts and scientific evidence. Safety interventions should be selected for implementation based on effectiveness not popularity. There is already evidence that this can be achieved for patients with sepsis1. The extent to which this can be applied to patients with surgical emergencies is still a challenge. The SURgical PAtient Safety System (SURPASS) checklist demonstrated a clear effect on mortality reduction. In the study2 using this comprehensive perioperative checklist, about 20 per cent of patients underwent emergency surgery. One year later, a specific SURPASS checklist for emergency surgery was developed that based the selection of checklist items on prevalence of intercepted incidents with the checklist. For emergency surgery this checklist is lean, focusing on the most frequent incidents. In emergency settings, checklists and safety management may seem trivial, although initiatives such as early warning scores (EWSs) and the medical emergency team (MET) enjoy considerable popularity, with the intention of improving healthcare quality, despite a lack of convincing evidence of effect on outcome. In the management of medical emergencies, EWSs are used to identify physiological deterioration in patients at an early stage, to alert the MET. The consistent effect of this system has been to reduce the number of patients who need resuscitation. A recent study3 of 1058 consultations triggered by EWSs in 981 patients over 5 years showed that 57 per cent of patients had their care escalated, usually through transfer to a critical care environment. Despite this, onequarter of patients died. A MET may already be part of daily patient care in many hospitals, but to what extent and in what form does a MET really contribute to better and safer care? There is limited evidence to prove that rapid response systems save lives. It is noteworthy that there is an extremely low number of MET consultations at night-time and relatively few during the day at weekends. These are times when supervision may be absent or suboptimal, but when a MET seems most needed. It seems improbable that patients would generally be healthier and less likely to deteriorate during the night or at weekends. Reliably identifying patients at risk before a life-threatening event occurs through an EWS system seems adequate, as many patients are transferred to the intensive care unit (ICU) after MET evaluation. EWSs are relatively good predictors of events such as death, unexpected ICU admission, emergency surgery and severe complications. The question remains whether the course of disease would have been any different without anEWS-triggeredMET intervention. Where are the data to support the hypothesis that a system of care that includes a MET produces better outcomes than standard care? Are mortality rates, for example, lower than the expectedmortality calculated from an organ failure score? In a clusterrandomized trial (MERIT study)4 in Australia, aMETwas introduced in 12 of 23 participating hospitals, whereas the remaining 11 hospitals continued functioning as usual and served as controls. This study showed that, although the team was consulted frequently, there was no in-hospital effect on the incidence of cardiac arrest, unplanned ICU admissions and unexpected mortality. The authors noted that almost half of calls to the teams were ‘early’ calls (before a cardiac arrest). As the proportion of early calls increased, the rate of cardiac arrests and unexpected deaths decreased. For every 10 per cent increase in the proportion of early calls, there was a 2·2 (95 per cent confidence interval (c.i.) –2·9 to –1·6) reduction per 10 000 admissions in overall cardiac arrests, and a 0·94 (–1·4 to –0·5) per 10 000 reduction in unexpected deaths. This was not reflected, however, in a statistical reduction in overall deaths or unplanned ICU admissions5. Systematic reviews6,7 on the subject have come to similar conclusions. The effect of a MET on patient outcome has not been demonstrated clearly. The most recent meta-analysis7 evaluated 18 studies involving METs and showed a 33·8 per cent reduction in rates of cardiopulmonary arrest outside the ICU (relative risk (RR) 0·66, 95 per cent c.i. 0·54 to 0·80), but this was not associated with lower hospital mortality rates (RR 0·96, 0·84 to 1·09). A large prospective study8 of