11 Uptick/downtick – the future of morbidity and mortality meetings

Introduction Morbidity and mortality (M and M) meetings are a national requirement for all departments in hospital. While they should confer improvement in education and patient care, they often ‘lack defined structure, resource support and…heterogeneity in case evaluation’.1 UK critical care has an embedded routine audit data collection on outcomes which is exported to a central body, Intensive Care National Audit and Research Centre (ICNARC), for cleaning and analysis. This data is submitted monthly but is returned on a quarterly basis preventing the timely identification of deaths or survivals that are unexpected (based on case-mix). In critical care, where the expected mortality is high, attention is fatigued if all mortality is reviewed and a more specific approach is required. A fundamental prerequisite for a good M and M meeting is appropriate, targeted case selection. Currently, time is wasted reviewing cases that are selected without systematic focus on ‘important variation’, so contemporaneous feedback on care is lost and learning and engagement are diminished. To combat this, we have developed a timely, targeted and transformative M and M meeting based on data-driven case selection. Methods Using code in R to generate an automated variable life-adjusted display (VLAD) chart using near real-time routine data from our electronic health record. This reveals the high yield outlying ‘uptick/downtick’ patients (patients who survived and should have died and vice-versa) (figure 1). Identifying crucial learning points from the clinical documentation of those selected ‘uptick/downtick’ patients using a specially developed Intensive Care Unit Structured Case Note Judgement Review format. Presentation of these targeted cases in a structured multi-disciplinary setting with key personnel highlighting the impact of timely narrative feedback, promoting staff engagement and morale and encouraging the growth of a proactive culture of safety and quality. Results The VLAD cumulatively charts the effect of each individual admission using the ICNARC generated probability of death. Each death represents a ‘downtick’ in the plot and each survivor represents an ‘uptick’. When a patient predicted to survive dies, then the downtick will be prominent (conversely a patient who survives despite a high predicted mortality will make a prominent uptick). Conclusion Our M and M model is an original in-depth and specific data mine of critical care patient morbidity and mortality. This unique process seeks to provides an enabled and focussed method of case selection and emphasise the potential marginal gains that can be made in patient care. Over time, it will improve future clinical decision making and therefore patient outcomes. Reference Vreugdenburg, et al. Morbidity and mortality meetings: gold, silver or bronze?ANZ Journal of Surgery2018;88:966–974.