WHAT SHOULD BE DONE TO REDUCE EMERGENCY DEPARTMENT CROWDING? – A DELPHI STUDY

Abstract:

Objectives & BackgroundEmergency department crowding is a serious public health problem throughout the developed world. In 2015 the UK Royal College of Emergency Medicine was charged by its executive with producing a toolkit to reduce crowding. This study was commissioned to inform said toolkit. The goal of the study was to develop a list of interventions to deal with crowding and exit block, based on consensus of a group of senior emergency physicians.MethodsThe Delphi technique was used to undertake a formal blinded consensus study. 33 experts were invited. These were committee members of the Royal College of Emergency Medicine who worked as emergency physicians in a diverse range of UK hospitals. Responses were collected using the web based survey tool SurveyMonkey over eight weeks in spring 2015. In round 1, experts suggested interventions to reduce crowding. In round 2, experts reviewed all proposals. Any proposal that was endorsed by at least 80% of experts (level chosen a priori) was included in the final recommendations. The authors then decided which professional role would be most able to deliver each intervention.ResultsOf 33 experts invited, 23 agreed to participate in round 1 and 18 in round 2. In round 1, respondents submitted 310 proposals which was reduced to 188 proposals when duplicates and irrelevant comments were removed. In round 1, 117 proposals reached the threshold of 80% agreement. These included 112 interventions recommended to reduce crowding (20 input, 40 throughput, 37 output and 15 multi-dimensional measures) as well as 5 interventions that should be avoided.ConclusionWe have developed a list of 117 expert-endorsed recommendations to deal with emergency department crowding. We not only identified which interventions to implement but also which to avoid. For each intervention we decided who should deliver it. Our findings provide useful advice to clinicians, commissioners, managers and policy makers.