United Kingdom resuscitation outcome study

Abstract

In 1982 the United Kingdom Resuscitation Council was founded to provide guidelines and improve the practice of cardiopulmonary resuscitation. As part of a wider programme a multicentre study was established to examine the circumstances, frequency and outcome of cardiopulmonary resuscitation in hospital. The aims of the study were varied. It would establish a baseline against which the effects of training and organisation could be assessed in the future. It would create standard records and develop a methodolgy for use within hospitals for purposes such as clinical audit and trials. It would provide a system to allow further study of the effectiveness of resuscitation in the community as a whole. This paper, on the methods and overall results, is delivered in the name of the research study group of the Resuscitation Council, whose principal investigator is Mrs Linda Bailey of the University of Sussex and whose research coordinator and epidemiological advisor is Prof Hugh Tunstall Pedoe of the University of Dundee. The study examined cardiac resuscitation in 15 United Kingdom hospitals ranging from large teaching institutions to small district hospitals. We believe that it is the largest study mounted of its kind. Each hospital had a local coordinator whose responsibility was to maintain a log of registered cases and forward the individual research results to the overall coordinator. The study methods were piloted in the first two hospitals. All cases of CPR attempted in hospital, with the exception of recurrent episodes within 24 h of the first were entered into the study. In addition, cases were entered where resuscitation had been commenced in the community and continued beyond arrival at hospital. The cases were usually identified from ‘crash calls’ received by the hospital switchboard or from records maintained by individual departments. The cases were logged for at least 12 months online, some hospitals being able to remain on line for considerably longer. Survivors were followed up for 1 year. Although over 6000 arrests were initially entered, comparative data were only available from 12 hospitals which were able to remain on line for 12 months. This constituted a core database of 4027 cases. A large survey such as this may suffer from both recording events that should not be included and of failing to record events that should have been. While the core database represents complete 1Zmonth monitoring of events in 12 hospitals, the extended database includes a small number of hospital months where there was an inadequate return of records. An analysis was performed to confirm that