Aims: To determine how emergency departments in England process laboratory investigation results, to identify risk, and to note examples of good practice.
Methods: A telephone survey was conducted, and data were entered anonymously into Excel spreadsheets. Fisher's exact test was used to test the independence of pairs of variables.
Results: Data were collected from 167 out of 193 (87%) emergency departments in England. The majority had nurse-requested blood tests. There was a statistical association between nurse-requesting and failure by the clinician seeing the patient to check results. Fourteen (8%) departments did not allow patients to leave until all their results were available. A senior doctor did a second 'safety' check of results in 83 (50%) departments. Many respondents were able to give examples of patients who had been recalled to hospital after a second check. Only a minority of departments had information systems that could identify high-risk patients.
Conclusion: A second 'safety' check by an experienced consultant, associate specialist or middle grade doctor identifies error. This is time-consuming, but could be supported and simplified by using intelligently designed information systems.