Evaluation of Checklists as a Diagnostic Tool in the Emergency Department: a pilot study

Abstract: Introduction: The Emergency Department (ED) is a complex environment. Patients present with many different complaints with alternating severity. This setting makes time efficiency and diagnostic accuracy crucial. The use of checklists in medicine has been successful in several medical fields, such as in reducing surgical complications and in decreasing catheter-related bloodstream infections in the intensive care unit, but has yet to be evaluated in the ED environment. This pilot study aimed to evaluate whether the use of checklists as a diagnostic tool in the ED could improve efficiency and be of benefit in the diagnostic process. Methods: Checklists were created for nine common patient complaints for use in the medical history uptake, the physical examination and for ordering relevant investigations. An evaluation period of six weeks after the checklist implementation was compared to a control period prior to the checklist implementation, and the subpopulation of patients on whom checklists were used was compared to the same control period. Time measures were used as indicators of efficiency. The proportion of patients leaving without being seen by a physician (LWBS) and admission rate were also used. Additionally, a substudy was conducted where five fairly inexperienced physicians were studied one week not using and one week using the checklists. Time measures, rate of admission and probability assessments for differential diagnoses and admission were used as outcome measures. Results: Checklists were used in 219 cases (7.9 % of eligible cases in the evaluation period). There was significantly longer patient length of stay (with time waiting for admission excluded) in the checklist subpopulation compared to the control period. Median times were 04:25 and 05:01, respectively (p<0.001). The checklist users consisted of a larger proportion of junior physicians than normal (64.4 % vs 41.4 % respectively). The substudy consisted of 61 cases, 36 in which checklists were used. The physicians were not less efficient and slightly better at probability assessments of differential diagnoses (p= 0.018) when checklists were used. Conclusion: Checklist usage appeared to be associated with impaired time efficiency, but the result was confounded by a high proportion of inexperienced physicians using them. The results in a more controlled setting showed no difference in time measures, and a benefit regarding probability assessment of differential diagnoses when physicians used checklists. This proves an interesting point for further investigation.