Quality of coding diagnoses in emergency departments: Effects on mapping the public's health

Background: Emergency department (ED) attendees reffect the health of the population served by that hospital and the availability of health care services in the community. Objectives: To examine the quality and accuracy of diagnoses recorded in the ED to appraise its potential utility as a guage of the population's medical needs. methods: Using the Delphi process, a preliminary list of health indicators generated by an expert focus group was converted to a query to the Ministry of Health's database. In parallel, medical charts were reviewed in four hospitals to compare the handwritten diagnosis in the medical record with that recorded on the standard diagnosis "pick list" coding sheet. Quantity and quality of coding were assessed using explicit criteria. results: During 2010 a total of 17,761 charts were reviewed; diagnoses were not coded in 42%. The accuracy of existing coding was excellent (mismatch 1%-5%). Database query (2,670,300 visits to 28 hospitals in 2009) demonstrated potential benefts of these data as indicators of regional health needs. conclusions: The fndings suggest that an increase in the provision of community care may reduce ED attendance. Information on ED visits can be used to support health care planning. A "pick list" form with common diagnoses can facilitate quality recording of diagnoses in a busy ED, profiling the population's health needs in order to optimize care. Better compliance with the directive to code diagnosis is desired.