Comparison of different methods assessing alcohol contribution to emergency room visits

The main objective of research was to compare different methods of assessing alcohol's contribution to injuries and medical conditions treated in ER in Poland. Four methods are discussed, including breathalyzer reading, clinical appraisal following Y91 codes of ICD 10, interviewer's observation and self-report.

In two Polish sites Warsaw and Sosnowiec probability samples of patients admitted to ER were selected and interviewed face-to-face by trained interviewers. Almost 1.5 thousand patients were interviewed with response rate of 67%. To identify injuries and medical conditions associated with drinking (preceded by alcohol consumption) four methods were used: clinical observations made by nurses trained in applying Y91 codes of ICD-10 (in the Sosnowiec ER only), observations by interviewers who did not receive such a training, breathalyzer readings and patient self-reports of drinking within six hours prior to the event.

A breathalyzer reading identified 4.5% under the influence of alcohol among all patients, and 5.7% among those who were breathalyzed. Clinical assessment in Sosnowiec and interviewers' observations in both cites identified almost the same proportion: 5.1% and 4.7%, respectively and self-reports almost 10%. All four measures for identifying potentially alcohol related ER visits found a total of 11.5% cases among sampled patients.