[Reliability of the information on the recent use of medication in a hospital based case-control study]

Accuracy of the information is essential to produce unbiased estimates of the association between exposure and outcome. We are carrying out a case-control study which aim is to investigate the association between the use of medication and falling injuries leading to hospitalisation in the elderly. As there is no gold-standard available, we estimated the reliability of the information on the use of these drugs within the 24 hours and two weeks before the fall using a test-retest strategy. Sixty-one individuals aged 60 years or more were re-interviewed within an interval of 5-7 days after the first interview. Kappa coefficients were high, showing a good consistency of collected data on medication recently used. Among the variables investigated, only gender showed an association with reliability of the information, which was more consistent among women compared to men.