Re: Benzodiazepines and risk of all cause mortality in adults: cohort study
We agree with Richard Kapit that such retrospective cohort studies have the inherent limitations of bias and especially recall bias: despite the availability of records we are not sure about their consumption. Only recording prescriptions does not properly reflect the intake status of patients. Above all, all cause morbidity estimation by specific disease categories might not even give a true picture of the co-morbidity since there might be patients with multiple co-morbid conditions, and those with multiple co-morbid conditions (e.g. two, three or more) might have increased mortality as compared to those with a single morbid condition. The study findings should be cautiously interpreted and randomized clinical trials might be another option.
Competing interests: No competing interests