Difficulties in assessing drug/food interactions from chart audit in an elderly Barbadian sample group

The aim of this study was to examine the prevalence of potential drug/food interactions among 290 elderly Barbadian outpatients. These findings indicate that chart audit is insensitive to definite interactions but was able to detect the prevalence of potential interactions (1.38 per outpatient). The risk of potential interactions was significantly higher among outpatients who were older, female, and those who used a larger number of prescribed medications. There were no significant differences in the prevalence of potential interactions among diabetics and nondiabetics. The findings indicate that the absence of relevant data on patients' symptoms, and anthropometric and laboratory measures in more than 80% of patients' charts makes it impossible to assess the prevalence of definite drug/food interactions from chart audit. Weight changes, nutrient levels, and symptoms were the most frequently missing data. Chart audit is therefore an invalid and unreliable approach to estimate definite drug/food medications.