Objective: This study compares agreement between telephone survey and ambulatory medical record data for an elderly patient population.
Methods: Medical records and telephone survey responses are used to compare health status (chronic medical condition, symptomatology, and functional status) of 142 elderly patients randomly selected from a family practice residency and a geriatric fellowship practice. Chart abstraction was performed by two resident and two faculty physicians after a training period designed to assure high inter-reviewer reliability. Telephone surveys were completed by two professional interviewers. Health status measures were taken from standard, published instruments. The kappa statistic was used to measure the agreement between medical record and survey data.
Results: Overall, there is little agreement between the medical record and the telephone survey results on chronic medical condition, symptomatology, and functional status.
Conclusions: Medical records abstractions and telephone survey methodologies did not yield comparable health status data when applied to the same elderly patient population. Functional status assessment and symptomatology are particularly problematic, but even the presence or absence of chronic diseases is often inconsistent in the two data sources.