Feasibility and Accuracy of a Postcard Diary System for Tracking Healthcare Utilization of Community-Dwelling Older Persons
Published In: The Journal of the American Geriatrics Society, v. 43, no. 5, May 1995, p. 550-552
Posted on RAND.org on December 31, 1994
OBJECTIVE: To test the feasibility of and to validate a self-report weekly postcard diary of health care utilization. DESIGN: Case-series and validation study. SETTING: Community-based sample of fee-for-service and health maintenance organization (HMO) patients. PARTICIPANTS: A convenience sample of 24 community-dwelling older persons who had failed a self-administered screen and were eligible for a study of outpatient comprehensive geriatric assessment consultation. MEASUREMENTS: Subjects completed and mailed in a weekly postcard diary documenting medical, counseling, or rehabilitation therapy visits. If a subject did not respond within 10 days after the end of the week, a telephone call was placed to gather the information. For a subset of 10 subjects who were HMO enrollees, all records were reviewed to determine accuracy of the postcard diaries. RESULTS: Of the 24 subjects enrolled, one HMO enrollee dropped out following hip surgery after 4 weeks of completing diaries. The remaining 23 subjects (96% of total entered) provided complete information for 12 weeks. Telephone follow-up to either obtain or clarify utilization information was required for 22% of subjects. For the subset of 10 HMO subjects, seven underreporting, two overreporting, and one incorrect day errors were detected-a total of 10 errors representing 9% of diaries and 1.3% of patient-days. Kappa as a measure of agreement between the two methods (self-report and chart review) was 0.82 (P < .0001). Compared with chart review, the diary report was 75.0% sensitive and 99.8% specific. Positive and negative predictive values were 91.3% and 99.2%, respectively. CONCLUSION: These findings suggest that a postcard diary system with telephone follow-up is both a feasible and reasonably accurate method of tracking health care utilization by community-dwelling older persons, although a small percentage will be unable to adhere to this method.