Review: Computerized reminders increase the rate of use of most preventive services
ACP J Club. 1997 May-Jun;126:80. doi:10.7326/ACPJC-1997-126-3-080
Shea S, DuMouchel W, Bahamonde L. A meta-analysis of 16 randomized controlled trials to evaluate computer-based clinical reminder systems for preventive care in the ambulatory setting. J Am Med Inform Assoc. 1996 Nov/Dec;3:399-409.
To determine the effectiveness of computer-based reminder systems to increase the use of preventive care measures in ambulatory care settings.
Studies were identified by searching MEDLINE (1966 to December 1995), Cumulative Index to Nursing and Allied Health (1982 to October 1995), and Health Planning and Administration database (1975 to November 1995) using the keywords reminder systems, software, computers, ambulatory care, preventive health services, primary prevention, HMO, family practice, professional practice, attitude to computers, automatic data processing, primary health care, and decision support systems or management. Reference lists of identified articles were also scanned.
Studies were selected if they were randomized controlled trials that compared practitioners who received computer-generated reminders with controls who received no intervention for preventive services in ambulatory care settings.
Data were extracted on the study setting, intervention, preventive services, number of patients in each treatment group, and success rate for the preventive service. Preventive services were divided into 6 categories: vaccinations, including influenza, pneumococcus, and tetanus; breast cancer screening, including clinical breast examination and mammography; cervical cancer screening, including pelvic examination and Pap test; colorectal cancer screening, including digital rectal examination, fecal occult-blood test, and sigmoidoscopy; cardiovascular risk reduction, including blood pressure check, hypertension follow-up, smoking assessment and counseling, dietary assessment and counseling, and cholesterol screening; and other preventive services, including glaucoma screening, dental screening, teaching self-examination techniques, weight monitoring, and teaching reporting of postmenopausal bleeding.
16 studies met the selection criteria. A weighted mixed-effects model regression analysis was done to analyze the success rates, which were defined as completion of the preventive maneuver. Compared with controls, computer-generated reminders increased the use of preventive practices for 4 of the 6 groups: vaccinations (adjusted odds ratio [OR] 3.09, 95% CI 2.39 to 4.00), breast cancer screening (OR 1.88, CI 1.44 to 2.45), colorectal cancer screening (OR 2.25, CI 1.74 to 2.91), and cardiovascular risk reduction (OR 2.01, CI 1.55 to 2.61). Manual reminders increased the use of the same 4 preventive practices compared with controls. Comparisons of computer-generated reminders with manual reminders did not show a statistically significant difference. Computer-generated reminders did not improve preventive practices for cervical cancer screening (OR 1.15, CI 0.89 to 1.49) or other preventive care (OR 1.02, CI 0.79 to 1.32) compared with controls.
Computer-generated reminders increase vaccinations, breast cancer screening, colorectal cancer screening, and cardiovascular risk reduction but do not increase cervical cancer screening.
Source of funding: The Commonwealth Fund.
For article reprint: Dr. S. Shea, Division of General Medicine, 622 West 168th Street, New York, NY 10032, USA. FAX 212-305-3302.
Previous systematic reviews have suggested that reminder systems are often more effective than other methods of changing the performance of practitioners, particularly if reminders occur during consultations (1, 2). However, more detailed evaluation of the potential benefits of computerized reminder systems is needed before widespread use can be justified.
This meta-analysis by Shea and colleagues is a helpful addition to the debate and meets most of the criteria recommended for assessing the quality of reviews (3). Although manual reminders were found to be almost as effective as computerized ones, the authors did not seek to include all high-quality studies of manual reminders; therefore, this assessment of their effectiveness is incomplete. 16 studies were included—15 were done in the United States, and 1 was done in Canada. Generalization of the conclusions to health care systems outside North America requires caution. Further, the findings apply only to preventive care in ambulatory care settings.
Computerized reminders can be effective. Shea and colleagues estimate that with a baseline compliance rate of 50% among practitioners, computerized reminders would increase the compliance rate to 64%. Even so, computerized reminders do not always work. It is necessary to determine which features of reminder systems are important for effectiveness, the costs of these systems compared with other methods of improving practitioner behavior, and adverse effects. Although computerized reminders seem to be promising, we cannot yet recommend them as "must buys" to providers of preventive care in ambulatory settings.
Richard Baker, MD
University of LeicesterLeicester, England, UK
Richard Baker, MD
University of Leicester
Leicester, England, UK