Am Fam Physician. 2002;65(6):1190-1195
Many proven and well-accepted preventive health measures are underused. Numerous systems have been proposed over the years to increase appropriate use of preventive care, but most meet with less than robust success. Dexter and associates conducted a randomized, controlled trial to determine the effects of computerized reminders on the rates at which preventive therapies are sought.
Over an 18-month period, the authors studied the use of a computerized reminder system in a general medicine service at an urban teaching hospital. They aimed to increase the use of four preventive health items (pneumococcal vaccination, influenza vaccination, prophylactic heparin for deep venous thrombosis prevention, and prophylactic aspirin). The study authors randomly assigned resident physicians and medical students to an intervention or a control group and followed their use of the targeted preventive health items on 10,065 consecutive hospital admissions.
All inpatient orders were made through a computerized order entry system at the study hospital. The intervention group received computer-generated reminders when inpatients were eligible for a given preventive health item. The ordering physicians and students could accept or decline the suggested items with one or two keystrokes on the computer. The control group was free to order any of the targeted items but received no reminders.
All four of the targeted preventive health items were ordered significantly more often by physicians and students in the intervention group. Pneumococcal vaccination was ordered for about 36 percent of eligible patients in the intervention group versus about 1 percent in the control group. Influenza vaccination rates were similar, with orders written for approximately 51 percent of appropriate patients in the intervention group versus 1 percent in the control group. The likelihood that ordering physicians and students would accept the reminders varied widely, with most accepting 11 to 40 percent of the suggested orders. Reminders were more likely to be accepted for older patients than for younger but appropriately eligible patients. The suggestion for use of prophylactic aspirin was more frequently accepted for patients hospitalized with myocardial infarction or unstable angina than other eligible diagnoses, such as cerebrovascular disease, peripheral vascular disease, and coronary risk factors.
The authors conclude that computer-generated reminders are successful in increasing the use of appropriate preventive health measures in hospitalized patients.
editor's note: In their discussion, the authors note a previous inpatient trial using computerized reminders at the same hospital that failed to increase the ordering of preventive health measures. Following that failure, the reminders in this trial were designed to generate pre-written orders that could be accepted or declined with a simple keystroke. The dictum that “an ounce of prevention is a ton of work” applies—interventions that allow physicians to easily incorporate preventive health practices in their usual workload are more likely to be successful. Computerization and an electronic medical record alone are not sufficient; a clever and efficient design is paramount.—b.z.