Development and pilot testing of computerized order entry algorithms for geriatric problems in nursing homes.

Objectives: To develop order entry algorithms for five common nursing home problems and to test their acceptance, use, and preliminary effect on nine quality indicators and resource utilization.

Design: Pre-post, quasi-experimental study.

Setting: Two Department of Veterans Affairs nursing homes.

Participants: Randomly selected residents (N=265) with one or more target conditions and 42 nursing home providers.

Intervention: Expert panels developed computerized order entry algorithms based on clinical practice guidelines. Each was displayed on a single screen and included an array of diagnostic and treatment options and means to communicate with the interdisciplinary team.

Measurements: Medical records were abstracted for the 6 months before and after deployment for quality indicators and resource utilization.

Results: Despite positive provider attitudes toward the computerized order entry algorithms, their use was infrequent and varied according to condition: falls (73.0%), fever (9.0%), pneumonia (8.0%), urinary tract infection (7.0%), and osteoporosis (3.0%). In subjects with falls, trends for improvements in quality measures were observed for six of the nine measures: measuring orthostatic blood pressure (17.5-30.0%, P=.29), reducing neuroleptics (53.8-75.0%, P=.27), reducing sedative-hypnotics (16.7-50.0%, P=.50), prescription of calcium (22.5-32.5%, P=.45), vitamin D (20.0-35.0%, P=.21), and external hip protectors (25.0-47.5%, P=.06). Little improvement was observed in the other conditions (documentation of vital signs, physical therapy referrals, or reduction of benzodiazepines or antidepressants). There was no change in resource utilization.

Conclusion: Computerized order entry algorithms were used infrequently, except for falls. Further study may determine whether their use leads to improved care.