The effects of selection defaults on the ordering practices of clinicians using an Electronic Medical Record

Abstract

The installation and use of computerized systems in health care, particularly Electronic Medical Records (EMRs), has rapidly expanded over the past decade. Overall, their implementation has led to positive outcomes. However, despite their many beneficial contributions to the medical field, EMR use has been shown to alter the way in which clinicians provide care to patients and has broadly resulted in changes to medical decision making. One possible explanation for why EMRs alter medical decision making relates to the structure of the choice environment they create for clinicians. For example, EMRs have the capability to list every avenue of patient care on a single screen to clinicians that can be ordered via simple checkboxes. This new choice environment might lead to changes in the ordering procedures of practitioners. Furthermore, when using choice environments of this nature, the options people request may be manipulated with pre-selected defaults (i.e. a checkbox already selected). This project examined how the ordering practices of clinicians changed as a result of selection defaults on admissions orders written for six fictitious patients. Clinicians were found to order more laboratory tests when all tests were pre-selected for them compared to when no laboratory tests were pre-selected. This increase in the number of tests requested led to increased cost of admitting the fictitious patients to the hospital. Moreover, practitioners were found to order significantly more tests that were deemed ‘Very Relevant’ by expert physicians when those tests alone were pre-selected for them, demonstrating that practitioner performance can be optimized with proper default utilization.