Pediatricians' Clinical Decision Making

Results of Two Randomized Controlled Trials of Test Performance Characteristics

This article explored how providing pediatricians with test accuracy information might impact patient management. Study participants were 653 pediatricians who completed surveys that were sent to a random selection of 1,502 pediatricians across the United States. Two clinical vignettes—“pertussis vignette” and “urinalysis vignette”—were presented to participants in a mailed questionnaire. Participants were randomly assigned questionnaires where both vignettes varied based on test information provided. Vignettes did not provide more information (control group), presented test characteristics (TC), or defined test characteristics (TC defined). Participants were to make patient management decisions based on the provided information.

Key Findings:

  • For the pertussis vignette, controls (23%) were significantly less likely to order tests than were the TC (73%) or TC defined (71%) groups.
  • There was no relationship between managing medication therapy or discharge plans and receiving test information on the pertussis vignette.
  • For the urinalysis vignette, there was no significant difference between TC and control groups on testing and referral.
  • The TC defined group referred and reviewed tests significantly more than the control group on the urinalysis vignette.