Computer predictions of abnormal test results. Effects on outpatient testing

Area of Science:

• Clinical Medicine
• Health Informatics
• Medical Decision Making

Background:

• Diagnostic test utilization significantly impacts healthcare costs.
• Physician decision-making in ordering tests can be influenced by various factors, including perceived probability of abnormality.
• Optimizing diagnostic test ordering is crucial for efficient patient care.

Purpose of the Study:

• To evaluate the impact of real-time predictive probability information on physician diagnostic test ordering.
• To assess the effect of this intervention on the utilization and cost of commonly ordered diagnostic tests.

Main Methods:

• Development of statistical equations to predict abnormalities for eight common diagnostic tests.
• Implementation of a microcomputer system to display predictive probabilities to physicians for intervention patients.
• A six-month controlled trial involving 112 physicians and over 15,000 patient visits, with half randomized to intervention.

Main Results:

• Patient charges for the eight study tests decreased by 8.8% for intervention patients compared to controls.
• The most significant cost reductions (over 10%) were observed for serum electrolyte tests and complete blood cell counts.
• Physicians in the intervention group ordered fewer low-probability tests, indicating a shift towards higher-risk patient targeting.

Conclusions:

• Providing timely predictive information enables physicians to make more targeted diagnostic test orders.
• This intervention can lead to reduced healthcare costs associated with diagnostic testing.
• Real-time predictive analytics show promise in improving the efficiency and cost-effectiveness of medical testing.