Prior Authorization of Medication and Its Influence on Provider Behavior: Latent Class Analysis

Area of Science:

• Health Services Research
• Medical Informatics
• Health Economics

Background:

• Insurance companies mandate prior authorization (PA) for prescriptions to ensure quality and safety.
• PA requirements can lead to provider dissatisfaction and negatively impact patient outcomes.
• Prescribers may alter their prescribing behaviors to circumvent PA processes.

Purpose of the Study:

• Identify distinct prescriber subgroups based on prior authorization (PA) behaviors using finite mixture modeling.
• Characterize these subgroups using relevant covariates.
• Examine the impact of subgroup membership on prescribing outcomes.

Main Methods:

• A cross-sectional, web-based survey of 1173 prescribers, oversampled for psychiatry.
• Latent class analysis of 12 indicators related to PA involvement, communication, and avoidance strategies.
• Analysis of covariates including demographics, provider role, specialty, and patient load.

Main Results:

• Four distinct prescriber classes emerged: high PA denial (25.15%), low volume PA (15.93%), problematic communication (19.96%), and low volume PA with problematic experiences (38.97%).
• Age, specialty type, and patient load were significant covariates for class membership.
• The largest class, reporting problematic PA experiences, showed significantly higher rates of altered prescribing and diagnostic behaviors.

Conclusions:

• Prescriber experiences with prior authorization (PA) are heterogeneous.
• Recognizing and accommodating diverse PA-related behaviors is essential for effective implementation.
• Systemic changes can help providers avoid altering medications, diagnoses, or delaying access to newer treatments.