Objective:

To assess the adoption of a modified headache-specific electronic health record-based (EHR-based) alert in primary care using an iterative design approach. 

Background:

A pilot study of an EHR-based alert designed to provide evidence-based diagnosis and management guidelines to primary care providers (PCP) for headaches reduced neurology referrals by 75% and MRI ordering by 35% but only marginally changed provider prescribing habits. The alert fired at the time of entering chief complaint or visit diagnosis. A subsequent cluster randomized control trial of the alert across Geisinger’s primary care saw the interventional alert fire 3738 times in twelve months. Results found poor adoption (12% acted upon) and no impact on provider prescribing habits, neurology referral or headache burden. 

Design/Methods:

Electronic and face to face feedback was collected from end-users regarding the alert. PCP’s identified the alert’s location, criteria for firing, and alert misfiring as major reasons for dismissal. The alert was redesigned and modified to present itself when a PCP ordered a neurology consultation on a patient who had not trialed one preventive and one abortive medication. Half our clinics began utilizing the alert in September 2022, the remainder in May 2023. We assessed the adoption of the alert in this study. 

Results:

Over the course of one year, the alert presented itself 1431 times. Providers utilized the modified alert 21% of the time which was significantly greater than the original alert (p<0.05). Neurology headache referrals also dropped overall from an average of 504 per month to 313 (38% decline) during this time. Only 24 neurology referrals were placed using the alert. The prescribing component of the alert, which was available with or without the alert, was used 496 times.  

Conclusions:

Iterative improvement of the EHR-based headache alert positively impacted primary care provider prescribing habits and may have reduced neurology referrals.