Background: Patient-provider communication through the patient portal has markedly increased in recent years. Some health care facilities implemented programs to enable providers to bill for responding to patient-initiated messages that require substantive medical decision making through an e-visit.
Objective: To evaluate the effect of billing eligible patient-initiated portal messages as e-visits using a mixed-methods approach.
Design: Retrospective observational pre-post comparison and prospective survey.
Setting: Large integrated health system with sites in 4 U.S. states.
Participants: Patients initiating portal message threads and health system providers completing an online survey.
Intervention: E-visit billing was implemented 18 August 2023.
Measurements: The volume of patient-initiated medical advice message threads pre- versus postimplementation of e-visit billing was compared. Health system provider perceptions of e-visit billing were assessed using an online survey.
Results: In the 6 months after e-visit billing implementation (18 August 2023 through 18 February 2024), the volume of patient-initiated medical advice message threads decreased by 8.8% (from 1 813 818 to 1 653 708) compared with the same dates the year prior (P = 0.002). A total of 5183 (0.3%) medical advice messages were billed. There was no difference in 7-day use of emergency services (emergency department visits or hospitalizations) in patients who proceeded with sending a message versus those that did not send a message after viewing the billing disclaimer on the patient portal. Providers reported overall acceptance of e-visit billing but expressed concerns about increased workload with the current process.
Limitation: Cannot assess changes in other forms of contact, including telephone calls.
Conclusion: These findings suggest that implementation of e-visit billing was associated with a modest decrease in patient-initiated portal message volume and was overall acceptable to providers in a large integrated health system.
Primary funding source: None.