UCLA Health Patient Cardiology Care Gaps

This is a prospective randomized clinical trial evaluating how two behaviorally-informed interventions (i.e., monthly report card and storyboard interventions) impact physician behavior, with the goal of closing care gaps in preventive care and disease management.

In particular, the monthly report card intervention seeks to elevate physicians' intentions to close their patients' care gaps, while the storyboard intervention seeks to prompt action by making patients' care gaps salient. The trial investigates the separate and joint impacts of the proposed behaviorally-informed interventions on encouraging physicians to close their patients' care gaps.

Despite the existence of evidence-based clinical guidelines on how to manage the health of patients with cardiovascular disease that have been shown to improve cardiac function and survival, significant gaps in care persist and optimal medical therapy is often under-utilized. For example, guideline-directed medical therapy (GDMT), statin and aspirin therapy, and blood pressure control are cornerstones of effective management of patients with cardiovascular disease or heart failure. However, current dashboard data indicate that within the UCLA Health system, these medical therapies remain under-utilized.

As part of the quality improvement (QI) initiative sponsored by the Division of Cardiology at UCLA Health, this trial will examine the independent and joint impacts on care gap closure of (1) sending monthly behaviorally-informed emails to physicians with personalized feedback on care gap performance and (2) enhancing the visibility of the open care gaps in the electronic health record upon patient encounter.

Eligible physicians (i.e., UCLA Health Cardiologists) will be randomized to one of four conditions based on a 2 (monthly report card intervention: receive vs. no receive) x 2 (storyboard intervention: receive vs not receive) between-subjects design:

• In the no intervention (control) condition, physician participants will receive quarterly emails with their performance report card (status quo).
• In the monthly report card only condition, behaviorally-informed monthly emails will be sent to inform physicians of their performance and elevate physicians' intentions to get more of their patients to close care gaps.
• In the storyboard only condition, the visibility of care gap banners in the electronic health record (EHR) will be enhanced, which will promptly remind physicians of each patient's care gaps at the start of a patient-physician encounter.
• In the monthly report card and storyboard condition, physicians will receive behaviorally-informed monthly emails and be reminded of each patient's care gaps via the EHR storyboard during patient-physician encounters.

The trial will include physicians participating in an existing UCLA incentive program, as of October 1, 2023, who have a panel size of above 50 patients. In August 2024, new physicians who are eligible for the incentive program and have at least 50 patients in the panel will be randomly assigned into one of the four conditions and become part of the study sample. Randomization will balance for physicians' baseline performance, subspecialty training, years out from training, and estimated panel size.

Monthly performance feedback emails will include physicians' performance on eight cardiology care gaps, their projected earnings in the current quarter and their earnings in the previous quarter. Care gaps include the following: (1) statin or PCSK9 inhibitor use, (2) aspirin use, (3) beta blocker use, (4) angiotensin-converting enzyme inhibitors, angiotensin II receptor blockers, or angiotensin receptor-neprilysin inhibitor use, (5) mineralocorticoid receptor antagonist use, (6) SGLT2i, (7) average HCC/RAF score, and (8) blood pressure control.

Analysis Plan:

• The investigators will use patient-level linear regression models, where the response is a 1/0 indicator of whether the patient has completed any of the open care gaps, with cluster-robust standard errors at the physician level.
• The primary regression model will look at the main effect of the report card intervention and the storyboard intervention.
• In a secondary regression, we will investigate the joint impact of the two interventions, including a report card x storyboard interaction.

Control variables include:

• Patient-level care gap closure (i.e., whether patients closed any of their open care gaps, excluding HCC/RAF score, during the six months before the intervention delivery). For patients who were not in a given physician's panel as of six months before the experiment, mean values across the panel will be imputed.
• Physician characteristics (gender, years out from training, subspecialty training within cardiology) and panel size.
• Patient age (If there will be a missing value, the investigators will replace it with the mean and add a dummy variable to indicate patients with missing age).
• Indicators for patient race/ethnicity (Black non-Hispanic, Hispanic, Asian non-Hispanic, white non-Hispanic, other/mixed, unknown).
• Indicators for patient gender (male, female, other/unknown).
• Indicators for patient 5-digit zip code (all zip codes outside of California will be collapsed into one group).
• Indicators for patient insurance type (Self-pay/uninsured, Medicaid, Medicare Advantage, Medicare Fee For Service, Commercial, VA, Other).

The investigators will conduct subgroup analyses based on:

• Whether the patient is female or male.
• Whether the patient is white non-hispanic or racial/ethnic minority.
• Whether the patient is 65+ (including 65) or below 65 years of age.
• Distance from patient's place of residence to their cardiologist's clinic.
• Whether the patient has commercial or non-commercial insurance.