Trial Outcomes & Findings for Informational Nudge to Improve Heart Failure Prescribing (NCT NCT05986695)

All employed providers within Cardiology and Primary Care units were enrolled into this study per agreement with management.

The leadership of both units agreed that all providers would participate. The 81 participants (providers) were randomly assigned to one of four study arms (control, Alert \[EHR note\], Pier Comparison Report \[Email notification\], or Both \[EHR and Email\]. The providers were numbered and blindly randomize by department, education, and employment site.

Informational Nudge to Improve Heart Failure Prescribing

The primary effectiveness outcome is the amount of SGLT2 or MRA prescriptions in eligible HF patients in the three intervention groups compared with the control group within 30 days of appointment. Thirty days is a common time interval for HF outcomes assessment and allows for chart documentation, care coordination, laboratory testing, and medication prescribing that may occur days after a patient encounter. The investigators will record all SGLT2 or MRA prescriptions, including those from non-targeted clinicians, given that nudge interventions, especially the informational alert, may impact other clinicians directly (e.g., view alert in EHR) or indirectly (e.g., referral from targeted clinician). Data represent a cumulative number of prescriptions filled.

The primary effectiveness outcome is the amount of SGLT2 or MRA prescriptions in eligible HF patients in the three intervention groups compared with the control group within 30 days of appointment. Thirty days is a common time interval for HF outcomes assessment and allows for chart documentation, care coordination, laboratory testing, and medication prescribing that may occur days after a patient encounter. The investigators will record all SGLT2 or MRA prescriptions, including those from non-targeted clinicians, given that nudge interventions, especially the informational alert, may impact other clinicians directly (e.g., view alert in EHR) or indirectly (e.g., referral from targeted clinician). Data represent a cumulative number of prescriptions filled.

Reach will be measured at the clinician level as the number of unique clinicians who received an informational alert or peer comparison report. The control group did not receive any alert.

The investigators will measure safety as the number of patients who were deemed by the participant (clinician) to discontinued prescribed medicine due to suspected adverse effects within 30 days of the nudge interventions.

Implementation will be assessed by clinician-directed survey of Acceptability of intervention Measure. The REDCap survey was on the Likert Scale: minimum 1 and maximum 5. With 5 being the highest. Acceptability was determine by the number of participants who were neutral or positive with their responses according the the Likert Scale. The anonymous surveys were sent to all in the three groups and did not collect any identifying information.

At the patient level, Reach will be measured as the number of unique patients for whom the participants received informational alerts

The investigators will measure the proportion of alerts relative to the total number of eligible patients with HF; this denominator will allow for comparisons of representativeness of the alert strategy.

Appropriateness will be assessed by clinician direct survey of Intervention Appropriateness Measure. Likert scale: minimum 1 and maximum 5. With 5 being the highest. The number of respondents who were neutral and positive were counted. The anonymous surveys were sent to all in the three groups and did not collect any identifying information.

Feasibility will be assessed by clinician direct survey of Feasibility of Intervention Measure. Likert scale: minimum 1 and maximum 5. With 5 being the highest. The survey participants responded used the Likert scale to determine if the intervention would be feasible. All responses neutral or positive were counted. The anonymous surveys were sent to all in the three groups and did not collect any identifying information.