High-risk Influenza Vaccine Alert

Massachusetts Institute of Technology, National Institute on Aging (NIA), National Bureau of Economic Research, Inc.

The purpose of this study is to assess, prospectively, the effect on flu vaccination rates of salient alerts in the electronic health record that indicate a patient's high risk for flu and its complications. The investigators hypothesize that the salient alerts will lead to increased flu vaccination compared with a standard flu alert.

The CDC (Centers for Disease Control) recommends a flu vaccination to everyone aged 6+ months, with rare exception; almost anyone can benefit from the vaccine, which can reduce illnesses, missed work, hospitalizations, and death. One barrier to vaccination is a lack of "cues to action," and, in particular, the lack of direct recommendation from medical personnel; this barrier is arguably the most effectively overcome by a simple nudge of clinicians, compared with barriers such as negative attitudes toward vaccination, low perceived utility of vaccination, and less experience with having received the vaccine. Geisinger partnered with Medial EarlySign (Medial) to develop a machine learning (ML) algorithm to help identify people at risk for serious flu-associated complications based on existing electronic health record data. Eligible at-risk patients will be randomized to an active control group (clinician will be shown a standard flu alert) or one of two experimental groups (clinician will be shown an alert indicating patient's high risk, with or without describing the patient's factors contributing to that risk).

Vaccination, Health Promotion, Health Behavior, Reminder Systems, Economics, Behavioral, Risk Assessment

Patient-participants will not be explicitly told about the different arms, although clinician-participants will see the different arms and may notice these differences.

Flu alert that indicates patient is at high risk for flu and its complications and presents the factors contributing to this high risk

Alert header indicates patient is at high risk for flu and its complications; alert body indicates the percentage of risk (e.g., in the top 3% of risk)

Patient received a flu diagnosis via a positive polymerase chain reaction [PCR]/antigen/molecular test (yes/no) during the 2022-23 flu season (from the patient's appointment date through April 30, 2023)

Received a "high confidence flu" diagnosis (with positive polymerase chain reaction [PCR]/antigen/molecular test) and/or "likely flu" diagnosis (as assessed via International Classification of Disease [ICD] codes or Tamiflu administration or positive PCR/antigen/molecular test) (yes/no) during the 2022-23 flu season (from the patient's appointment date through April 30, 2023) Note that "likely flu" is a superset of the "high confidence flu" diagnoses.

Diagnosed with flu-related complications (yes/no) during the 2022-23 flu season (from the patient's appointment date through April 30, 2023)

Received at least one COVID-19 vaccination (yes/no) during the 2022-23 flu season (from the patient's appointment date through April 30, 2023)

Patient receives a flu vaccine (yes/no) during the 2022-23 flu season (from the patient's appointment date through April 30, 2023)

Patient Inclusion Criteria: Age 18+ Have been determined to be in the top 20% of risk through Medial's ML algorithm Attend an appointment where the flu alert fires (Geisinger sets when flu alerts start and end--between ~9/1/2022 and ~4/30/2023, as well as the trigger conditions for the alert, which includes valid departments and visits and excludes contraindications like Guillain-Barre syndrome) Clinician Inclusion Criteria: Any Geisinger clinician who sees patient-participants in our study for an appointment where their flu shot alert fires

Data with no personally identifiable information will be made available to other researchers on the Open Science Framework for transparency. This will include the essential data and code needed to replicate the analysis that yielded reported findings.

By the paper's online publication date. Data will remain available for as long as the Open Science Framework hosts it.