"Multicomponent nudge" interventions were successful in prompting an increase in clinicians ordering, and patients obtaining influenza vaccination, in a trial of the interventions against usual care conducted in 47 primary care outpatient clinics in Philadelphia and Seattle.1
"In this trial, we show that a multicomponent nudge to patients and clinicians that includes previsit text messaging, peer comparison to clinicians, and an automatic pended order result in an increase in influenza vaccine completion of 5.1 percentage points (in risk difference for probability of completion) during the visit," report lead author Shivan Mehta, MD, MBA, MSHP, Perelman School of Medicine, University of Pennsylvania, PA, and colleagues.
The BE IMMUNE (Behavioral Economics to Improve and Motivate Vaccination in Primary Care Using Nudges through the Electronic Health Record) pragmatic, cluster-randomized trial was conducted at primary care clinics affiliated with Penn Medicine (Philadelphia, PA) and UW Medicine (Seattle, WA) from September 25, 2023 to February 20, 2024. Funding from the National Institute of Aging facilitated evaluating the interventions in different health systems and different areas of the Country.
Patients were randomized at a 2:1 ratio to the intervention (n=52,526) or usual care (n=20,142) groups. The previsit text messages to patients randomized to the intervention group were sent 3 days and 1 day prior to the clinic appointment. The texts had endowment framing, with the first advising that the influenza vaccine was available, and the second indicating that it was on reserve for them.
In addition, intervention patients deemed high-risk for not completing vaccination were randomized at a 1:1 ratio to receive either the standard texting (n=20,4112), or an "intensification nudge" (n=20,440) with a bidirectional text message that included a menu of barriers and concerns about the vaccines. The patients receiving the bidirectional text could respond and receive tailored information and links.The groups deemed at high-risk patient were the elderly, of non-Hispanic Black race, in residence in lowest quartile zip code median household income, or who did not receive vaccination in prior year.
"Prior work has shown that patient-directed nudges before visits can boost vaccine completion," Mehta told Contagion."We wanted to build upon this and further engage with primary care clinicians to have conversations with patients about getting vaccinated, since we know they are a highly trusted source of health information."
The two clinician-facing interventions were an automated pended order for the vaccination, as the default for the clinician to sign or cancel; and monthly emailed feedback comparing the clinician's influenza vaccination completions rates to that of peers.The investigators noted that peer comparison feedback to clinicians has generally been effective at reducing "low-value services," but that there has been limited evidence for its utility in preventative health.
The trial primary outcome was completion of vaccination at the clinic visit. Secondary outcomes were completion of the vaccination within 3 months of the visit, and the influenza vaccine ordering rate.
What You Need to Know
Combining patient texts, default pended orders, and clinician peer comparison raised influenza vaccine completion during visits by 5.1 percentage points versus usual care.
Automated pended orders and peer comparison feedback significantly increased vaccine ordering and completion, reinforcing the influence of clinicians as trusted messengers in preventive care.
Bidirectional texts did not outperform standard reminders among high-risk groups, underscoring that nudges help overcome inertia or mild hesitancy but are less effective against entrenched resistance or misinformation.
Mehta and colleagues reported that the adjusted odds ratio (AOR) for vaccination at the clinic visit with the nudge intervention compared to usual care was 1.28 (97.5% CI, 1.13-1.45), with a probability of 31.4% compared to 26.4%.Ordering of vaccination was also higher with the intervention, with probability of 34.7% compared to 27.6% with usual care. There was a statistical difference between the intervention and usual care for vaccine completion within 3 months.
The bidirectional text messaging did not have evidence advantage over standard texting for the high-risk group.The investigators suggest this might be related to its use of automated, rather than clinician responses.In addition, they posit, "high-risk patients may have had strong hesitation or barriers to vaccination that could not be overcome by text messaging alone."
Mehta elaborated on this concern."There are a lot of reasons for declining flu vaccine rates, and misinformation is one of them. Our prior work has shown that social norms in the form of endorsement by primary care clinicians can increase participation in health prevention, as they can have a conversation with patients about the benefits of getting vaccinated, particularly for older and higher-risk individuals.
"While these types of nudges can help when there is hesitancy or inertia, they will not typically work if there is active resistance," Mehta conceded.
Reference
1. Mehta SJ, Waddell KJ, Linn KA. Nudges to clinicians and patiets for influenza vaccines during visits. The BE IMMUNE Randomized Clinical Trial. JAMA Intern Med. 2026. Published online January 5, 2026. doi:10.1001/jamainternmed.2025.7133.
