Communication, Vaccination Refusal
Conditions
Keywords
Immunization, Persuasive Communication, Electronic Mail, Economics, Behavioral, Psychology
Brief summary
In this evaluation, two versions of emails will be sent to Geisinger employees who have not been vaccinated against COVID-19. A delayed-contact control group will be emailed after a delay of at least two days. The researchers hypothesize that either of the emails - which use content informed by behavioral nudge theory - will lead to more COVID-19 vaccinations than the delayed-contact control group.
Detailed description
The researchers aim to evaluate the effectiveness of email reminders in promoting COVID-19 vaccinations among Geisinger employees. The project will compare two emails, which will vary in the subject line and email content. Both emails will test different approaches that draw from principles in behavioral science: social proof, reframing, and scarcity. Employees receiving these emails will be compared with employees in a delayed-contact group, who will not receive emails. Employees in the control group will later receive one of the emails.The responses to the emails will help determine the communications later sent to the control group. All emails will ask recipients if they want to receive a COVID-19 vaccination. If they click on a hyperlink indicating yes, they will be forwarded to an online registration and scheduling homepage. If they click on a hyperlink indicating no, they will be forwarded to an online questionnaire surveying people about their main reason for declining the vaccine at this time. The questionnaire will automatically present information intended to assuage the specific concern that people endorse, followed by a hyperlink to the scheduling site. The data will be analyzed with logistic regression models with the control group as the reference group, to compare the two email conditions versus the control group. This set of analyses will only be conducted for scheduling a vaccination appointment as opposed to email engagement outcomes (e.g., number of emails opened), which will not be applicable for the control group. A second set of logistic regression models predicting scheduling a vaccination appointment and email engagement will also be run, comparing the two email conditions against each other. The data will be examined two days later (to help decide which email to send to the delayed-control group) and four days later (when the policy for employee vaccination will change, making the content of the email irrelevant). Update: The researchers looked at 3 days of data - as the emails to the delayed contact group were only sent at that time - and 4 days of data when the policy for employee vaccinations changed.
Interventions
Sponsors
Geisinger Clinic
Study design
Allocation
RANDOMIZED
Intervention model
PARALLEL
Primary purpose
PREVENTION
Masking
DOUBLE (Subject, Caregiver)
Masking description
Participants will not know what group they are in. The relevant care provider (either their primary care provider or provider administering the vaccine) will not know about this study or assignment into groups.
Intervention model description
1 x 3 design
Eligibility
Sex/Gender
ALL
Healthy volunteers
Yes
Inclusion criteria
* Employees at Geisinger Health System who have not gotten a COVID-19 vaccination
Exclusion criteria
* None
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Number of Employees Who Scheduled Vaccination Appointments After 3 Days | 3 days | Binary variable of a scheduled vaccination appointment 3 days after the emails were sent |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Number of Employees Who Opened Emails After 3 Days | 3 days | Binary variable indicating whether the employee opened the email 3 days after the emails were sent |
| Number of Employees Who Clicked Scheduling Links After 3 Days | 3 days | Binary variable indicating whether the employee clicked the link to schedule a vaccination appointment 3 days after the emails were sent |
Other
| Measure | Time frame | Description |
|---|---|---|
| Number of Employees Who Scheduled Vaccination Appointments Until Policy Change (4 Days) | 4 days | Binary variable of a scheduled vaccination appointment up to the end of January 19, when new distribution policies were enacted. |
| Number of Employees Who Opened Emails Until Policy Change (4 Days) | 4 days | Binary variable indicating whether the employee opened the email up to the end of January 19, when new distribution policies were enacted. |
| Number of Employees Who Clicked the Scheduling Link Until Policy Change (4 Days) | 4 days | Binary variable indicating whether the employee clicked the link to schedule a vaccination appointment up to the end of January 19, when new distribution policies were enacted. |
Countries
United States
Participant flow
Participants by arm
| Arm | Count |
|---|---|
| Delayed Contact This group will not receive an email for at least two days. | 3,178 |
| Social Proof This group will receive an email that points out that vaccine availability for employees will soon be limited. It also emphasizes how many fellow health care workers have been vaccinated. Seeing the behaviors of other people might encourage recipients to copy that behavior.
Social Proof: Email
Scarcity Message: Email | 3,198 |
| Reframing Side Effects and Adverse Reactions This group will receive an email that points out that vaccine availability for employees will soon be limited. It also addresses concerns about the side effects and adverse reactions of the vaccine. Due to potentially overblown concerns about the vaccine caused by the salience of side effects in the (social) media, the email attempts to reframe the risks by explicitly noting the small possibility of being affected by serious side effects and by contrasting that to the more severe effects of COVID-19. As a result of this reframing, recipients might recalibrate their perception of risks and benefits and opt for vaccination.
Reframing: Email
Scarcity Message: Email | 3,190 |
| Total | 9,566 |
Baseline characteristics
| Characteristic | Total | Delayed Contact | Social Proof | Reframing Side Effects and Adverse Reactions |
|---|---|---|---|---|
| Age, Categorical <=18 years | 0 Participants | — | — | — |
| Age, Categorical >=65 years | 0 Participants | — | — | — |
| Age, Categorical Between 18 and 65 years | 0 Participants | — | — | — |
| Race and Ethnicity Not Collected | 0 Participants | — | — | — |
| Region of Enrollment United States | 9566 participants | 3178 participants | 3198 participants | 3190 participants |
| Sex: Female, Male Female | 0 Participants | — | — | — |
| Sex: Female, Male Male | 0 Participants | — | — | — |
Adverse events
| Event type | EG000 affected / at risk | EG001 affected / at risk | EG002 affected / at risk |
|---|---|---|---|
| deaths Total, all-cause mortality | 0 / 0 | 0 / 0 | 0 / 0 |
| other Total, other adverse events | 0 / 0 | 0 / 0 | 0 / 0 |
| serious Total, serious adverse events | 0 / 0 | 0 / 0 | 0 / 0 |
Outcome results
Primary
Number of Employees Who Scheduled Vaccination Appointments After 3 Days
Binary variable of a scheduled vaccination appointment 3 days after the emails were sent
Time frame: 3 days
| Arm | Measure | Value (COUNT_OF_PARTICIPANTS) |
|---|---|---|
| Delayed Contact | Number of Employees Who Scheduled Vaccination Appointments After 3 Days | 101 Participants |
| Social Proof | Number of Employees Who Scheduled Vaccination Appointments After 3 Days | 207 Participants |
| Reframing Side Effects and Adverse Reactions | Number of Employees Who Scheduled Vaccination Appointments After 3 Days | 220 Participants |
Comparison: We conducted a logistic regression with the two experimental conditions as dummy variables with delayed contact as the reference group.p-value: <0.00195% CI: [1.65, 2.69]Regression, Logistic
Comparison: We conducted a logistic regression with the two experimental conditions as dummy variables with delayed contact as the reference group.p-value: <0.00195% CI: [1.77, 2.87]Regression, Logistic
Comparison: We conducted a logistic regression with the two experimental conditions as dummy variables with delayed contact as the reference group.p-value: <0.00195% CI: [0.88, 1.3]Regression, Logistic
Secondary
Number of Employees Who Clicked Scheduling Links After 3 Days
Binary variable indicating whether the employee clicked the link to schedule a vaccination appointment 3 days after the emails were sent
Time frame: 3 days
| Arm | Measure | Value (COUNT_OF_PARTICIPANTS) |
|---|---|---|
| Delayed Contact | Number of Employees Who Clicked Scheduling Links After 3 Days | 562 Participants |
| Social Proof | Number of Employees Who Clicked Scheduling Links After 3 Days | 382 Participants |
Secondary
Number of Employees Who Opened Emails After 3 Days
Binary variable indicating whether the employee opened the email 3 days after the emails were sent
Time frame: 3 days
| Arm | Measure | Value (COUNT_OF_PARTICIPANTS) |
|---|---|---|
| Delayed Contact | Number of Employees Who Opened Emails After 3 Days | 1481 Participants |
| Social Proof | Number of Employees Who Opened Emails After 3 Days | 1398 Participants |
Other Pre-specified
Number of Employees Who Clicked the Scheduling Link Until Policy Change (4 Days)
Binary variable indicating whether the employee clicked the link to schedule a vaccination appointment up to the end of January 19, when new distribution policies were enacted.
Time frame: 4 days
Other Pre-specified
Number of Employees Who Opened Emails Until Policy Change (4 Days)
Binary variable indicating whether the employee opened the email up to the end of January 19, when new distribution policies were enacted.
Time frame: 4 days
Other Pre-specified
Number of Employees Who Scheduled Vaccination Appointments Until Policy Change (4 Days)
Binary variable of a scheduled vaccination appointment up to the end of January 19, when new distribution policies were enacted.
Time frame: 4 days