[Skip to Content]
Access to paid content on this site is currently suspended due to excessive activity being detected from your IP address 54.205.176.107. Please contact the publisher to request reinstatement.
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
Download PDF
Figure. Flow of Study Participants
Figure. Flow of Study Participants
1.
Choi JJ, Laibson D, Madrian B, Metrick A. Optimal defaults.  Am Econ Rev. 2003;93(2):180-185Article
2.
Halpern SD, Ubel PA, Asch DA. Harnessing the power of default options to improve health care.  N Engl J Med. 2007;357(13):1340-1344PubMedArticle
3.
Kressel LM, Chapman GB, Leventhal E. The influence of default options on the expression of end-of-life treatment preferences in advance directives.  J Gen Intern Med. 2007;22(7):1007-1010PubMedArticle
4.
Johnson EJ, Goldstein D. Do defaults save lives?  Science. 2003;302(5649):1338-1339PubMedArticle
5.
Thaler RH, Sunstein CR. Nudge: Improving Decisions About Health, Wealth, and Happiness. New Haven, CT: Yale University Press; 2008
Research Letter
July 7, 2010

Opting In vs Opting Out of Influenza Vaccination

JAMA. 2010;304(1):43-44. doi:10.1001/jama.2010.892

To the Editor: Changes in how a choice is presented can affect the actions of decision makers, who have a tendency to stick with the default option.13 For example, organ donation rates are much higher in an opt-out system (donor status is the default, explicitly opting out is required if a person does not want to donate) than in an opt-in system (nondonor status is the default, explicitly opting in is required if a person wants to be a donor).4 Both systems give decision makers autonomy to choose according to their personal principles, but the opt-out system provides a “nudge”5 toward donation.

Although influenza vaccination may help prevent morbidity and mortality from seasonal or other pandemic influenza (such as 2009 influenza A [H1N1]), many people decline to receive an annual flu shot even when it is available for free at the workplace. We assessed whether modifying the default option could influence seasonal influenza vaccination.

Methods

In September 2009, after institutional review board waiver of signed consent, 480 faculty and staff employees at Rutgers University (stratified to achieve equal assignment by sex and employment category) were randomly assigned to 1 of 2 conditions (Figure). Those in the opt-out condition received an e-mail from the university occupational health department explaining that the participant had been scheduled for a flu shot appointment, with the day, time, and location provided; hyperlinks allowed participants to change or cancel the appointment. For those in the opt-in condition, the e-mail explained that free seasonal flu shots were available and provided a link to a Web page where participants could schedule an appointment for the following week. Five days later, all participants with an appointment (opt-out participants who changed or did not cancel their appointment and opt-in participants who made an appointment) were each sent a reminder e-mail about the appointment. Participants without an appointment could also be vaccinated as walk-ins. Two participants were excluded due to an e-mail error.

In January 2010, coders blinded to group assignment abstracted vaccination records at the occupational health department. χ2 analyses conducted in SAS version 9.2 (SAS Institute, Cary, North Carolina) had an 80% power to detect an effect size of at least 0.13 with α = .05. Logistic regression analyses tested mediation.

Results

In the opt-out condition, 108 of 239 participants (45%; 95% confidence interval [CI], 39%-52%) were vaccinated at the occupational health department, compared with 80 of 239 participants (33%; 95% CI, 27%-39%) in the opt-in condition (P = .008), a 36% relative increase. Six participants (3 in each group) were vaccinated after randomization but before initial e-mails were sent.

This difference was mediated by participant appointment status: only 18 opt-out participants (8%) canceled appointments, and only 50 opt-in participants (21%) made appointments. Consequently, opt-out participants were much more likely than opt-in participants to have an appointment (92% [95% CI, 89%-96%] vs 21% [95% CI, 16%-26%]; P < .001). Participants with an appointment were more likely than those without to get a flu shot, although not necessarily at the appointment time (Figure): 148 of 271 participants (55%; 95% CI, 49%-61%) vs 40 of 207 participants (19%; 95% CI, 14%-25%) (P < .001). Statistically controlling for appointment status eliminated the default effect (Sobel test = 7.64, P < .001).

Comment

Both opt-in and opt-out conditions allow decision makers to select the option they want, but the opt-out condition increased the probability of a flu shot appointment, which in turn increased the likelihood of getting vaccinated. The study was limited to a university workplace sample, and vaccination records were limited to vaccinations received at the occupational health department. Nevertheless, the results suggest that automatic scheduling of flu shot appointments may be an effective way to increase vaccination rates.

Back to top
Article Information

Author Contributions: Dr Chapman had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Chapman, Li, Colby, Yoon.

Acquisition of data: Chapman, Li.

Analysis and interpretation of data: Chapman.

Drafting of the manuscript: Chapman.

Critical revision of the manuscript for important intellectual content: Li, Colby, Yoon.

Statistical analysis: Chapman.

Administrative, technical, or material support: Li, Colby, Yoon.

Study supervision: Chapman.

Financial Disclosures: None reported.

Additional Contributions: We acknowledge the cooperation and support of the staff in the Rutgers Department of Occupational Health. They did not receive additional compensation for their contributions.

References
1.
Choi JJ, Laibson D, Madrian B, Metrick A. Optimal defaults.  Am Econ Rev. 2003;93(2):180-185Article
2.
Halpern SD, Ubel PA, Asch DA. Harnessing the power of default options to improve health care.  N Engl J Med. 2007;357(13):1340-1344PubMedArticle
3.
Kressel LM, Chapman GB, Leventhal E. The influence of default options on the expression of end-of-life treatment preferences in advance directives.  J Gen Intern Med. 2007;22(7):1007-1010PubMedArticle
4.
Johnson EJ, Goldstein D. Do defaults save lives?  Science. 2003;302(5649):1338-1339PubMedArticle
5.
Thaler RH, Sunstein CR. Nudge: Improving Decisions About Health, Wealth, and Happiness. New Haven, CT: Yale University Press; 2008
×