Gregory D. Zimet1, Dena S. Cox
2, Anthony D. Cox
2, Hana Anderson
3, Janet N. Arno
4, Edward J. Brizendine
5, and Barry P. Katz
5. (1) Department of Pediatrics, Indiana University School of Medicine, Indiana University, 410 W. 10th Street, HS 1001, Indianapolis, IN, USA, (2) Kelley School of Business, Indiana University, Indianapolis, IN, USA, (3) Chicago Department of Public Health, Chicago, IL, USA, (4) Marion County Health Department, Bell Flower Clinic, 1101 West Tenth Street, Indianapolis, IN, USA, (5) Biostatistics, Indiana University School of Medicine, Indianapolis, IN, USA
Background:
Although prior research and theory suggest that message framing and provider recommendations may be effective approaches to influencing discrete health behaviors such as immunization, systematic evaluations of these intervention approaches has been lacking.
Objective:
To evaluate, via a randomized, controlled design, two sets of interventions designed to increase HBV vaccine acceptance: 1. Three message framing conditions (gain-framed; loss-framed; control) and 2. Two provider communication conditions (vaccine offered; vaccine recommended).
Method:
Patients with no history of HBV infection or vaccination attending three STD clinics were recruited to complete an audio computer-assisted self-interview (ACASI). Within the ACASI, participants were randomized to the three framing conditions. Upon completion of the ACASI, participants saw a health care provider and were re-randomized to the two communication conditions. The outcome was acceptance of the first dose of HBV vaccine.
Result:
1,288 patients completed the ACASI. They were 60% male, 81% African-American and 18-80 years old (median=29). 54% overall received the first dose of vaccine. The breakdown of vaccine receipt by intervention was as follows: Gain-Framed (55.6%); Loss-Framed (55.6%); Control (49.8%); Vaccine Offered (49.8%); Vaccine Recommended (57.6%). Multiple logistic regression analysis indicated that compared to the control condition, neither the gain-framed (OR=1.26, 95%CI=.96-1.65) nor the loss-framed (OR=1.26, 95%CI=.96-1.65) condition increased uptake over the control condition. In contrast, participants in the vaccine-recommended condition were 1.4 times more likely to receive the vaccine than those in the vaccine-offered condition (OR=1.37, 95%CI=1.10-1.71).
Conclusion:
There was no effect of message-faming on HBV vaccine acceptance, but simply recommending rather than offering vaccine resulted in a significant increase in vaccination rates.
Implications:
Brief behavioral interventions can have a substantial impact on public health by influencing discrete health behaviors such as vaccination. However, the negative results related to message framing suggest that there is limited value to that particular approach to behavior change.