The Effect of Postcard Reminders on Vaccinations Among the Elderly: A Block-Randomized Experiment

The behavioral sciences suggest that reminders can be an effective way to encourage behavior change. Reminders may be particularly helpful in encouraging the elderly to vaccinate; while children maintain well-recorded vaccination schedules, adults often do not and thus miss important vaccinations. We analyze the effect of a postcard reminder sent by the Louisiana Department of Health to 208,867 residents of Louisiana aged 65-70 who are listed as overdue at least one of four vaccines in the Louisiana Immunization Information System. We use block randomization and a stepped wedge design to study the effect of the postcard reminder. Elderly individuals are blocked by their current vaccine record, and randomized to receive the postcard in one of four consecutive months. Our design addresses heterogeneous effects across time and between individuals with different vaccine records. We compare between months when the postcard is sent out and find a larger effect of the reminder card earlier in the flu season. We also compare between individuals with different vaccines records. Finally, we compare immediate and lagged effects of the reminder. More generally, our study shows different ways to study an intervention when pure parallel trials are not possible.