Time-Efficient Methods of Measuring the Impact of Targeted Dietary Intervention Campaigns: The Polyp Prevention Trial Experience

Abstract

LEARNING OUTCOME: To describe assessment tools that measure the impact of targeted campaigns within a dietary intervention trial. Dietary intervention trials face the challenge of maintaining changes over long periods of time. In addition to supporting behavior change, interventions need timely ways to assess the impact of both short- and long-term strategies in order to efficiently address problems. These considerations were pertinent in the Polyp Prevention Trial (PPT), a randomized trial that tests the effect of a low-fat (20% kcal from fat), high-fiber (18 gms fiber/1000 kcal), high fruit/vegetable (5-8 daily servings based on caloric intake) eating plan on the recurrence of large bowel adenomatous polyps, the precursor lesion for most colorectal cancers. Half of the 2079 PPT participants were randomized to the intervention group and provided with ongoing nutrition/behavioral counseling to help them maintain the eating plan for 4 years. Participants complete Food Frequency Questionnaires (FFQ) and 4-Day Food Records annually to measure dietary changes, but to prevent study bias, only trial averages are shared with participants and staff. At the end of Year 1, participants made marked progress toward goals, but plateaued in Year 2 with respect to fruits/vegetables(F/V) and showed a slight drift upward in fat percent. To encourage participants to reach and maintain goals, 3 targeted campaigns (each lasting 6 to 8 months) were launched sequentially: first a fat campaign (to lower fat); then the Fruit & Veg-a-thon (to increase F/V intake); and finally the PPT On My Mind campaign (to support progress toward all 3 goals). Campaign messages were delivered via phone, in-person sessions, and mail using a variety of print and non-print materials. While post-fat campaign FFQ data showed slight reductions in fat, it was difficult to attribute this decrease directly to the campaign. Furthermore, there was no way to give participants feedback about individual progress. Therefore, time-efficient methods to evaluate campaign impact were incorporated into the second and third campaigns. First, FFQ data were used to identify food group sources that differentiated goal achievers from those not reaching goals. On routine campaign contacts, previous day's intake information related only to these specific foods was collected and tracked and a simple impact analysis was performed monthly. This focused data collection provided information about the short-term impact of targeted intervention efforts (e.g., monthly changes in mean F/V intake during the Fruit & Veg-a-thon varied from +1.5 to 2 servings over pre-campaign levels) while taking less than 10 minutes/participant on average (vs. 60 to 90 minutes for an FFQ). These measures also offered a way to provide frequent feedback to participants to encourage further behavior changes.