Log2Lose: Development and Lessons Learned From a Mobile Technology Weight Loss Intervention
Ryan Shaw, PhD, RN ; Erica Levine, MPH ; Martin Streicher, MS ; Elizabeth Strawbridge, MS, RD ; Jennifer Gierisch, PhD ; Jane Pendergast, PhD ; Sarah Hale, PhD ; Shelby Reed, PhD ; Megan McVay, PhD ; Denise Simmons, BS ; William Yancy, MD ; Gary Bennett, PhD ; Corrine Voils, PhD
Background: Providing financial incentives has gained popularity as a strategy to promote weight loss, but questions remain about how best to utilize them. A promising mobile health strategy provides users with near-real-time financial incentives based on both the process of weight loss (behavioral modification) and actual weight loss. To maximize the impact of this strategy, a methodology is needed to close the gap between the desired behavior and the financial incentive. Leveraging mobile health tools—such as mobile phone apps, cellular body weight scales that transmit data to physicians and researchers, and text messaging for instructions and encouragement—has the potential to close this gap.
Objective: This study aimed to describe the development of an innovative technology-based solution and lessons learned from a feasibility trial—Log2Lose—that encouraged individuals to lose weight by providing near-real-time financial incentives for weight loss and/or dietary self-monitoring.
Methods: We recruited participants (N=96) with a body mass index greater than or equal to 30 kg/m2 for a 24-week weight loss trial. Participants received a behavioral intervention of biweekly, in-person group sessions and were instructed to log a minimum number of daily calories in MyFitnessPal and to step on the BodyTrace cellular scale at least twice per week. In a 2×2 design, participants were randomized into 4 groups to receive financial incentives for the following: (group 1) weekly weight loss and dietary self-monitoring, (group 2) dietary self-monitoring only, (group 3) weekly weight loss only, or (group 4) no financial incentives. Diet and weight data from the devices were obtained through application programming interfaces. Each week, we applied algorithms to participants’ data to determine whether they qualified for a monetary incentive (groups 1-3). A text message notified these participants of whether they met weight loss and/or self-monitoring requirements to earn an incentive and the amount they earned or would have earned. The money was uploaded to a debit card.
Results: Our custom-engineered software platform analyzed data from multiple sources, collated and processed the data to send appropriate text messages automatically, and informed study staff of the appropriate incentives. We present lessons learned from the development of the software system and challenges encountered with technology, data transmission, and participants (eg, lost connections or delayed communication).
Conclusions: With consistent and constant validation checks and a robust beta test run, the process of analyzing data and determining eligibility for weekly incentives can be mostly automated. We were able to accomplish this project within an academic health system, which required significant security and privacy safeguards. Our success demonstrates how this methodology of automated feedback loops can provide health interventions via mobile technology.
Trial Registration: ClinicalTrials.gov NCT02691260; https://clinicaltrials.gov/ct2/show/NCT02691260
JMIR Mhealth Uhealth 2019;7(2):e11972