Robert Jamison’s research while affiliated with Harvard Medical School and other places

Background We evaluated the outcomes of a pilot SMS text messaging–based public health campaign that identified social networking nodes and variations of response rates to develop a list of variables that could be used to analyze and develop an outreach strategy that would maximize the impact of future public health campaigns planned for Kuwait. Computational analysis of connections has been used to analyze the spread of infectious diseases, dissemination of new thoughts and ideas, efficiency of logistics networks, and even public health care campaigns. Percolation theory network analysis provides a mathematical alternative to more established heuristic approaches that have been used to optimize network development. We report on a pilot study designed to identify and treat subjects at high risk of developing type 2 diabetes mellitus in Kuwait. Objective The aim of this study was to identify ways to optimize efficient deployment of resources and improve response rates in a public health campaign by using variables identified in this secondary analysis of our previously published data (Alqabandi et al, 2020). This analysis identified key variables that could be used in a computational analysis to plan for future public health campaigns. Methods SMS text message screening posts were sent inviting recipients to answer 6 questions to determine their risk of developing type 2 diabetes mellitus. If subjects agreed to participate, a link to the Centers for Disease Control and Prevention prediabetes screening test was automatically transmitted to their mobile devices. The phone numbers used in this campaign were recorded and compared to the responses received through SMS text messaging and social media forwarding. Results A total of 180,000 SMS text messages through 5 different campaigns were sent to 6% of the adult population in Kuwait. A total of 260 individuals agreed to participate, of which 153 (58.8%) completed the screening. Remarkably, 367 additional surveys were received from individuals who were not invited by the original circulated SMS text messages. These individuals were invited through forwarded surveys from the original recipients after authentication with the study center. The original SMS text messages were found to successfully identify influencers in existing social networks to improve the efficacy of the public health campaign. Conclusions SMS text messaging–based health care screening campaigns were found to have limited effectiveness alone; however, the increased reach through shared second-party forwarding suggests the potential of exponentially expanding the reach of the study and identifying a higher percentage of eligible candidates through the use of percolation theory. Future research should be directed toward designing SMS text messaging campaigns that support a combination of SMS text message invitations and social networks along with identification of influential nodes and key variables, which are likely unique to the environment and cultural background of the population, using percolation theory modeling and chatbots.

BACKGROUND A SMS based public health program was undertaken in Kuwait to decrease the risk of developing type 2 diabetes. We analyzed our data to refine and develop a wide-ranging campaign with the goal of reducing the risk of developing type 2 diabetes in a population-based initiative. OBJECTIVE We evaluated the outcomes of a pilot short messaging service (SMS)-based public health campaign that identified social networking nodes and variations of response rates to develop a list of variables that could be used to analyze and develop an outreach strategy that would maximize impact of future public health campaigns planned for Kuwait and even other countries that would benefit from this approach. Percolation theory network analysis provides a mathematical alternative to more established heuristic approaches that have been used to optimize network development [1, 2]. We report on a pilot study that identified the potential impact of variables that could be optimized allowing for efficient deployment of resources and improve response rates in a public health campaign, that was designed to identify and treat subjects at high risk of developing type 2 diabetes mellitus (T2DM) in Kuwait. METHODS Text messaging screening (TMS) posts were sent inviting recipients to answer 6 questions to determine risk of developing T2DM. If subjects agreed to participate, a link to the Centers for Disease Control (CDC) prediabetes screening test was automatically transmitted to their mobile devices. The phone numbers used in this campaign where recorded and compared to the responses received through SMS responses and social media forwarding. RESULTS A total of 180,000 SMSs through 5 different campaigns were sent to 6% of the adult population in Kuwait. Two-hundred-sixty individuals agreed to participate while 153 (58.8%) completed the screening. Remarkably, 367 additional surveys were received from individuals who were not invited by the original circulated SMS messages. These individuals were invited by forwarded surveys from the original recipients after authentication with the study center suggesting that the original SMS messaging campaign identified influencers or nodes in existing social networks. CONCLUSIONS SMS-based healthcare screening campaigns were found to have limited effectiveness alone but the increased reach through shared second party forwarding suggests the potential of exponentially expanding not only the reach of the study, but actually identified a higher percentage of eligible candidates through the use of percolation theory network analysis. Future research should be directed towards designing TMS campaigns that support a combination of SMS invitations and social networks along with identification of influential nodes using percolation theory modeling and chat bots. CLINICALTRIAL IRBs of Harvard Medical School, Boston US and Dasman Diabetes Institute, Kuwait City, Kuwait