LOS ANGELES u2013 December 2019 – mPulse, the leader in Conversational AI solutions for health activation, announces the results of a two year study that assessed the influence of social determinants of health (SDOH) on the prescription refill rates of Medicare (Part D) patients and the benefits of Conversational AI text messaging solutions. The study analyzes a large cross-sectional dataset with results from over 99,000 patients and over 270,000 refill reminders across a 2-year period.
Social determinants are linked to non-adherence. Challenges such as financial stress and limited transportation options can impact an individualu2019s ability to get the prescriptions they need. These influences are commonly not shared with physicians and care teams, so they canu2019t provide patients with help and additional support services. This study shows that patients with low SDOH impact are more likely to request a refill than those with high SDOH impact.
The study used an SDOH Index, developed by mPulse, which quantifies SDOH impacts at a census tract level as well as a predictive model that can identify patients least likely to engage with the text messaging solution. Automated text conversations in Spanish and English were initiated with patients to allow them to refill their prescriptions conveniently. Once members impacted by SDOH barriers engaged with the conversational text message, they refilled at equivalent rates to those with lower SDOH barriers.
u201cUnderstanding health needs and challenges on both the individual and community levels is crucial to understanding and appropriately addressing them,u201d said Chris Nicholson Co-Founder and CEO of mPulse. u201cOur Behavioral Data Science team identified the opportunity to model the impact of SDOH on refill adherence, and the results demonstrate how programs and services can be better targeted to deliver the greatest impact.u201d
The study measured refill-rates in partially adherent and nonadherent Kaiser Permanente member populations. The results confirmed previously published pilot results but addressed a much larger population and leveraged mPulseu2019s multicultural Conversational AI capabilities to deliver text messages to Spanish speakers.
u201cWe continue to include more elements to address the cultural linguistic and SDOH-related impacts reported in this study, so that more patients can benefit from our solutionu201d said the paper’s corresponding author, Rena Brar Prayaga, Director of Behavioral Data Science at mPulse.
For this program, Conversational AI was used to automate conversations in both English and Spanish that helped members review and confirm their prescriptions and asked members about barriers they were facing to direct them to appropriate support services. Over the study period, 307,484 responses were received from members, and over 92% of these messages were successfully understood and handled by Conversational AI. The ability for automation to manage replies from members significantly reduced the demand on pharmacy staff resources, enabling staff to focus on higher level tasks and more complex patient needs.
Authors of the study are Erwin Jeong and Harmony Noble from Kaiser Permanente Southern California, Rena Brar Prayaga, Andrew Paster and Ram Prayaga from mPulse, and Ridhika Agrawal and Benjamin Nguyen from mPulse and Grinnell College.
