In a peer-reviewed study recently published in Nature Digital Medicine, Waymark co-founder and Head of Clinical Sanjay Basu and VP of Data Science and AI Sadiq Patel – along with co-authors Joseph Rigdon and Dean Schillinger –  found that LLMs that use a SMART (Sequential Multiple Assignment Randomized Trial) design can help Medicaid programs more reliably personalize communication strategies for their patients.

Historically, health plans and provider groups have relied on A/B testing different outreach strategies to deliver more targeted patient communications aimed at initiatives like HEDIS gap closure campaigns. However,despite its widespread use, traditional A/B testing is limited in personalizing interventions and determining why some patients are more or less responsive to different communication strategies. The findings of this study offer insights into the implications for and impact on patient communications LLM tools can have when leveraged strategically.

Read the full study here.