Engagement Analytics in Healthcare

The biggest challenge of using Digital Health isn’t technology - it’s keeping participants engaged. A 2025 Lancet Digital Health study reveals a key finding: 𝐚𝐠𝐞 𝐢𝐬 𝐭𝐡𝐞 𝐬𝐭𝐫𝐨𝐧𝐠𝐞𝐬𝐭 𝐩𝐫𝐞𝐝𝐢𝐜𝐭𝐨𝐫 𝐨𝐟 𝐚𝐝𝐡𝐞𝐫𝐞𝐧𝐜𝐞 & 𝐫𝐞𝐭𝐞𝐧𝐭𝐢𝐨𝐧 𝐨𝐟 𝐝𝐢𝐠𝐢𝐭𝐚𝐥 𝐡𝐞𝐚𝐥𝐭𝐡 𝐩𝐫𝐨𝐝𝐮𝐜𝐭𝐬. 📊 𝐓𝐡𝐞 𝐝𝐚𝐭𝐚 𝐭𝐞𝐥𝐥𝐬 𝐚 𝐜𝐥𝐞𝐚𝐫 𝐬𝐭𝐨𝐫𝐲: ➤ Younger adults (18-29) have highest dropout rates, often early on. ➤ Older adults (60+) are the most consistent and engaged users. ➤ Novel multistate modeling algorithms can be used to predict dropout points and identify low engagement periods 💡 𝐊𝐞𝐲 𝐭𝐚𝐤𝐞𝐚𝐰𝐚𝐲𝐬 𝐟𝐨𝐫 𝐩𝐡𝐚𝐫𝐦𝐚 & 𝐝𝐢𝐠𝐢𝐭𝐚𝐥 𝐡𝐞𝐚𝐥𝐭𝐡 𝐥𝐞𝐚𝐝𝐞𝐫𝐬: ➤ Tailored engagement works. Strong onboarding, personalisation, and gamification help younger users, while simplified, educational, and user-friendly design benefit older adults. ➤ Timing matters. Participation fluctuates between weekdays and weekends - adjust strategies for content & notifications accordingly. ➤ Proactive intervention is key. Predictive insights can improve retention in real time. Product value & engagement is the foundation for creating effective digital health products that truly work. With the right design and personalisation we can drive positive behavioural change and improve outcomes with real-world impact. 🔗 Link to study in comments.

A voice- and text-enabled conversational agent was primarily used for health information, casual interactions, and clinical data entry — yet over half of users discontinued after a single session, underscoring barriers to sustained digital engagement. 1️⃣ Among 24,537 users of the Albert Health app, 58% engaged in only one session. 2️⃣ The most frequent intents were health information (32%), small talk (20%), and clinical parameter logging (16%). 3️⃣ Voice input dominated casual (64%) and medication-related (53%) interactions; screen-based input was preferred for clinical tasks (61%). 4️⃣ Participants in disease-specific programs exhibited higher sustained engagement than general health users (OR = 0.67). 5️⃣ A higher proportion of voice-based interactions was positively associated with continued use (OR = 1.005); screen-based interaction predicted attrition (OR = 0.994). 6️⃣ Engagement was more likely to be sustained when users employed a balanced mix of clinical and non-clinical intents (OR = 1.56). 7️⃣ Unexpectedly, higher system confidence scores in chatbot responses were associated with reduced user retention (OR = 0.43). 8️⃣ Users aged 15–45 were less likely to sustain engagement compared to pediatric or older adult cohorts. 9️⃣ Fall-back responses (13% of interactions) were frequently due to non-standard speech, slang, or recognition errors, highlighting limitations in natural language processing. 🔟 A modest engagement peak on day 8 aligned with reminder notifications, but overall retention remained low beyond initial use. ✍🏻 Selahattin Colakoglu, Mustafa Durmus, Zeynep Pelin Polat, Asli Yildiz, Emre Sezgin. User Engagement with A Multimodal Conversational Agent for Self-Care and Chronic Disease Management: A Retrospective Analysis. Journal of Medical Systems. 2025. DOI: 10.1007/s10916-025-02202-2

AI is poised to significantly shift one of healthcare’s most persistent and costly challenges: medication non-adherence. Despite decades of interventions, 50–70% of patients still don’t take medications as prescribed, increasing avoidable hospitalizations, complications, and costs. But recent research shows that AI-powered patient experience systems are changing the game, not by reminding harder but by engaging smarter. These tools deliver personalized, adaptive support at scale, improving adherence, reducing provider burden, and unlocking better outcomes. Here are 4 takeaways for building the next generation of digital engagement: 🎯 𝟭. 𝗥𝗲𝗶𝗻𝗳𝗼𝗿𝗰𝗲𝗺𝗲𝗻𝘁 𝗹𝗲𝗮𝗿𝗻𝗶𝗻𝗴 𝗶𝘀𝗻’𝘁 𝗷𝘂𝘀𝘁 𝗳𝗼𝗿 𝗮𝗱 𝘁𝗲𝗰𝗵 Static reminders can only do so much. In the REINFORCE trial, AI-personalized text messages boosted medication adherence by up to 36.6% in people with diabetes — a powerful example of how adaptive messaging can drive real behavior change. ➡️ These systems learn what motivates each patient, then adjust in real time to optimize impact. 🤖 𝟮. 𝗖𝗼𝗻𝘃𝗲𝗿𝘀𝗮𝘁𝗶𝗼𝗻𝗮𝗹 𝗔𝗜 𝗰𝗮𝗻 𝗯𝗿𝗶𝗱𝗴𝗲 𝘁𝗵𝗲 𝘀𝘂𝗽𝗽𝗼𝗿𝘁 𝗴𝗮𝗽 Tools like Wysa and Roberto use CBT-based dialogue to address psychological barriers and sustain engagement. ➡️ In one trial, patients averaged over 33 sessions — a strong signal of lasting value. 📊 𝟯. 𝗛𝘂𝗺𝗮𝗻-𝗔𝗜 𝗰𝗼𝗹𝗹𝗮𝗯𝗼𝗿𝗮𝘁𝗶𝗼𝗻 𝘀𝘁𝗶𝗹𝗹 𝘄𝗶𝗻𝘀 Collaborations where physicians refined AI outputs consistently outperformed both AI alone and traditional physician-only decision-making. ➡️ The future isn’t full automation. It’s a smart augmentation of clinical expertise. 🧠 𝟰. 𝗔𝗜’𝘀 𝘃𝗮𝗹𝘂𝗲 𝗹𝗶𝗲𝘀 𝗶𝗻 𝗽𝗿𝗼𝗮𝗰𝘁𝗶𝘃𝗲 𝗿𝗶𝘀𝗸 𝗺𝗮𝗻𝗮𝗴𝗲𝗺𝗲𝗻𝘁 AI models can predict medication non-adherence with up to 82% accuracy, including in high-risk transplant patients, enabling proactive intervention before problems escalate. ➡️ Adaptive AI systems don’t just improve adherence. They reduce downstream healthcare costs and enhance treatment ROI. For pharma and digital health leaders, this is a call to shift from reminders to responsive systems: ✔️ Tailored content, not templates ✔️ Adaptive cadence, not static schedules ✔️ Predictive insights, not retrospective reporting Research clearly demonstrates: Combining AI, behavioral design, and clinical oversight transforms patient experience into a core strategic advantage. And building evidence-backed, regulatory-compliant engagement systems ensures patient safety, trust, and long-term success. 📥 Curious how these systems could be applied in your therapeutic area. Let’s talk about designing engagement that scales and performs. #AIinHealthcare #DigitalHealth #PatientEngagement #MedicationAdherence #ReinforcementLearning #HumanAICollaboration #HealthOutcomes Sources: DOI: 10.1038/s41746-024-01028-5 DOI: 10.2196/37302 DOI: 10.3390/pharmacy11010018 DOI: 10.3389/fdgth.2021.669869 DOI: 10.1001/jamanetworkopen.2020.37107

Pharma has finally moved from static NPI lists to dynamic, behavior-based segmentation. AI now lets us predict engagement, personalize content, and automate campaigns with compliance built in. But here’s the hard truth: even the most tailored content is wasted if it doesn’t land during the narrow window when an HCP is actually making a treatment decision. Traditional provider engagement models are designed to maximize reach and frequency. However, they struggle to deliver relevance during the increasingly narrow windows when clinical consideration can be influenced. That window? It rarely lasts more than 7 days. Any engagement, no matter how personalized, delivered before or after is just noise. The real breakthrough? Using prospective diagnostics data, lab and genomic signals that reveal real-time clinical need. Unlike claims data, which lags behind, diagnostics data tells you when a doctor is about to make treatment recommendations for a patient that is eligible for a specific therapy. You can now: 1. Engage HCPs before the treatment decision is made (think within hours, not weeks) 2. Leverage your content as relevant information not spam 3. Ensure eligible patient has a chance to receive the right therapy Having access to prospective diagnostics data may not be enough though. You need a partner who can activate that data almost instantly, across multiple channels. The brands that win are those who don’t just collect insights, but can trigger compliant, real-time engagement. Whether it’s digital, field force, or peer-to-peer - right when the clinical decision window opens. Choosing a partner with both the data and the activation muscle means your message reaches the right HCP, on the right channel, at the exact moment it matters. That’s how you move from missed opportunities to measurable impact. Bottom line: Sophisticated segmentation is useless if your content isn’t timely and relevant. The brands that win are those who deliver the right message, at the right moment, when the HCP is actually making a decision for the patient in front of them. #PharmaLeadership #HCPengagement #Diagnostics #PrecisionMedicine #CommercialExcellence #PatientCentric #PharmaExec #Innovation #Pharma #HCPMarketing #HCPTargeting Diaceutics PLC

"𝗔𝗻𝗼𝘁𝗵𝗲𝗿 𝗴𝗲𝗻𝗲𝗿𝗶𝗰 𝗲𝗺𝗮𝗶𝗹 𝗮𝗯𝗼𝘂𝘁 𝘆𝗼𝘂𝗿 𝗱𝗿𝘂𝗴?"  𝘋𝘦𝘭𝘦𝘵𝘦. Your perfectly crafted message just joined thousands of others in a doctor's trash folder. Reps and MSLs spend 60% of their time on admin instead of building the relationships that matter. Healthcare professionals are drowning in noise. And everyone's losing. When engagement isn't personalized, everyone loses. HCPs miss critical information that could impact patient care. Sales teams waste time on low-value activities. Meanwhile, competitors who adapt faster are winning mindshare and trust. But what if AI could transform each interaction from 𝘨𝘦𝘯𝘦𝘳𝘪𝘤 to 𝘨𝘦𝘯𝘪𝘶𝘴? It's already happening: A global team uses an AI powered omnichannel platform to tune content and outreach for each HCP in real time. The system watches what each HCP actually clicks and when they prefer to learn. It notices formats too. Articles. Webinars. Short summaries. Case studies. If a cardiologist reads real-world evidence late at night and ignores promos, the platform suppresses standard emails. It serves non-promotional insights instead. It lines up invites to medical webinars that fit their schedule. It keeps the focus on value. For each HCP, the AI drafts instant clinical briefs and suggests a next best action for the rep or MSL. That might be a follow up call on the practical use of a new therapy. Or a peer-reviewed paper on a topic they've shown interest in. Reps and MSLs see dynamic, prioritized call lists with context cards. Recent activity. Preferred times. Content consumed. Suggested openers. They show up relevant and right on time. Chatbots answer common questions 24/7. That reduces repetitive in-person detailing and clears the runway for deeper talks. Does this replace reps or MSLs? No. It upgrades them: • Less admin. More science and trust. • AI finds the signal. Humans bring judgment and nuance. • Field teams decide what matters for this patient, this clinic, today. • Feedback from the field teaches the system what good looks like. The loop gets smarter. Teams using this approach report: • 3x faster information delivery • 78% higher HCP engagement rates • 92% of reps saying relationships are stronger 𝗧𝗵𝗲 𝗔𝗜 𝗿𝗲𝘃𝗼𝗹𝘂𝘁𝗶𝗼𝗻 𝗶𝗻 𝗽𝗵𝗮𝗿𝗺𝗮 𝗲𝗻𝗴𝗮𝗴𝗲𝗺𝗲𝗻𝘁 𝗶𝘀𝗻'𝘁 𝗰𝗼𝗺𝗶𝗻𝗴. 𝗜𝘁'𝘀 𝗵𝗲𝗿𝗲.  Look at what industry leaders are already achieving: • 𝗡𝗼𝘃𝗮𝗿𝘁𝗶𝘀 transformed their commercial model, achieving 2x better HCP engagement with AI-guided timing • 𝗠𝗲𝗿𝗰𝗸 reduced sales rep admin tasks by 30% through AI automation • A 𝘁𝗼𝗽-𝟭𝟬 𝗽𝗵𝗮𝗿𝗺𝗮 doubled their email open rates with AI-powered smart timing 𝘈𝘯𝘥 𝘵𝘩𝘪𝘴 𝘪𝘴 𝘫𝘶𝘴𝘵 𝘵𝘩𝘦 𝘣𝘦𝘨𝘪𝘯𝘯𝘪𝘯𝘨. 𝗪𝗵𝗮𝘁'𝘀 𝘆𝗼𝘂𝗿 𝘁𝗮𝗸𝗲? 🔄 How do you think this will change the rep-HCP relationship? 𝘚𝘩𝘢𝘳𝘦 𝘺𝘰𝘶𝘳 𝘵𝘩𝘰𝘶𝘨𝘩𝘵𝘴 𝘣𝘦𝘭𝘰𝘸. 𝘈𝘴 𝘴𝘰𝘮𝘦𝘰𝘯𝘦 𝘪𝘯 𝘵𝘩𝘦 𝘵𝘳𝘦𝘯𝘤𝘩𝘦𝘴, 𝘺𝘰𝘶𝘳 𝘱𝘦𝘳𝘴𝘱𝘦𝘤𝘵𝘪𝘷𝘦 𝘮𝘢𝘵𝘵𝘦𝘳𝘴.

Why does patient engagement drop after appointments? 🤔 Patients feel disconnected once they leave the clinic. Here’s how AI can help bridge that gap: 👉 Continuous, personalized follow-ups = better outcomes. Just like behavior change, patient engagement is in small, consistent touchpoints. When patients don’t receive regular support, they lose track of their treatment plan. How can AI drive engagement? ✅ Timely reminders. Subtle nudges to take medications, follow exercise routines, or schedule check-ups to keep patients on track. ✅ Personalized content. AI tailors follow-ups to each patient’s needs, making sure they receive relevant health information. ✅ Ongoing communication. Chatbots and virtual assistants can check in with patients between appointments, answering questions and providing guidance. ✅ Data-driven insights. AI monitors progress and provides feedback to both patients and healthcare providers, improving the care plan over time. ✅ Improved patient satisfaction. Continuous support leads to better experiences and long-term engagement. Consistency is key. Small, timely interactions can make a big difference in patient outcomes. Reference: CareSignal® – Lightbeam's Deviceless Remote Patient Monitoring

Patient engagement in healthcare is fundamentally broken: - ~50% of patients don’t adhere to treatment plans - Up to half drop out of care over time - Poor adherence is estimated to contribute to ~10% of hospitalizations and $100–300B in avoidable costs annually Most outcomes are driven between visits, but that’s where the system does almost nothing. A new layer is emerging: continuous, AI-driven patient engagement reinforcing care plans in real time: - Daily check-ins (symptoms, adherence) - Medication and measurement reminders - Context-aware behavioral nudges - Escalation to care teams when needed For example, with the Nozomi AI companion, instead of hoping patients follow instructions, you continuously support and guide patient behavior. Result: - Up to ~20–30% improvement in adherence (program-dependent) - Higher engagement and program completion - Fewer missed visits More in the scheme ↓ P.S. If you're working on chronic care or care programs, let’s discuss a pilot. 👉 𝗝𝘂𝘀𝘁 𝗰𝗼𝗻𝗻𝗲𝗰𝘁 𝘄𝗶𝘁𝗵 𝗺𝗲 𝗮𝗻𝗱 𝗰𝗼𝗺𝗺𝗲𝗻𝘁 “𝗣𝗶𝗹𝗼𝘁”. #engagement #adherence #patientengagement

Kicking off 2026 with a question many digital health programs are still figuring out: what happens when engagement drops and the intervention doesn’t adapt? This week’s #ResearchwithPillar review points to an important direction for the future of digital health: adaptive, whole-person coaching that responds in real time to user needs. A pilot trial published in Springer Nature examined LvL Up (Level Up), a holistic mobile health lifestyle coaching program, tested whether combining app-based coaching, peer support, and targeted human coaching using an adaptive SMART trial design. Researchers recruited 123 adults in Singapore between the ages of 21 and 59 who were at risk for chronic conditions. Participants enrolled in either the intervention or control group. What’s intersting isn’t the LvL UP program itself, but how support shifted when participants stopped engaging. Here are a few key takeaways from the study: ✨ Adaptive coaching matters. Nearly half of participants were initially non-responders; targeted, MI-informed human coaching helped re-engage many. ✨ Blended models are feasible. Over 95% of participants were paired with a peer supporter, and retention exceeded 90%. ✨ Whole-person outcomes showed promise. Positive trends were observed in mental well-being, psychological distress, and sleep duration. ✨ Engagement is still the bottleneck. Digital session adherence lagged, reinforcing the role humans can play when motivation dips. TL;DR: Digital health works best when it’s designed to respond, not just deliver. Programs that adapt to real engagement patterns by blending technology with human support may be better positioned to drive meaningful outcomes at scale. #digitalhealth #lifestylemedicine #healthcoach -------- Pillar’s digital engagement platform enables organizations to deliver and scale care programs that equip patients with the necessary resources to engage proactively with their health. Our white-label infrastructure programmatically guides patients from outreach and activation to digital pathways with on-demand education, services and forms. With an in-house network of over 1,200 vetted providers, Pillar is an industry leader in health coaching with enterprise partnerships ranging from chronic condition management and care navigation to AI evaluation and quality assurance.