Enterprise AI Analysis
A community-codesigned LLM-powered chatbot for primary care: a randomized controlled trial
With a global shortage of primary healthcare physicians—particularly in resource-limited settings—large language models (LLMs) have the potential to support and enhance patients’ health awareness. Here we developed P&P Care (Population Medicine and Public Health), an LLM-powered primary care chatbot using a dual-track role-play codesign framework where community stakeholders and researchers simulated each another’s perspectives across four phases: contextual understanding; cocreation; testing and refinement; and implementation and evolution. The codesigned chatbot was integrated with e-learning modules and tested in a randomized controlled trial. The trial included 2,113 participants (1,052 women and 1,061 men) from urban and rural areas across 11 Chinese provinces who were randomly assigned to receive a consultation either with preparatory e-learning via the P&P Care or without. The study met its primary endpoint with the e-learning group showing significantly higher objective health awareness (mean score 2.95 ± 1.22) compared with the consultation-only group (mean score 2.34 ± 1.02; P<0.001). Codesign offers a scalable solution for deploying LLMs in resource-limited settings. Chinese Clinical Trial Registry identifier: ChiCTR2500098101.
Executive Impact Summary
We developed P&P Care, a patient-facing LLM-powered primary care chatbot, using a dual-track role-play codesign framework designed to enable equitable LLM deployment. In a RCT across 11 provinces in China, this community-codesigned chatbot, which integrates literacy-bridging e-learning alongside technical and cultural adaptations, provided high-quality primary care consultations. The codesign approach effectively engaged underserved populations by-passed by conventional telemedicine and hospital-centric LLM applications, proving essential for harnessing LLMs in resource-limited settings. While tested in China, this approach may offer a globally scalable model to augment primary care capacity, addressing the critical gap between the rapid AI advances and their equitable real-world deployment.
2.95
Objective Health Awareness (mean score)
1.03
Consultation Efficiency (mean difference)
98.43%
Network Completion Rate (Rural)
Deep Analysis & Enterprise Applications
Select a topic to dive deeper, then explore the specific findings from the research, rebuilt as interactive, enterprise-focused modules.
The dual-track role-play codesign framework enabled community-aligned chatbot development, synergistic integration of multimedia e-learning modules, and the design of a pragmatic RCT. Community stakeholders and researchers simulated each other's perspectives across four phases: contextual understanding; cocreation; testing and refinement; and implementation and evolution.
This approach ensured usability, cultural appropriateness, and effective deployment in resource-limited settings by addressing key barriers such as AI health literacy deficits and infrastructural constraints. Key outputs included patient-centric symptom articulation protocols, adaptive dialogue algorithms, short video tutorials, and modular e-learning content.
The randomized controlled trial included 2,113 participants across 11 Chinese provinces. The e-learning plus group showed significantly higher objective health awareness (mean score 2.95 ± 1.22) compared with the consultation-only group (mean score 2.34 ± 1.02; P<0.001).
Subgroup analyses revealed disproportionately greater benefits for older participants, women, and those requiring multidisciplinary consultations. The e-learning intervention significantly reduced the health awareness gap, aligning perceived and actual knowledge, crucial for judicious health decisions. P&P Care also outperformed usual care and telemedicine in primary care quality domains.
The codesign process shaped a pragmatic RCT design, context-aligned trial arms, patient-centred evaluation rubrics, and culturally appropriate recruitment. Challenges in implementation included infrastructural limitations in rural areas, leading to engineering optimizations like tiered service backbones and offline caching, achieving a 98.43% completion rate in ablation testing.
The success in rural China, despite physician shortages, highlights the model's potential for global scalability. Future directions include addressing linguistic heterogeneity, public hesitancy, and regulatory uncertainties to ensure equitable and sustainable LLM deployment.
66.7%
Reduction in AI health literacy demands during cocreation
Codesign Framework Steps
Contextual Understanding
→
Cocreation
→
Testing & Refinement
→
Implementation & Evolution
| Feature | P&P Care (e-learning plus) | Usual Care | Telemedicine |
|---|---|---|---|
| Diagnostic Precision |
|
|
|
| Test Ordering Appropriateness |
|
|
|
| Long-Term Disease Management |
|
|
|
| Accessibility in Rural Areas |
|
|
|
Empowering Older Adults
The P&P Care chatbot, especially with its integrated e-learning modules, demonstrated a disproportionately greater benefit for older participants (≥40 years) compared to younger individuals. This group showed a substantially higher increase in objective health awareness and significant improvements across all consultation quality metrics, including listenability, attention, conciseness, integrity, and empathy.
This finding challenges the stereotype of low digital adaptability in older adults, positioning them as a high-impact demographic for this e-learning integrated AI solution. By narrowing the health awareness gap and enhancing self-assessment accuracy, P&P Care helps older adults make more judicious health decisions, fostering greater agency in managing their health.
Advanced ROI Calculator
Estimate the potential return on investment for your enterprise by leveraging AI-powered solutions to enhance efficiency and reclaim valuable team hours.
Annual Savings
$0
Annual Hours Reclaimed
0
Implementation Roadmap
A strategic phased approach to integrate P&P Care into your operations, ensuring smooth adoption and maximized impact.
Phase 1: Pilot Deployment & Feedback Integration (Months 1-3)
Conduct initial pilot studies in selected community settings. Gather user feedback to refine chatbot and e-learning modules. Optimize network performance and speech recognition for local dialects.
Phase 2: Regional Expansion & Localization (Months 4-9)
Expand deployment to additional regions, incorporating cultural and linguistic adaptations. Establish local support teams and community engagement strategies. Begin initial impact assessment on health awareness.
Phase 3: Advanced Feature Integration & Scalability Assessment (Months 10-18)
Integrate advanced features such as enhanced disease management protocols and personalized health pathways. Assess scalability and resource efficiency across diverse settings. Prepare for broader national or international adoption.
Phase 4: Long-Term Monitoring & Policy Advocacy (Months 19-24+)
Implement continuous monitoring of user outcomes and system performance. Conduct longitudinal studies on health impacts and cost-effectiveness. Engage with policymakers to establish regulatory frameworks and secure sustainable funding for equitable AI healthcare.
Ready to Transform Your Enterprise?
Leverage the power of community-codesigned AI to address critical healthcare gaps and empower your population with enhanced health awareness.
Ready to Get Started?
Book Your Free Consultation.
Let's Discuss Your AI Strategy!
Lets Discuss Your Needs
Select a Date
Sun
Mon
Tue
Wed
Thu
Fri
Sat