Voice-Driven AI Interface for Collecting Patient-Reported Outcomes

Use Case:

Application of AI and voice technology to collect patient-reported outcomes (PROMs) via phone-based interfaces improving accessibility and maintaining data validity for populations with limited digital literacy.

Outcome:

Full alignment with technical and business goals defined in the StairwAI support program; Feasibility of AI-based answer mapping from voice responses to structured data confirmed; AI integration roadmap designed for PROMs use in real-world medical and pharmaceutical contexts; Risk assessment and mitigation strategies established for IVR-based data collection; Project ready for market launch and commercialization phase.

Ecosystem Support:

Support from the StairwAI project enabled Promptly Health to refine its AI roadmap, validate its feasibility study, and receive expert mentoring for both technical development and business strategy.

AI Relevance:

This success story illustrates the AI-on-Demand value proposition by: I) enabling voice-based access to digital health services for underserved or elderly populations; II) designing explainable AI models for clinical data mapping; III) demonstrating integration of AI into voice interfaces for health research and care; IV) promoting interoperability with existing PROM standards and healthcare IT systems.

Summary:

“Promptly Health, a Portuguese digital health company, tackled the challenge of collecting structured clinical data from patients who may lack digital proficiency or internet access. They developed and validated an AI-enhanced system for collecting Patient-Reported Outcome Measures (PROMs) using Interactive Voice Response (IVR) technology. This innovative solution allows patients to answer standardized health questionnaires by phone, with AI algorithms mapping their open-ended, voice-provided responses to predefined clinical answer sets. The system includes a pipeline that converts scientific PROM questionnaires to speech using text-to-speech, receives voice input through IVR, transcribes the input via speech-to-text, and applies AI to map responses accurately – without compromising scientific validity.
Risks—such as ambiguous answers or speech-to-text conversion failures—were carefully addressed with fallback mechanisms, including offering hints, repeat prompts, or keypad input. Promptly also developed a business model focused on offering this voice-AI service to healthcare providers, research organizations, and pharmaceutical companies through licensing, subscription models, and integration services. With TRL7+ reached and the technical feasibility proven, Promptly is now preparing for commercial launch, setting the stage for broader adoption of AI-driven voice tools in digital health ecosystems.”