SMARThealth ChatGPT : Supporting community health workers to provide guideline-based maternal care in rural India
Background
- In India, approximately 1 million community health workers, known as ASHAs,
care for 25 million pregnant women and their babies each year. - Detecting high-risk conditions in women before complications arise is crucial for
averting maternal and newborn mortality and morbidity. In rural India, factors like
anaemia, hypertension and gestational diabetes persist as significant risk factors. - SMARThealth Pregnancy improves community-based screening and
management of anaemia, diabetes and hypertension during pregnancy and the
postnatal period. Integrating our large language model (LLM) chatbot into the
SMARThealth Pregnancy platform will enable real-time assistance for ASHAs
through the provision of guideline-based information.
Aims
- To collaboratively develop an LLM to help ASHAs deliver guideline-based care for pregnant and postpartum women in rural India.
Research Methodology
- This pioneering initiative aims to collaborate with ASHAs to co-create an LLM-based chatbot customised to their needs that will provide contextually precise, clinically accurate and gender-equitable responses tailored for maternal healthcare. Multiple rounds of user testing and feedback has already been gathered from ASHAs on the utility, accuracy, biases and limitations of the approach.
- Machine learning techniques will be implemented on a pre-trained ChatGPT 4.0 model, after which a repository of gold-standard questions will be developed with clinically validated answers. A user-friendly interface will be created in Hindi and Telugu with text-to-text, text-to-voice and voice-to-voice input/output options, and user personas tailored to ASHAs’ preference
