The Challenge:
The project was commissioned by a charity that supports a world-renowned children’s hospital and it’s patients by, amongst other initiatives, funding research.
Ensuring the best patient care is a priority for the charity, therefore White Swan was asked to analyse the millions of online conversations focused on paediatric health, relating to their experience of certain conditions, to inform research strategies and improve care.
As a pilot project, this would then determine if this source of data could complement others in helping the hospital to understand patients’ experiences better in the future.
What We Did:
We created a natural language processing (NLP) based filter to identify conversations related to paediatric conditions and children's health within the millions of conversations online.
After identifying 900k documents likely to be relevant to the paediatric experience, we segmented this to focus on documents related to congenital heart disease and paediatric cardiomyopathy and their associated conditions, symptoms, and treatment experiences. The data was sourced from patient-centric social sites (e.g. forums, reviews, Reddit) over a two-year period. Quantitative and qualitative techniques were used to analyse this curated data to identify key themes and insights into the experience of patients and their carers.
The Results:
Informative and confirmative insights were uncovered and shared with the teams at the hospital. The project proved that social data has the potential to inform research strategies and provide significant insights into the treatment and symptom experience of paediatric patients. Further roll-outs are therefore planned across other clinical specialisms.
