Mental health re-imagined: Deep data science approach

I was honored to take the stage at the National Medical Research Council’s Awards Ceremony and Research Symposium in Singapore recently to speak about ongoing challenges in behavioral health, and the momentum we have in the market to address these challenges.

There was lively engagement from the audience, a testament to the fact that the problems surrounding behavioral health resonate with many stakeholders here in Singapore, as well as around the globe.

Despite the rise in prevalence and cost associated with mental illness, innovations in behavioral health are limited. The key contributors to this problem include:

  • When compared to other disease areas, the field has less access to aggregated, longitudinal real-world data.
  • The majority of insights are captured in unstructured clinical notes within patient electronic health records and are not readily usable for research, meaning there is much untapped potential.
  • Outcomes collected in clinical trials are not translatable to routine clinical care.
  • There are siloes between physical and mental health, with limited understanding of how comorbidities impact a patient’s health.

Overall, the endpoints for behavioral and mental health conditions are shockingly ill-defined, with no standardized set of quality metrics. In the absence of a standard set of common vocabulary and metrics at representative scale, there is little we can develop to improve outcomes of patients who need the most help. The good news is that we are at an inflection point for behavioral health with the converging rise in data science maturity, remote digital technology coming into fruition, and a shift in health care toward value-based care models. 

At Holmusk, we are harnessing these forces to make the provision of evidence the basis for decision making in behavioral health, starting with launching and commercializing the following solutions:

  • The richest clinical database in behavioral health, which reflects how behavioral health care is delivered in routine clinical practice
  • A curated behavioral health data model that ingests data from across a range of settings and treatment centers, and maps it to enable easy comparison and analysis
  • Proprietary natural language processing models powered by machine learning that unlock valuable information that is currently inaccessible and stored within unstructured clinical notes in the EHR
  • A trusted research environment that facilitates analysis and addresses pressing questions in behavioral health
  • A clinical analytics platform with built-in AI-powered predictive analytics for crisis risk and resource utilization management that is delivering outcomes for the NHS in the UK

The future of behavioral health care starts with data and the science that enables that data to guide us toward better decision making. With the right partners across the ecosystem, Holmusk has embarked on the path to start standardizing the data endpoints to enable data-driven care transformation. This will be the first step into enabling clinical decision support tools for routine care settings, as well as helping patients across the entire continuum of care. For example, the development of digital biomarkers will help in detecting early psychiatric symptoms and precision psychiatry tools for early intervention. 

I believe Singapore, as well as other countries in Asia Pacific with strong data infrastructure, are ready to take on this challenge to fuel data-driven innovation in behavioral health. We look forward to building this future with our healthcare partners in the region.

Visit our website to learn more about our work and how you can get involved.

Leave a Comment