Creating a Learning Mental Health System using Data-Driven Technology
As a child and adolescent psychiatrist, I spend a large proportion of my day typing information into an Electronic Patient Record (EPR) System. This is time that could be spent seeing patients, so I expect that it should be of some value. Unfortunately, when I try to find out how my patients compare to others, whether my interventions work, or how much the care that I deliver costs compared to other services, it proves almost impossible.
At the same time, mental health suffers from huge gaps in the evidence base. We often don’t know whether commonly prescribed medications are effective because large trials have not been done on the complex patients who we meet in clinic. Mental health is not unique, but some of the issues are more acute because of a legacy of stigma and underfunding.
Challenges in direct care are mirrored in research. Annual UK Mental Health research funding is barely one fifth of that spent on cancer research. Meanwhile, spending is dwarfed by the suffering caused by mental illness and the huge costs to services and the economy. to be public and professional trust, information governance, cooperation and adoption, but ultimately, these data-driven technologies promise a mental health system that learns from every patient who is treated.
The recently published Long Term Plan for the NHS acknowledged these challenges and prioritised their resolution through increased funding and better use of data and technology. Indeed, the Plan pledges that digitally enabled care will become a mainstream feature across the NHS.
Some of the foundations for better data-driven care are already in place. Most mental health trusts use EPRs and our national Mental Health Data Collections are among the most advanced in the world. This provides a source of rich data at a local level and comprehensive data at a national level.
New sources of data, from outside the healthcare system, such as smartphones, sensors, social media and other commercial and public sector organisations are increasingly available. These can augment healthcare data but raise ethical and regulatory questions. Many data-driven technologies remain unproven and new methodologies are required to demonstrate their efficacy, cost effectiveness and safety.
Research in genomics and neuroimaging, in combination with advanced data analytics, finally seem set to have a major impact on frontline care, with fundamental implications for disease classification, prognostics and our ability to reduce the harm caused by treatments. Gaps in our knowledge about mental health interventions will increasingly be filled by low-cost research using data from EPR Even large traditional trials will be augmented by routinely collected data.
Speech recognition might reduce the time that I spend typing notes and natural language processing might allow us to analyse the free text that I enter. None of these technologies cause me to fear for my job, but I am hopeful that they will help me to make better decisions with patients and to spend more time providing care. In some cases, they will help us to prevent the development of disorders, in others they might help patients to take control of their own recovery, with access to the right treatment, at the right time and in the right setting.
With a new focus on improving mental health services come a huge opportunity to reimagine every stage of care, aided by data driven technologies. This report from Reform outlines those opportunities and the associated challenges
Among the biggest challenges are likely to be public and professional trust, information governance, cooperation and adoption, but ultimately, these data-driven technologies promise a mental health system that learns from every patient who is treated.
None of these technologies cause me to fear for my job, but I am hopeful that they will help me to make better decisions with patients and to spend more time providing care