Unblocking the one-way system in health information – it’s time for two-way traffic
The idea of the information highway may be an outmoded cliché, but here we are, with health information starting to whiz its way to eager recipients including care providers, policy makers, industry and academia. The flow of this data is like a fleet of lorries travelling down a motorway. This is the direct result of valiant efforts by visionary individuals who understand the humanitarian potential of this data and took the necessary steps to move forward, despite taking a wrong slip road or two. However, one look at the motorway reveals that the flow is mainly in one direction, with a notable lack of inbound traffic. It’s true that data sharing has resulted in returns such as new policies, research papers, new programs and even (some) change in clinical practice, but these alone do not capture the full value of destination generated output.
This matters for three fundamental reasons:
Firstly, how can we fundamentally improve current practice efficiently without real-time learning? If change is based on the one-way exchange, then time to deliver healthcare improvement is wholly dependent on long-term cycles of traditional narrative dissemination, interpretation and adaptation. The cycle of research, policy and resulting change programs can take months to years to implement. Two-way traffic can help bridge this time lag, so that data can be acted on as it is generated, with real-time changes based on real-time information. This reflects the goals of rapid quality improvement initiatives crucial to delivering change at the point of care.
Secondly, let’s consider the brave new world of digital prescriptions, where applications will be as common as medications and artificial intelligence (AI) will contribute to clinical decision making. These interventions will offer a benefit that traditional practice cannot, transforming over time and possibly becoming a new intervention by learning across millions of patient records. This will be (hopefully) unsupervised adaptation, so we need to ensure the long-term safety and efficacy of these interventions. When a prescription is created it is trialled and once proven effective, deployed to the masses. If the formulae changes, the process is repeated. This process is not scalable for digital interventions and digicovigilance may require much more resource than pharmacovigilance. Sending data without integrating return data will lead inevitably to a gap between availability and safety.
Last but not least, despite the myriad of solutions in the pipeline that will revolutionise the delivery of care, there will continue to be a requirement for a real clinician to assess a patient. The clinician needs information to make assessments. Collecting that information from the patient is fraught with difficulty. For instance, a patient with arthritis will not necessarily remember the frequency, severity and under which circumstances pain occurs. An application may collect that information for them, but any number of factors may render the information unavailable during a clinical encounter. Repeating information recorded in an application to a physician manually is inefficient. The application vendor may also hold the data but may be unable to share it due to lack of infrastructure.
The patient should not BE the personal health record, but instead should HAVE a personal health record, a way for any clinician or service they interact with to use their data to improve their care and the system as a whole. Therefore, each procurement contract with a digital provider who is creating an application for patient care should have an agreement that data generated within the application should be interoperable and integrated with clinical records at a patient level, preferably via a personal health record. An information highway should have two-way traffic to enable and empower patients and clinicians. Therefore, the government should make it a priority to facilitate this two-way flow for the improvement of quality and safety across the entire digital health landscape.
Joshua Symons is Director of the Big Data & Analytical Unit (BDAU) at the Centre for Health Policy (CHP), Institute of Global Health Innovation (IGHI), Imperial College London.
Hutan Ashrafian is Chief Scientific Advisor and Honorary Senior Clinical Fellow at the Institute of Global Health Innovation (IGHI), Imperial College London