By Xiaoxuan Liu, medical doctor and a PhD student in ophthalmology at University Hospitals Birmingham NHS Foundation Trust and; Professor Alastair Denniston, Consultant Ophthalmologist (Uveitis and Medical Retina), University Hospitals Birmingham NHSFT & Hon Professor, University of Birmingham, UK
In today’s world, healthcare data is being generated and stored in vast quantities. The digitisation of health records has already demonstrated huge benefits for improving patient care, by enabling efficient information handling by healthcare practitioners and better communication between medical teams. We write this from our perspective as doctors who are both ‘coal-face’ clinicians and health data researchers. In our clinical role, we need a system of health data that is integrated, efficient (high speed and high quality), secure, and with clear accountability and data ownership. As researchers who work with health data, many of the needs are similar but there is also a requirement for reliable anonymisation, whereas the speed of responsiveness and transmission is less of a priority.
A health data system that is integrated
Unfortunately, the current setup remains inadequate for the level of information fluidity that we have become accustomed to. As it stands, electronic health record (EHR) data is stored in centralised, static databases. Even in a healthcare system such as the UK National Health Service (NHS), EHR data is siloed in a disjointed way and cannot be freely shared between individual units. The formats and structures for data storage are non-standardised and built to suit to the operations of each healthcare provider. Whilst there are some sharable forms of EHR data, the existing systems (such as EMIS, PACs, NORSE) tend of be purpose specific (i.e. specialty specific or data type specific) and require a significant degree of manual oversight to ensure secure transfer of data outside the host institution. An individual’s lifetime health record is therefore stored in pieces at each provider, often behind an opaque wall to which the individual has no access. In its current form, EHR data remains an ever-expanding resource with huge untapped potentials.
A health data system that is efficient
The current method of EHR storage is highly inefficient and not conducive to effective clinical care. Clinicians rely upon accurate and up-to-date health information to make informed decisions and treatment plans. This information comes in all forms and are generated in multiple places, such as general practitioners, hospitals, community clinics, dentists, pharmacies, nursing homes, high street optometrists, and so on. Patients interact with many different healthcare providers in a lifetime, with each interaction representing one facet of an individual’s health profile. Much of this data is held electronically, for example blood tests, imaging, histology samples, medication history, operation notes, correspondence from different medical specialties, etc. Currently, EHR data split across independent systems either cannot be readily shared, or even if it can be shared, requires time and manual facilitation, with the result being a duplicated copy of the existing source data. This current method is highly inefficient, and translates to poorly informed decision-making resulting in misdiagnoses, delays to treatment and avoidable harm to the patient. Implementing a shared data source, could bring huge benefits to clinical care delivery.
A health data system that is accountable and transparent
Healthcare data is currently hosted by providers, rather than the patient. What’s more, the patient tends not to have access to their own data, as the mechanisms for this are not in place. In many aspects of healthcare, the relationship between healthcare professionals and patients has evolved from a paternalistic one, to one of patient centricity. This shift in authority towards the patient is yet to be reflected in health data ownership. For now, healthcare professionals remain as the gateway for clinical information, releasing what is deemed appropriate to the patient.
Might it be technically, clinically and operationally feasible, in a health blockchain, to enable access to the EHR data controlled by a patient-managed decryption key? This would mean that access to the EHR data could be controlled by the patient, or whoever the patient shares the responsibility with (such as a care-giver).
A health data system that is research-ready
Health data research has made significant leaps in recent years with the availability of rich and diverse data from EHR systems. Health data driven studies improve our understanding of disease, risk factors, success rates of treatments through analysis of routinely collected clinical data. A movement of EHR to a blockchain seems to offer real potential to improve access to medical data for research. The authority to grant access could then be managed by patients, allowing them to maintain full control over who has access, what they have access to, and how long they can access it for. What’s more, this process of data sharing could be continuous and automatically updated throughout an individual’s lifetime, allowing prospective studies of healthy ageing/disease progression at the individual patient level.
Blockchain technology has the potential to bring about significant improvements in healthcare delivery, and ultimately better care for patients. But key to its successful deployment will be user trust. Both healthcare providers and patients must believe that the EHR data held within the blockchain will not be compromised. As such, continuous regulatory oversight and early involvement of users in the development process will be crucial for establishing and maintaining user trust.
Keywords
Healthcare, medicine, health data, electronic medical record, electronic health record, inter-operability