By Phil McSweeney, Executive Chair, PatientSource
Research published by Imperial College Institute of Global Health Innovation last week revealed just how widespread and potentially dangerous the interoperability of electronic medical records (EMR) is in the NHS. It demonstrates an over-reliance of archaic systems that don’t support the cross organisational and cross-geographical boundaries that modern healthcare requires.
The 152 hospital trusts surveyed use 21 separate electronic systems to record patient health care, and 23% of those trusts had only paper records.
The study revealed that in one year, on more than 11m occasions (9% of attendances) patients attended a hospital using a different health record system to their previous hospital attendance. There were several pairs of trusts that commonly cared for many of the same patients over the one-year study period. For example, just over 2m patients that attended two or more trusts had consecutive encounters shared between 20 pairs of hospitals. Only two of these pairs of trusts used the same EHR systems. This leads to inadequate data sharing that ultimately results in extremely costly inefficiency that could potentially put patients at risk of harm.
This study only begins to scratch the surface of a much bigger interoperability problem, more and more in evidence since the collapse of National Programme for IT (NPfIT). Any analysis of the application landscape in and around many acute hospitals shows the complex eco-system and frustrations that exist. In many locations there is inadequate digital data transfer available to and from primary care to acute, from ambulance services to acute, between A&E and the rest of the hospital, and so on. Many hospitals have their own stand-alone EMRs in specialities, like ophthalmology, and there is inadequate interoperability between laboratory systems, pathology systems and a main EMR. These challenging interoperability problems represent a significant barrier to the efficient sharing of digital records for tens of millions of patients, whether they move between trusts or not. They can also seriously hamper the efficient workflow of pressured clinical staff.
At the outset at PatientSource we began to try to address these problems. We’ve developed an intuitive, cloud-hosted and interoperable EMR. It can sit atop incumbent legacy systems and, via HL7, FHIR and open APIs, it can open incumbent systems from within the PatientsSource clinical record. We might call it ‘leading edge’ or ‘next generation’ or any other marketing hype language, but we prefer just to say it is fit for purpose. Interoperability should be ‘this generation’ not next. The PatientSource system could be interoperable with any of the 21 different EMR systems used across the 152 trusts.
Worryingly, it is still possible to procure an EMR that doesn’t give a user full interoperability. According to the BBC report on the research, the NHS said it was “working to ensure different systems could work together’, with a spokesperson for NHSX saying: “NHSX is setting standards, so hospital and general practitioner IT systems talk to each other and quickly share information, like X-ray results, to improve patient care.” In August this year we saw the framework list of eight EMR vendors who, among other criteria, have shown how they would interoperate with other systems to ensure that data is available to clinicians at the point of need. They were also asked to demonstrate not only how they would deliver enterprise wide solutions but also how they could provide thinner deployments that provide a basis for modular solutions (with or without SMEs and other partners). Even if they do these things well, they represent less than half of the 21 incumbents identified by Imperial College.
Back in 2018, Health 2.0 released an update to its EMR API survey which revealed the very slow movement by EMR vendors in allowing small health tech vendors to integrate their solutions with them. In 2018 the Health 2.0 survey stated that things were better than in their 2016 survey but the situation is still not good. We are continuing to inch in the right direction, but is this pace of change acceptable?
The Imperial College study provides further warnings and suggests some policy considerations. It advises we need to prepare for the ‘baby boomer bump’, a 33% increase in the population aged over 65. The concomitant increase in chronic illnesses including diabetes, heart disease, cancer and dementia together with increasing service specialisation and centralisation will make the lack of interoperability and data-sharing all the more problematic for patients, clinicians and delivering on long-term plans.
Their study suggests that an accurate, contemporaneous overview of the current use and spatial distribution of health record systems in the NHS in England is required, which is something the government needs to consider. Overlaying the use and distribution of these record systems with empirical data on patient movement between healthcare organisations can provide a valuable tool to guide better data sharing where it is most needed. The Imperial College research team also offers a metric – the proportion of patients with consecutive encounters at trusts using different health record systems. This information could be used to guide quality improvement in interoperability at a national or regional system level.
Lastly, they suggest that trusts should be encouraged to consider the systems in use at other trusts with which they commonly share patients when adopting new health record systems. I can’t see that happening easily but I do agree that all trusts should be urged to use open standards and develop suitable APIs to better link data between their different systems – both in new deployments and legacy systems. PatientSource already offers that – for this generation. It is time that all EMR providers follow suit.
