By Tom Foley.
Dr Afzal Chaudhry, Consultant Nephrologist, Chief Clinical Information Officer and Associate Lecturer, Cambridge University Hospitals
In 2014, CUH became the first UK healthcare provider to implement Epic, the market leading EHR. The Trust now has 2.5 years of rich structured clinical data that has been recorded through the EHR. The Trust is using this data for reporting and real time operational analysis and to monitor and reengineer pathways to ensure that patients are treated in the most effective and efficient way. They are also exploring other ways of using this data to improve services.
One key strand is to develop predictive models and decision support systems. They are beginning to roll these out across the major pathways. One example is a model that helps to identify if a patient potentially has sepsis, accompanied by a decision support system that guides clinicians towards appropriate investigations and antibiotics. Last year, 55% of sepsis patients arriving in the ED received antibiotics within 90 minutes, this year it is 100% and the time to first antibiotic dose continues to fall even further – this will lead to better outcomes. These decision support systems can take into account, care plans, severity scores, allergies, etc. and provide links to the underlying guidelines, to ensure that clinicians do not become deskilled.
There is potential for these predictive models and decision support systems to be shared between providers. Even if they do not use Epic, the clinical understanding, logic and algorithms can still be shared and implemented on other systems. For example, CUH is currently collaborating with a provider in Iowa, to create a predictive model for surgical wound closure.
Despite being one of the UK’s most digitally mature providers, CUH is “only scratching the surface” of what might be possible by using this data. Constraints include, funding, data quality, skills (e.g. data curation capabilities), need for partner trusts and IG complexities. The Trust is working towards a hybrid centralised / distributed model – employing both central data analysts / scientists along with analysts who are based in the specialty / business departments. Although this will require additional investment, it will not be on the scale required to develop the existing infrastructure and in part will be offset by a rationalisation of data management staff trust-wide.
Visualising the data at the right time and at the right organisational level is an on-going challenge. The Trust is hoping to allow clinicians to query the data themselves using tools without the need for direct informatics support. This would allow them to tailor treatment to patients based on the outcomes of previous similar patient cohorts. This requires consistent recording by clinicians. Operational buy in and enforcement is key to achieving this.
There are additional opportunities to provide information to patients and clinicians in a more timely fashion and to pre-empt needs, e.g. with a more educated and informed approach to discharge planning it should be possible to reduce length of stay. Quality of care can be improved further by continuing to automate administrative functions that currently reduce the amount of time that clinicians can spend with patients.
CUH and the University of Cambridge, through its NIHR biomedical research centre (BRC), have created an academic University Lecturer in Clinical Informatics post (Dr Lydia Drumright). Through the BRC, the Trust has built a research portal that will enable it to make its data available to academics, within an appropriate ethical and IG framework. They believe that there is scope to undertake a wide range of research on such data and have had significant interest from within the University of Cambridge, wider collaborators and industry. Historically, this has been limited by the lack of robust data and more work is needed on data provenance. It is essential that routinely collected and derived NHS data is seen to be as robust as academic / research data and that it is analysed and interpreted for actionable outcomes in the same way.
The Trust also believes that it has a responsibility to lead, to share its developments and to collaborate with other providers. It has been named as an NHS England Global Digital Exemplar. There is significant potential in regional data sharing networks that link primary and secondary providers.
CUH would welcome help in spreading best practice, promoting exemplars, funding for innovative developments, policy development and advocacy.