Using patients’ electronic health records to improve health and care

Coronary heart disease (CHD) is responsible for around 73,000 deaths in the UK every year and is the leading cause of avoidable deaths nationwide. Deaths are considered avoidable if good quality care or health interventions could have altered the outcome.

For the NHS, understanding the role of different healthcare services throughout a patient’s journey helps it identify shortfalls in care for CHD. Changes intended to reduce heart attacks and hospital admissions can then be made, with the potential to improve survival for all CHD patients. However, few large studies have investigated how national variation in patients’ journeys through primary care and hospital services - including measures used to prevent, diagnose and treat CHD - affects their outcome.

One way to look for opportunities to improve care is to analyse a patient’s interactions with healthcare providers through their electronic health records (EHRs). EHRs are kept for all patients using their unique NHS number and exist for primary care, hospitals, and disease and death registries. They contain rich data about patients’ symptoms, referrals, prescriptions and procedures.
No single EHR resource can fully illustrate the patient journey because resources are restricted to records for specific healthcare providers. However, linking several sources together could provide insights into the quality of care patients receive and how this affects their outcomes.

NIHR Programme Grants for Applied Research (PGfAR) awarded £2 million funding to a team of researchers at University College London (UCL) to carry out a major study investigating how linking EHR data sources could identify opportunities to improve cardiac patients’ care.

Led by Professor Harry Hemingway, Professor of Clinical Epidemiology at UCL, this programme of research realised the potential of linking different EHR data sources to study missed opportunities for heart disease patients. Using one of the world’s only specific EHRs for patients with cardiovascular disease (CVD), Professor Hemingway and his team accessed linked Clinical Practice Research Datalink (CPRD) EHRs to develop a novel research platform, CALIBER (ClinicAl disease research using LInked Bespoke studies and Electronic health Records). The new platform was designed so EHR data could be securely organised and analysed by approved researchers within a trusted research environment.

Our aim was to create a data resource that was potentially scalable or adaptable to the entire UK population of 65 million.
- Professor Harry Hemingway, lead researcher on the CALIBER study

CALIBER was created by linking EHRs from general practices (provided by CPRD) with records from hospitals (Hospital Episode Statistics, provided by NHS Digital) and data from the Office for National Statistics. The linked data covers patients’ diagnoses, prescriptions, operations, level of social deprivation and cause of death.

Different coding systems (e.g. ICD-10, SNOMED CT, Read codes) are used to describe the same health condition in different parts of the healthcare system, so in the past EHR data has been difficult and time-consuming for researchers to use. CALIBER is a unique research platform that supports open science by providing researchers with access to characteristics of health conditions (phenotypes) derived from linked EHRs.

During the early years of CALIBER, the platform held anonymised data from 2 million initially healthy adults in primary care, 100,000 people with angina and up to 300,000 patients who had experienced a heart attack. The NIHR funding award supported a programme of projects to evaluate different stages of patients’ medical journeys for angina and heart attack. This allowed the team to better understand patients’ risk factors by tracking healthy patients who went on to develop CVD.

The projects also identified opportunities for earlier diagnosis of CVD and preventative care to improve patients’ health. They investigated how variations in care across the country and different settings can affect patients’ outcomes and, in one study, showed that where opportunities for care were frequently missed, more deaths were recorded within 30 days of a heart attack.

Alongside development of the CALIBER platform, one of the team’s projects focused on improving the diagnosis of angina in hospital-based chest pain clinics. They developed the 2010 NICE chest pain guideline into a web-based decision making tool designed to give evidence-based support to clinicians in clinics. The tool was tested in four clinics and although it did not alter the decision-making behaviour of clinicians, it provided a good understanding of how new tools could be integrated into EHR systems in the future.

The creation of CALIBER has facilitated national and international research and directly influenced policy and practice within cardiac care in the UK. The 2016 NICE guideline on assessing and diagnosing recent chest pain drew on the study’s findings to change the way in which patients with suspected stable angina are diagnosed. This also included recommendations that psychosocial factors, including depression, are assessed in people with a heart attack.

A CALIBER analysis of more than one million patients’ blood pressure measurements to assess the risk of developing different types of CVD influenced clinical practice guidelines in the UK (published by Public Health England, now the UK Health Security Agency) and the US (published in the Journal of the American College of Cardiology). Public Health England incorporated results from the team’s analysis of the cost of angina care into a tool to help commissioners provide cost-effective interventions to prevent CVD.

The UK government’s 2013 transparency in healthcare agenda was also influenced by the CALIBER research programme. The government committed to making clinician performance data available and linking primary and secondary care data ‘to allow new insights into the quality of services and better understanding of the way services interact’.

CALIBER has supported development of the Health Data Research UK (HDR UK) Phenotype Library, an open access resource providing the research community with information, tools and nearly 800 phenotyping algorithms for UK EHR research. The Phenotype Library is combined with health dataset information in HDR UK's Innovation Gateway, hosting content from numerous contributing organisations. This resource aims to drive the next generation of research methods.

CALIBER supported vital research on the impact of COVID-19, including a study published in The Lancet that showed the UK government’s infection control strategy prior to lockdown could lead to a high number of extra deaths. The research estimated that over 1 year, between 35,000 and 70,000 extra deaths would occur from COVID-19 among extremely vulnerable people with multiple underlying health conditions including CHD, kidney disease and diabetes. The results informed the government’s decision to enforce the UK lockdown in March 2020 and data were included in UK policy recommendations.

CALIBER also supported another important study looking at the impact of the COVID-19 pandemic on cancer care services. The research estimated there could be a 20% increase in deaths among people newly diagnosed with cancer as a result of changes to their care during the pandemic. Published in BMJ Open, these data informed policy recommendations from SAGE, NERVTAG and Independent SAGE.

Other projects using CALIBER have spanned drug discovery, cost-effectiveness studies and disease prevention. For example, a study on weight gain over 10 years, published in The Lancet Diabetes and Endocrinology, used the CALIBER platform to identify young adults as most at risk of weight gain. The findings were used to provide the public with an online tool to calculate their risk of weight change based on individual weight and height, age, sex, ethnicity and socioeconomic area characteristics.

The CALIBER programme demonstrates that NHS data collected over time in primary care holds an important key to unlocking new insights for public health action.
- Professor Harry Hemingway, lead researcher on the CALIBER study

CALIBER has become a key platform for scientific research using EHRs to understand population health. It continues to be an important vehicle for vital research within HDR UK and the NIHR University College London Hospital Biomedical Research Centre, informing national and international public health policy.

• Details about CALIBER and information on how to access the CALIBER resource is available at CALIBER, UCL Institute of Health Informatics.
• This study was funded by NIHR Programme Grants for Applied Research. The NIHR provided additional funding for specific substudies through its Research for Patient Benefit Programme and an NIHR Research Methods Fellowship award.
• More information about the study is available on the NIHR Funding & Awards website.
• Read more Making a Difference stories.