Assessing the impact of the COVID-19 pandemic on the accuracy, completeness and agreement of stroke cases across a national registry and whole population electronic health records
James Farrell, John Nolan, Roger Lambert, Ana Torralbo, Steffen E Petersen, Mevhibe Hocaoglu, Chris Tomlinson, Reecha Sofat, Qi Huang, Evan Kontopantelis, Martin James, Sarah Lessels, Jacqueline A L MacArthur, Angela Wood, William Whiteley, Spiros Denaxas, on behalf of the CVD-COVID-UK/COVID-IMPACT Consortium
Stroke is a leading cause of death and disability in the UK, affecting over 100,000 people annually, with more than 1.3 million stroke survivors living with its lasting effects. High-quality stroke care is essential to improve patient outcomes, with prompt diagnosis and treatment being crucial to minimise brain damage and enhance recovery. To improve the quality of care of people affected by stroke, it is important to have a good understanding of the numbers of people that are affected by stroke, the quality of care that they receive and their outcomes. Equally, it’s important to understand how the COVID-19 pandemic impacted the way stroke is recorded in the NHS.
Data is recorded in a person’s medical record whenever they interact with the health service, such as during a GP appointment or hospital visit. This data is recorded by healthcare professionals for use by other healthcare professionals, and includes information relevant to diagnosis, monitoring and treatment of disease.
Different healthcare providers, for example hospitals and GPs, record this information in different ways, using different codes and different computer systems. There are also disease audits or registries, which are designed to capture information on a particular disease that is not routinely captured in primary care or hospital records.
Information on stroke is recorded in multiple different datasets, including primary care, hospitalisations and death records. Stroke information is also captured by the Sentinel Stroke National Audit Programme (SSNAP), an audit that aims to capture information to judge the quality of care of stroke patients admitted to stroke units. The SSNAP dataset provides more detailed information that is not available in other NHS records, however it is possible that it misses some people affected by stroke, for example those not admitted to hospital. We also do not know if the number of people missed in SSNAP and other sources changed during the COVID-19 pandemic due the disruption in healthcare and data recording.
Combining data across different datasets can provide a more complete picture. It can also be used to provide additional details, for example by supplementing information from GP records with the results of tests that can determine stroke type, carried out at a hospital. It can also help us plan and prepare better for any future pandemics and understand how data are recorded during periods of healthcare disruption.
Carrying out these analyses across the entire population of England is complex and requires the development of complex computational instructions to efficiently extract, combine, and analyse relevant data across different sources.
Citation details to follow. Preprint available at: https://doi.org/10.64898/2025.12.03.25340732
- View the analysis code used in NHS England's SDE for England
- View the phenotyping algorithms and codelists used in NHS England's SDE for England
This is a sub-project of project CCU005 approved by the CVD-COVID-UK / COVID-IMPACT Approvals & Oversight Board (sub-project: CCU005_08).
Licensed under the Apache License, Version 2.0 (the "License"); you may not use this software except in compliance with the License. You may obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0. Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.