COVID-19 is a new illness that can affect your lungs and airways. It's caused by a type of virus called a coronavirus. The virus has rapidly spread around the world, and is associated with a high risk of serious complications, including the need for specialised hospital care (for example, use of a breathing machine, or ventilator) and death.
We know that older people and individuals with pre-existing health problems who develop COVID-19 are at higher risk of serious disease or death. It is therefore important that we understand exactly which patients are at the greatest risk of harm, so that we can ensure those individuals can try and minimise their exposure to the virus (called “social distancing”), treat people earlier if possible, and make sure treatments are prioritised for those patients that are most likely to benefit.
We will aim to find out which existing health problems, drug treatments, or other factors (for example, smoking or pregnancy) are most strongly associated with people being admitted to intensive care or dying as a complication of COVID-19. We will do this by looking at routinely collected data from GP practices, hospital intensive care and national death records.
We intend to undertake the research extremely quickly, so that the findings can be passed on to the NHS and public health authorities to help guide care for the wider population as soon as possible.
COVID-19 is a rapidly evolving problem, presenting huge challenges to health services. There is an urgent need for tools to help clinicians, managers and policymakers decide how to optimise social distancing measures and to best deliver services to the most critically ill patients.
The proposed research will take advantage of a new linkage between the Clinical Practice Research Datalink (CPRD) and Intensive Care National Audit and Research Centre (ICNARC), COVID-19 Hospitalisation in England Surveillance System (CHESS) and Second Generation Surveillance (SGSS) datasets and aims to develop a risk stratification tool to determine which patients in primary care are at highest risk of admission to intensive care and which pre-morbid factors predict survival. We will specifically look for causal associations with cardiovascular drug therapy as this is amenable to change and thus a potential opportunity for intervention.
Survival models will be developed to describe the association between key pre-morbid clinical factors (e.g. sociodemographics, comorbidities, prescribing, other clinical factors) and three key outcomes: intensive care admission, intensive care survival, and length of intensive care stay. Standard established methodological approaches to data processing and analysis will be undertaken to minimise the time taken to develop the necessary models. Advanced methodologies to account for known biases (confounding by indication and selection bias) will be used to explore pharmacoepidemiological relationships between cardiovascular medication and COVID-19 outcomes.
The resulting risk stratification tools will have value in identifying individuals at greatest risk of severe illness, enabling more tailored social distancing, facilitating early pre-emptive care, targeting any future vaccine delivery, and planning for use of intensive care facilities. The linked dataset would also facilitate other epidemiological analyses of COVID-19.
Health Outcomes to be Measured:
- Admission to intensive care with confirmed COVID-19
- Survival of intensive care admission with confirmed COVID-19
- Duration of intensive care admission with confirmed COVID-19
- Positive test for SARS-CoV-2 (antibody or antigen) result or a coded diagnosis of COVID (GP coded record)
- Unplanned admission to hospital with confirmed COVID-19
- Non-invasive ventilation with confirmed COVID-19
- Duration of mechanical ventilation with confirmed COVID-19
- All death with COVID-19
2011 Rural-Urban Classification at LSOA level;HES Admitted Patient Care;ICNARC (COVID-19 Intensive Care National Audit and Research Centre);ONS Death Registration Data;Patient Level Index of Multiple Deprivation;SGSS (Second Generation Surveillance System)