As the pandemic developed, growing numbers of patients began reporting new health problems many weeks or months after being infected with COVID-19. Post-acute, or long COVID, is widespread but little understood – raising a multitude of issues for researchers, policymakers and clinicians. The development of a long COVID phenotype is a significant step forward, allowing a better understanding of who was most at risk of chronic post-COVID-19 symptoms.


Long COVID is clearly an important phenomenon – some estimates suggest 40% of COVID-19 patients experience ongoing health problems. However the symptoms (anything from respiratory problems to “brain fog” and muscle aches) and their severity are highly varied.

The condition seems to have been significantly under-reported in primary care and data, other than surveys by healthcare professionals, is limited.

Tools are needed to better understand and define long COVID so researchers can identify it from routinely collected patient data, so healthcare professionals are more able to diagnose new cases and policymakers and clinicians can take action on its prevention and management.


Research part-funded by HDR UK, allowed work to begin using data (from the Primary Care Sentinel Cohort (PCSC) of the Oxford Royal College of General Practitioners (RCGP)) on 7.4 million patients in England, to develop a long COVID phenotype.

Of 428,479 people identified as having had acute COVID-19, 7,471 (1.8%) were recorded as having long COVID (a figure that probably reflects significant under-reporting).

The researchers explored who was at most risk from long COVID, their socioeconomic circumstances, comorbidities and symptoms (and how these compared with acute COVID-19). They also compared hospitalised people with long COVID with those not hospitalised.

Their work involved close working with GPs – and their willingness to help in identifying long COVID.


The research results were detailed in a paper entitled Developing a Long COVID Phenotype for Postacute COVID-19 in a National Primary Care Sentinel Cohort: Observational Retrospective Database Analysis (JMIR Public Health and Surveillance, August 2022).

Co-author Bernardo Meza-Torres, a researcher at the University of Oxford’s Nuffield Department of Primary Care Health Sciences, said: “This research helped with recognition of long COVID as a condition. In the beginning there was some resistance to the term long COVID because it refers to a heterogeneous group of conditions.”

Long COVID patient profiles often varied from those of people (frequently male and older) most at risk from acute COVID-19.

Tendencies among long COVID patients included:

  • being female (64.7%)
  • being aged 20-70
  • having been hospitalised and/or admitted to ICU with acute COVID-19
  • being obese
  • living in a conurbation
  • comorbidities like depression, anxiety, asthma, and hypertension.

The team’s phenotype enables long COVID cases to be identified from routinely collected patient information and has been made publicly available to facilitate research.

This will allow retrospective comparisons between different groups, for example in vaccine exposure between long COVID and non-long COVID groups.

Lead author Dr Nikhil Mayor, an NIHR Academic Clinical Fellow, said: “This was about identifying people with chronic problems. People can present with all manner of symptoms affecting all different body systems, which makes it difficult to classify.

“As a complex condition, long COVID require further research. There are many questions to answer – we want to know what prevents long COVID and whether vaccination be protective.”

What the Impact Committee said:

The committee highlighted the paper’s contribution to understand of the human phenome as well as its realised or potential impact on public health including decision-making and economics.


Email nikhil.mayor@nhs.net or bernardo.meza-torres@phc.ox.ac.uk.