Imperial College London

Academic Institute

Jennifer Quint - Chief Investigator - Imperial College LondonJennifer Quint - Corresponding Applicant - Imperial College LondonAnn Morgan - Collaborator - Imperial College LondonArdita Koteci - Collaborator - Imperial College LondonChukwuma Iwundu - Collaborator - Imperial College LondonClaudia Gulea - Collaborator - Imperial College LondonConstantinos Kallis - Collaborator - Imperial College LondonHannah Whittaker - Collaborator - Imperial College LondonMark Weeks - Collaborator - Imperial College LondonRikisha Shah Gupta - Collaborator - Imperial College London

CPRD681

It is increasingly understood that there is a wide range of symptoms and consequences amongst people with COVID-19. Whilst those who have been hospitalised often have ongoing symptoms due to severity of illness, there is increasing concern that those who have “mild disease” (i.e. managed in the community) may also be at risk of ongoing symptoms and development of new diseases beyond initial illness. We aim to describe the symptom profiles, healthcare use and new diseases following COVID-19 infection in a group of people who test positive and are not hospitalised, those who are hospitalised, and if numbers allow, those who may have possible COVID-19 infection.

Little is known about ongoing symptoms, healthcare utilisation and incidence of new disease among people with confirmed or possible COVID-19, and how this differs relative to severity of COVID-19 illness (i.e. those who are hospitalised). This will be a retrospective cohort study using routinely collected primary care records from CPRD Aurum in the first instance, followed by use of linked data from HES and ONS when more recent data are available. The study cohort will include adults who tested positive for COVID-19 or were diagnosed with possible COVID-19 between August - October 2020, with subsequent follow up until the end of data collection. We will describe outcomes between patients who were hospitalized with COVID-19 and patients who tested positive for COVID-19 but did not require hospitalization and if numbers allow, in those with possible COVID-19. We will estimate incidence of predefined symptoms, diseases, prescriptions and healthcare utilisation (including referral to hospital). We will use negative binomial regression to compare incidence rates based on aggregate level data taking into account the magnitude of both the numerator and denominator of interest. We plan to repeat analyses with a longer follow up period and with linked hospital and mortality data as they become available.

16/03/2021

Release