How Many People Are Still Ill from COVID?

Long COVID — defined by NICE as symptoms persisting four or more weeks after infection — peaked at 2.1 million sufferers across Great Britain in January 2022. The ONS COVID Infection Survey, which tracked prevalence monthly until its discontinuation in March 2023, was the primary measurement tool; subsequent estimates put the figure at approximately 1.5 million in 2024. The REACT-2 study found 1 in 14 infected people developed persistent symptoms. Women are affected more than men, and prevalence concentrates heavily among working-age adults aged 35–69, making this one of the UK's largest chronic illness cohorts.

The workforce damage is substantial. Some 800,000 people report long COVID limits their daily activities; 50,000 are unable to work at all. Long-term sickness inactivity has risen from 2.1 million to 3.2 million since 2019, and OBR estimates attribute roughly 20% of that increase to long COVID. The most reported symptoms — fatigue at 72%, difficulty concentrating at 44%, breathlessness at 38% — are precisely those that erode productivity. The estimated economic cost runs to £5 billion per year in lost output, from both complete absence and reduced capacity among those still working.

NHS England established 85 long COVID assessment clinics across the country, down from 91 at peak, with an average 12-week wait for an initial appointment. Scotland and Wales run similar networks; Northern Ireland remains comparatively under-resourced. There is no licensed treatment — care is symptom-based, combining respiratory physiotherapy, occupational therapy, and psychological support. The NIHR has funded over £50 million of research including the PHOSP-COVID hospitalised cohort study. Multiple clinical trials of antivirals, anti-inflammatories, and metabolic interventions were running in 2024–25. Vaccination halves the risk of developing long COVID.

Long COVID's occupational distribution mirrors pandemic exposure patterns. Healthcare workers, care home staff, transport workers, and others unable to work from home were disproportionately affected during the first two waves before vaccination. The leading pathophysiology hypotheses centre on viral persistence in tissue reservoirs, dysregulated autoimmune responses triggered by infection, and microclot formation impairing oxygen delivery to organs. Children were not spared — multisystem inflammatory syndrome (MIS-C and PIMS-TS) hospitalised hundreds, with some experiencing lasting cardiac and neurological effects. Post-exertional malaise, where physical or cognitive effort triggers symptom relapse lasting days, makes conventional rehabilitation counterproductive and traps patients in enforced inactivity. Clinical trials running in 2024–25 are testing antivirals such as ensitrelvir, low-dose naltrexone for neuroinflammation, and apheresis for microclots, though none has yet produced definitive results. The £50 million NIHR research programme remains modest relative to the scale of the condition.

The ONS COVID Infection Survey, which provided the most robust prevalence estimates through random household sampling, was discontinued in March 2023 due to funding cuts — creating a significant measurement gap during a period when the condition was still evolving. There is no internationally agreed case definition: NICE defines long COVID as symptoms persisting beyond four weeks, while the WHO requires three months, producing substantially different prevalence counts from the same population. The widely cited 1.5 million figure for 2024 is extrapolated from older survey data and modelling assumptions rather than directly measured. Self-reported symptom surveys cannot reliably distinguish long COVID from other health conditions — depression, deconditioning, or thyroid disorders — that emerged or worsened during the same period. Economic cost estimates ranging from £2.5 billion to £5 billion per year are highly sensitive to assumptions about whether affected individuals experience reduced productivity or complete workforce absence.

Long COVID's occupational distribution mirrors pandemic exposure patterns. Healthcare workers, care home staff, transport workers, and others unable to work from home were disproportionately affected during the first two waves before vaccination. The leading pathophysiology hypotheses centre on viral persistence in tissue reservoirs, dysregulated autoimmune responses triggered by infection, and microclot formation impairing oxygen delivery to organs. Children were not spared — multisystem inflammatory syndrome (MIS-C and PIMS-TS) hospitalised hundreds, with some experiencing lasting cardiac and neurological effects. Post-exertional malaise, where physical or cognitive effort triggers symptom relapse lasting days, makes conventional rehabilitation counterproductive and traps patients in enforced inactivity. Clinical trials running in 2024–25 are testing antivirals such as ensitrelvir, low-dose naltrexone for neuroinflammation, and apheresis for microclots, though none has yet produced definitive results. The £50 million NIHR research programme remains modest relative to the scale of the condition.

The ONS COVID Infection Survey, which provided the most robust prevalence estimates through random household sampling, was discontinued in March 2023 due to funding cuts — creating a significant measurement gap during a period when the condition was still evolving. There is no internationally agreed case definition: NICE defines long COVID as symptoms persisting beyond four weeks, while the WHO requires three months, producing substantially different prevalence counts from the same population. The widely cited 1.5 million figure for 2024 is extrapolated from older survey data and modelling assumptions rather than directly measured. Self-reported symptom surveys cannot reliably distinguish long COVID from other health conditions — depression, deconditioning, or thyroid disorders — that emerged or worsened during the same period. Economic cost estimates ranging from £2.5 billion to £5 billion per year are highly sensitive to assumptions about whether affected individuals experience reduced productivity or complete workforce absence.