Abstract
BACKGROUND: The radiological trajectory of post-COVID-19 is uncertain. We present a prospective, observational, multicentre cohort study using multimodality imaging to describe the pulmonary sequelae of patients hospitalised with COVID-19, predictors of persistent abnormal radiology and implications on health status. METHODS: In survivors of COVID-19, we performed convalescent CT pulmonary angiogram and high-resolution CT imaging as part of the CISCO-19 study (ClinicalTrials.gov ID NCT04403607). This included serial blood biomarkers and patient-reported outcomes 28-60 days following discharge from hospital. RESULTS: Of the COVID-19 cohort, 88 (56%) patients of the COVID-19 cohort (n = 159; mean age, 55 years; 43% female) had persisting radiological abnormalities at 28-60 days postdischarge. This included ground-glass opacification (45%), reticulation/architectural distortion (30%) or mixed pattern (19%). These features were very infrequent among a group of age-matched, sex-matched and cardiovascular risk factor-matched controls (n=29). The majority of COVID-19 cohort (68%) had less than 20% persisting radiological abnormalities, with 67% demonstrating overall improvement compared with admission imaging. Older age, premorbid performance status, typical acute COVID-19 radiological features, markers of severe acute COVID-19, convalescent ICAM-1 and P-selectin were associated with persisting lung abnormalities (all p<0.05). Patients with persisting abnormalities were shown to have lower levels of physical activity and predicted maximal oxygen utilisation (derived VO(2)) (both p<0.05). Higher percentage of abnormal lung parenchyma was associated with lower patient-assessed quality of life (EQ-5D-5L) score (p=0.03). CONCLUSIONS: Persistent radiological abnormalities post-COVID-19 were common at 28-60 days postdischarge from hospital, although most improved. Patients with persisting radiological abnormalities 28-60 days postdischarge are at risk of persisting health impairment in the longer term and represent a population for targeted intervention. TRIAL REGISTRATION NUMBER: NCT04403607.