A large study has found that half of long COVID patients see substantial improvement within three years of their initial infection.
The NIH-funded RECOVER-Adult study tracked 13,000 adults with a history of SARS-CoV-2 infection. Researchers focused on 1,019 patients who had a Long COVID Research Index (LCRI) score of 11 or higher at the 12-month mark. This index, ranging from 0 to 30, assesses the presence and severity of 11 symptoms, including postexertional malaise, fatigue, and brain fog, providing a full measure of long COVID severity.
Tracking Recovery Over Time
The study employed a time-to-event analysis to monitor symptom improvement over three years. Patients were assessed at 12, 18, and 24 months post-infection, with follow-ups to evaluate changes in their condition. This approach allowed researchers to capture individual disease trajectories, distinguishing it from previous studies that focused on symptom prevalence at specific time points.
The primary outcome was defined as two consecutive visits with an LCRI score below 11. Secondary outcomes included a 50% reduction in LCRI score coupled with good self-reported quality of life, measured using the PROMIS-GH survey, and complete remission (LCRI score of 0). By 24 months, 15% (12.6-17.6) of participants achieved complete remission, rising to 26.9% (23.3-30.5) by 36 months.
Promising Results
By 36 months, 50.1% of patients met the primary outcome, with a median time of 36 months (33- >51 months) to achieve this milestone. The probability of a 50% LCRI reduction alongside good quality of life reached 56.4%. These findings highlight the potential for significant recovery over time, even for those with persistent symptoms.
A Step Forward
While the study does not establish causal relationships between individual characteristics and recovery timing, it provides critical insights into the natural history of long COVID. The findings, published in Ann Intern Med (2026, Epub November 3. doi:10.7326/ANNALS-25-05275), offer a more subtle understanding of recovery patterns, moving beyond prevalence estimates to prognostic information.