Design: Phase 2, randomised, double-blind, placebo-controlled trial, 69 sites
Population: 959 US adults with long COVID of cognitive, autonomic or exercise phenotype
Primary outcome: Clinically significant change in patient-reported outcomes at day 90
Effect: No significant difference in any phenotype; e.g. cognitive 25-day regimen 3.2% (95% CI −10.4 to 16.8)

RECOVER-VITAL tested whether long COVID is driven by persisting virus. It randomised 959 adults with at least 12 weeks of cognitive, autonomic or exercise-related symptoms at 69 US sites. Participants received 15 or 25 days of nirmatrelvir-ritonavir or placebo with ritonavir. The primary outcome was a clinically significant change in patient-reported scores at day 90.

Neither course helped any phenotype. Adjusted differences against placebo ranged from −7.8% to 3.2%, with no significant results and no difference in secondary performance measures. There were no deaths and no safety signal.

This follows earlier negative trials of shorter antiviral courses. Prescribing antivirals for long COVID outside a trial is not supported. Management remains symptom-based: pacing for post-exertional malaise, assessment for postural tachycardia, and support for cognitive symptoms.

Do not prescribe nirmatrelvir-ritonavir for long COVID
Screen for postural tachycardia and orthostatic intolerance in autonomic symptoms
Advise pacing rather than graded exercise where post-exertional malaise is present
Review drug interactions if ritonavir has been started elsewhere

Why it matters

It weakens the case that persisting virus drives long COVID symptoms enough to treat with antivirals.

The statistics, in plain English

Each confidence interval crosses zero and is wide, for example 3.2% (95% CI −10.4 to 16.8) in the cognitive group. The trial cannot exclude a small benefit in a subgroup, but it found no signal of one in 959 patients.