Causal Inference
Effect of Nirmatrelvir-ritonavir During Acute COVID-19 on Long COVID Prevention: A Trial Emulation Ning Zhang* Ning Zhang Zhang Zhang Zhang Zhang Zhang Zhang Zhang Zhang Department of Epidemiology, UNC Chapel Hill
Background Evidence on Nirmatrelvir-ritonavir (NMV/r) use during acute COVID-19 on long COVID prevention is mixed, and NMV/r initiation timing and completion have not been accounted for.
Methods We evaluated the effect of NMV/r treatment during acute COVID-19 on long COVID onset by emulating six nested trials aligned with treatment initiation 0 to 5 days after symptom onset using data from the VISION study, a longitudinal prospective cohort of adults with acute COVID-19 in North Carolina. We compared two protocols: (1) initiation of NMV/r within 5 days of symptom onset with completion as prescribed, and (2) no initiation of NMV/r within 5 days of symptom onset. We estimated the per-protocol effect of NMV/r on long COVID from 8 to 72 weeks post treatment initiation using g-computation, adjusting for age, sex, race, ethnicity, vaccination recency, prior infection, comorbidity, and urbanicity. We also assessed whether earlier NMV/r initiation (within 3 days of symptom onset) was associated with greater long COVID risk reduction.
Results Among 2,274 participants, 1,061 initiated NMV/r within 5 days of symptom onset (median age 58 years (IQR: 43, 67), 69% female, 46% vaccinated ≤6 months), and 1,213 did not (median age 43 years (IQR: 34, 56), 75% female, 31% vaccinated ≤6 months). The estimated 72-week risks of long COVID were 35.7% (95% CI: 32.4%, 39.1%) of initiators treated within 5 days with completion as prescribed, and 39.8% (95% CI: 36.1%, 43.3%) who did not, corresponding to a risk difference of -4.0 percentage points (95% CI: -8.8, 1.3). Earlier initiation was associated with a 72-week risk difference of -4.6 percentage points (95% CI: -9.8, 0.5).
Conclusions Long COVID risk was substantial in this population. Timely initiation and completion of NMV/r was associated with a modest but imprecise reduction in long COVID risk. Our findings align with prior evidence of limited benefit of NMV/r use for acute COVID-19 in long COVID prevention.
