Clinical Scorecard: May long COVID improve without full recovery?

At a Glance

Key Highlights

• Among adults with likely long COVID at 12 months, the estimated probability of sustained improvement was 36% by 24 months and 50% by 36 months.
• Complete symptom resolution was estimated in 15% by 24 months and 27% by 36 months.
• By 36 months, adjusted improvement probabilities ranged from 60% for predominantly smell or taste disturbances to 19% for multisystem symptoms.
• Improvement probabilities were similar for female and male participants; estimates differed by age and infection period.
• The observational design cannot establish causes of improvement, and the primary outcome represented improvement below the likely long COVID threshold, not complete recovery.

Guideline-Based Recommendations

Diagnosis

• The study defined likely long COVID as an LCRI score of at least 11.

Management

• The study did not evaluate treatments received during follow-up and provides no treatment recommendations.

Monitoring & Follow-up

• Participants completed symptom assessments at 3-month intervals.
• Sustained improvement was defined as an LCRI score below 11 at two consecutive visits.

Risks

• Participants were selected for likely long COVID at specific time points, limiting generalizability to all people with a history of COVID-19.
• Symptom-based classification could not establish whether all reported symptoms were attributable to infection.
• Missing follow-up assessments required statistical imputation, and the observational design precluded causal conclusions.

Patient & Prescribing Data

The primary analysis included 1,019 adults with likely long COVID 12 months after first infection; additional analyses included participants meeting criteria at 18 or 24 months.

Treatments during follow-up were not evaluated. The authors called for placebo-controlled therapeutic intervention trials.

Clinical Best Practices

• Interpret the reported improvement estimates as cohort-level observational findings, not individual prognoses.
• Distinguish sustained improvement below the LCRI threshold from complete symptom resolution.
• Do not infer that age, infection period, or symptom subtype caused differences in improvement.

Related Resources & Content

• Annals of Internal Medicine study, as summarized in the supplied article