Clinical Report: Response on External Clinical Validation of the RECOVER Research Index

Overview

The authors clarify that their study evaluated the RECOVER PASC score in a specialty referral cohort with clinically diagnosed Long COVID, compared with people who recovered after SARS-CoV-2 infection. It was not designed to assess population-level performance, disease prevalence, or discrimination from other conditions with overlapping symptoms. They describe the cohort and comparator as defining the clinical question and call for evaluation in additional settings and populations.

Background

The RECOVER PASC score is a research index, and its performance may vary across populations with different symptom burdens and clinical severity. The authors state that their study assessed classification in a specialty-clinic setting after comprehensive evaluation, using expert diagnosis based on the 2024 NASEM consensus definition as the clinical reference standard. They note that Long COVID lacks a biological gold standard and characterize this consensus-based diagnosis as a pragmatic comparator. The 2024 revision of the RECOVER-Adult index incorporated additional participant data, broader symptom ascertainment, and changes to symptom weighting and threshold.

Data Highlights

The source provides no numerical performance estimates. It describes the study population and design qualitatively.

Key Findings

• The authors characterize their investigation as an external clinical validation in a well-characterized specialty referral cohort, not a population-based validation.
• The study compared people clinically diagnosed with Long COVID after comprehensive assessment with recovered controls selected to represent uncomplicated recovery after SARS-CoV-2 infection.
• The authors identify expert clinical diagnosis using the 2024 NASEM consensus definition as the most appropriate currently available clinical reference standard, while noting that it is not a biological standard.
• Enrollment required at least one neuropsychiatric manifestation because of the parent study design and the Brain Health Program’s clinical focus; enrollment did not depend on RECOVER PASC scores or score-threshold criteria.
• Brain fog was not an enrollment requirement, and the authors state that the neuropsychiatric criterion defines the clinical spectrum represented in the cohort.
• The authors report that the 2024 RECOVER-Adult index revision incorporated additional participant data, broader symptom ascertainment informed by patient-community input, and a changed symptom-weighting model and threshold; they call for future evaluation in community-based populations, primary care, and broader clinical phenotypes.

Clinical Implications

The authors state that the findings should be interpreted in relation to the specialty referral population studied and should not be used to infer population-level prevalence or performance. They identify assessment in community-based populations, primary care, and broader Long COVID clinical phenotypes as necessary future work to characterize performance across settings.

Conclusion

The response distinguishes specialty-clinic external validation from population-based validation and explains how cohort selection, enrollment criteria, and comparator choice define the study’s scope. The authors frame the RECOVER PASC score as an evolving research index requiring assessment across varied clinical populations.

Related Resources & Content

1. Azola, Veenhuis, Rubin, et al., Journal of General Internal Medicine, 2026 -- Response to Goldman and Martin on the External Clinical Validation of the RECOVER Research Index
2. RECOVER-Adult Research Consortium, JAMA, 2024 -- 2024 Update of the RECOVER-Adult Long COVID Research Index
3. BMC Psychiatry (Springer) — Evaluation of a pro-recovery training intervention (REFOCUS-RETAFORM) in specialist mental health services across France: stepped-wedge cluster randomised controlled trial protocol
4. npj Digital Medicine — Reply to: Utilizing foundation models for developing clinical tools
5. BMC Psychiatry (Springer) — Translation, cross-cultural adaptation, reliability, and validity of the Persian version of the Recovery Assessment Scale-Domains and Stages (P-RAS-DS) for people with mental illness
6. BMJ Mental Health — Validation and Adjustment of the OxSATS Suicide Risk Model Following Self-Harm in an English Context
7. 2024 Update of the RECOVER-Adult Long COVID Research Index
8. 2024 Update of the RECOVER-Adult Long COVID Research Index - PMC
9. Long COVID Clinical Guidance | Long COVID | CDC
10. RECOVER-VITAL clinical trial finds no benefit of extended duration of PAXLOVID for Long COVID | RECOVER COVID Initiative