Melles · Annals of internal medicine 2023 · cohort study · n=3325

Hydroxychloroquine Dose and Risk for Incident Retinopathy : A Cohort Study.

Cited 57 times in the scientific literature.

Level 3 - non-randomized controlled study

Non-randomized longitudinal cohort study

PubMed 36645889 · doi:10.7326/M22-2453 · record verified 2026-08-30

What was done

Cohort study evaluating adults aged 18 years or older within a U.S. integrated health network who received hydroxychloroquine for 5 or more years between 2004 and 2020 and underwent serial retinopathy screening. Hydroxychloroquine weight-based dose was determined from pharmacy dispensing records during the first 5 years of treatment (categorized as ≤5, 5 to 6, or >6 mg/kg per day). Incident hydroxychloroquine retinopathy was assessed through central adjudication of spectral-domain optical coherence tomography, and 15-year risk was estimated using the Kaplan-Meier method.

What was found

Among 3,325 patients, 81 developed hydroxychloroquine retinopathy (56 mild, 17 moderate, and 8 severe). The overall cumulative incidences were 2.5% at 10 years and 8.6% at 15 years. At 15 years, cumulative retinopathy incidence was 2.7% for ≤5 mg/kg per day, 11.4% for 5 to 6 mg/kg per day, and 21.6% for >6 mg/kg per day. Corresponding 15-year risks for moderate-to-severe retinopathy were 1.1%, 2.4%, and 5.9%, respectively.

Why it matters

This study provides long-term risk estimates demonstrating a clear dose-dependent increase in incident retinopathy, supporting clinical recommendations to keep long-term hydroxychloroquine dosing at or below 5 mg/kg per day.

Limits

Exposure assessment relied on pharmacy dispensing records rather than verified pill ingestion, introducing potential dose misclassification from nonadherence. The study is observational, restricted to patients who remained on therapy for at least 5 years, and conducted within a single integrated healthcare network.

Cited by