Do Vitreous Floaters Predict Retinal Detachment? Retrospective Cohort Study in Primary Care
In primary care, new-onset floaters—especially many or recent—signal higher retinal detachment risk than flashes alone.
*Retrospective cohort study; Level 2b (OCEBM).
*Retrospective cohort study; Level 2b (OCEBM).
Citation
van Zon B, Spoelder M, Peters HJ, Akkermans R, Crama N, van de Laar FA. Do Vitreous Floaters Predict Retinal Detachment? Retrospective Cohort Study in Primary Care. Annals of Family Medicine. 2026;24:111-116. doi:10.1370/afm.240149.
Background
Floaters are often treated as harmless, but they may be the first warning sign of a retinal tear or early retinal detachment. This study estimated the absolute and relative risk of retinal detachment after floaters or flashes in primary care.
Patients
Adults (18 years or older) with a new episode of floaters and/or flashes (2012–2021) from 7 family practices in the Netherlands (1,181 eye-specific episodes; 184,885 patient-years). Episodes were limited to one eye; key symptom details were sometimes missing.
Intervention
Presentation with floaters (including “many” floaters: 10 or more, or described as cloud/haze/curtain; and “acute” duration: 14 days or less).
Control
Flashes alone (reference group for relative risk).
Outcome
Retinal detachment (the study also counted retinal tears as retinal detachment for analysis).
Follow-up Period
Until final diagnosis recorded for the episode.
Results
Absolute risk of retinal detachment was 6.1% with floaters alone, 4.7% with flashes alone, and 8.4% with both.
| Higher-risk symptom group (vs flashes alone) | Absolute risk | Relative risk (95% confidence interval) | Number needed to harm |
|---|---|---|---|
| Many floaters | 19.8% | 4.20 (1.87–9.40) | 7 |
| Acute floaters + flashes | 11.3% | 2.39 (1.11–5.15) | 15 |
| Many floaters + flashes | 29.2% | 6.20 (2.47–15.55) | 4 |
Number needed to harm = patients with the symptom pattern needed for one additional retinal detachment compared with flashes alone.
Limitations
Retrospective record review with substantial missing symptom detail. Counting retinal tears as retinal detachment may overstate detachment risk. Main group comparisons were not clearly different after adjusting for age and sex.
Funding
No external funding reported; authors reported no conflicts.
Clinical Application
In primary care, urgently refer patients with many new floaters, or acute floaters with flashes; floaters should weigh more than flashes alone in risk assessment.
Discussion
Sign in to join the discussion.
In this retrospective cohort study, new-onset floaters were associated with retinal detachment risk of 6.1% versus 4.7% for flashes alone; given the observational design, how should we weigh validity and applicability before changing primary-care referral thresholds? In this retrospective cohort, retinal detachment occurred in 6.1% of episodes with floaters alone and 4.7% with flashes alone, with an adjusted relative risk of 1.29. Which statement best interprets this result?