This narrative mini-review (position paper by APMO) critically appraised iridology — the alternative practice of diagnosing systemic disease via iris inspection — examining its historical origins, biological plausibility, and diagnostic accuracy across 9 controlled studies and 3 computer-aided/machine-learning studies, through March 2026.
No controlled study demonstrated diagnostic accuracy beyond chance. The most rigorous trial (Münstedt et al., n=110) found iridology sensitivity of just 4% for cancer detection (3/68 correct). Cancer detection rates of 50–53% (Herber et al.) equaled coin tossing. The 2024 Australian NHMRC systematic review (GRADE methodology) confirmed low-certainty evidence that iridology is not an effective diagnostic tool. Computer-aided iridology studies reporting ~90% internal accuracy were methodologically circular and lacked external validation.
- The evidence base is small in volume and largely poor in methodological quality (case-control designs with healthy comparators inflate apparent accuracy). - Most studies carried high QUADAS-2 risk of bias; only a small number of prospective blinded trials exist. - This review itself is a narrative mini-review by a patient advocacy group (APMO), not an independent systematic review with meta-analysis.
Ophthalmologists should clearly tell patients that iridology has no diagnostic validity and risks delaying real medical care — while distinguishing it from legitimate ocular signs of systemic disease (e.g., Kayser-Fleischer rings, diabetic retinopathy). When patients raise iridology, address it with empathy and scientific clarity, and reinforce evidence-based screening pathways.
Explore related topics
