Observation burden: why surveillance in ophthalmology deserves the same scrutiny as treatment.

Ophthalmology manages a substantial share of its patients through prolonged observation, spanning nearly every subspecialty, from choroidal nevi and glaucoma suspects to optic disc drusen and diabetic retinopathy. Yet surveillance schedules rarely face the scrutiny of cost, benefit and harm demanded of treatment. This article reframes observation as an intervention in its own right, with direct costs, patient burden, psychological harm and downstream testing cascades that are seldom quantified when monitoring protocols are created. It introduces the concept of surveillance inertia, whereby protocols specify when to start watching but rarely when to stop, and proposes a Surveillance Value Assessment built around six questions: number needed to observe (NNO), incremental cost-effectiveness, patient burden, downstream cascade, stopping rules and actionability and time-sensitivity. NNO is offered as a surveillance analogue of number needed to treat. Drawing on models from oncology and radiology, where structured risk-stratified surveillance with explicit exit criteria is already established, the article contends that ophthalmology should connect its existing risk-prediction tools to explicit surveillance-value metrics and define its own standards before external payers impose them.