A report worth reading slowly
More than 8,000 peer-reviewed papers on dry eye appeared in the six years after DEWS II. That flood of evidence is why the Tear Film and Ocular Surface Society convened DEWS III in late 2023, and why its reports, published open access in the American Journal of Ophthalmology in 2025, deserve the attention of anyone who manages ocular surface disease. I have spent much of my academic life on dry eye, from teaching it to our residents to studying oxidative stress of the ocular surface in the laboratory, so I read these reports with one question in mind: what should I do differently on Monday morning?
DEWS III at a glance
- Symptoms are now required: signs without symptoms are ocular surface disease, not dry eye disease.
- One short questionnaire screens, and diagnosis requires at least one objective homeostasis marker.
- The staged treatment ladder is abandoned in favor of mechanism-matched prescribing by etiological driver.
- There is no severity grading scheme; sign-to-symptom correlations proved too weak to support one honestly.
The definition sharpens
The new definition states: "Dry eye is a multifactorial, symptomatic disease characterized by a loss of homeostasis of the tear film and/or ocular surface, in which tear film instability and hyperosmolarity, ocular surface inflammation and damage, and neurosensory abnormalities are etiological factors."
Three edits matter. "Symptomatic" now sits inside the definition itself: signs without symptoms are ocular surface disease, not dry eye disease. Homeostasis loss extends to "the tear film and/or ocular surface," acknowledging that the surface microenvironment, not only the tears, can fail. And the drivers "are etiological factors," meaning any one of them, in any combination, may dominate in a given patient. That last point quietly prepares the ground for the new treatment philosophy.
Diagnosis: one questionnaire, three sign groups
DEWS III ends the era of choosing between questionnaires. The , a six-question short form with a simple summed score, is now the single screening instrument; a score of 4 or more is a positive screen and a prompt for further testing. Diagnosis then requires at least one objective homeostasis marker:
- Non-invasive under 10 seconds
- of 308 mOsm/L or higher in either eye, or an inter-eye difference greater than 8 mOsm/L
- Ocular surface staining above defined thresholds: more than 5 corneal spots, more than 9 conjunctival lissamine green spots, or lid-margin staining at least 2 mm long and covering a quarter of its width
Fluorescein breakup time is demoted to a fallback for clinics without non-invasive imaging, and its cut-off tightens from 10 to 5 seconds. Notably, there is no severity grading scheme: sign-to-symptom correlations proved too weak to support one honestly. I find that humility refreshing; it matches what we all see in practice.
Beyond evaporative versus aqueous deficient
The old dichotomy survives only as a teaching simplification. DEWS III subclassifies by etiological drivers grouped under tear film component deficiencies (lipid, aqueous, mucin and glycocalyx), eyelid anomalies (blink and closure problems, lid margin disease including meibomian gland dysfunction and ocular rosacea), and ocular surface abnormalities. Most dry eye is acknowledged to be (hyper)evaporative, and meibomian gland dysfunction, present in roughly half to two-thirds of cases, rightly sits at the center of the map.
The most consequential scientific correction, for me, concerns inflammation. Evaporative disease often shows muted or absent tear inflammatory markers. Inflammation cannot be assumed to be active in every patient, which finally explains a familiar frustration: why some patients flourish on or lifitegrast while others feel nothing. Mechanism-matched prescribing is no longer a slogan; it is the architecture of the report.
Ternary complex: cyclosporine (green) sandwiched between cyclophilin and calcineurin — the actual drug–target assembly (PDB 1M63).
Cyclosporine and its protein targets, rendered from Protein Data Bank structures. Drag to rotate, scroll to zoom — and compare the drug on its own with the complexes it forms inside a T cell.
Management: drivers, not stages
DEWS III explicitly abandons the staged treatment ladder, calling staging systems a suboptimal architecture for a disease with multiple, shifting drivers. In its place stands a prescribing algorithm: identify the clinically relevant drivers, then match each to an intervention with evidence for that mechanism, expecting that several treatments together will usually be appropriate. First-line care remains humble and universal: lifestyle modification, tear supplements, and environmental adjustment.
Two additions to my own counseling come straight from the epidemiology chapters. Depression and anxiety accompany dry eye in roughly 40 percent of patients, so now I ask. And screen use of even one to two hours a day is a consistent, modifiable risk factor, so I prescribe blink habits as seriously as drops.
What I take back to my clinic
DEWS III will not make dry eye simple; nothing will. But it gives us a cleaner definition, a leaner diagnostic battery, and an honest, mechanism-first way to treat. For a disease that hundreds of millions of people live with, that is real progress.