Ask ten clinicians how they diagnose dry eye and you may get ten different routines. The surveys Tear Film and Ocular Surface Society (TFOS) ran around DEWS II and again ahead of DEWS III confirmed it: clinicians were using, on average, nine separate tests to reach a diagnosis. That is not thoroughness, it is drift, and it undermines the one thing a diagnosis needs to be, namely consistent. When a patient is assessed for diabetes, the threshold on a blood test settles the question regardless of who orders it. Dry eye deserves the same discipline, and TFOS DEWS III (Jones et al., 2025) now provides it with a short, repeatable pathway built on the DEWS II foundation.
Step one: screen with the OSDI-6
The pathway opens with a symptom screen, because dry eye is, by its refined definition, a symptomatic disease. DEWS II offered a choice between the 12-item OSDI and the DEQ-5, but subsequent research showed the two instruments did not reliably agree, so the same patient could screen positive on one and negative on the other. That’s a problem when the whole point for a diagnostic gate is consistency.
DEWS III adopts the OSDI-6, a shortened, simplified form. The troublesome items have been resolved, including the blurred-vision question that left patients wondering whether to answer for spectacles-on or spectacles-off, and there are no longer any non-applicable responses. The practical outcome is that you simply sum the scores, and a total of four or more is a positive screen. It takes two to three minutes, the patient can complete it in the waiting area rather than the consulting room, and there is evidence that short, simple questionnaires perform best in children, so the same tool stretches across a wide age range. You remain free to use longer questionnaires later to inform management. For the diagnosis itself, the OSDI-6 keeps the front door simple.
Step two: confirm with one sign
A positive screen tells you it is worth proceeding. To confirm, DEWS III asks for a single positive sign rather than a battery. You need either a non-invasive tear breakup time under ten seconds or elevated tear osmolarity. If the non-invasive breakup time is short and the screen was positive, the diagnosis is made. It really is that direct.
The reason only one of these two is required is instructive. TFOS took a large population already diagnosed under DEWS II criteria and asked what would have happened with fewer tests. Non-invasive breakup time and osmolarity proved largely interchangeable as the confirming sign, so requiring both adds cost and chair time without adding diagnostic yield. Osmolarity remains in the framework because it is part of the pathophysiology of dry eye, but the report is candid that we lack a convenient, inexpensive way to measure it well, with sampling limited to the tear meniscus on the one device still widely available. Non-invasive breakup time, by contrast, is quick, repeatable and, increasingly, well served by modern instruments.
If you have no non-invasive option, fluorescein breakup time can substitute with a different cut-off, but it comes with a caveat worth remembering. The tear film is less than a tenth of the thickness of a human hair, so introducing fluorescein changes both the volume and the chemistry of the very thing you are trying to measure, and even performed well it is less diagnostic than the non-invasive method.
Step three: do not skip ocular surface staining
The single most important message in the diagnostic section is about staining. The same population analysis showed that ocular surface staining is the test you cannot leave out: rely on a single sign without it and a meaningful proportion of patients go undiagnosed, while staining combined with one other sign approaches the full diagnosed population. Limit yourself to corneal staining alone, and your diagnostic yield falls by roughly half.
Staining therefore has to be done properly and across all the relevant surfaces. That means fluorescein for the cornea, lissamine green for the conjunctiva, and attention to the lid margin. Technique matters: to get a useful concentration of lissamine green onto the surface, hold the strip in place for at least five seconds before instilling the drop, whereas with fluorescein you want the minimum volume that gives an even film. Done this way, staining is both a diagnostic anchor and a window onto where the disease is acting.
A word on severity, and on what comes next
It is tempting to fold these results into a single severity score, but DEWS III is clear that the science does not support it. You can grade symptoms, you can grade inflammation or redness, you can grade staining, but there is no validated way to add a level-four staining picture, a level-two symptom score and a level-one inflammation finding into one number. Report severity feature by feature, and resist the urge to manufacture a composite.
What the pathway does deliver, almost for free, is the raw material for management. Run the screen, then move from least to most invasive: tear meniscus height and blink rate, non-invasive breakup time, redness and inflammatory markers, interferometry for the lipid layer, lid expression, staining, and meibography where available. The same sequence that confirms the diagnosis also populates the driver-based subclassification that steers treatment, the subject of the next article in this series.
A diagnostic platform such as me-check® can carry much of this load in one place, capturing OSDI and meibography and feeding a consistent record that makes the diagnose-then-classify workflow practical in a busy clinic. The instruments matter less than the discipline behind them: screen, confirm with one sign, and always stain. Used consistently, the DEWS III pathway turns a scattered nine-test habit into a three-step routine you can defend, repeat, and act on.
Educational content for eye care professionals. Product indications and availability vary by region. Refer to the device Instructions for Use and follow local regulations. This material is for professional education and does not replace clinical judgement.
