The instruction is handed out at the end of appointments all over the world. Warm flannel, five minutes, twice a day. It is free, it is sensible, and almost nobody keeps it up past a fortnight.
Part of that is boredom. A larger part is that done the way it is usually described, it does not work very well, and people can tell.
Where the oil is, and why that is awkward
Around thirty meibomian glands run vertically through each eyelid, opening in a line just behind the lashes. They secrete the oily layer that caps the tear film and slows evaporation. When the oil thickens and the openings block, the film breaks up early and the eye stings, waters at inconvenient moments, and feels gritty by late afternoon.
Warming is meant to soften what is sitting in those glands so that a blink, or a deliberate press, can move it out. The obstacle is geography. The glands are in the tarsal plate, on the inner side of the lid, and anything warm you hold against your face is on the outer side.
The oil itself also changes. Hetman and Borchman reported in 2020 that the ratio of cholesteryl esters to wax esters in meibum was 36% lower in donors with dry eye from meibomian gland dysfunction than in donors without, and that shifting that ratio substantially alters how the lipid mixture behaves as it warms and cools. The material being melted is not the same material in everyone.
Somebody put a probe on the inside of the lid
In 2008 Blackie and colleagues published the study this article exists because of. They applied warm compresses under three different protocols and measured the temperature of both the outer and the inner eyelid surface throughout, and then watched it fall for ten minutes afterwards.
| Method | Outer lid | Inner lid |
|---|---|---|
| 15 min, no reheating (n=10) | 41.2°C at 1 min | 38.8°C at 4 min, then falling |
| 30 min, reheated every 2 min (n=10) | 43.3°C at 6 min | 40.4°C, took the full 30 min |
| 30 min, reheated, contact optimised (n=12) | 42.2°C at 4 min | 40.8°C, took 20 min |
Blackie CA, Solomon JD, Greiner JV, Holmes M, Korb DR. Inner eyelid surface temperature as a function of warm compress methodology. Optom Vis Sci. 2008;85(8):675-83. Compresses were heated to 45 ± 0.5°C in all three groups.
Read the first row again. A compress that is not reheated puts the outer lid past 41°C inside sixty seconds, which feels like something is working, while the inner surface never gets past 38.8°C and starts cooling after four minutes.
The outer lid is not where the glands are. The sensation of warmth and the delivery of heat come apart, and the sensation is the one people use to judge whether they have done enough.
The authors' own conclusion was that a precise, customised and labour-intensive procedure is necessary to treat gland obstruction using warm compresses at all.
What the numbers turn into, in a bathroom
- Start hotter than feels necessary. The compresses in the study were heated to about 45°C. Not hot enough to hurt, and considerably hotter than a flannel under a running tap ends up.
- Reheat every two minutes, or use something that holds heat. This is the step that separates the three rows in the table. A wet cloth is a heat reservoir with a small capacity and a large surface area, which is a description of something that cools fast.
- Four minutes is the floor. That is the minimum the authors give to reach 40°C on the inner lid, and they add that longer is needed for more severe obstruction. Five minutes without reheating does not reach it at all.
- Press the lid into the shape of your eye. The third group did better in less time purely by optimising contact. Air gaps are insulation.
Heat first. Cleaning second. Always that way round.
Lid hygiene and warmth are usually recommended together and almost never in a stated order, which is how most people end up doing them in whichever order is convenient.
The sequence follows from the mechanism. Heat softens what is sitting in the gland. Cleaning the lid margin afterwards clears the openings so that softened material has somewhere to go. Reverse it and you have cleaned the outside of a pipe that is still blocked, then applied heat to a margin you are no longer going to touch.
What to clean with is a separate question with its own literature, and the products differ from each other far more than the category name suggests. That is covered in what eyelid wipes and cleansers actually do.
Three to four weeks, and what happens in them
Nothing happens on day two. Gland function is not a switch, and the improvement people report is the accumulation of many small clearances rather than one.
Three to four weeks of near-daily use is the interval worth committing to before forming a view. Most people who conclude that warm compresses do not work for them have run the experiment for nine days with a flannel that was never reheated.
It is also worth being honest about the strength of the evidence underneath all of this. The TFOS DEWS II Management and Therapy Report in 2017 reviewed dry eye treatments comprehensively and concluded that many of the available options lack the Level 1 evidence needed to support recommending them. Warm compresses are widely advised and mechanically well described. The measurements above are about how to do the thing properly, not proof of how much it delivers.