Demodex blepharitis (DB) is a chronic lid margin disease caused by overpopulation of Demodex mites. Its pathognomonic sign, collarettes at the base of the lashes, was present in 57.7% of 1,032 patients who were seen in US eyecare clinics for any reason.1
“I look at Demodex as something that lives with us, and I think it actually lives with all of us.”
Ocular surface disease is not the most glamorous thing an ophthalmologist treats, but a patient who leaves my office no better than when they arrived has not been served. An untreated lid margin is one of the most common reasons that happens, especially in surgical practices where the focus is on the lens or the cornea.
Why Demodex blepharitis is underrecognized
When I first started treating DB, I looked for the dramatic cases: lash margins so dense with collarettes that I hesitated to touch the lids. That threshold misses most of the patients who need treatment. I now weigh symptoms and treatment history as heavily as lid findings.
The data support that shift. In the Titan study, 44.0% of patients with collarettes had never been diagnosed with blepharitis. DB prevalence was also higher among patients already using tea tree oil than among those who were not (74.5% vs. 56.7%).1
In a prospective study of 199 patients at two tertiary clinics, we epilated one lash from each lid of every patient. Demodex was newly identified in 55.3%, and only 1% of those patients had previously been diagnosed and treated.2
Those findings match what I see in-clinic. Patients arrive after cycling through artificial tears or tea tree oil products, which reduce the problem without resolving it. Some are young women with excellent lid hygiene who are still symptomatic because no one has considered Demodex.
Patient survey data describe the same pattern: delayed diagnosis, multiple provider visits, and persistent symptoms despite warm compresses, artificial tears, and tea tree oil–based lid hygiene.3 These patients end up doctor shopping to find the right answer, and that is a failure on our end.
Demodex Blepharitis: Treatment Outcomes Calculator
Use this calculator to compare the cost and efficacy of treatments for Demodex blepharitis to help counsel patients.
$65
One bottle covers the full 6-week course, so cost is the same at both timepoints. Manufacturer reports most insured patients pay $30 or less with coverage and available assistance.
$125
Based on OTC retail prices for tea tree oil products, in whole packs, assuming once-daily use.
Bars show additional patients achieving mite eradication beyond each treatment's own control arm. Shaded bands show the 95% CI.
Cost per additional patient with mite eradication
Patient cost divided by the added eradication rate beyond the control arm.
At $65 per course, lotilaner costs $134 per additional patient eradicated vs vehicle. For tea tree oil, this figure is not calculated because its effect vs sham was not statistically significant.
NNT — lotilaner vs vehicle
3
Treat 3 for 1 additional eradication
NNT — TTO vs sham
n.s.
Not statistically significant vs sham
Total spend — lotilaner
$13,000
200 patients
Total spend — TTO
$25,000
200 patients, 6–8 weeks
What is NNT? Number needed to treat: how many patients must be treated for one additional patient to achieve mite eradication compared with the control arm. Lower is better. NNT is not shown for tea tree oil because its effect vs sham was not statistically significant.
At 6–8 weeks, treating 200 patients with lotilaner 0.25% would result in an estimated 97 additional patients achieving mite eradication compared with vehicle. The estimate for tea tree oil vs sham (+3) was not statistically significant. These are separate comparisons against different controls, not a head-to-head result.
Methodology: Bars show the estimated additional patients achieving mite eradication (0 mites/lash) beyond each treatment's own control arm, derived by applying pooled odds ratios from a sponsor-funded systematic review and meta-analysis of 9 RCTs (N=1,513) to pooled control-arm rates: vehicle 12.9% (2–4 weeks) and 15.8% (6–8 weeks) for lotilaner 0.25%; sham 28.0% (2–4 weeks) and 95.7% (6–8 weeks) for tea tree oil/terpinen-4-ol. Control rates are crude totals of control-arm events. Bands show the 95% confidence interval of each pooled odds ratio and do not account for uncertainty in control rates. Tea tree oil did not differ significantly from sham for mite eradication at either timepoint; the 6–8 week estimate is based on 2 trials in patients with low baseline mite density. Treatments were compared with different controls, and no direct or indirect comparison between lotilaner 0.25% and tea tree oil was performed; results should not be interpreted as head-to-head. Source: Gaddie IB, et al. Front Ophthalmol. 2026;6:1978403. doi:10.3389/fopht.2026.1978403
Demodex is a question of balance, not presence
I believe most of us, if not all of us, carry Demodex. DB is not an on/off diagnosis. Like cataract, it exists on a continuum, and the clinical question is whether the mite population is under control and whether the patient is symptomatic.
“It becomes less of a question of how much or how many versus are your patients bothered, and what are the mites doing that's actually affecting the ecosystem of the eye.”
The analogy I use is Clostridioides difficile. Patients say they "caught C. diff" in the hospital, but C. diff is part of normal gut flora. Antibiotics disrupted the surrounding ecosystem and allowed it to overgrow. The lid margin works the same way. It hosts native bacteria and mites, and disease begins when that balance tips.
Considerations for treating Demodex blepharitis
The first is that mites do not travel alone. Demodex acts as a vector for bacteria, and my colleagues and I set out to characterize that burden in the Pandora study, a prospective, multicenter study that cultured swabs from the lid margin and conjunctival fornices.
Preliminary results compared 24 patients with at least 10 collarettes against 11 patients with no signs or symptoms of blepharitis. Methicillin-resistant Staphylococcus epidermidis was isolated from 29% of DB patients vs. 18% of controls, and 63% of DB patients carried more than one bacterial type, compared with 45% of controls.4
The less pathogenic, methicillin-sensitive coagulase-negative S. epidermidis showed the reverse pattern: it was more common in controls than in DB patients (45% vs. 21%).4 These are early numbers from a small cohort, and enrollment was planned to reach 60 DB patients and at least 40 controls. Still, the direction is consistent: the Demodex-affected lid margin hosts a more pathogenic bacterial population.4
The second is that disrupting the local environment may shift mite populations. In a prospective study of 62 cataract patients, Demodex colonization rose from 22.6% before surgery to 32.3% 3 weeks after surgery and post-operative steroid therapy.5 A 2025 prospective study of 100 cataract patients found no significant change. That study excluded patients with ocular surface or dry eye disease, so the relationship is not yet settled.6
Host response also varies. Patients who react strongly to ordinary triggers can be highly symptomatic with a modest mite load, and dermatographism is a useful clue here. Other patients with heavily coated lashes report few complaints. For that reason, I base treatment decisions on what the mites are doing to that patient's ocular surface, not on counts alone.
Treating the underlying pathology
For years, dry eye therapy meant managing symptoms with anti-inflammatories. DB is an underlying pathology that drives inflammation, blepharitis, and dry eye, and we now have a targeted treatment for it.
The Demodex Expert Panel on Treatment and Eyelid Health (DEPTH) recommends lotilaner ophthalmic solution 0.25% (XDEMVY, Tarsus Pharmaceuticals) as first-line therapy.7 Lotilaner selectively inhibits the mite's GABA-gated chloride channels, which paralyzes and kills the mite.3
The pivotal trial results are unusually strong for ophthalmology, where we are often satisfied with single-digit improvements.
After 43 days of twice-daily dosing:
- In SATURN-1 (N = 421), complete collarette cure was 44.0% vs. 7.4%, and mite eradication was 67.9% vs. 17.6%.8
- In SATURN-2 (N = 412), complete collarette cure was 56.0% vs. 12.5%, and mite eradication was 51.8% vs. 14.6%.9
- Lid margin erythema cure was 19.1% vs. 6.9% in SATURN-1 and 31.1% vs. 9.0% in SATURN-2.8,9
In the pooled analysis of 833 patients, 85.1% of treated patients reached 10 or fewer collarettes, compared with 28.0% on vehicle.10 This matters when colleagues ask about the patients who did not reach mite eradication. Eradication was defined as zero mites per lash, a strict endpoint. The clinically meaningful reduction in disease burden was far more widespread.
Demodex feed on the lipids our lids produce. My view is that removing those competing consumers should improve both the tear film and the inflammatory load. Data in patients with concomitant meibomian gland disease support that view. In the pooled ERSA and RHEA studies, collarette grade 0 was achieved in 44.7% vs. 0% at day 43 and 65.8% vs. 0% at day 85. Meibum quality, gland function, and patient-reported outcomes also improved at 6 and 12 weeks.11
When a patient has both DB and MGD, I treat the Demodex first. I call it resetting the ocular surface. If the patient improves but is not fully resolved after the course, I add dry eye therapy at that point. Demodex and dry eye frequently coexist, and neither should be pigeonholed as the only explanation for symptoms.
Why the full 6 weeks matters
Most dry eye drugs are twice daily indefinitely. The first question patients ask about any prescription is how long they have to take it, and a 6-week course with a defined end date is a real advantage.
“It's really nice to be able to put somebody on a medication that's twice a day for six weeks, instead of putting them on a lifelong drug.”
The course only works if it is completed. The 6-week regimen is built around the mite's life cycle, including eggs that hatch after the adult population has been treated.12 When a patient does not resolve, the cause is almost always adherence: missed days, or a week off for vacation before restarting. I stress that an incomplete course delivers little benefit. In patients who missed a few days, I sometimes extend therapy by a week or two, which is off-label.
My counseling focuses on the most common reason patients quit, which is feeling no difference at 4 weeks. I tell them in the exam lane that most people do not notice improvement until weeks 5 or 6. When I ask patients at follow-up when they started feeling better, most say the same thing.
When the drug was new, I saw patients at 2, 4, and 6 weeks to watch for anything the trials might have missed. I am now comfortable with a single follow-up at 3 months.
I also have mascara wearers discard their current mascara, buy a new one, and use disposable wands so they do not reintroduce eggs from the tube. I don't ask them to replace their other cosmetics.
The small group that relapses
A very small subset of my patients respond to treatment and then recur. They tend to be fair-complexioned, light-haired, and light-eyed, and many have dermatographism. My working theory is that they lack whatever mechanism normally keeps the mite population in check.
The extension data are consistent with gradual reaccumulation in some patients, alongside a durable effect in most. After a single 6-week course in SATURN-1, 39.8% of treated patients had collarette grade 0 at day 180 and 23.5% at day 365, compared with 2.7% and 2.9% of controls. Erythema cure in the treatment arm continued to rise, reaching 28.7% at 1 year.12
For patients who relapse, I re-treat off-label. I then add maintenance measures such as tea tree oil to keep the population low, so repeat prescriptions are rarely needed.
Judging success in practice
I am one of the few clinicians who still epilate lashes to look at mites under a microscope. I don't expect most eyecare providers to do that, and the pivotal trials already established the drug's effect on mite counts under controlled conditions.8,9 I don't need to reprove those results in my office.
In clinical practice, I judge success by symptom resolution, confirmed by a cleaner lid margin at the slit lamp. The difference now is that we reach symptom resolution by treating the underlying pathology rather than suppressing the symptoms indefinitely.
Addressing the ocular surface before surgery
Patients come in expecting routine cataract or refractive surgery. Unhappy post-operative patients tell their friends, so optimizing the ocular surface protects the surgeon as well as the patient.
In the Titan data, 55.7% of patients with cataract had collarettes.1 Ocular surface disease with fluctuating vision can make pre-operative measurements unreliable and reduce satisfaction after cataract surgery, as the ERSA/RHEA investigators note.11 The Pandora investigators also raised concern that the pathogenic bacterial burden seen in DB could compromise outcomes in patients undergoing cataract or refractive surgery or wearing contact lenses.4
I compare it to taking a car to a mechanic for a clicking sound when the car also has a flat tire. A good mechanic tells you about both problems. Our job is to identify everything that could affect the outcome, explain it, and let the patient make an informed decision about sequencing.
Prior authorization requirements have made lotilaner harder to prescribe, but that is not a reason to skip it. When coverage is denied, I encourage patients to call their insurer directly and explain that the denial is preventing their doctor from treating them appropriately. Patients have more leverage in that conversation than we often assume.
Building the “Do I Have Demodex?” calculator
During my final year of residency, I wanted to know how common Demodex was in general ophthalmology clinics. At that point, prevalence had not been well characterized in that setting. We enrolled every adult patient who came through the clinic, including cataract, glaucoma, retina, and pterygium patients. We epilated one lash from each of the four lids and examined each under light microscopy.
Demodex presence was significantly associated with:2
- Age (P = 0.005)
- History of pterygium surgery (P = 0.01)
- Dermatochalasis (P = 0.001)
- Blepharitis (P < 0.001)
- Dry eye disease (P = 0.001)
Presence was independent of sex, number of clinic visits, self-administered lid hygiene, and prescription ophthalmic drops. Patients using alpha-agonist drops had a lower risk of colonization.2 Our 55.3% prevalence, determined by epilation, closely matches the 57.7% Titan figure, which was determined by collarettes.1,2 Two different methods produced the same answer.
We used those variables to build a multivariate model that estimates the likelihood of finding Demodex on a patient's lashes, available at doihavedemodex.com. Sampling one lash per lid almost certainly undercounts mites. In hindsight, "Would I find Demodex" would be a more accurate name for the calculator. I suspect nearly everyone has them, and the real question is whether we detect them and whether the patient is symptomatic.
Key takeaways
- Nearly 6 in 10 eyecare patients have collarettes, and many with DB have never been diagnosed, including patients already using tea tree oil products.1,2
- DB exists on a continuum. Treatment decisions should reflect the patient's symptoms and ocular surface impact, not mite counts alone.
- Mites are associated with a heavier and more pathogenic bacterial burden. Surgery and post-operative steroids may shift mite populations, although findings are mixed.4-6
- In pivotal trials, lotilaner 0.25% twice daily for 6 weeks reduced collarettes to 10 or fewer in 85.1% of patients, with benefits extending to patients with concomitant meibomian gland disease.10,11
- Incomplete courses are the most common cause of treatment failure. Patients should expect symptom relief around weeks 5 to 6 and should not stop at 4 weeks.
- A small subset of patients relapse. Extension data show a durable but gradually waning effect over 1 year after a single course.12
- DB is present in more than half of patients with cataract and should be treated before biometry and surgical planning.1,11
Closing thoughts
“It's not the sexiest thing to treat as an ophthalmologist, but it's our job as clinicians to make sure that the person leaving your office is better than when they came in.”
Patients take hours out of their lives to see us, and they deserve our full attention at the level of the ocular surface, not a reflexive recommendation for artificial tears. Treating DB to completion is one of the simplest ways to stack the odds in our favor before surgery. It is also one of the few ocular surface treatments with a defined endpoint, which patients appreciate.
This article was written by Mariel Mohns, MS based on clinical insights from Eric Rosenberg, DO.
