Crisaborole on the Face: Why 98.8% Irritation-Free Assessments Were Not an Eczema Clearance Rate
AI-created illustration: a fictional adult with dermatitis-like facial dryness. This is not a photograph of a trial volunteer or a result from Dr. Arsy #003. The study discussed below enrolled healthy volunteers.
A number such as “98.8% without irritation” can sound reassuring when choosing a face cream for eczema. But what was counted: patients, treatment visits, individual skin assessments or clear faces? A crisaborole study shows why that denominator and the starting condition of the skin matter.
The study tested healthy skin, not active facial eczema
The 2016 primary phase 1 paper describes a randomized, double-blind, single-center tolerability study in 32 healthy adults. Twenty-four received crisaborole ointment 2% and eight received its vehicle. Actual ages were 18–53, half were women, and all participants were White. These features limit extrapolation to children, different populations or actively inflamed skin.
Treatment was twice daily for 21 days across 13 anatomical areas. Importantly, the face/hairline application was a fingertip-width strip below the hairline from one ear toward the other side. It was not a study of applying medicine throughout the eyelids, around the eye surface or to every form of facial rash.
What the 98.8% figure actually counted
Investigators repeatedly scored burning or stinging, redness and itch at multiple sites and visits. Across active and vehicle groups, 98.8% of all local tolerability assessments had a score of zero. Ten of 8,697 assessments, about 0.1%, had scores above one. Thousands of repeated measurements from 32 people remain measurements from 32 people.
| Study finding | Appropriate interpretation | Incorrect interpretation |
|---|---|---|
| 98.8% of assessments scored zero | Most repeated local checks found no irritation signs or symptoms | 98.8% of eczema patients were cured |
| Healthy adults treated for 21 days | Short-term tolerability evidence in that population | Universal facial tolerance at every age |
| Face testing included a hairline strip | Evidence for the tested application area | Permission to put ointment in the eyes |
The group-specific proportions of zero-score assessments were 99.0% with crisaborole and 98.2% with vehicle. All 32 participants completed the study. One active-treatment participant reported three mild treatment-related application-site pain events, involving behind-the-ear areas and hairline stinging; each lasted less than a day. Those details prevent “well tolerated” from becoming “never stings.” Anacor Pharmaceuticals sponsored the study.
Why the eczema label can still list application-site pain
Healthy intact skin and eczema-affected skin answer different questions. The current Eucrisa prescribing information reports application-site pain, including burning or stinging, in 45 of 1,012 participants receiving twice-daily crisaborole in two four-week eczema trials, versus six of 499 receiving vehicle. The label rounds these to 4% and 1%. These are people with adverse reactions, rather than the healthy-skin study’s repeated individual assessments.
These findings are not contradictory. The populations, duration and outcome definitions differ. Our crisaborole eczema-trial guide discusses disease efficacy separately. The healthy-volunteer experiment cannot supply a facial-eczema clearance percentage.
Face-care decisions still require specific instructions
The current U.S. label indicates crisaborole for mild-to-moderate atopic dermatitis in adults and children aged three months or older, but this 32-person study did not establish infant facial tolerability. The product is for skin use, not ophthalmic, oral or intravaginal use. Hypersensitivity, including contact urticaria, requires prompt action according to the label and clinical advice.
Ask which exact facial areas should receive the medicine, what local sensations can occur and when symptoms warrant review. Do not persist through severe swelling or worsening just because a healthy-skin study sounded reassuring. Our stinging-product discussion helps frame that conversation.
For comparison of study design rather than a ranking of treatments, see the regional ruxolitinib face-and-neck trial and the tacrolimus–fluticasone facial trial. Each studied a different question.
Dr. Arsy Moisturizer #003, 4 oz provides cosmetic hydration, moisture-barrier support and ordinary dry-skin comfort. It was not the ointment in this study, does not inherit crisaborole’s results and is not an eczema medicine. Discuss moisturizer suitability separately from prescription treatment.
Adult educational content; primary full paper and current prescribing information checked November 9, 2026.