Topical Metronidazole Versus Ivermectin for Low-density Demodex Rosacea: A Rater-blinded, Randomized, Split-face Trial

Marvin Chia-Han Yeh1, Jerry Tsai2, Yu-Chen Huang1,3,4* and Hsiao-Han Wang1,4*

1Department of Dermatology, Wan Fang Hospital, Taipei Medical University, 111, Hsing-Long Road Sec 3, Wenshan District, Taipei City 116, Taiwan, 2Johns Hopkins University School of Medicine, Baltimore, MD, USA, 3Department of Dermatology, School of Medicine, College of Medicine, Taipei Medical University, Taipei and 4Research Center of Big Data and Meta-analysis, Wan Fang Hospital, Taipei Medical University, Taipei, Taiwan. *E-mails:;

Accepted Nov 1, 2022; Epub ahead of print No 1, 2022

Acta Derm Venereol 2022; 102: adv00811. DOI: 10.2340/actadv.v102.4391

Topical metronidazole and ivermectin are the standard treatments of choice for papulopustular rosacea (PPR). In addition to PPR, recent studies have also shown the efficacy of these treatments against erythematotelangiectatic rosacea (ETR) (1). Their therapeutic effects arise from eradication of overpopulated Demodex mites on the human face. However, few trials have measured the density of Demodex for each patient prior to treatment. Even though clinical improvement was sometimes observed in ETR, the rationale for using anti-parasitic agents on patients with low-or-zero density Demodex was not straightforward. The efficacies of topical metronidazole and ivermectin in patients with rosacea and low-density Demodex, especially those who present predominantly with the ETR subtype, remains unclear. Hence, the aim of this study was to evaluate the efficacy of topical metronidazole and ivermectin on ETR patients with low Demodex counts.


This study screened 33 patients who presented with ETR and had not received topical ivermectin and metronidazole in the last 3 months. The study was conducted between January 2021 and December 2021. To confirm each patient’s Demodex density 4 locations were sampled randomly (forehead, left and right cheek, chin) using 2 different techniques (squeeze method (2) and standardized skin surface biopsy (SSSB) (3)). Patients with Demodex counts higher than the proposed threshold (thumbnail: 11/cm2, SSSB: 5/cm2) were excluded. An inclusion/exclusion diagram is shown in Table SI. Topical ivermectin and metronidazole were randomly assigned for split-face treatment using computer-generated permuted blocks of 4 in random order (random allocation was performed by YCH, and patient enrollment HHW by patient assignment by MCHY). The frequency of the treatment was based on the drug label: metronidazole twice daily and ivermectin before bedtime. Efficacy and tolerance were assessed at baseline and after 1, 2 and 3 months of treatment. Evaluation included clinical erythema assessment (CEA, 5-grade scale on persistent erythema) performed by a dermatologist blinded to treatment conditions and Flushing ASsessment Tool (FAST (4), 10-point scale on redness, warmth, itchiness, tingling, and skin roughness) by the patient. Finally, at the end of the 3-month study, patients also graded their overall improvement using patient self-assessment survey (PSA).

Statistical analysis

Wilcoxon signed-rank test and χ2 test were used to evaluate statistical significance, defined as p < 0.05. Study sample size was determined by assuming 20% difference and standard deviation (SD) of 50% between the 2 treatments, using a 2-sided test with α=0.05, 27 patients was required to reach 80% power.


Twenty-seven out of 33 patients screened completed the trial (4 were excluded due to high Demodex counts, 2 dropped out because of a surge of COVID-19 infections) The baseline demographics are shown in Table I. Mean Demodex counts were 0.6 vs 0.7 mites/cm2 for SSSB and squeeze methods, respectively (p = 0.68, Table SII). Both ivermectin and metronidazole significantly improved patient erythema after 1 month of treatment, but no difference was found between metronidazole and ivermectin over the 3-month course (change in CEA of metronidazole vs ivermectin: –1.07 vs –1.04, p = 0.782 (Table II); the negative sign denotes improvement compared with baseline). On subjective FAST scores, patients reported better improvement with ivermectin on erythema and warmth after 2-months treatment (–1.37 vs –1.96, p = 0.048, and –1.37 vs –1.85, p = 0.041, respectively), and itchiness and roughness after 3-months treatment (–2.03 vs –2.70, p < 0.05 and –1.33 vs –2.14, p < 0.05). However, after 3-months treatment, patients reported no difference between metronidazole and ivermectin on overall improvement (p = 0.17) (Table I and Figs S1–S7). No major adverse effects were reported.


This study is limited by its single-blinded design. Due to the difference in texture and frequency of the medication, blinding of the participants was not possible. Although bias in clinical assessments was reduced by blinding the rater from treatment conditions, it is unclear how patients’ knowledge of medications used on each side of the face may have affected self-assessed improvement. Further double-blinded study is required to investigate the efficacy of metronidazole and ivermectin.

This single-blinded, randomized, split-face study showed that both topical metronidazole and ivermectin were effective treatment against persistent erythema, even for patients with low or zero Demodex. Ivermectin may be more effective in improving patients’ subjective symptoms, such as warmth, itchiness, and skin roughness. Considering the low Demodex count in the current cohort, possible mechanisms of treatment, including the anti-inflammatory effect (including the effect caused) of the anti-parasitic agents (5). The low Demodex patient cohort may be why no statistical difference was found between metronidazole and ivermectin, as the difference in their anti-parasitic effect was not shown. More studies are necessary to elucidate the anti-inflammatory effect of these anti-parasitic agents in patients with rosacea, especially those with low or zero Demodex counts.


The authors would like to thank the Ministry of Science and Technology (MOST) grants MOST 109-2314-B-038-148 for funding this research.

The authors have no conflicts of interest to declare.


  1. Huang HP, Hsu CK, Lee JY. Rosacea with persistent facial erythema and high Demodex density effectively treated with topical ivermectin alone or combined with oral carvedilol. Dermatol Ther 2021; 34: e14899.
  2. Huang HP, Hsu CK, Lee JY. Thumbnail-squeezing method: an effective method for assessing Demodex density in rosacea. J Eur Acad Dermatol Venereol. 2020; 34: e343–e345.
  3. Aşkin U, Seçkin D. Comparison of the two techniques for measurement of the density of Demodex folliculorum: standardized skin surface biopsy and direct microscopic examination. Br J Dermatol 2010; 162: 1124–1126.
  4. Kawata AK, Revicki DA, Thakkar R, Jiang P, Krause S, Davidson MH, et al. Flushing ASsessment Tool (FAST): psychometric properties of a new measure assessing flushing symptoms and clinical impact of niacin therapy. Clin Drug Investig 2009; 29: 215–229.
  5. Ventre E, Rozières A, Lenief V, Albert F, Rossio P, Laoubi L, et al. Topical ivermectin improves allergic skin inflammation. Allergy 2017; 72: 1212–1221.