SEARCH

SEARCH BY CITATION

Summary

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. What’s already known about this topic?
  7. References

Rosacea is a common chronic skin disease affecting the face. There are numerous treatment options, but it is unclear which are the most effective. The aim of this review was to assess the evidence for the efficacy and safety of treatments for rosacea. Searches included the Cochrane Skin Group Specialised Register, the Cochrane Central Register of Controlled Trials in The Cochrane Library, MEDLINE, EMBASE, Science Citation Index, and Ongoing Trials Registers (updated February 2011). Randomized controlled trials in people with moderate to severe rosacea were included. Fifty-eight trials, including 27 from the original review, comprising 6633 participants were included in this updated review. Interventions included topical metronidazole, oral antibiotics, topical azelaic cream or gel, topical benzoyl peroxide and/or combined with topical antibiotics, sulphacetamide/sulphur, and others. There was some evidence that topical metronidazole and azelaic acid were more effective than placebo. Two trials indicated that doxycycline 40 mg was more effective than placebo. There was no statistically significant difference in effectiveness between doxycycline 40 mg and 100 mg but there were fewer adverse effects. One study reported that ciclosporin ophthalmic emulsion was significantly more effective than artificial tears for treating ocular rosacea. Although the majority of included studies were assessed as being at high or unclear risk of bias, there was some evidence to support the effectiveness of topical metronidazole, azelaic acid and doxycycline (40 mg) in the treatment of moderate to severe rosacea, and ciclosporin 0·05% ophthalmic emulsion for ocular rosacea. Further well-designed, adequately powered randomized controlled trials are required.

Rosacea is a chronic skin disease that can affect the cheeks, nose, eyes, chin and forehead. It is characterized by recurrent episodes of flushing, erythema, papules, pustules and telangiectasia.1–3 Although there is no standard clinical definition of the condition, rosacea is generally classified into four subtypes and one variant.4,5

Subtype 1 Erythematotelangiectatic rosacea. The clinical features include flushing and persistent central facial erythema with or without telangiectasia.

Subtype 2 Papulopustular rosacea. This is characterized by persistent central facial erythema with transient, central face papules or pustules, or both.

Subtype 3 Phymatous rosacea. This may occur on the nose (rhinophyma), chin, forehead, cheeks or ears.

Subtype 4 Ocular rosacea. It may be found in up to 58% of cases, but it is frequently undiagnosed.

Variant Granulomatous rosacea. This is noninflammatory and characterized by hard, brown, yellow or red cutaneous papules, or nodules of uniform size.

Progression from one subtype to another is possible; however, each individual characteristic may change from absent to severe.1,3,5

As with most chronic skin diseases, rosacea requires long-term treatment, therapies are numerous, and their use is frequently based on anecdotal evidence.2,3,6 Management strategies for people with rosacea can often be tailored to the specific subtype of rosacea, but because rosacea can have a significant impact on quality of life, these strategies should also be directed towards achieving improvement in general well-being.3,7–10

This is a summary of a Cochrane systematic review that was conducted to examine the different management options and determine the most effective strategy in the treatment of rosacea. Systematic reviews, which represent the uppermost level of reliability in the hierarchy of evidence, are a valuable source of information for healthcare providers, consumers, researchers and policymakers. They provide one of the best opportunities for clinicians to understand and translate the current best evidence of the effects of healthcare interventions into their daily clinical practice. Cochrane reviews are internationally recognized for their high quality and are updated regularly.

Methods

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. What’s already known about this topic?
  7. References

A systematic review of randomized controlled trials (RCTs) was conducted following a prespecified protocol.11

Search strategies

We searched 11 electronic databases and trial registers (Table 1) from inception to February 2011 for relevant studies. We scanned the bibliographies of included studies, published reviews and articles that cited the included studies. Attempts were made to locate unpublished and ongoing trials through correspondence with authors and pharmaceutical companies. No language restrictions were imposed and several studies were translated. Two reviewers (E.J.v.Z. and M.A.G.) independently reviewed all studies from searches for eligible RCTs.

Table 1.   Electronic databases and trial registers searched
Electronic databases
 The Cochrane Skin Group Specialised Register
 The Cochrane Central Register of Controlled Trials
 MEDLINE
 EMBASE
 Science Citation Index
 BIOSIS
Trial registers
 The metaRegister of Controlled Trials, http://www.controlled-trials.com
 The U.S. National Institutes of Health Ongoing Trials Register, http://www.clinicaltrials.gov
 The Australian and New Zealand Clinical Trials Registry, http://www.anzctr.org.au
 The World Health Organization International Clinical Trials Registry platform, http://www.who.int/trialsearch
 The Ongoing Skin Trials Register, http://www.nottingham.ac.uk/ongoingskintrials

Inclusion criteria

We included RCTs that compared any type of intervention used to treat rosacea, either as stand-alone or as a combined intervention, vs. placebo or active treatment. We also considered the effects of avoidance of some foods, e.g. spicy food, as well as the use of certain cosmetics and sunscreens.

Outcome measures

Our two primary outcomes were (i) impact on quality of life and (ii) participant-assessed changes in rosacea severity. Secondary outcomes were physician-assessed changes in rosacea severity, drop-out rates and adverse events.

Data extraction and synthesis

Details of eligible trials were extracted and summarized using a structured data extraction form (E.J.v.Z., S.F.K., Z.F.). Disagreements were resolved by discussion. The review authors (E.J.v.Z., S.F.K., Z.F. or B.R.C.) independently assessed risk of bias in the included studies using the Cochrane Collaboration’s domain-based evaluation tool as described in Chapter 8, Section 8.5, in the Cochrane Handbook for Systematic Reviews of Interventions.12 We presented continuous outcomes where possible on the original scale as reported in each individual study. Dichotomous outcomes were presented as relative risk ratios. All outcomes were reported with their associated 95% confidence intervals. A meta-analysis was carried out if we identified a sufficient number of studies ( 3) investigating similar treatments that reported data that were amenable to pooling.13 In view of a lack of demonstrable statistical heterogeneity between the studies (I> 60%), a fixed-effect model was used to combine the results of individual studies in this review.

Results

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. What’s already known about this topic?
  7. References

Description of the included studies

The results of the literature search are shown in Figure 1. Of the 5434 references retrieved from the searches, 58 studies met our inclusion criteria (see Table 2 for characteristics of included studies).11 The trials were grouped into six categories of interventions: topical metronidazole (= 28), oral antibiotics (= 15), topical azelaic cream or gel (= 11), topical benzoyl peroxide and/or combined with topical antibiotics (= 4), sulphacetamide/sulphur (= 3), and other therapies (= 14). Eighteen trials evaluated comparisons in more than one category, e.g. a topical plus an oral agent. The duration of treatment ranged between 2 and 3 months, and only two studies addressed interventions for ocular rosacea.14,15 Diversity in study design, skewed data, missing standard deviations, and a mix of different comparators and dosing regimens did not permit pooling of the data or allow the authors to make accurate and direct comparisons of some of the interventions. Only two of the included studies15,16 reported assessments of change in ‘quality of life’ as a result of the interventions, and half (29) of the remaining studies evaluated participant-assessed changes in rosacea severity, both of which were the primary outcomes for this review. Most of the studies gave preference to clinician-assessed numbers of papules or pustules as outcomes rather than patient reported outcomes.

image

Figure 1.  Search results.

Download figure to PowerPoint

Table 2.   Characteristics of included RCTs in the review
StudyMethodsParticipantsInterventionsOutcomes
  1. BID, twice daily; BZP, benzoyl peroxide; IGA, Investigator Global Assessment; IPL, intense pulsed light; N, number; PDL, pulsed dye laser; PGA, Physician’s Global Assessment; PHA, polyhydroxy acid; QD, once daily; RCT, randomized controlled trial; TID, three times daily; VAS, visual analogue scale; yrs, years.

Akhyani et al. (2008)55RCT open-label Teheran, Iran67 (37 male/30 female) Mean age 47·93 yrs Papulopustular and erythematotelangiectatic rosacea3 months A: Azithromycin – 1st month 500 mg 3 times/week, 2nd month 250 mg 3 times/week, 3rd month 250 mg twice/week B: Doxycycline – 100 mg once dailyDecrease in inflammatory lesions (mean %) Participants’ own assessment of treatment at the end of the 3rd month Side-effects
Bamford et al. (1999)54RCT double-blind Minnesota, U.S.A44 (sex unspecified) Mean age 56·9 yrs (group A), 58·9 yrs (group B) Active rosacea tested positive for Helicobacter pylori (UBT RWBT tests)2 weeks A: Clarithromycin – 500 mg TID and omeprazole 40 mg QD B: Placebo – QDExtent and intensity of rosacea at follow-up as measured by the number of papules and pustules Extent and intensity of erythema and telangiectasia
Barnhorst et al. (1996)14RCT within-patient comparison Cleveland, U.S.A.13 (7 male/6 female) Age 72·8 yrs Ocular rosacea and previous diagnosis of facial rosacea12 weeks A: Eyelid hygiene plus warm compresses plus metronidazole (0·75%) gel BID B: Eyelid hygiene and warm compresses BIDEye and eyelid grading by physician Score: 1–5 (higher score is worse)
Beutner and Calvarese (2005)19RCT investigator-blinded Multicentre, U.S.A.1299 (340 male/959 female) Mean age 48·4 yrs (group A), 48·3 yrs (group B), 47·8 yrs (group C) Papulopustular rosacea10 weeks A: Metronidazole gel 1%– QD B: Metronidazole cream 1%– QD C: Metronidazole gel vehicle – QD% reduction lesion count % rated as success Noninferiority of metronidazole gel to metronidazole cream Safety and tolerability
Bitar et al. (1990)20RCT double-blind Quebec, Canada only weeks 1–4100 (41 male/59 female) Mean age 50·3 yrs (group A), 50·8 yrs (group B) Papulopustular rosacea2 months A: Metronidazole cream 1%– BID B: Placebo cream – BIDImprovement in clinical evaluation by physician Improvement of global impression
Bjerke et al. (1989)21RCT double-blind Trondheim, Norway97 (44 male/53 female) Mean age 47 yrs Papulopustular and erythematotelangiectatic rosacea2 months A: Metronidazole cream 1%– BID B: Placebo cream – BIDSelf-assessed changes in rosacea severity, physician’s global evaluation, lesion count reduction, reduction in erythema and telangiectasia, adverse events
Bjerke et al. (1999)31RCT double-blind Oslo, Norway116 (57 male/59 female) Age 48·4 yrs (group A), 50·3 yrs (group B) Papulopustular rosacea3 months A: Azelaic acid cream 20%– BID B: Placebo (vehicle) – BIDSelf-assessed changes in rosacea severity, decrease in N of lesions, physician’s global impression of improvement, decrease in erythema and telangiectasia
Bleicher et al. (1987)17RCT double-blind Boston, U.S.A.40 (16 male/24 female) Mean age 48·7 yrs Moderate to severe rosacea9 weeks A: Metronidazole 0·75% gel – BID B: Placebo (vehicle) – BIDSelf-assessed changes in rosacea severity, physician’s global evaluation, decrease in lesion counts, erythema and telangiectasia
Breneman et al. (1998)22RCT double-blind Ohio, U.S.A.156 (33% male/67% female) Mean age 48·5 yrs (group A), 46·9 yrs (group B) Papulopustular rosacea10 weeks A: Metronidazole 1% cream – QD B: Placebo (vehicle) – QDChange in lesion count, physician’s global evaluation score of very good improvement (> 75%), decrease in erythema, telangiectasia
Breneman et al. (2004)39RCT double-blind 2 centres, U.S.A.53 (18 male/35 female) Mean age 43·1 yrs (group A), 45·7 yrs (group B) Papulopustular rosacea12 weeks A: Benzoyl peroxide 5% and clindamycin 1% gel – QD B: Placebo (vehicle) – QD% change in N of papules and pustules, change in severity of erythema, telangiectasia, flushing, burning/stinging, overall rosacea severity assessment
Carmichael et al. (1993)32RCT within-patient comparison Cardiff, U.K.33 (15 male/18 female) Mean age 56·9 yrs (male), 52·8 yrs (female) Papulopustular rosacea13 weeks A: Azelaic acid cream 20%– BID B: Placebo (vehicle) – BIDSubjective severity score, decrease in lesion count, erythema and telangiectasia, physician’s overall rating of improvement
Dahl et al. (1998)23RCT double-blind (second phase) Multicentre, U.S.A88 (32 male/56 female) Mean age 48·6 yrs (group A), 43·7 yrs (group B) Moderate to severe rosacea6 months A: Metronidazole 0·75% gel – BID B: Placebo (vehicle) – BIDRelapse (appearance of papules and pustules)
Dahl et al. (2001)24RCT investigator-blinded Three centres, U.S.A.72 (21 male/51 female) Mean age 45 yrs (group A), 47 yrs (group B) Papulopustular and erythematotelangiectatic rosacea12 weeks A: Metronidazole 0·75% cream – QD B: Metronidazole 1% cream – QDMedian % change in lesion counts, % change in total erythema score, physician’s assessment of global severity
Del Rosso et al. (2007) (study a)18RCT double-blind Multicentre, U.S.A.251 (91% female group A, 95% female group B) Mean age 46·8 yrs (group A), 47·6 yrs (group B) Papulopustular rosacea16 weeks A: Doxycycline 40 mg capsule – QD B: Placebo capsule – QDMean change in total lesion count, mean change in Clinician’s Erythema Assessment, mean change in IGA scale, adverse events
Del Rosso et al. (2007) (study b)18RCT double-blind Multicentre, U.S.A.286 (94% female treatment group, 95% female control group) Mean age 46·3 yrs (group A), 47·6 yrs (group B) Papulopustular rosacea16 weeks A: Doxycycline 40 mg capsule – QD B: Placebo capsule – QDMean change in total lesion count, mean change in Clinician’s Erythema Assessment, mean change in IGA scale, adverse events, 4-week post-treatment evaluation
Del Rosso et al. (2008)53RCT double-blind Multicentre, U.S.A.91 (27 males/64 females) Age 44·3 yrs (group A), 45·2 yrs (group B) Papulopustular and erythematotelangiectatic rosacea16 weeks A: Doxycycline 40 mg QD + metronidazole gel 1%– QD B: Doxycycline 100 mg QD + metronidazole gel 1%– QDMean change in total lesion count, mean change in Clinician’s Erythema Assessment, mean change in IGA scale, adverse events
Draelos and Fuller (2005)40RCT double-blind Winston-Salem, U.S.A.30 (male/female) Age 20–65 yrs Papulopustular and erythematotelangiectatic rosacea4 weeks A: Lotion vehicle + 1% 4-ethoxybenzaldehyde – BID B: Lotion vehicle – BIDOrdinal assessment erythema, desquamation, dermatitis, uneven skin tone, overall disease severity, tolerability, adverse events
Draelos et al. (2006)41RCT investigator-blinded Winston-Salem, U.S.A.30 (male/female) Age 19–66 yrs Papulopustular and erythematotelangiectatic rosacea12 weeks A: Azelaic acid 15% gel + self-selected skin cleanser and moisturizer – BID B: Azelaic acid 15% gel BID + standardized PHA containing cleanser and anti-ageing moisturizerNumber of lesions, IGA, participant’s assessment of rosacea severity
Dreno et al. (1998)25RCT investigator-blinded Multicentre, France100 (age and sex unspecified) Moderate to severe rosacea12 weeks A: Metronidazole 0·75% cream – BID B: Metronidazole 0·75% gel – BIDDecrease in lesions and IGA, patient’s preference, adverse events
Elewski et al. (2003)36RCT double-blind Multicentre, U.S.A.251 (56 male/195 female) Age 49 yrs (group A), 46 yrs (group B) Papulopustular rosacea15 weeks A: Azelaic acid 15% gel – BID B: Metronidazole 0·75% gel – BIDChange in lesion count, % change in lesion count, change in severity for erythema and telangiectasia, IGA, participant’s overall improvement ratings and cosmetic acceptability
Fowler (2007)52RCT double-blind Multicentre, U.S.A.72 (16 male/56 female) Age unclear Papulopustular and erythematotelangiectatic rosacea16 weeks A: 40 mg doxycycline QD + metronidazole gel 1% BID B: Placebo capsules + metronidazole gel 1%– BID After 12 weeks metronidazole gel discontinued but oral medication or placebo continued until week 16Mean change in lesion count, IGA, % change in lesion count, change in Clinician’s Erythema Assessment score
Grosshans et al. (1997)63RCT double-blind Strasbourg, France34 (6 male/28 female) Mean age 44 yrs (group A), 49 yrs (group B) Papulopustular and erythematotelangiectatic rosacea4 months A: Rilmenidine 1 mg – QD B: Placebo tabletsN of participants with a decrease of at least 50% in lesion count, decrease in lesion count and erythema, Physician’s global investigation
Guillet et al. (1999)26RCT investigator-blinded Multicentre, Europe114 (sex unspecified) Age 22–82 yrs Papulopustular and erythematotelangiectatic rosacea12 weeks A: Metronidazole 0·75% gel B: Metronidazole 0·75% lotion – application frequency unclearCompare efficacy and safety of two formulations, reduction in lesion count, physician’s global evaluation
Jorizzo et al. (1998)27RCT double-blind Multicentre, U.S.A. (distribution of participants unclear)277 (age and sex unclear) Papulopustular rosacea10 weeks A: Metronidazole 1%– QD B: Metronidazole 1%– BID C: Placebo (vehicle) – QD D: Placebo – BIDDecrease in N of lesions and erythema, physician’s global evaluation
Karsai et al. (2008)68RCT double-blind within-patient comparison Karlsruhe, Germany20 (14 male/6 female) Mean age 62 yrs Erythematotelangiectatic rosaceaOne treatment A: 959-nm PDL + 1064-nm Nd:YAG laser (sequential application) B: 959-nm PDL C: 1064-nm Nd:YAG If no effect treatment was repeated up to three times in the same sessionImprovement assessed by review of standardized photographs, side-effects
Koca et al. (2010)42RCT open-label Zonguldak, Turkey Baseline imbalance49 (29 male/20 female) Mean age 50·7 yrs (group A), 48·4 yrs (group B) Papulopustular rosacea12 weeks A: Metronidazole cream 1%– BID B: Pimecrolimus cream 1%– BIDChange in N of lesions, severity of rating of erythema and telangiectasia
Koch and Wilbrand (1999)64RCT double-blind Kassel, Germany40 (11 male/29 female) Age unclear Facial rosacea6 weeks A: Dark sulphonated shale oil 200 mg 2 tab TID – after 2 weeks 2 tab BID B: PlaceboReduction in lesions, reduction in erythema, reduction of scaling, IGA
Koçak et al. (2002)28RCT double-blind Ankara, Turkey63 (15 male/48 female) Mean age 51 yrs Papulopustular rosacea2 months A: Permethrin 5% cream – BID B: Metronidazole 0·75% gel – BID C: Placebo – BIDMean difference in erythema, telangiectasia, oedema and phyma, mean difference in N of lesions and Demodex folliculorum
Lebwohl et al. (1995)43RCT investigator-blinded Multicentre, U.S.A.63 (21 male/42 female) Age 25–80 yrs Papulopustular and erythematotelangiectatic rosacea8 weeks A: Sodium sulphacetamide 10%/sulphur 5%– BID B: Metronidazole 0·75% gel – BIDPGA, overall severity of rosacea (papulopustules, erythema, telangiectasia), adverse events, patient evaluated global response, cosmetic acceptability
Maddin (1999)37RCT double-blind within-patient comparison Vancouver, Canada40 (11 male/29 female) Mean age 52·2 yrs (males), 49·6 yrs (females) Papulopustular rosacea15 weeks A: Azelaic acid 20% cream – BID B: Metronidazole 0·75% cream – BIDSelf-assessed changes in rosacea severity, decrease in redness, decrease in lesion count, physician’s global evaluation
Marks and Ellis (1971)50RCT double-blind London, U.K.56 (27 male/29 female) Mean age 47·8 yrs Papulopustular and erythematotelangiectatic rosacea6 weeks A: Tetracycline TID 250 mg for 1 week and then BID in weeks 2–6 B: Ampicillin dosage unspecified C: PlaceboLesion count, N of participants with > 50% improvement, amount of erythema, participant’s opinion, adverse events
Monk et al. (1991)59RCT double-blind Multicentre, U.K.33 (17 male/16 female) Mean age 46·9 yrs (group A), 50·7 yrs (group B) Papulopustular and erythematotelangiectatic rosacea9 weeks A: Metronidazole gel 0·75% + placebo capsules – BID B: Placebo gel + oxytetracycline 250 mg – BIDLesion count, assessment of erythema, physician’s global evaluation, adverse events
Montes et al. (1983)44RCT double-blind Buenos Aires, Argentina Only first 4 weeks included; bias after 4 weeks64 (19 male/39 female/6 not known) Age unclear Papulopustular rosacea4 weeks followed by 4 weeks for participants who showed improvement A: BZP acetone gel 5% QD first 4 weeks and 10% last 4 weeks B: Placebo (acetone gel vehicle)Lesion count, overall response, erythema and telangiectasia
Mostafa et al. (2009)45RCT double-blind, within-patient comparison Zagazig, Egypt24 (1 male/23 female) Mean age 51·08 yrs Facial rosacea15 weeks A: Azelaic acid 20% cream – BID B: Metronidazole 0·75% cream – BID C: Permethrin 5% cream – BIDPhysician’s assessment, photographic assessment, participant’s assessment, side-effects
Neuhaus et al. (2009)69RCT investigator-blinded within-patient comparison San Francisco, U.S.A.30 (9 male/21 female) Mean age 45·8 yrs Erythematotelangiectatic rosaceaThree treatment sessions each month A: PDL B: IPL C: No treatmentErythema (spectrophotometer), erythema grade and telangiectasia grade, quantitative telangiectasia counts, VAS for signs and symptoms
Nielsen (1983) (study a)29RCT double-blind Boden, Sweden81 (32 male/49 female) Mean age 47 yrs Rosacea in different degrees2 months A: Metronidazole cream 1%– QD B: Placebo (vehicle) – QDPhysician’s global evaluation, lesion counts, reduction in erythema and telangiectasia, photographic evaluation, participant’s opinion
Nielsen (1983) (study b)60RCT double-blind Boden, Sweden51 (17 male/34 female) Mean age 44 yrs Rosacea2 months A: Placebo cream QD and oxytetracycline BID – 250 mg B: Metronidazole cream 1% QD and placebo tablets – BIDReduction in erythema, lesions and telangiectasia, physician’s global evaluation, photographic evaluation, participant’s opinion
Pye and Burton (1976)56RCT double-blind Bristol, U.K.29 (sex unspecified) Age 24–86 yrs Variable degrees of rosacea6 weeks A: Metronidazole 200 mg BID combined with hydrocortisone 1% cream B: Lactose BID combined with hydrocortisone 1% creamClinical severity assessed by physician
Rigopoulos et al. (2005)46RCT double-blind Multicentre, Europe246 (70 male/176 female) Mean age 48·9 yrs Facial rosacea12 weeks A: Cream containing 1% extract of a flavonoid-rich plant, Chrysanthellum indicum– BID B: Placebo – BIDSeverity of erythema, erythema surface, investigator’s overall assessment, investigator’s final efficacy assessment
Saihan and Burton (1980)57RCT double-blind Bristol, U.K.40 (age and sex unspecified) Papulopustular rosacea12 weeks A: Oxytetracycline 250 mg – BID B: Metronidazole 200 mg – BIDClinical improvement assessed by participant and two doctors
Sanchez et al. (2005)58RCT double-blind Multicentre, U.S.A.40 (8 male/32 female) Age 41·6 yrs (group A), 38·8 yrs (group B) Papulopustular and erythematotelangiectatic rosacea12 weeks A: Metronidazole 0·75% lotion BID + doxycycline hyclate 20 mg BID (followed by 4 weeks monotherapy of doxycycline hyclate) B: Metronidazole 0·75% lotion BID + placebo tablets BID (followed by 4 weeks placebo tablets)Change in lesion count at 12- and 16-week visits, changes in Clinician’s Global Severity Score and Clinician’s Global Erythema Assessment, adverse events
Sauder et al. (1997)47RCT double-blind Multicentre, Canada and U.S.A.; distribution of participants unclear103 (40 male/63 female) Mean age 50 yrs Papulopustular and erythematotelangiectatic rosacea8 weeks A: Topical sodium sulphacetamide 10% and sulphur 5% lotion BID B: Placebo (vehicle)Physician’s global evaluation, lesion count reduction, participant’s assessment of improvement, erythema and adverse events
Schachter et al. (1991)61RCT double-blind Multicentre, Canada125 (40 male/61 female/24 not known) Mean age 45·4 yrs Papulopustular rosacea2 months A: Metronidazole 1% cream BID and placebo capsules – TID B: Placebo cream BID and tetracycline 250 mg – TIDClinical evaluation, adverse events, global evaluations, participant assessed global improvement
Schechter et al. (2009)15RCT double-blind Florida, U.S.A.37 (24 male/13 female) Age 65·7 yrs (group A), 69·6 yrs (group B) Ocular rosacea3 months A: Ciclosporin 0·05% ophthalmic emulsion – BID B: Artificial tears – BIDOcular Surface Disease Index (OSDI) to determine the impact on quality of life, Schirmer test, measurement of corneal staining, tear breaking-up time, corneal staining, N of meibomian glands expressed, quality of the excreta
Sharquie et al. (2006)65RCT double-blind Baghdad, Iraq25 (9 male/16 female) Mean age 48·2 yrs Rosacea in different degrees3 months (thereafter cross-over) A: Zinc sulphate 100 mg capsules – TID B: Placebo – TIDDisease severity, number of lesions and telangiectasia, and the presence or absence of rhinophyma, photographic assessment, side-effects, ocular response
Sneddon (1966)51RCT double-blind Rotherham, U.K.85 (26 male/52 female/7 not known) Mean age 47 yrs Papulopustular and erythematotelangiectatic rosacea4 weeks A: Tetracycline 250 mg – BID B: Placebo – BIDAssessable improvement after 1 month
Tan et al. (2002)30RCT double-blind Multicentre, Canada120 (31 male/89 female) Mean age 51 yrs (group A), 47·7 yrs (group B) Moderate to severe rosacea12 weeks A: Metronidazole 1% + sunscreen SPF 15 – BID B: Placebo – BIDPhysician’s global improvement, reduction in lesion counts, reduction facial erythema, local tolerance, reduction facial telangiectasia, safety and tolerability, self-assessed global evaluation
Thiboutot et al. (2003) (study a)33RCT double-blind Multicentre, U.S.A.329 (84 male/245 female) Age 48 yrs (group A), 49 yrs (group B) Papulopustular rosacea12 weeks A: Azelaic acid 15% gel – BID B: Vehicle – BIDIGA, change in N of lesions, facial erythema, telangiectasia, participant’s assessment of rosacea severity, adverse events
Thiboutot et al. (2003) (study b)33RCT double-blind Multicentre, U.S.A.335 (93 male/242 female) Age 48 yrs (group A), 47 yrs (group B) Papulopustular rosacea12 weeks A: Azelaic acid 15% gel – BID B: Vehicle – BIDIGA, change in N of lesions, facial erythema, telangiectasia, participant’s assessment of rosacea severity, adverse events
Thiboutot et al. (2008)34RCT double-blind Multicentre, U.S.A. Exclusions (20) in one centre due to protocol deviation92 (28 male/64 female) Mean age 48·5 yrs (group A), 49·6 yrs (group B) Papulopustular and erythematotelangiectatic rosacea12 weeks A: Azelaic acid 15% gel QD + placebo gel – QD B: Azelaic acid 15%– BIDIGA, treatment response, change in lesion count, erythema, telangiectasia, investigator’s and participant’s assessment of overall improvement, participant’s opinion on cosmetic acceptability and tolerability
Thiboutot et al. (2009)35RCT double-blind (only second phase) Multicentre, U.S.A.136 (35 male/101 female) Mean age 46·4 yrs (group A), 47·5 yrs (group B) Papulopustular and erythematotelangiectatic rosacea24 weeks A: Azelaic acid 15% gel – BID B: Vehicle – BIDRelapse rate, adverse events, inflammatory lesion count, IGA, investigator’s rating of overall improvement, rating by subject of cosmetic acceptability
Torok et al. (2005)48RCT investigator-blind Multicentre, U.S.A.152 (43 male/109 female) Mean age 47 yrs Papulopustular and erythematotelangiectatic rosacea12 weeks A: Sodium sulphacetamide 10% and sulphur 5% cream including sunscreen SPF 15 – BID B: Metronidazole 0·75% cream – BIDTotal N of lesions, erythema grade, investigator global severity, participant’s assessment of global improvement, adverse events, tolerance
Veien et al. (1986)62RCT double-blind Multicentre, Denmark76 (36 male/40 female) Mean age 42·4 yrs Papulopustular and erythematotelangiectatic rosacea8 weeks A: Metronidazole 1% cream and placebo tablets – BID B: Tetracycline tablets 250 mg BID and placebo creamReduction in lesion count, intensity of erythema
Verea Hernando et al. (1992)49RCT double-blind La Coruña, Spain40 (13 male/27 female) Mean age 57·8 yrs (group A), 62·2 yrs (group B) 3 months A: Erythromycin gel 2%– BID B: Metronidazole gel 0·75%– BIDNumber of lesions, erythema and teleangiectasia, global assessment by physician, patient’s assessment
Weissenbacher et al. (2007)16RCT double-blind (first 4 weeks only) Munich, Germany40 (25 male/15 female) Mean age 58 yrs Papulopustular rosacea4 weeks A: Pimecrolimus 1% cream – BID B: Vehicle cream – BIDRosacea severity score, subjective severity assessment, quality of life assessment, photographic assessment
Wilkin (1989)66RCT double-blind (only first study period) Richmond, U.S.A.15 (4 male/11 female) Age 41–60 yrs Erythematotelangiectatic rosacea53 days A: Placebo for 18 days (period A), placebo for 17 days (period B), followed by nadolol 40 mg QD for 18 days (period C) B: Placebo for 18 days (period A), placebo for 17 days (period B), and then nadolol 40 mg BID for 18 days (period C) C: Nadolol 40 mg for 18 days (period A), placebo for 17 days (period B), and then placebo for 18 days (period C) D: Nadolol 40 mg BID for 18 days (period A), placebo for 17 days (period B), and then placebo for 18 days (period C)Reduction of flushing intensity, number and duration of flushes and intensity as assessed by participant
Wilkin and de Witt (1993)67RCT investigator-blinded U.S.A. (Data largely unclear)43 (sex distribution unclear) Age 25–70 yrs Papulopustular and erythematotelangiectatic rosacea12 weeks A: Clindamycin 1% lotion BID + placebo capsules – 4 times daily during first 3 weeks and thereafter BID B: Vehicle lotion BID + tetracycline 250 mg – 4 times daily first 3 weeks and thereafter BID% change in mean lesion count, skin tolerance, physician’s and participant’s assessment of result
Wolf et al. (2006)38RCT investigator-blind Multicentre, U.S.A.160 (44 male/116 female) Mean age 51·1 yrs (group A), 51·1 yrs (group B) Papulopustular rosacea5 weeks A: Metronidazole 1% gel – QD B: Azelaic acid 15% gel – BIDLesion counts, investigator global severity score, erythema severity, tolerability, adverse events, participant’s satisfaction

Risk of bias of included studies

Only three of the studies met all of the criteria across all of the domains in the Cochrane Collaboration’s tool for assessing the risk of bias, and therefore these studies were considered to be at ‘low risk of bias’ (plausible bias unlikely to seriously alter the results).17,18 Thirty studies were categorized as ‘unclear risk of bias’ (plausible bias that raises some doubt about the results) because one or more criteria were assessed as unclear, while the remaining 25 studies were assessed as ‘high risk of bias’ (plausible bias that seriously weakens confidence in the results) because one or more of the criteria were not met.

Effects of interventions

A summary of the most important results is provided in Table 3.

Table 3.   Summary of results of included studies
StudyInterventionsSummary outcomesComments
  1. BID, twice daily; BZP, benzoyl peroxide; CI, confidence interval; GEE, generalized estimating equations; IGA, Investigator’s Global Assessment; IPL, intense pulsed light; ITT, intention to treat; MD, mean difference; N, number; PDL, pulsed dye laser; PGA, Physician’s Global Assessment; PHA, polyhydroxy acid; PP, per protocol; QD, once daily; RR, relative risk; VAS, visual analogue scale.

Akhyani et al. (2008)55A: Azithromycin B: Doxycycline29/37 vs. 24/30 considered themselves improved (RR 0·98, 95% CI 0·77–1·25) N of lesions at baseline 19·24 ± 9·67 and after 3 months 1·90 ± 3·28 for group A, and 18·86 ± 8·95 at baseline and after 3 months 2·34 ± 3·47 for group BBoth treatments reduce the N of lesions from baseline, no evidence of differences between the two groups. However, some data were skewed
Bamford et al. (1999)54A: Clarithromycin + omeprazole B: PlaceboN of pustules (SD) 6·2 (8·3) vs. 12·6 (19·3). Total score (SD) 6·3 (2·5) vs. 7·9 (4·9)No evidence of a difference between the treatment and placebo groups. Data on pustules quite skewed. Data on total score very skewed
Barnhorst et al. (1996)14A: Lid hygiene plus warm compresses + metronidazole gel B: Lid hygiene and warm compressesEye and eyelid grading pre-post mean (SD) 1·5 (1·7) vs. 1·0 (1·7) (higher score is worse)Small group, limited data. Participants not blinded. Data skewed
Beutner and Calvarese (2005)19A: Metronidazole gel B: Metronidazole cream C: Metronidazole gel–vehicleReduction in lesion count 66·7% vs. 58·3% vs. 46·2% Subjects rated as success according to physicians 38·4% vs. 35·4% vs. 27·5%Large ‘vehicle’ effect
Bitar et al. (1990)20A: Metronidazole cream B: Placebo creamN papules (SD) 4·5 (4·24) vs. 6·5 (4·95). N pustules 1·5 (1·41) vs. 3·4 (4·94). Erythema and telangiectasia, no statistical difference Mean change in severity (SD) was 2·80 (1·41) vs. 3·30 (1·41) with MD of −0·50 (95% CI fixed −1·05 to 0·05)Data on papules and pustules are skewed
Bjerke et al. (1989)21A: Metronidazole cream B: Placebo cream43/50 in metronidazole group considered themselves improved compared with 24/47 in placebo group (RR 1·68, 95% CI 1·25–2·28) Lesion count reduction 78% vs. 48%No SDs reported for lesion counts
Bjerke et al. (1999)31A: Azelaic acid cream B: Placebo (vehicle)62/76 in azelaic acid group considered themselves improved compared with 22/38 in placebo group (RR 1·41, 95% CI 1·05–1·89). Physician’s global evaluation of improvement: 61/76 vs. 21/38 (RR 1·45 95% CI 1·07–1·97). Decrease in lesions 73·4% vs. 50·6%, in erythema 47·9% vs. 37·9%, in telangiectasia 22·3% vs. 23·5%No SDs reported for lesion counts
Bleicher et al. (1987)17A: Metronidazole gel B: Placebo (vehicle)Self-assessment of improvement, 28/37 vs. 4/37 Physician’s global evaluation of improvement, 29/37 vs. 1/37. Decrease in lesion counts, 65·1% vs. 14·9%. Reduction in erythema, 0·8 vs. 0·3 (erythema rating 0–3, higher is worse)No SDs reported. Within-patient comparison
Breneman et al. (1998)22A: Metronidazole cream B: Placebo (vehicle)Physician’s global evaluation of improvement 26/104 vs. 6/52 (RR 2·17, 95% CI 0·95–4·93). Decrease in lesion count 8 vs. 3No SDs reported for lesion counts
Breneman et al. (2004)39A: Benzoyl peroxide and clindamycin gel B: Placebo (vehicle)Patient’s global assessment 1·54 (much better to slightly better) vs. 2·50 (slightly better). Physician’s global assessment 1·85 (marked improvement) vs. 2·96 (minimal improvement). Scale, ranging from 0, meaning ‘clear’, to 5, meaning ‘worse’. Reduction in lesion counts 71·3% (25·3) vs. 19·3% (89·6), no statistically significant difference in erythema score, telangiectasia, burning or stinging. Overall rosacea severity assessment improvement 29·3% vs. 10·6%Some SDs were missing, and most data were skewed
Carmichael et al. (1993)32A: Azelaic acid cream B: Placebo (vehicle)On VAS 16/33 participants were improved in the azelaic cream-treated side vs. only 1/33 in the placebo-treated side. Physician’s overall rating of improvement 30/33 improved vs. 11/33. Papule count 2·5 (2·87) vs. 6·3 (4·6), pustule count 0·0 (0·17) vs. 0·4 (0·57)Data were skewed. Within-patient comparison
Dahl et al. (1998)23A: Metronidazole gel B: Placebo (vehicle)Lesion count 3·3 vs. 5·5, relapse rate 23% vs. 42%, free of lesions 53% vs. 32%. Erythema (74% had no or mild erythema vs. 55%), telangiectasia (no significant difference or effect) and dryness were also assessedNo SDs were provided N of adverse events unclear
Dahl et al. (2001)24A: Metronidazole 0·75% cream B: Metronidazole 1% creamMedian % change, inflammatory lesion counts from baseline to end point −62% for group A vs. −60% for group B. % change in total erythema severity score from baseline to end point −26% for group A vs. −30% for group B 
Del Rosso et al. (2007) (study a)18A: Doxycycline 40 mg B: Placebo58/127 in group A achieved a two-point or greater improvement in IGA score compared with 32/124 in group B (RR 1·77, 95% CI 1·24–2·52); 39 participants in group A achieved an IGA score of 0 (clear) or 1 (near clear) vs. 24 in group B (RR 1·59, 95% CI 1·02–2·47). Mean change in Clinician’s Erythema Assessment scale (0 = none and 4 = severe redness) −2·7 vs. −1·8 (authors state = 0·017) Dropouts: 26/127 vs. 21/124 (RR 1·21, 95% CI 0·72–2·03). Adverse events: 56 in group A vs. 48 in group B (RR 1·14, 95% CI 0·85–1·53)SDs were missing, some were calculated by B.R.C.
Del Rosso et al. (2007) (study b)18A: Doxycycline 40 mg B: Placebo32/142 in group A had achieved a two-point or greater improvement in IGA score compared with 23/144 in group B (RR 1·41, 95% CI 0·87–2·29). More than twice as many participants achieved an IGA score of 0 (clear) or 1 (near clear) in group A; 21 participants in group A achieved an IGA score of 0 or 1 vs. 9 in group B (RR 2·37, 95% CI 1·12–4·99). Mean change in Clinician’s Erythema Assessment scale: no statistically significant difference. Mean total lesion count at week 20 was 10·3 vs. 15·3 Dropouts: 27/142 vs. 26/144 (RR 1·05, 95% CI 0·65–1·71). Adverse events: 93 vs. 74 (RR 1·27, 95% CI 1·04–1·55)SDs were missing, some were calculated by B.R.C.
Del Rosso et al. (2008)53A: Doxycycline 40 mg + metronidazole gel B: Doxycycline 100 mg + metronidazole gelIGA mean change from baseline, −0·01 (95% CI fixed −0·12 to 0·10). Change in total inflammatory lesion count, −0·30 (95% CI fixed −3·03 to 2·43). Change in Clinician’s Erythema Assessment from baseline. Scores range from 0 to 4, 0 = no redness present, 4 = severe redness. −4·2 for 40 mg group vs. −4·0 for 100 mg group. Adverse events: 6/44 participants treated with 40 mg vs. 26/47 participants in 100 mg group (RR 0·25, 95% CI 0·11–0·54)Unclear how many people were screened N of lesions at baseline unclear SDs were missing. Some were calculated by B.R.C.
Draelos and Fuller (2005)40A: Lotion (vehicle) + 1% 4-ethoxybenzaldehyde B: Lotion (vehicle)% improvement in erythema: 43·7% in the active treatment group vs. 16·7% in vehicle group (authors state difference between the two groups is not statistically significant). % improvement in overall severity; 49% vs. 15% (authors state < 0·05)Percentage improvement in dermatitis is not addressed Data of subject evaluation not provided
Draelos et al. (2006)41A: Azelaic acid gel + self-selected skin cleanser and moisturizer B: Azelaic acid gel + standardized PHA-containing cleanser, and anti-ageing moisturizerNo statistically significant treatment difference in change in median lesion count, no exact data are provided, only figure. Statistically significant treatment difference in erythema severity in favour of PHA group after 8 and 12 weeks, no exact data are provided, only figureUnclear how many patients were randomized in each group and because very limited outcomes data were reported no reliable conclusions could be drawn
Dreno et al. (1998)25A: Metronidazole 0·75% cream B: Metronidazole 0·75% gelPhysician’s global evaluation was good to excellent in 77·5% of participants in cream group vs. in 73·2% in gel group. Reduction in lesion counts was 61·3% in cream group vs. 63·5% in gel groupStated as ITT analysis but the only data that were reported was PP ‘analysis’ at 12 weeks. 67 out of 100 were analysed
Elewski et al. (2003)36A: Azelaic acid gel B: Metronidazole gel50/124 participants in the azelaic acid gel group considered themselves improved vs. 43/127 in the metronidazole gel group (RR 1·19, 95% CI 0·68–1·65). 86/124 participants were considered to be improved by the physicians vs. 70/127 in the placebo group (RR 1·26, 95% CI 1·03–1·53). Decrease in lesions is 12·9 vs. 10·7 (= 0·003), decrease in % lesions 72·7% vs. 55·8% (< 0·002). Erythema improvement 56% vs. 42% (= 0·02)Difference of two lesions is statistically significant, but not clinically relevant, authors themselves stated that a difference of five lesions was considered relevant. SDs are lacking. Metronidazole seemed to be better tolerated with fewer side-effects
Fowler (2007)52A: 40 mg doxycycline + metronidazole gel B: Placebo capsules + metronidazole gelMean change in physician’s global evaluation score −1·3 vs. −0·8 (authors state = 0·01). Mean change in lesion count −13·86 vs. −8·47 (authors state = 0·002). Mean % change in lesions count −66·4% vs. −48·2% (authors state = 0·008). Mean change from baseline in erythema score −1·3 vs. −0·7 (authors state no statistically significant difference)No SDs were provided
Grosshans et al. (1997)63A: Rilmenidine B: Placebo tablets6/15 vs. 6/19 considered their rosacea improved (RR 1·27, 95% CI 0·51–3·14). The physicians reported that 5/15 vs. 1/19 showed improvement (RR 6·33, 95% CI 0·83–48·59), and in line with the participants’ assessments rilmenidine was not considered to be effective. 69·2% vs. 57·1% had at least 50% decrease in lesion count. Decrease in lesion count 1 vs. 2No SDs were provided
Guillet et al. (1999)26A: Metronidazole 0·75% gel B: Metronidazole 0·75% lotionPGA: 55% was markedly improved or clear in the gel group vs. 61% in the lotion group. Quote: ‘... reducing inflammatory lesion counts by more than 70% on gel and lotion’Frequency of application not reported. Poster, data incomplete
Jorizzo et al. (1998)27A: Metronidazole 1% QD B: Metronidazole 1% BID C: Placebo (vehicle) QD D: Placebo BIDRates of improvement assessed by physicians 79% vs. 72% vs. 39% vs. 45%. Decrease in lesion count 58% vs. 58% vs. 30% vs. 40%. Physician’s global evaluation of improvement 79% vs. 72% vs. 39% vs. 45%Unclear how many participants were initially recruited. Unclear how many participants started in each group, no SDs, dropout rate unclear. Data seem very skewed
Karsai et al. (2008)68A: PDL + Nd:YAG laser B: 959-nm PDL C: 1064-nm Nd:YAGAfter 4 weeks the N of participants and treated sites with dual wavelength, grade 1 = 1 (participant), grade 2 = 1, grade 3 = 10 and grade 4 = 8 In the PDL group, grade 1 = 5 (participants), grade 2 = 3, grade 3 = 2, and grade 4 = 0 In the Nd:YAG group, grade 1 = 2 (participants), grade 2 = 6, grade 3 = 2, and grade 4 = 0 Adverse events were transient purpura and immediate post-treatment erythema. Authors state no statistical significance between group differencesOnly nose was treated, no numbers of adverse events are provided
Koca et al. (2010)42A: Metronidazole cream B: Pimecrolimus cream25/25 showed a measure of improvement compared with 22/25 (RR 1·13, 95% CI 0·96–1·33). Change in N of lesion counts from 16·0 ± 4·6 to 0·6 ± 1·5 vs. from 26 ± 14·4 to 3·7 ± 6·8. No significant difference from baseline to endpoint in both groups for erythema or telangiectasiaN of lesion counts is massively skewed and the baseline data for number of lesions in the two groups are not comparable (26 vs. 16)
Koch and Wilbrand (1999)64A: Dark sulphonated shale oil B: PlaceboLesion counts 15·9 to 4·3 in treatment group and 16·1 to 14·1 in placebo (< 0·0001). No side-effects were notedTotal N randomized 30, unclear how many in each group. Poster, incomplete data. SD not reported
Koçak et al. (2002)28A: Permethrin 5% cream B: Metronidazole gel C: PlaceboMD in erythema score (SD) 1·30 (0·76) vs. 1·45 (0·69) vs. 0·05 (0·22). MD in papules 4·30 (5·51) vs. 5·10 (5·44) vs. 0·25 (1·29). MD in pustules 1·78 (3·74) vs. 3·50 (3·55) vs. 0·25 (0·72). No effects on rhinophyma and telangiectasiaMost data are skewed. Except for pustules, permethrin 5% cream showed same results as metronidazole
Lebwohl et al. (1995)43A: Sodium sulphacetamide/sulphur B: Metronidazole gelNo statistically significant difference in decrease of papule count (MD −1·10, 95% CI −3·19 to 0·99). However, there was a statistically significant difference in decrease in the pustule count in favour of group A (MD −1·90, 95% CI −3·01 to −0·79). No statistically significant treatment differences in any of the patient evaluations. Adverse events: 6 (concerning five patients) vs. 3. The five patients discontinued treatment in group A vs. one in group B due to adverse events that were rated as related to treatmentThere are also data of physician’s global evaluation; however, no baseline data are provided. No exact data are provided for patient evaluations
Maddin (1999)37A: Azelaic acid cream B: Metronidazole creamMean score on the azelaic acid-treated side was according to the patients 1·87 (SD 0·76) compared with 2·33 (SD 0·95) on the metronidazole-treated side (= 0·02, higher score is worse). Regarding the physicians: 2·7 (SD 1·0) compared with 3·1 (SD 1·0) on the metronidazole-treated side (= 0·05). Decrease in lesion count 78·5% vs. 69·4%Large SDs, data skewed
Marks and Ellis (1971)50A: Tetracycline B: Ampicillin C: Placebo14/20 vs. 14/17 vs. 9/19 considered themselves improved. Lesion count post-treatment 4·60 (6·20) vs. 9·53 (8·79) vs. 16·63 (12·81). N improved > 50% 17/20 vs. 9/17 vs. 4/19 were considered to be improved according to the physiciansLarge SDs, data skewed. Dropouts unclear
Monk et al. (1991)59A: Metronidazole gel + placebo capsules B: Placebo gel + oxytetracycline8/16 vs. 12/17 considered themselves improved (RR 0·71, 95% CI 0·40–1·26). 9/16 vs. 12/17 were improved according to physicians (RR 0·80, 95% CI 0·47–1·35). Reduction in lesion count of more than 50% in both groups, with 100% clearing in 75% vs. 66%The data were analysed using a χ2test comparing proportion of patients with no lesions at week 9 vs. those with them. The Ns are small (27 in total) and the analysis was a completers’ analysis. Analysis should have been a ‘change from baseline’ in line with all the other papers presented this time, rather than a χ2 test. Some SD are missing
Montes et al. (1983)44A: BZP acetone gel B: Placebo19/31 vs. 9/27 were considered improved (RR 1·84, 95% CI 1·01–3·36). Overall response 2·69 vs. 3·71, papule score 0·89 vs. 1·91, pustule score 0·46 vs. 1·31 in favour of BZP (higher score is worse)Unclear how many participants started, only first 4 weeks included
Mostafa et al. (2009)45A: Azelaic acid cream B: Metronidazole cream C: Permethrin creamAt end of treatment azelaic acid was statistically significantly more effective on inflammatory lesions than both metronidazole and permethrin creams, while there was no statistically significant difference between metronidazole and permethrin (authors state). There was a statistically significant difference in the mean score of erythema at end of treatment for the three groups (< 0·05). The effect of the three creams on erythema was not statistically significantly different from each other. (Side-effects included itching, burning sensation, oedema and scales. No significant differences in side-effects between the three groups)The study analysis has been carried out on skewed data
Neuhaus et al. (2009)69A: PDL B: IPL C: No treatmentPatient assessment – repeated-measures analysis using GEEs revealed statistically significantly better results after IPL or PDL in all of the following patient-rated signs and symptoms than in untreated controls: erythema, flushing, dryness, burning and pruritus (all < 0·05). Blinded investigator rating – statistically significant lower erythema and telangiectasia grades were seen in IPL- and PDL-treated groups (all < 0·01) than the nontreated group, but no statistically significant differences were seen between IPL and PDL (> 0·6)No exact data were provided, making the data unusable
Nielsen (1983) (study a)29A: Metronidazole cream B: Placebo (vehicle)Physician’s global evaluation of improvement 24/41 vs. 8/40 (RR 2·93, 95% CI 1·50–5·73). Papule count 8·6 vs. 16·6, and pustule count 0·3 vs. 0·8. Reduction on erythema from 3·8 to 2·5 for metronidazole group and from 3·7 to 3·1 in placebo group. Authors state < 0·05. No effects on telangiectasiaNo SDs were provided
Nielsen (1983) (study b)60A: Placebo cream and oxytetracycline B: Metronidazole cream and placebo tablets24/22 vs. 24/26 patients considered themselves improved (RR 0·99, 95% CI 0·84–1·17). 24/25 vs. 25/26 were improved according to the physicians (RR 1·0, 95% CI 0·78–1·12). No adverse events in both groups 
Pye and Burton (1976)56A: Metronidazole orally combined with hydrocortisone 1% cream B: Lactose combined with hydrocortisone 1% creamThe physicians considered that 10/15 vs. 2/14 improved (RR 4·64, 95% CI 1·23–17·68)Only limited data were given in this study
Rigopoulos et al. (2005)46A: Cream containing 1% extract of a flavonoid-rich plant, Chrysanthellum indicum B: Placebo98/125 vs. 92/121 patients considered themselves improved (RR 1·03, 95% CI 0·90–1·18). 68/125 showed improvement vs. 54/121 according to physicians (RR 1·22, 95% CI 0·94–1·57). Mean change in rosacea severity score from baseline: difference was −0·34 in favour of the flavonoid-rich cream (fixed 95% CI −0·50 to −0·18) 
Saihan and Burton (1980)57A: Oxytetracycline orally B: Metronidazole orallyAccording to the combined rating of participants and physicians, no evidence of a difference between treatments was found at end of study. Scale 1–3, a higher rating indicating lesser severity. The mean scores were 2·30 (SD 1·00) vs. 2·60 (SD 0·70), difference −0·30 (95% CI −0·83 to 0·23)Limited data were provided in this study
Sanchez et al. (2005)58A: Metronidazole lotion + doxycycline hyclate 40 mg B: Metronidazole lotion + placebo tabletsReduction in lesions 15 vs. 6. Authors state that reduction in doxycycline group was still statistically significant at week 16 (< 0·01) Clinician’s global severity score (0–4, lower is better): decrease from baseline of 0·9 vs. 0·4Data from graphs, SDs are missing
Sauder et al. (1997)47A: Topical sodium sulphacetamide and sulphur lotion B: Placebo (vehicle)Self-assessed improvement in rosacea severity 90% vs. 58%. Physician’s evaluation of improvement 98% vs. 68%, participants and physicians 7-point improvement score both showed < 0·001 using nonparametric test. Lesion count reduction (SEM) 78% (4·5) vs. 36% (8·9), < 0·001 using nonparametric test. Improvement in erythema 83% vs. 31% (authors state < 0·001)Overall reporting quality was inadequate: the number of participants in each treatment arm was not reported, improvement as an outcome was ill defined, and the data reported as continuous outcomes were skewed and largely unusable
Schachter et al. (1991)61A: Metronidazole cream and placebo capsules B: Placebo cream and tetracycline orallyThe trialists concluded that topical metronidazole and oral tetracycline both improve the papule and pustule counts and, that ‘metronidazole applied topically as a 1% cream is well tolerated and is as effective as standard therapy with tetracycline 250 mg 3 times a day’No exact data are provided. Data estimated from figures. 125 were randomized, unclear how many in each group, there were 24 withdrawals. The study is at high risk of bias due to the significant number of dropouts and based on the data reported we are unable to concur with the conclusions of equivalent effectiveness reached by the investigators
Schechter et al. (2009)15A: Ciclosporin 0·05% ophthalmic emulsion B: Artificial tearsAssessment of changes in quality of life were carried out with the Ocular Surface Disease Index (scale 0–100, 100 = worst). Baseline scores were 19·1 (SD 13·9) vs. 16·9 (SD 15·8). Difference between the change scores at completion of the study was 8·6 in favour of topical ciclosporin (= 0·022, 95% CI 1·78–15·42). Data of physicians’ assessments confirm efficacy of topical ciclosporin (Schirmer score, tear break-up time, corneal staining and number of unoccluded expressible meibomian glands) 
Sharquie (2006)65A: Zinc sulphate orally B: PlaceboLesions showed a statistically significant decrease 1 month after starting treatment with zinc sulphate and become 0 after 3 months of therapy in all participants on zinc sulphate, but not in placebo group. Erythema was also statistically significantly reduced after 3 months in zinc sulphate group but not in placebo group. There was no significant effect of zinc sulphate on telangiectasia. No important side-effects were reported apart from mild gastric upset in three patients who were on zinc sulphateSDs at 3 months missing. Numbers of papules and pustules are not provided
Sneddon (1966)51A: Tetracycline orally B: Placebo28/36 improved vs. 19/42 in placebo (RR 1·72, 95% CI 1·18–2·50). Assessable improvement after 1 month 78% vs. 45%Dropouts unclear, adverse events unclear
Tan et al. (2002)30A: Metronidazole 1% + sunscreen SPF 15 B: Placebo – BIDWithout any reference to data it was reported that patient’s self-assessment showed significant improvement (= 0002). Reduction in N of lesions 13·57 (15·2) in metronidazole group vs. 4·56 (12·3) in placebo group. Reduction in erythema 0·89 (0·52) vs. 0·58 (0·13). No statistically significant difference in local toleranceData are skewed. High dropout rate: 17/61 metronidazole compared with 14/59 in the placebo group, and these were excluded from the efficacy analysis
Thiboutot et al. (2003) (study a)33A: Azelaic acid gel B: Vehicle100/164 patients in azelaic acid group considered themselves improved compared with 71/165 in placebo group (RR 1·42, 95% CI 1·14–1·76). Physician’s global evaluation of improvement: 110/164 vs. 79/165 (RR 1·40 95% CI 1·16–1·70). Change in N of lesions from 17·5 to 6·8 vs. 17·6 to 10·5. Overall improvement in erythema: 44% vs. 29% (= 0·0017). Overall improvement in telangiectasia: 77% vs. 80% (not statistically significant). Overall improvement according to investigator: 51% vs. 27% (< 0·001)SDs not provided but can be estimated from figures
Thiboutot et al. (2003) (study b)33A: Azelaic acid gel B: Vehicle98/169 patients in azelaic acid group considered themselves improved compared with 65/166 in placebo group (RR 1·72, 95% CI 1·34–2·20). Physician’s global evaluation of improvement: 120/169 vs. 91/166 (RR 1·30 95% CI 1·09–1·53). Change in N of lesions from 17·8 to 8·9 vs. 18·5 to 12·1. Overall improvement in erythema: 46% vs. 28% (= 0·0005). Overall improvement in telangiectasia: 73% vs. 78% (not statistically significant). Overall improvement according to investigator: 46% vs. 31% (< 0·0048)SDs not provided but can be estimated from figures
Thiboutot et al. (2008)34A: Azelaic acid gel QD + placebo gel – QD B: Azelaic acid gel BIDNo statistically significant difference was found in any efficacy end point in the comparison azelaic acid 15% gel QD vs. BID (authors state > 0·05 for all comparisons) 
Thiboutot et al. (2009)35A: Azelaic acid gel B: VehicleRelapses: 13 in azelaic acid group and 20 in vehicle group. The increase in mean lesion count in maintenance phase was 5·5 in azelaic acid group and 7·5 (data from figure), authors state = 0·03. No change in erythema was observed nor in telangiectasia in both groupsPatients had to achieve a 75% reduction in lesion count to enter the maintenance phase. Data cannot be extrapolated to all patients with papulopustular rosacea. Difference of two lesions is not clinically relevant
Torok et al. (2005)48A: Sodium sulphacetamide and sulphur cream including sunscreen SPF 15 B: Metronidazole cream group59/75 were considered to be improved vs. 45/77 (RR 1·35, 95% CI 1·08–1·68). Reduction in lesion counts: 80% vs. 72%. Proportion of subjects with improvement of erythema score 69% vs. 45%. 48 subjects reported adverse events in group A vs. 41 in group B. Of these 38 in group A and 20 in group B were related to treatment and this difference was statistically significant (= 0·003)No exact data were provided for lesions counts or erythema
Veien et al. (1986)62A: Metronidazole cream and placebo tablets B: Tetracycline tablets and placebo creamMedian number of lesions after 8 weeks 11·1 vs. 0. Percentage of no improvement of erythema after 8 weeks 11·1 vs. 12·5Tetracyclines had a more rapid onset on lesions. Dropout rate 1, but unclear in which group
Verea Hernando et al. (1992)49A: Erythromycin gel B: Metronidazole gel16/22 vs. 17/18 patients considered themselves improved (RR 0·77, 95% CI 0·58–1·02). The physicians’ assessments were largely in agreement with these dataLesion counts: baseline imbalance with respect to N of lesions placed the study at a serious risk of bias, so are not listed here
Weissenbacher et al. (2007)16A: Pimecrolimus cream B: Vehicle creamDermatology Life Quality Index (score 0–30, higher score = more impairment) showed a reduction of the mean absolute value from 5·50 to 3·10 vs. 6·70 to 3·70 (authors state = 0·75). VAS (0–100 mm, higher = worse) from 53·45 to 48·95 vs. from 64·75 to 43·35 (authors’ reported value = 0·48). Total rosacea severity score reduced from 6·88 to 4·68 vs. 7·00 to 4·33 (authors state = 0·59)Although the data reported in this study were sparse, no statistically significant differences in efficacy could be demonstrated between treatment arms for any of the outcomes
Wilkin (1989)66A: Placebo B: Placebo C: Nadolol 40 mg D: Nadolol 40 mg BIDThere were no statistically significant differences between nadolol and placebo pretreatments for flushing reactions provoked in the laboratory, either globally or within challenge agent or dosage categoriesWe included only first period (A). Small groups, unclear what dropout rate was. Data unusable
Wilkin and de Witt (1993)67A: Clindamycin lotion + placebo capsules B: Vehicle lotion + tetracycline capsules81·8% considered themselves improved in the clindamycin group vs. 90·5% in the tetracycline group. Physicians’ rating of success 94·7% vs. 94·4% improvedUnclear how many patients were randomized in each group, making these data unusable. Total number randomized 43
Wolf et al. (2006)38A: Metronidazole gel B: Azelaic acid gel57/78 vs. 65/82 considered themselves improved (RR 0·92, 95% CI 0·77–1·10). 44/78 were considered by the physicians to be cleared or nearly cleared vs. 44/82, with no statistically significant difference between the groups (RR 1·05, 95% CI 0·79–1·39). Reduction in lesion count: 80% vs. 77%Unclear if the questionnaire that had been used to assess ‘Patient satisfaction’ had been validated, tested and piloted. Of questionable face/construct validity, and with a wide range of very subjective and overlapping variables to be measured
Studies with only topical metronidazole

Fourteen trials provided data on the effectiveness of topical metronidazole.14,17,19–30 Data from three studies could be pooled (see Fig. 2).21,22,29 Topical metronidazole was more effective than placebo and the results were both statistically significant [relative risk (RR) 1·95, 95% confidence interval (CI) 1·48–2·56] and clinically important. The pooled data from these studies was largely in agreement with the participant-assessed outcomes for this comparison and were confirmed and supported by the other studies.14,17,19,20,23–28,30 There were no statistically significant differences between the two concentrations of topical metronidazole (0·75% and 1%), or comparisons using different vehicles and topical metronidazole was also shown to be effective in maintaining remission.23–26 There were no significant differences in the number of dropouts and adverse events across the intervention groups in these studies.

image

Figure 2.  Physician’s Global Evaluation of improvement of rosacea.

Download figure to PowerPoint

Studies with only azelaic acid

Six studies evaluated the effect of azelaic acid out of which three studies compared the effectiveness of azelaic acid vs. placebo.31–35 Pooled participant-assessed data from these studies indicated an improvement in rosacea severity rate of complete remission or marked improvement of 70–80% in the azelaic acid group compared with 50–55% in the placebo group (RR 1·52, 95% CI 1·32–1·76, see Fig. 3).31,33 These results were largely in agreement with the physician-assessed outcomes in these studies. A single daily dose of azelaic acid appears to be as effective as the twice daily dose.34 There was no statistically significant difference during maintenance phase between the azelaic acid group and vehicle-only group (RR 0·73, 95% CI 0·43–1·23).35

image

Figure 3.  Participant-assessed improvement of rosacea.

Download figure to PowerPoint

Studies comparing topical metronidazole and azelaic acid

Three studies provided data for this comparison, one of which had a within-patient study design; therefore pooling of data with the other two studies was not possible.36–38 In two of the studies there was no statistically significant difference between the treatment groups in the patient-assessed outcomes.36,38 However, in the Maddin37 study the patients considered azelaic acid to be more effective. The investigators in both the Elewski et al.36 and the Maddin37 studies were more satisfied with the outcomes of treatment with azelaic acid compared with metronidazole.

Studies with other topical treatments

Most of the other comparisons consisted of single studies and investigated drug, e.g. permethrin, benzoyl peroxide, benzoyl peroxide combined with clindamycin, sodium sulphacetamide 10%/sulphur 5%, pimecrolimus, topical erythromycin, topical ciclosporin, ethoxybenzylaldehyde and adjunctive benefit of a polyhydroxy acid (PHA) regimen and flavonoid-rich cream.15,16,28,39–49 The results are reported in Table 3, but are not discussed as most of these studies were judged to be at high risk of bias and had skewed or unusable data. However, two studies assessed our primary outcome ‘quality of life’.15,16 In Weissenbacher et al.16 the efficacy of pimecrolimus vs. vehicle was investigated and although the data reported in this study were sparse, no statistically significant differences in efficacy could be demonstrated between treatment arms for any of the outcomes. The study of Schechter et al.15 demonstrated a moderate improvement in quality of life in the group with ocular rosacea which had been treated with ciclosporin 0·05% ophthalmic emulsion compared with artificial tears. The data based on physicians’ assessments provide further supportive evidence for the effectiveness of topical ciclosporin 0·05% in the treatment of ocular rosacea.

Studies with oral antibiotics

Two studies which investigated the effect of tetracycline were included.50,51 Based on the participant-assessed outcomes in Marks and Ellis (1971) there was insufficient evidence to demonstrate that tetracycline was more effective than placebo.50 In contrast with the participant-assessed changes, the physicians’ assessments indicated that tetracycline appeared to be significantly more effective than placebo. Seventeen of 20 participants in the tetracycline group were considered to be improved vs. four of 19 in the placebo group in the study of Marks and Ellis50 (RR 4·04, 95% CI 1·66–9·83). The results of the study by Sneddon (1966) showed that 28 of 36 participants in the tetracycline group were improved vs. 19 of 42 in the placebo group (RR 1·72, 95% CI 1·18–2·50).51

Two studies which compared doxycycline vs. placebo were included in this review.18 The data from the Investigator’s Global Assessment (IGA) in Del Rosso et al.18 (2007, study a) indicated that doxycycline 40 mg was more effective than placebo. Fifty-eight participants out of the 127 in the doxycycline 40 mg group achieved a 2-point or greater improvement in IGA score compared with 32 out of 124 in the placebo group (RR 1·77, 95% CI 1·24–2·52). Thirty-nine participants in the doxycycline group achieved an IGA score of 0 (clear) or 1 (near clear) vs. 24 in the placebo group (RR 1·59, 95% CI 1·02–2·47).

In Del Rosso et al.18 (2007, study b) there was no statistically significant difference in IGA; 32 participants of the 142 in the doxycycline 40 mg group had achieved a 2-point or greater improvement in IGA score compared with 23 of 144 in the placebo group (RR 1·41, 95% CI 0·87–2·29). However, more than twice as many participants achieved an IGA score of 0 (clear) or 1 (near clear) in the doxycycline group; 21 participants in the doxycycline group achieved an IGA score of 0 or 1 vs. 9 in the placebo group (RR 2·37, 95% CI 1·12–4·99). The dropout rate was comparable between both groups in each of the studies but the number of participants reporting adverse events was higher in the doxycycline groups. The study of Fowler (2007)52 supported the efficacy data of doxycycline.

Physician-based assessments provided no evidence of any difference in efficacy between the 40 and 100 mg in Del Rosso et al. (2008)53 but only a quarter of the side-effects were reported for the lower as opposed to the higher dose.

The data reported for treatment with clarithromycin and omeprazole were to a large extent considered unusable.54 One study compared doxycycline (100 mg) with azithromycin and although both treatments appeared equally effective the standard deviations were large and the data were skewed.55 Two studies which included treatment with oral metronidazole provided limited data.56,57 Doxycycline 40 mg was compared with topical metronidazole in one study and most data had to be estimated from graphs; however, they indicated that doxycycline might be more effective than topical metronidazole.58

Four studies compared topical metronidazole and oral (oxy)tetracycline, but pooling of data was not possible due to clinical heterogeneity between the studies.59–62 Although the quality of reporting of these studies was poor they did nevertheless indicate that there was no statistically significant difference in effectiveness between metronidazole cream and (oxy)tetracycline.

We included studies which examined topical clindamycin vs. tetracycline, rilmenidine, dark sulphonated shale oil, zinc sulphate and nadolol but the data from these studies were considered largely unusable.63–67

Studies with laser- and/or light-based treatment

In one study the effectiveness of dual-wavelength 595-nm pulsed-dye laser (PDL) and 1064 nm Nd:YAG was investigated, but this was only on the nose.68 In another study (PDL vs. intense pulsed light therapy vs. control) the data were limited and unusable.69

Discussion

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. What’s already known about this topic?
  7. References

Fifty-eight studies were included in this updated Cochrane review. It was somewhat disappointing to see that our principal outcome, ‘quality of life’, was assessed in only two of the studies15,16 and that only half the studies addressed participants’ assessments of improvement in rosacea severity. The majority of studies focused on numbers of papules and pustules which, although they may provide a quantifiable, objective, and more readily intelligible outcome, are generally considered to be clinician-centred rather than patient-preferred. Rosacea is a chronic skin disease and the importance of self-assessments by participants of the effectiveness of interventions should not be underestimated.

Pooling of data was not feasible for most of the comparisons, and was possible only for several outcomes in the trials which evaluated topical metronidazole and azelaic acid.

Regarding treatments for erythematotelangiectatic rosacea (subtype 1), the use of different scoring systems to assess improvement of erythema and telangiectasia, and the paucity and variability of evidence on the effects of interventions on this subtype of rosacea did not in most cases permit firm conclusions to be made. Although topical treatments with metronidazole, azelaic acid and sodium sulphacetamide/sulphur as well as oral treatment with doxycycline appeared to be effective in reducing erythema (not telangiectasia), further research into their effects on erythema is needed to confirm these findings.18,24,28,29,33,36,47,48,52,59 Lasers and light therapies appear to have a major role to play in the treatment of erythematotelangiectatic rosacea but these treatment modalities are still largely under-researched. There was some evidence that PDL and intense pulsed light therapy are capable of reducing erythema and telangiectasia on the face.69 The effects of laser therapy for rosacea on the nose were investigated in only one study.68

For the treatment of papulopustular rosacea (subtype 2), pooled data for topical metronidazole and azelaic acid indicate that both are effective treatments for rosacea.21,22,29,31,33 However, based on physicians’ assessments in Elewski et al. (2003)36 and both physician and participant assessments in Maddin (1999)37 azelaic acid appears to be more effective than metronidazole albeit with more side-effects and, therefore, further evidence is required.

Topical metronidazole was also shown to be effective in maintaining remission.23

A single daily dose of azelaic acid appears to be as effective as the twice-daily dose and is also likely to result in improved compliance.34 RCTs investigating the effectiveness of azelaic acid and metronidazole used in maintenance therapy are still required.

The effectiveness of benzoyl peroxide (alone or combined with clindamycin) in the treatment of papulopustular rosacea is unclear. Inadequate study design, and incomplete and skewed data did not enable definite conclusions to be drawn.39,44

Sodium sulphacetamide 10% in combination with sulphur 5% appeared to be more effective than metronidazole, but further research is warranted, as all the relevant included studies were assessed as being at ‘high risk of bias’.43,47,48

The evidence for the effectiveness of permethrin was inconclusive and therefore further trials with a rigorous study design are required.28,45

There was no evidence to support the effectiveness of pimecrolimus; however, this was based on limited and largely unusable data presented in two studies.16,42

There was no evidence that 4-ethoxybenzaldehyde, flavonoid-rich plant cream or adjunctive PHA skin regimen were effective.40,41,46 Although the use of skin products that contain no irritating ingredients are often recommended for the treatment of papulopustular rosacea, no studies on the use of these cosmetics could be included in this review. No eligible studies were identified for dapsone or topical tretinoin although these treatments are still fairly commonly used.8,70,71

Tetracyclines are used extensively for the treatment of rosacea and although their efficacy may be widely accepted by clinicians, this is currently not substantiated by high-level evidence from methodologically sound clinical trials.50,51

While a number of studies included in this review demonstrated the efficacy of an anti-inflammatory dose of doxycycline as a reduction in physician-assessed lesion counts, quite significantly, the participants’ views and satisfaction with the effects of this intervention were not assessed.18,52,58

There is evidence that the 40-mg dose is at least as effective as the 100-mg dose, has a correspondingly lower risk of adverse effects, and that, albeit these events may be mild to moderate, more were reported with the 100-mg dose of doxycycline than the 40-mg dose.53

Only one study which examined the effects of azithromycin was included but the data were skewed, and consequently more methodologically robust research is required on the effects of this specific intervention.55

Several studies examined other interventions such as rilmenidine and ampicillin and although the latter showed some evidence of effectiveness, neither of these are currently considered as treatment options by clinicians.50,63

For treatment of phymatous rosacea (subtype 3), surgical therapies as well as ablative laser therapies have been used with reportedly good results, but no eligible randomized controlled trials were identified for this systematic review.

The symptoms of ocular rosacea (subtype 4) are often mild but can also be severe and debilitating, and although ocular involvement occurs in up to 58% of people with rosacea, only two trials included in this review examined the treatment of ocular rosacea.14,15 Although there was insufficient evidence to support the efficacy of topical metronidazole for ocular rosacea, there was some evidence of a consistent improvement in all outcomes with ciclosporin 0·05% ophthalmic emulsion and that it was more effective than artificial tears in the treatment of ocular rosacea.14,15

For all treatments, the adverse events reported were mostly mild and transient and comprised skin irritation, pruritus and stinging/burning or dry skin. In most of the studies the number and type of adverse events did not differ significantly between active treatment and the placebo groups; however, these were not always reported adequately or completely.

In conclusion, evidence of treatment effect could be demonstrated for only a limited number of the interventions studied. These were for interventions with topical metronidazole, azelaic acid and doxycycline (40 mg) in the treatment of moderate to severe rosacea, and ciclosporin 0·05% ophthalmic emulsion for ocular rosacea.

There is insufficient evidence to support either the effectiveness or lack of effectiveness of interventions for the management of erythematotelangiectatic rosacea. Clearing of the ‘redness of the face’ in patients with rosacea can have a significant impact on their quality of life but the evidence for the efficacy of light-based therapies, which are commonly used for erythematotelangiectatic rosacea, is lacking and further studies addressing the efficacy of these treatment modalities are warranted.

For papulopustular rosacea, topical metronidazole, azelaic acid and anti-inflammatory dose doxycycline (40 mg) appear to be effective and safe for short-term use, with similar rates of adverse events as in the placebo groups except for doxycycline 40 mg, which showed an increased risk of side-effects. There is evidence that 40 mg is at least as effective as 100 mg with evidence of fewer adverse effects and there is some evidence that tetracycline is effective. No clear evidence is available demonstrating that any one of these treatments, or any combination of treatments, has a particular advantage in terms of higher remission rates and/or fewer adverse effects.

No studies could be included that addressed treatment of phymatous rosacea. Well-designed RCTs addressing which is the most effective treatment for phymatous rosacea are therefore still required.

For ocular rosacea ciclosporin 0·05% ophthalmic emulsion showed some evidence of benefit over artificial tears. The impact of available treatment on ocular rosacea warrants further examination as up to 58% of patients with rosacea suffers from this subtype.

Finally, there is an urgent need for high-quality, well-designed and rigorously reported studies of the more widely used treatments for rosacea such as tetracycline, minocycline, azithromycin, isotretinoin, topical retinoids and light-based therapies. Less direct interventions, such as dietary adjustments, avoidance measures for trigger factors, the use of sunscreens, and patient education are further areas of much-needed research. Outcomes collected in future trials should be primarily based on a standardized scale of the participant’s assessment of the treatment efficacy and also have a greater emphasis on changes in quality of life as a result of the interventions. Standardized and uniform scales should be developed and used for physicians’ assessments, and these should reliably reflect global evaluation, lesion counts and assessment of telangiectasia. Furthermore, to ensure improved clinical decision-making, future research should place a greater emphasis on the management and treatment of rosacea based on the staging pattern of the disease.

It is hoped that clinicians treating patients with rosacea find this evidence-based review helpful and that it will alter their prescribing to improve patient outcomes. Clinical decision-making on the choice of intervention for rosacea should be based on high-level evidence if available, but in the absence of such evidence for any specific intervention these decisions should continue to be guided by clinical experience and patients’ individual characteristics and preferences until further evidence becomes available.

What’s already known about this topic?

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. What’s already known about this topic?
  7. References
  • • Rosacea is an important dermatological condition with a significant impact on quality of life.

  • • There are various options available for treatment but it is unclear which are the most effective.

What does this study add?

  • • This study updates the existing Cochrane systematic review with additional trials and applies intention-to-treat principles to the meta-analyses.

  • • Topical metronidazole, azelaic acid and doxycycline (40 mg) appear to be effective and safe for papulopustular rosacea.

  • • There is evidence that doxycycline 40 mg is at least as effective as 100 mg with evidence of fewer adverse effects.

  • • Ciclosporin 0·05% ophthalmic emulsion is effective for ocular rosacea.

References

  1. Top of page
  2. Summary
  3. Methods
  4. Results
  5. Discussion
  6. What’s already known about this topic?
  7. References
  • 1
    Crawford GH, Pelle MT, James WD. Rosacea: I. Etiology, pathogenesis, and subtype classification. J Am Acad Dermatol 2004; 51:32741.
  • 2
    Korting HC, Schöllmann C. Current topical and systemic approaches to treatment of rosacea. J Eur Acad Dermatol Venereol 2009; 23:87682.
  • 3
    Powell FC. Clinical practice. Rosacea. N Engl J Med 2005; 352:793803.
  • 4
    Wilkin J, Dahl M, Detmar M et al. Standard classification of rosacea: report of the National Rosacea Society Expert Committee on the Classification and Staging of Rosacea. J Am Acad Dermatol 2002; 46:5847.
  • 5
    Wilkin J, Dahl M, Detmar M et al. National Rosacea Society Expert Committee. Standard grading system for rosacea: report of the National Rosacea Society Expert Committee on the Classification and Staging of Rosacea. J Am Acad Dermatol 2004; 50:90712.
  • 6
    Pelle MT, Crawford GH, James WD. Rosacea: II. Therapy. J Am Acad Dermatol 2004; 51:499512.
  • 7
    Baldwin HE. Oral therapy for rosacea. J Drugs Dermatol 2006; 5:1621.
  • 8
    Jansen T, Plewig G. Rosacea: classification and treatment. J R Soc Med 1997; 90:14450.
  • 9
    Bikowski JB, Goldman MP. Rosacea: where are we now? J Drugs Dermatol 2004; 3:25161.
  • 10
    Elewski BE, Draelos Z, Dréno B et al. Rosacea – global diversity and optimized outcome: proposed international consensus from the Rosacea International Expert Group. J Eur Acad Dermatol Venereol 2011; 25:188200.
  • 11
    van Zuuren EJ, Kramer S, Carter B et al. Interventions for rosacea. Cochrane Database Syst Rev 2011; 3:CD003262.
  • 12
    Higgins JPT, Green S (eds). Cochrane Handbook for Systematic Reviews of Interventions. Version 5.1.0 [updated March 2011]. Available from http://www.cochrane-handbook.org, 2011.
  • 13
    Treadwell JT, Tregear SJ, Reston JT et al. A system for rating the stability and strength of medical evidence. BMC Med Res Methodol 2006; 6:52.
  • 14
    Barnhorst DA, Foster JA, Chern KC et al. The efficacy of topical metronidazole in the treatment of ocular rosacea. Ophthalmology 1996; 103:18803.
  • 15
    Schechter BA, Katz RS, Friedman LS. Efficacy of topical cyclosporine for the treatment of ocular rosacea. Adv Ther 2009; 26:6519.
  • 16
    Weissenbacher S, Merkl J, Hildebrandt B et al. Pimecrolimus cream 1% for papulopustular rosacea: a randomized vehicle-controlled double-blind trial. Br J Dermatol 2007; 156:72832.
  • 17
    Bleicher PA, Charles JH, Sober AJ. Topical metronidazole therapy for rosacea. Arch Dermatol 1987; 123:60914.
  • 18
    Del Rosso JQ, Webster GF, Jackson M et al. Two randomized phase III clinical trials evaluating anti-inflammatory dose doxycycline (40-mg doxycycline, USP capsules) administered once daily for treatment of rosacea. J Am Acad Dermatol 2007; 56:791802.
  • 19
    Beutner K, Calvarese B. A multi-center, investigator-blind clinical trial to assess the safety and efficacy of metronidazole gel 1% as compared to metronidazole gel vehicle and metronidazole cream 1% in the treatment of rosacea. J Am Acad Dermatol 2005; 52 (Suppl. 3):P10.
  • 20
    Bitar A, Bourgouin J, Doré N et al. A double-blind randomized study of metronidazole (Flagyl) 1% cream in the treatment of acne rosacea, a placebo controlled study. Drug Invest 1990; 2:2428.
  • 21
    Bjerke JR, Nyfors A, Austad J et al. Metronidazole (Elyzol) 1% cream vs. placebo cream in the treatment of rosacea. Clin Trials J 1989; 26:18794.
  • 22
    Breneman DL, Stewart D, Hevia O et al. A double-blind, multicenter clinical trial comparing efficacy of once-daily metronidazole 1 percent cream to vehicle in patients with rosacea. Cutis 1998; 61:447.
  • 23
    Dahl MV, Katz HI, Krueger GG et al. Topical metronidazole maintains remissions of rosacea. Arch Dermatol 1998; 134:67983.
  • 24
    Dahl MV, Jarratt MJ, Kaplan D et al. Once-daily topical metronidazole cream formulations in the treatment of papules and pustules of rosacea. J Am Acad Dermatol 2001; 45:72330.
  • 25
    Dreno B, Dubertret L, Naeyaert JM et al. Comparison of the clinical efficacy and safety of metronidazole 0·75% cream with metronidazole 0·75% gel in the treatment of rosacea. J Eur Acad Dermatol Venereol 1998; 11 (Suppl. 2):S2723.
  • 26
    Guillet B, Rostain E, Powell F et al. Metronidazole 0·75% gel and lotion are both effective in the treatment of rosacea. J Eur Acad Dermatol Venereol 1999; 12 (Suppl. 2):S145.
  • 27
    Jorizzo JL, Lebwohl M, Tobey RE. The efficacy of metronidazole 1% cream once daily compared with metronidazole 1% cream twice daily and their vehicles in rosacea: a double blind clinical trial. J Am Acad Dermatol 1998; 39:5024.
  • 28
    Koçak M, Yagli S, Vahapoglu G et al. Permethrin 5% cream versus metronidazole 0·75% gel for the treatment of papulopustular rosacea. Dermatology 2002; 205:26570.
  • 29
    Nielsen PG. Treatment of rosacea with 1% metronidazole cream. A double-blind study. Br J Dermatol 1983; 108:32732.
  • 30
    Tan JKL, Girard C, Krol A et al. Randomized placebo-controlled trial of metronidazole 1% cream with sunscreen SPF 15 in treatment of rosacea. J Cutan Med Surg 2002; 6:52934.
  • 31
    Bjerke R, Fyrand O, Graupe K. Double-blind comparison of azelaic acid 20% cream and its vehicle in treatment of papulo-pustular rosacea. Acta Derm Venerol 1999; 79:4569.
  • 32
    Carmichael AJ, Marks R, Graupe KA et al. Topical azelaic acid in the treatment of rosacea. J Dermatol Treat 1993; 4 (Suppl. 1):S1922.
  • 33
    Thiboutot D, Thieroff-Ekerdt R, Graupe K. Efficacy and safety of azelaic acid (15%) gel as a new treatment for papulopustular rosacea: results from 2 vehicle-controlled, randomized phase III studies. J Am Acad Dermatol 2003; 48:83645.
  • 34
    Thiboutot DM, Fleisher AB, Del Rosso JQ et al. Azelaic acid 15% gel once daily versus twice daily in papulopustular rosacea. J Drugs Dermatol 2008; 7:5416.
  • 35
    Thiboutot DM, Fleischer AB, Del Rosso JQ et al. A multicenter study of topical azelaic acid 15% gel in combination with oral doxycycline as initial therapy and azelaic acid 15% gel as maintenance therapy. J Drugs Dermatol 2009; 8:63948.
  • 36
    Elewski BE, Fleisher AB, Pariser DM. A comparison of 15% azelaic acid gel and 0·75% metronidazole gel in the topical treatment of papulopustular rosacea: results of a randomized trial. Arch Dermatol 2003; 139:144450.
  • 37
    Maddin S. A comparison of topical azelaic acid 20% cream and topical metronidazole 0·75% cream in the treatment of patients with papulopustular rosacea. J Am Acad Dermatol 1999; 40 (6 Pt 1):9615.
  • 38
    Wolf JE Jr, Kerrouche N, Arsonnaud S. Efficacy and safety of once-daily metronidazole 1% gel compared with twice-daily azelaic acid 15% gel in the treatment of rosacea. Cutis 2006; 77 (Suppl. 4):311.
  • 39
    Breneman D, Savin R, VandePol C et al. Double-blind, randomized, vehicle-controlled clinical trial of once-daily benzoyl peroxide/clindamycin topical gel in the treatment of moderate to severe rosacea. Int J Dermatol 2004; 43:3817.
  • 40
    Draelos ZD, Fuller BB. Efficacy of 1% 4-ethoxybenzaldehyde in reducing facial erythema. Dermatol Surg 2005; 31 (7 Pt 2):8815.
  • 41
    Draelos ZD, Green BA, Edison BL. An evaluation of a polyhydroxy acid skin care regimen in combination with azelaic acid 15% gel in rosacea patients. J Cosmet Dermatol 2006; 5:239.
  • 42
    Koca R, Altinyazar HC, Ankarali H et al. A comparison of metronidazole 1% cream and pimecrolimus 1% cream in the treatment of patients with papulopustular rosacea: a randomized open-label clinical trial. Clin Exp Dermatol 2010; 35:2516.
  • 43
    Lebwohl MG, Medansky RS, Russo CL et al. The comparative efficacy of sodium sulfacetamide 10%/sulfur 5% (Sulfacet-R) lotion and metronidazole 0·75% (Metrogel) in the treatment of rosacea. J Geriat Dermatol 1995; 3:1835.
  • 44
    Montes LF, Cordero AA, Kriner J et al. Topical treatment of acne rosacea with benzoyl peroxide acetone gel. Cutis 1983; 32:18590.
  • 45
    Mostafa FF, El Harras MA, Gomaa SM et al. Comparative study of some treatment modalities of rosacea. J Eur Acad Dermatol Venereol 2009; 23:228.
  • 46
    Rigopoulos D, Kalogeromitros D, Gregoriou S et al. Randomized placebo-controlled trial of a flavonoid-rich plant extract-based cream in the treatment of rosacea. J Eur Acad Dermatol Venereol 2005; 19:5648.
  • 47
    Sauder DN, Miller R, Gratton D et al. The treatment of rosacea: the safety and efficacy of sodium sulfacetamide 10% and sulfur 5% lotion (Novacet) is demonstrated in a double-blind study. J Dermatol Treat 1997; 8:7985.
  • 48
    Torok HM, Webster G, Dunlap FE et al. Combination sodium sulfacetamide 10% and sulfur 5% cream with sunscreens versus metronidazole 0.75% cream for rosacea. Cutis 2005; 75:35763.
  • 49
    Verea Hernando M, Margusino Framiñán L, Seco Vilariño C et al. Comparative study of topical erythromycin and topical metronidazole in the treatment of rosacea. Farm Clin 1992; 9:4729.
  • 50
    Marks R, Ellis J. Comparative effectiveness of tetracycline and ampicillin in rosacea: a controlled trial. Lancet 1971; 2:104952.
  • 51
    Sneddon IB. A clinical trial of tetracycline in rosacea. Br J Dermatol 1966; 78:64952.
  • 52
    Fowler JF Jr. Combined effect of anti-inflammatory dose doxycycline (40-mg doxycycline, usp monohydrate controlled-release capsules) and metronidazole topical gel 1% in the treatment of rosacea. J Drugs Dermatol 2007; 6:6415.
  • 53
    Del Rosso JQ, Schlessinger J, Werschler P. Comparison of anti-inflammatory dose doxycycline versus doxycycline 100 mg in the treatment of rosacea. J Drugs Dermatol 2008; 7:5736.
  • 54
    Bamford JT, Tilden RL, Blankush JL et al. Effect of treatment of Helicobacter pylori infection on rosacea. Arch Dermatol 1999; 135:65963.
  • 55
    Akhyani M, Ehsani AH, Ghiasi M et al. Comparison of efficacy of azithromycin vs. doxycycline in the treatment of rosacea: a randomized open clinical trial. Int J Dermatol 2008; 47:2848.
  • 56
    Pye RJ, Burton JL. Treatment of rosacea by metronidazole. Lancet 1976; 1:121112.
  • 57
    Saihan EM, Burton JL. A double-blind trial of metronidazole versus oxytetracycline therapy for rosacea. Br J Dermatol 1980; 102:4435.
  • 58
    Sanchez J, Somolinos AL, Almodóvar PI et al. A randomized, double-blind, placebo-controlled trial of the combined effect of doxycycline hyclate 20-mg tablets and metronidazole 0·75% topical lotion in the treatment of rosacea. J Am Acad Dermatol 2005; 53:7917.
  • 59
    Monk BE, Logan RA, Cook J et al. Topical metronidazole in the treatment of rosacea. J Dermatol Treat 1991; 2:913.
  • 60
    Nielsen PG. A double-blind study of 1% metronidazole cream versus systemic oxytetracycline therapy for rosacea. Br J Dermatol 1983; 109:635.
  • 61
    Schachter D, Schachter RK, Long B et al. Comparison of metronidazole 1% cream versus oral tetracycline in patients with rosacea. Drug Invest 1991; 3:2204.
  • 62
    Veien NK, Christiansen JV, Hjorth N et al. Topical metronidazole in the treatment of rosacea. Cutis 1986; 38:20910.
  • 63
    Grosshans E, Michel C, Arcade B et al. Rilmenidine in rosacea: a double blind study versus placebo. Ann Dermatol Venereol 1997; 124:68791.
  • 64
    Koch R, Wilbrand G. Dark sulfonated shale oil versus placebo in the systemic treatment of rosacea. J Eur Acad Dermatol Venerol 1999; 12 (Suppl. 32):S1434.
  • 65
    Sharquie KE, Najim RA, Al-Salman HN. Oral zinc sulfate in the treatment of rosacea: a double-blind, placebo-controlled study. Int J Dermatol 2006; 45:85761.
  • 66
    Wilkin JK. Effect of nadolol on flushing reactions in rosacea. J Am Acad Dermatol 1989; 20 (2 Pt 1):2025.
  • 67
    Wilkin JK, De Witt S. Treatment of rosacea: topical clindamycin versus oral tetracycline. Int J Dermatol 1993; 32:657.
  • 68
    Karsai S, Roos S, Raulin C. Treatment of facial telangiectasia using a dual-wavelength laser system (595 and 1,064 nm): a randomized controlled trial with blinded response evaluation. Dermatol Surg 2008; 34:7028.
  • 69
    Neuhaus IM, Zane LT, Tope WD. Comparative efficacy of nonpurpuragenic pulsed dye laser and intense pulsed light for erythematotelangiectatic rosacea. Dermatol Surg 2009; 35:9208.
  • 70
    Thiboutot DM. Acne and rosacea. New and emerging therapies. Dermatol Clin 2000; 18:6371.
  • 71
    Wilkin JK. Rosacea. Pathophysiology and treatment. Arch Dermatol 1994; 130:35962.