Oral flucloxacillin is less effective than local antibiotics for impetigo in limited disease (level of evidence [LOE] 1a).
Topical metronidazole and azelaic acid are effective for rosacea (LOE 1a).
Betadine is effective for minor infections following partial thickness burns (LOE 1b).
Terbinafine is effective against fungal infections of the nail (LOE 1a).
Miconazole is effective against oral thrush (LOE 1a).
Level of evidence (LOE)
1a: Systematic reviews (with homogeneity) of randomized controlled trials (RCTs). 1a-: Systematic review of randomized trials displaying worrisome heterogeneity. 1b: Individual RCT (with a narrow confidence interval). 1b-: Individual RCT (with a wide confidence interval). 1c: All or none RCTs. 2a: Systematic reviews (with homogeneity) of cohort studies. 2a-: Systematic reviews of cohort studies displaying worrisome heterogeneity. 2b: Individual cohort study or low-quality RCTs (<80% follow-up). 2b-: Individual cohort study or low-quality RCTs (<80% follow-up/wide confidence interval). 2c: “outcomes” research; ecological studies. 3a: Systematic review (with homogeneity) of case-control studies. 3a-: Systematic review of case-control studies with worrisome heterogeneity. 3b: Individual case-control study. 4: Case series (and poor-quality cohort and case-control studies). 5: Expert opinion without explicit critical appraisal, or based on physiology, bench research, or “first principles.”
Source: Essential Evidence Plus. Levels of evidence.1
Do you use silver sulfadiazine for partial-thickness burns? If you do, you may be surprised to learn that the evidence for its use in this situation is conflicting. This was just one of the findings of our systematic review of the methodologic quality and statistical and clinical relevance of current therapies for minor dermatologic ailments.
Given that minor ailments, frequently dermatologic, account for 40% to 70% of all consultations in family medicine,2,3 guidelines based on better research are needed. This need is underscored by the increasing delegation of minor treatments to staff nurses, nurse practitioners, and physician assistants, who should undergo comprehensive training, preferably based on valid guidelines.4,5 Moreover, consultations for prevalent minor ailments often lead to prescriptions for medications, thereby generating considerable costs.6,7
Methods
The starting point for this review was the textbook, Minor Ailments in Primary Care: An Evidence-Based Approach,6 which describes 119 minor ailments, selected mainly on the basis of disease prevalence. We selected all dermatologic ailments (International Classification of Primary Care-code ‘S’) (N=42) (TABLE).5
We searched the online databases PubMed, Cochrane Controlled Trials Register, and Clinical Evidence for articles relating to the treatment of these conditions. For each ailment, we used various search terms for indication and treatment.8 (See note at end of Methods section.) We excluded alternative (nonallopathic) and most preventive therapies because they are unusual in the daily practice of family medicine.
We searched only for trials in which treatments were compared with placebo or a reasonable, accepted usual therapy. The search followed a hierarchy of evidence:8 systematic reviews (SRs), then randomized controlled trials (RCTs), then other research articles (nonrandomized clinical trials, case series). When we found a relevant SR published in 2004 or later, we did not search for a lower level of evidence (LOE). Instead, we restricted our subsequent search to RCTs published after the publication date of the SR.8 Two of the authors (SPG and JAHE) selected articles independently, based on article title and abstract. Disagreements in selection were discussed and consensus was reached. If an article contained relevant first-line therapy, we also used the “related articles” option in PubMed to check for more sources. (See note at end of Methods section.)
To evaluate the methodologic quality of SRs and trials, we ranked articles according to the method of infoPOEMs.8 (See key.) Two experienced researchers (JAHE and AKN) scored all articles independently. Consensus was reached in cases of disagreement.9 We deemed evidence convincing if the study showed the intervention was effective and if the LOE of the study was high (levels 1a, 1b, or 2a).
Evaluating breadth of treatment application. To explore whether a treatment for a certain minor ailment could be applied to other ailments with similar symptoms and thus increase the strength of the treatment’s rationale, we clustered ailments, where possible, into bacterial infection, fungal infection, itch, and pain.
We classified the efficacy of therapies as yes, likely (if the result was not convincingly effective or based on small studies, or if the study objective was unclear), or no. Treatments with no trials to support them are so identified. As to whether the evidence was convincing, we indicated yes, no, or conflicting.
Post hoc analysis. For trials with a wide confidence interval and for therapies described as not clearly effective, we performed a post hoc power analysis to explore if the trial was underpowered.10 We compared the number of subjects in the study (n1) with the number we calculated as necessary for the study to have sufficient power (n2). For all studies, we used standardized values (α=0.05 and β=0.20). If n1≥n2 we considered the study design accurate, and if n1< n2 we concluded that the power was insufficient for the study to be able to answer its objectives.
