Background: Asthma is a high-prevalent capricious disease with multiple triggering factors and frequent medical admissions. Steroids are the first line treatment, but cause important adverse effects. The American National Heart Lung Blood Institute, together with WHO, promoted a global plan of International Guidelines (IG) supporting steroids use to control asthma. Selective disclosure of the results of Cochrane Reviews (CR) appeared in IG. Objectives: Balancing risks and benefits of different steroids routes to control asthma using a meta-ethnography synthesis to state third order interpretations for future independent research. Methods: Meta-ethnography is an interpretive method clarifying concepts, expressing synthesis and lighting new hypotheses to policy-makers. We searched, matched and categorized concepts and outcomes from CR and IG, focusing on different steroid routes for treating asthma, and synthesizing third order interpretations in the attempt to state hypotheses, not elicited in original papers. Results: 32 CR and 7 IG were retrieved and categorized. The concept of risk showed conflicting understanding. In IG meant exacerbation requiring corticosteroids while in CR it referred to adverse event with steroid use. In IG the evidence was limited to small, single-centre studies yielding recommendations based on extrapolation from trials of add-on therapy and hiding adverse effects. In CR evidence showed similar net benefit with different steroid routes or doses. These findings, in IG, guided to mismatches of benefits, including earlier discharge and fewer relapses, far from being balanced with the evidence of adverse events. Third order interpretations showed reducing height in children with inhaled steroids, incompletely reported in IG. Cooperative management with families and steroid-sparing drugs, assessed in dated IG, reappeared in third order interpretations. Conclusions: Bewildering conflict of interests and low quality exists in authoritative IG showing selective disclosure of recommendations derived from CR. Third order interpretations elicited innovative approaches stifled in IG and stimulate testable hypotheses for future CR.