Aim: To assess the agreement between Cochrane Neonatal Group reviews and clinical guidelines of a University Neonatology Department, to evaluate the reasons for potential disagreements and to ascertain whether Cochrane reviews were considered for the guidelines development.
Methods: The recommendations in the reviews and guidelines were compared and classified as being in 'agreement', 'partial agreement' or 'disagreement'. The guideline authors were interviewed for reasons about disagreement and whether Cochrane reviews were considered during the guideline development.
Results: Agreement between reviews and guidelines was found for 133 interventions (77%), partial agreement for 31 interventions (18%) and disagreement for nine interventions (5%). Six interventions were recommended in the guidelines, but not in the reviews. Three interventions were recommended in the reviews, but not in the guidelines. Use of consensus statements, evidence on surrogate markers, observational studies, basic immunology and pathophysiological knowledge, expert opinion, economical constraints, reservations about the external validity and unawareness of reviews were reasons for disagreement. Cochrane reviews were rarely (22%) used during the guideline development.
Conclusion: We found agreement between more than three quarters of Cochrane reviews and neonatal guidelines. However, few important disagreements occurred. Reviews were only used for guideline development in about a fifth of cases.