RESUMO
OBJECTIVE: To verify the usefulness and reliability of a software tool we developed to help to apply the American Academy of Pediatrics (AAP) Guidelines 2004 on hyperbilirubinaemia according to the infant's age in hours and their clinical risk factors. DESIGN: Randomised, cross-over, controlled trial with 20 simulated clinical cases comparing the "manual" application of the guidelines with our software application. PARTICIPANTS: Fifteen doctors (eight final-year residents and seven consultants) from two hospitals in Spain. MAIN OUTCOME MEASURES: Major errors (defined a priori as any deviation from the AAP guidelines that involve a risk of morbidity or mortality for the patient), minor errors (those that cause discomfort and/or, in extremely rare cases, morbidity) and time spent. RESULTS: Fifteen doctors each managed 20 simulated cases, half by using the guidelines alone and half using the software tool. Without the software application, 42 "minor" errors were made. With it, only 25 errors were made. "Major" errors also decreased from 10 to 2 with the software. As a group, the residents benefited most; they made an average of 1.8 errors fewer per 10 cases. Use of the software reduced the time taken by the residents to resolve the cases, although the mean reduction in time was not significant for the group of consultants. CONCLUSIONS: The use of simulated clinical cases revealed many errors in the routine management of hyperbilirubinaemia. The software helped clinicians make fewer errors and saved time for residents, but not consultants.