Article type
Year
Abstract
Background: Because of the huge amount of medical literature the importance of review articles for clinical decision making increases. Methodological flaws often affect the validity of reviews. One major problem is the identification of all relevant trials. In the project presented here, we analysed how far the database for paediatric review articles can be improved by handsearching German paediatric journals. Objectives: How many controlled trials are published in the three of the most relevant German paediatric journals in the past 30 years? What are the characteristics of these trials concerning their methodological quality and their representation in Medline?
Methods: Three journals (Kinderarzt, Monatszeitschrift Kinderheilkunde, P diatrische Praxis) were handsearched for controlled trials between 1970 and 1998. The representation of the identified trials in Medline was checked and the methodological quality of the trials was assessed with respect to randomisation, blinding and controlling for drop-outs.
Results: "P diatrische Praxis" (1970-1998), not indexed in Medline, contained 8 controlled trials, 4 were randomised. None of them were found by a Medline search. "Kinderarzt" (1970 - 1998), not indexed in Medline, contained 39 controlled trials, 20 were randomised. None of them were found by a Medline search. "Monatszeitschrift Kinderheilkunde" (1970 - 1998), indexed in Medline, contained 319 controlled trials, 136 were randomised. Only 22 of the controlled and 14 of the randomised controlled trials werde found by a Medline search. All together we found 366 controlled clinical trials by handsearching, 160 of them were randomised. Only 22 of all the controlled clinical trials and 14 of all the randomised trials were found in Medline. The majority of the identified trials was of medium quality.
Conclusions: A high number of controlled clinical trials published in German paediatric journals is not represented in Medline. Thus they are not easily accesible, neither for individual clinicians nor for persons working on systematic reviews in the field of paediatrics. We conclude that handsearching in German paediatric journals helps to avoid biases in paediatric systematic reviews by completion of the database.
Methods: Three journals (Kinderarzt, Monatszeitschrift Kinderheilkunde, P diatrische Praxis) were handsearched for controlled trials between 1970 and 1998. The representation of the identified trials in Medline was checked and the methodological quality of the trials was assessed with respect to randomisation, blinding and controlling for drop-outs.
Results: "P diatrische Praxis" (1970-1998), not indexed in Medline, contained 8 controlled trials, 4 were randomised. None of them were found by a Medline search. "Kinderarzt" (1970 - 1998), not indexed in Medline, contained 39 controlled trials, 20 were randomised. None of them were found by a Medline search. "Monatszeitschrift Kinderheilkunde" (1970 - 1998), indexed in Medline, contained 319 controlled trials, 136 were randomised. Only 22 of the controlled and 14 of the randomised controlled trials werde found by a Medline search. All together we found 366 controlled clinical trials by handsearching, 160 of them were randomised. Only 22 of all the controlled clinical trials and 14 of all the randomised trials were found in Medline. The majority of the identified trials was of medium quality.
Conclusions: A high number of controlled clinical trials published in German paediatric journals is not represented in Medline. Thus they are not easily accesible, neither for individual clinicians nor for persons working on systematic reviews in the field of paediatrics. We conclude that handsearching in German paediatric journals helps to avoid biases in paediatric systematic reviews by completion of the database.