[Differences in the quality of Spanish clinical trials published in international periodicals and of the ones presented in general medicine periodicals with wide readership].

BACKGROUND

The assessment of the methodological quality of controlled clinical trials (CCTs) carried out in Spain and published in international journals and the comparison with those published in widely read general journals of medicine (N Engl J Med, Lancet, JAMA, BMJ) may help establishing their limitations and improving quality in future studies.

METHODS

Search in Medline of CCTs. Studies were evaluated according to a structured questionnaire (J Clin Epidemiol 1992; 45:225-265). The odds ratio and their 95% confidence intervals were used to compare spanish studies with the remaining studies. The following variables were considered as confounders and controlled for by logistic regression analysis: number of participating centers, sample size, funding source, and the inclusion of epidemiologists in the research team. The journal's impact factor was taken into account.

RESULTS

Spanish CTTs had a smaller sample size, were mostly monocentric, reported les frequently the source of funding, and exhibited a lower participation of epidemiologists. The informed consent and the approval by the ethics committee were more frequently omitted. The major methodological differences with the other studies were: lack of pre-study sample size and statistical power estimation, lack of inclusion criteria, poor explanation of patients' flow in the selection process, lack of explanation for unmasked procedures, poor description of methods to assess intervention compliance, under-reporting confidence intervals, and less frequent use of the intention-to-treat principle. The quality score of spanish studies was lower (9.4 +/- 1.7 vs 10.7 +/- 1.5; p < 0.001). These limitations improved with higher journal's impact factor, the quality score was 8.5 in journals with an impact factor < 1.5, and 10.6 in those with an impact factor > 4.5. This last figure is almost identical to the average of CCTs published in general journals of medicine.

CONCLUSIONS

Most spanish studies achieved a good quality score. Nevertheless, there are still issues which can be overcome easily. If this goal is achieved, the results will obtain a higher repercussion, as a better quality is observed in those published in journals with the highest impact factor.