The article discusses the detection of bias in meta-analyses using a graphical test proposed by Egger et al. The authors of the meta-analysis of inpatient geriatric consultations argue that the bias observed was not due to retrospective detection but to true heterogeneity, which they used to identify program elements that might have influenced outcomes. They emphasize that meta-analytical methods should not be solely relied upon to determine the quality of meta-analyses, especially for complex interventions. The authors also highlight the importance of considering methodological issues unique to each trial, particularly in complex interventions. They critique the mega-trial used by Egger et al., noting its differences from the trials in their meta-analysis and the potential for unreported variability. The article concludes by emphasizing the need for expert interpretation of meta-analysis results and the importance of addressing heterogeneity.The article discusses the detection of bias in meta-analyses using a graphical test proposed by Egger et al. The authors of the meta-analysis of inpatient geriatric consultations argue that the bias observed was not due to retrospective detection but to true heterogeneity, which they used to identify program elements that might have influenced outcomes. They emphasize that meta-analytical methods should not be solely relied upon to determine the quality of meta-analyses, especially for complex interventions. The authors also highlight the importance of considering methodological issues unique to each trial, particularly in complex interventions. They critique the mega-trial used by Egger et al., noting its differences from the trials in their meta-analysis and the potential for unreported variability. The article concludes by emphasizing the need for expert interpretation of meta-analysis results and the importance of addressing heterogeneity.