RT Journal
T1 Error in data extraction and analyses in: Treatment-related mortality with bevacizumab in cancer patients: a meta-analysis
JF JAMA
JO JAMA
YR 2011
FD June 8
VO 305
IS 22
SP 2294
OP 2294
DO 10.1001/jama.2011.763
UL http://dx.doi.org/10.1001/jama.2011.763
AB 
Error in Data Extraction and Analyses:  In the Clinical Review entitled “Treatment-Related Mortality With Bevacizumab in Cancer Patients: A Meta-analysis,” published in the February 2, 2011, issue of JAMA (2011;305[5]:487-494), data were inappropriately extracted from specified fatal adverse events instead of total fatal adverse events for the studies by Hurwitz et al18 and Escudier et al.29 An error was also made in swapping sample sizes between the bevacizumab group and the control group in the study by Saltz et al.20 These errors affect incidence rates and associated relative risks, confidence intervals, and P values in the abstract, text, Tables 2 and 3, and Figure 2 of the article. Overall, the final results did not change substantially and remain consistent with the conclusions of the article. The article was corrected online and an accompanying letter describes these errors fully.