Copyright 2008 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2008
We would like to raise the following issues regarding the interesting article by Hollander et al.1 The high dropout rates (176 of the 630 randomized subjects) reported in the study may have reduced power and comparability and thus limit the validity of the results.2 Furthermore, including sample size estimation in the manuscript would facilitate assessment of the impact of follow-up losses on the study results.
Lopez CA, Kanna B. Low-Dose Rosiglitazone in Patients With Type 2 Diabetes Mellitus Requiring Insulin Therapy. Arch Intern Med. 2008;168(5):548. doi:10.1001/archinte.168.5.548