A retrospective survey of quality of reporting on randomized controlled trials of metformin for polycystic ovary syndrome
Department of Traditional Chinese Medicine, The First People’s Hospital of Shunde, Foshan, PR China
Trials 2014, 15:128 doi:10.1186/1745-6215-15-128Published: 17 April 2014
From previous reviews, there still have been controversies over the effect of metformin (MET) on reproductive function in PCOS patients. The reasons for the inconsistent findings especially lie in the transparency and accuracy of randomized controlled trials (RCTs) reports. However, we could find no data about the quality of RCTs reporting in MET for PCOS. Thus, a retrospective survey related to the quality of reporting in MET for PCOS was conducted.
A retrospective survey was conducted by two investigators. Two investigators assessed the quality of overall reporting and key methodological factors reporting using items from the CONSORT 2010 statement.
A total of 39 RCTs were included in full text. The median overall quality score was 9, with a minimum of 2 and a maximum of 13. Good or general reporting existed in 11 items with positive rate of more than or equal to 50%. The median score of key methodological items was 4 with a minimum of 0 and a maximum of 5. Randomization, allocation concealment, blinding, baseline characteristics and intention-to-treat (ITT) analysis were reported in 26 (67%), 19 (49%), 20 (51%), 38 (97%) and 17 (44%) of the 39 RCTs, respectively. After adjustment, the mean overall score increased by about 1.71 for manuscripts with funding source (95% CI, 0.18 to 3.24), while it increased by about 3.51 for manuscripts published in one year increment (95% CI, 1.82 to 5.19). There was a relatively close, significant correlation (r = 0.589, P < 0.001) between the score of overall reporting quality and year of publication.
Although the overall reporting quality of RCTs in MET for PCOS has improved over time, reporting of key methodological items remains poor. Reporting of RCTs on MET for PCOS should keep up with the standards of the CONSORT statement.