Table_4.xls (25.5 kB)
Download file

Adverse events individually and by category.

Download (0 kB)
posted on 2013-03-12, 02:05 authored by Glen I. Spielmans, Margit I. Berman, Eftihia Linardatos, Nicholas Z. Rosenlicht, Angela Perry, Alexander C. Tsai

Trials with no events in either study arm are not included in summary OR calculations.


The clinical registry report indicated that statistically significant differences emerged between drug and placebo in glucose, total cholesterol, fasting LDL cholesterol, nonfasting and fasting triglycerides, and prolactin. These differences were not reported quantitatively and were described as not “clinically meaningful.”


Median levels of change in fasting total cholesterol, triglycerides, HDL cholesterol, LDL cholesterol, and fasting plasma glucose were reported. Categorical measures (i.e., numbers of patients who had abnormal scores) were not reported. The clinical trial registry noted that there was a statistically significant but clinically nonmeaningful difference between drug and placebo on nonfasting LDL cholesterol.


Data on metabolic parameters were reported in terms of median change, but no categorical reporting of laboratory abnormalities was provided. Differences between drug and placebo were reported as not statistically significant in terms of median change on glucose, cholesterol, and triglycerides.


Because the total number of events in the OFC group was higher than the sample size of the group, an effect size could not be calculated, and it was thus not factored into the overall effect size estimate for sedation. Given the very small sample of the study, this makes virtually no difference in the overall effect size estimate.


The number of participants providing data differed substantially across metabolic testing parameters. The average sample size across the metabolic testing groups provided the denominator for the pooled number of abnormal metabolic test results in each trial, with the total number of participants who experienced an abnormal metabolic testing result comprising the numerator. A participant may have experienced more than one event. Also, boundaries of abnormal tests were defined by standard Lilly reference ranges, a resource not available to our research team.


Boundaries of abnormal tests were defined by standard Lilly reference ranges, a resource not available to our research team.


Triglycerides and unclearly described laboratory tests were completed in this study, but the results were described only as yielding “no clinically significant differences” between groups.


Abnormal metabolic laboratory values were defined as follows: fasting glucose ≥126 mg/dl, LDL cholesterol ≥160 mg/dl, HDL cholesterol ≤40 mg/dl, total cholesterol ≥240 mg/dl, and triglycerides ≥200 mg/dl.


The clinical trial registry entry noted that approximately 17% of patients taking quetiapine met this criterion, compared to 6% of placebo patients. We extracted numbers of patients based on these percentages.


Defined as >20 ng/ml for males and >30 ng/ml for females.


These parameters were apparently not measured.


Weight gain was provided in terms of means and standard deviations; however, no categorical measure of significant weight gain was reported.

AIMS, Abnormal Involuntary Movement Scale; Barnes, Barnes Akathisia Scale; HbA1c, glycated hemoglobin; SAS, Simpson-Angus Scale.