Mészáros Agnes, Czobor Pál, Bálint Sára, Komlósi Sarolta, Simon Viktória, Bitter István
Semmelweis University, Department of Psychiatry and Psychotherapy, Budapest, Hungary.
Int J Neuropsychopharmacol. 2009 Sep;12(8):1137-47. doi: 10.1017/S1461145709990198. Epub 2009 Jul 7.
Our objective was to conduct a meta-analysis of therapeutic efficacy of pharmacological treatment of adult ADHD based on data from controlled clinical trials. We used the search engines PubMed and Medline to identify relevant clinical trials. Short-term studies with double-blind parallel-group design were selected for the analysis. Altogether, we identified 11 trials that met the criteria, and investigated a total of 1991 subjects, 694 and 1297 of whom were treated with placebo or active medication, respectively. In order to pool efficacy data from studies with different characteristics, including different number of participants, different trial duration and measures of efficacy, the statistical effect sizes for each study had to be calculated. Our findings showed that the pooled effect size across all treatments was in the medium-to-high range (Cohen's d=0.65, p<0.0001 vs. placebo), and the effect size for stimulants (Cohen's d=0.67, p<0.0001 vs. placebo) was somewhat higher than for non-stimulant medications (Cohen's d=0.59, p<0.0001 vs. placebo). The current database of controlled trials for adult ADHD is relatively small, and does not include data for many of the potentially important agents. In addition, effect-size estimates for different classes of medications (i.e. stimulant and non-stimulant medications) were based on separate studies; head-to-head comparisons of various agents are severely lacking. Nonetheless, results of this meta-analysis across all ADHD medications in adult subjects demonstrated statistically significant and clinically robust improvement in symptom severity compared to placebo treatment.
我们的目标是基于对照临床试验的数据，对成人注意力缺陷多动障碍（ADHD）药物治疗的疗效进行荟萃分析。我们使用搜索引擎PubMed和Medline来识别相关的临床试验。选择采用双盲平行组设计的短期研究进行分析。我们总共确定了11项符合标准的试验，共调查了1991名受试者，其中694名和1297名分别接受了安慰剂或活性药物治疗。为了汇总来自具有不同特征（包括不同参与者数量、不同试验持续时间和疗效测量方法）的研究的疗效数据，必须计算每项研究的统计效应量。我们的研究结果表明，所有治疗方法的汇总效应量处于中高范围（Cohen's d = 0.65，与安慰剂相比，p < 0.0001），兴奋剂的效应量（Cohen's d = 0.67，与安慰剂相比，p < 0.0001）略高于非兴奋剂药物（Cohen's d = 0.59，与安慰剂相比，p < 0.0001）。目前成人ADHD对照试验的数据库相对较小，并且不包括许多潜在重要药物的数据。此外，不同类别药物（即兴奋剂和非兴奋剂药物）的效应量估计是基于单独的研究；各种药物的直接比较严重缺乏。尽管如此，这项针对成人受试者所有ADHD药物的荟萃分析结果表明，与安慰剂治疗相比，症状严重程度在统计学上有显著改善且在临床上有明显改善。
