Doyle Orla, McGlanaghy Edel, O'Farrelly Christine, Tremblay Richard E
UCD School of Economics and UCD Geary Institute for Public Policy, University College Dublin, Ireland.
Clinical Psychology, School of Health in Social Science, University of Edinburgh, Edinburgh, United Kingdom.
PLoS One. 2016 Jun 2;11(6):e0156397. doi: 10.1371/journal.pone.0156397. eCollection 2016.
This study examined the impact of a targeted Irish early intervention program on children's emotional and behavioral development using multiple methods to test the robustness of the results. Data on 164 Preparing for Life participants who were randomly assigned into an intervention group, involving home visits from pregnancy onwards, or a control group, was used to test the impact of the intervention on Child Behavior Checklist scores at 24-months. Using inverse probability weighting to account for differential attrition, permutation testing to address small sample size, and quantile regression to characterize the distributional impact of the intervention, we found that the few treatment effects were largely concentrated among boys most at risk of developing emotional and behavioral problems. The average treatment effect identified a 13% reduction in the likelihood of falling into the borderline clinical threshold for Total Problems. The interaction and subgroup analysis found that this main effect was driven by boys. The distributional analysis identified a 10-point reduction in the Externalizing Problems score for boys at the 90th percentile. No effects were observed for girls or for the continuous measures of Total, Internalizing, and Externalizing problems. These findings suggest that the impact of this prenatally commencing home visiting program may be limited to boys experiencing the most difficulties. Further adoption of the statistical methods applied here may help to improve the internal validity of randomized controlled trials and contribute to the field of evaluation science more generally.
ISRCTN Registry ISRCTN04631728.
本研究使用多种方法检验结果的稳健性，考察了一项有针对性的爱尔兰早期干预项目对儿童情绪和行为发展的影响。对164名参与“为生活做准备”项目的参与者的数据进行了分析，这些参与者被随机分配到干预组（从孕期开始进行家访）或对照组，以检验干预对24个月时儿童行为检查表得分的影响。我们使用逆概率加权法来处理不同的损耗，使用置换检验来解决小样本量问题，并使用分位数回归来描述干预的分布影响。结果发现，少数治疗效果主要集中在最有可能出现情绪和行为问题的男孩中。平均治疗效果显示，进入总问题临界临床阈值的可能性降低了13%。交互作用和亚组分析发现，这一主要效果是由男孩驱动的。分布分析表明，处于第90百分位的男孩外化问题得分降低了10分。未观察到对女孩或总问题、内化问题和外化问题的连续测量指标有影响。这些发现表明，这个从产前开始的家访项目的影响可能仅限于遇到最多困难的男孩。进一步采用此处应用的统计方法可能有助于提高随机对照试验的内部效度，并更广泛地为评估科学领域做出贡献。
ISRCTN注册库ISRCTN04631728。
