Department of Anesthesiology, Duke University Medical Center, 315 Trent Dr., Hanes House Rm. 285, Durham, NC 27710, USA.
Anesth Analg. 2013 Feb;116(2):448-54. doi: 10.1213/ANE.0b013e318277dd7d. Epub 2013 Jan 9.
A top priority for research studies is to ensure that potential participants receive adequate information to make a truly informed decision. Understanding patient experiences with the recruitment process may identify areas for improvement in the consent process. We examined which factors were associated with the decision to consent in a clinical research study.
Patients scheduled for elective surgery were asked to complete a questionnaire about the consent process, immediately after being approached to participate in an anesthesia-related research study. Sociodemographic characteristics, preoperative levels of anxiety and depression, medical comorbidities, factors that may affect decision to participate in a research study, and study design features were collected. A multivariable logistic regression model was estimated to identify factors associated with providing consent. Performance of the prediction model was assessed using the receiver operating characteristic curve. Internal validity was assessed by a bootstrap analysis.
In all, 282 participants completed the questionnaire. Of those, 179 (63%) had consented to participate in research, and 103 (37%) had declined to participate. In the multivariable logistic regression model, the odds of providing consent were higher for males (odds ratio [OR] [95% confidence interval]=2.49 [1.29-4.79]) and for patients with higher levels of patient comfort (OR=1.84 [1.22-2.78]). The odds of providing consent were lower for protocols that require additional testing (OR=0.15 [0.06-0.39]) and patients with higher levels of concern about blood sampling (OR=0.70 [0.54-0.90]) or worry about study risks (OR=0.72 [0.55-0.95]). Bootstrap analysis revealed a stable model with high internal validity.
The 2 strongest predictors of consent were male gender and comfort; predictors of refusal were protocol type that requires additional testing, greater concern about blood sampling and study risks, and lower overall patient comfort with the study. These patient and study characteristics may inform modification of the consent process for clinical research studies and facilitate the development of more accurate enrollment projections and strategies.
研究的首要任务是确保潜在参与者获得足够的信息，以便做出真正知情的决定。了解患者对招募过程的体验可能会发现同意过程中需要改进的地方。我们研究了哪些因素与临床研究中同意的决定有关。
在安排择期手术的患者中，在被邀请参加与麻醉相关的研究时，要求他们完成一份关于同意过程的问卷。收集社会人口统计学特征、术前焦虑和抑郁水平、合并症、可能影响参与研究决定的因素以及研究设计特征。使用多变量逻辑回归模型来确定与同意相关的因素。使用接收者操作特征曲线评估预测模型的性能。通过自举分析评估内部有效性。
共有 282 名参与者完成了问卷。其中，179 名（63%）同意参与研究，103 名（37%）拒绝参与。在多变量逻辑回归模型中，男性同意的可能性更高（比值比[OR] [95%置信区间]=2.49 [1.29-4.79]），对患者舒适度较高的患者也是如此（OR=1.84 [1.22-2.78]）。需要额外测试的方案（OR=0.15 [0.06-0.39]）和对采血的担忧程度较高的患者（OR=0.70 [0.54-0.90]）或对研究风险的担忧程度较高的患者（OR=0.72 [0.55-0.95]）同意的可能性较低。自举分析显示模型具有较高的内部有效性和稳定性。
同意的两个最强预测因素是男性和舒适度；拒绝的预测因素是需要额外测试的方案类型、对采血和研究风险的担忧程度更高，以及患者对研究的整体舒适度降低。这些患者和研究特征可能会影响临床研究同意过程的修改，并有助于制定更准确的入组预测和策略。
