Using item banks to construct measures of patient reported outcomes in clinical trials: investigator perceptions
Using item banks to construct measures of patient reported outcomes in clinical trials: investigator perceptions
复制标题
DOI：
10.1177/1740774508098414
复制
发表时间：
2008-01-01
期刊：
影响因子：
2.7
通讯作者：
Weinfurt, Kevin P.
中科院分区：
文献类型：
--
作者：
Flynn, Kathryn E.;Dombeck, Carrie B.;Weinfurt, Kevin P.
背景项目反应理论(IRT)承诺比传统方法更敏感和有效地测量患者报告的结果(PRO)；然而，从基于IRT的题库中选择和使用PRO措施不同于当前使用PRO措施的方法。目的预测IRT题库进入临床试验的障碍。方法我们对42名临床研究人员进行了半结构化电话或面对面访谈，他们在2005年7月至2006年5月期间在《美国医学会杂志》、《新英格兰医学杂志》或其他领先临床杂志上发表了临床试验结果。结果42名受试者中有39人理解了IRT题库中的新产品，即定制的短表和计算机化的适应性测验。大多数参与者(38/42)认为题库在他们的临床试验中可能是有用的，但他们提到了几个潜在的采用障碍，包括经济和后勤限制，对题库是否比目前的PRO措施更好的担忧，对如何说服研究人员或统计学家使用题库的担忧，对FDA或赞助商接受度的担忧，以及缺乏在特定疾病人群中验证的题库的可用性。结论临床研究人员对IRT题库提供的一种新的PRO测量方法持开放态度，但在被广泛采用之前，银行开发者必须解决研究人员和利益相关者的担忧。临床试验2008；5：575-586。Http://cti.sagepub.com
Background Item response theory (IRT) promises more sensitive and efficient measurement of patient-reported outcomes (PROs) than traditional approaches; however, the selection and use of PRO measures from IRT-based item banks differ from current methods of using PRO measures.Purpose To anticipate barriers to the adoption of IRT item banks into clinical trials.Methods We conducted semistructured telephone or in-person interviews with 42 clinical researchers who published results from clinical trials in the journal of the American Medical Association, the New England Journal of Medicine, or other leading clinical journals from July 2005 through May 2006. Interviews included a brief tutorial on IRT item banks.Results After the tutorial, 39 of 42 participants understood the novel products available from an IRT item bank, namely customized short forms and computerized adaptive testing. Most participants (38/42) thought that item banks could be useful in their clinical trials, but they mentioned several potential barriers to adoption, including economic and logistical constraints, concerns about whether item banks are better than current PRO measures, concerns about how to convince study personnel or statisticians to use item banks, concerns about FDA or sponsor acceptance, and the lack of availability of item banks validated in specific disease populations. Limitations Selection bias might have led to more positive responses to the concept of item banks in clinical trials.Conclusions Clinical investigators are open to a new method of PRO measurement offered in IRT item banks, but bank developers must address investigator and stakeholder concerns before widespread adoption can be expected. Clinical Trials 2008; 5: 575-586. http://cti.sagepub.com