A Study of Sources and Types of Queries in Drug Clinical Trials

Yinhua Yu,Zhu Sainan,Xiaoyan Yan,Yao Chen
2011-01-01
Abstract:Objective:To reveal the status of the current data quality of clinical trials by the investigation of the amount and class of queries,so as to provide a reference to investigator and CRA on their data collection and monitoring works.Methods: Query databases of 41 phaseⅡand phaseⅢ clinical trials were collected,together with the basic information of these trials such as number of subjects,centers,etc.Then all of these query databases were joined together and analysis was conducted after coding which classified the source and type of queries.Results: The data quality is quite similar not only between different phase,but between Chinese and Western medicine.Among all the CRF models,library test is the largest source of query,which made up 35.67% of all the quires,followed by time date data(15.16%) and efficiency data(10.13%).Study summary page,time date data and treatment compliance data have the highest query index(21.83,17.62 and 16.80 query/1000 data points).The type of queries is depended on which part of CRF they were raised: missing data is the most common type of queries;extreme value is often found in quantitative data;mistake would happen when some record on CRF had to be computed by investigator;data incomplete is special problem of the records of medical history,combination drugs and adverse event.The different in the query constitution between Chinese and Western trials is focus on efficiency data,which has a proportion of 6.70% in Western medicine trials and 17.42% in Chinese medicine trials.Conclusion: Characters of different CRF model and frequency of kinds of queries provides a good reference to trial monitoring and auditing.Time Date data,which is not only more likely to raise queries but also have a large amount,deserve the most attention during the data capture and monitoring procedure.And special attention also should be paid to laboratory data since query raised from these parts contribute a large proportion.Furthermore,data in study summary page and treatment compliance data are also easy to raise problems during the data capture procedure.The difference in constitution of queries is result from the difference of trial design and conducting.
What problem does this paper attempt to address?