The Optimization Of Edit Checks & Derivations Specification Design In Datalabs System

Clinical Data Management (CDM) is a profession with increasing importance in pharmaceutical research and development. The primary objective of Clinical Data Management (CDM) is to ensure timely delivery of high-quality data which are necessary to satisfy both good clinical practice (GCP) requirements and the statistical analysis and reporting requirements. Data quality is directly related to that whether we can make a proper evaluation about the safety and effectiveness of drugs. Data cleaning refers to a collection of activities used to assure the completeness, validity and accuracy of data. Data cleaning activities may include manual reviews of data; Programmed edit checks that identify inaccurate or invalid data using ranges, missing data, protocol violations and consistency checks; or aggregate descriptive statistics that reveal unusual patterns in data. Programmed edit checks and manual review are two main methods of Data Validation which should ensure that the database is accurate, consistent and a true representation of the patient’s profile. This paper, mainly focusing on the Edit Checks & Derivations Specification (EDS), discusses how to design a more user-friendly EDS template matched with DataLabs system; The purpose is to ensure the data quality, as well as improve work efficiency of Data Managers.Objective:This paper describes the clinical data management requirements as well as one data validation method-Programmed edit checks, and discusses the disadvantages of traditional EDS template for OC system while being used in the new DataLabs system; The objectives of this paper are to design a new EDS template based on DataLabs system, to compare it with OC EDS template by collecting the time of creating EDS using these two different templates, to explore potential application possibility of DataLabs EDS template, and to provide the basis for future application.Method:(1) To design a new EDS template based on the characteristics of DataLabs system through literature search and clinical data management experience in a large pharmaceutical company, especially processes and methods of EDS design.;(2) The two evaluation indicators are defined as time for creating EDS and quality of EDS. The data is collected from 18 Data Managers with similar experiences of creating EDS, who are divided into two groups. Based on 8 core modules which will be used in most studies, one group uses DataLabs EDS template, and the other group uses OC EDS template to create EDS for 9 DataLabs studies;(3) To choose time for creating EDS using two templates as the primary variable, Two Sample t Test statistical model is chosen to compare the OC EDS template and DataLabs EDS template. The nonparametric test-Wilcoxon test is also used to analyze the modules which do not meet the requirements of Two Sample t Test.Result:(1) Comparing the time difference of using two different templates, the average time of DataLabs EDS template (114.3 minutes) is shorter than that of OC EDS template (308.2 minutes), the total time for creating EDS of 8 core modules show normal distribution and homogeneity of variance, and P-value of Two Sample t test< 0.0001. Five Modules including Vital Signs, Medical History, Adverse Event, Medication Error, Subject Summary show normal distribution and homogeneity of variance, the P values of t test are all< 0.05; For other three modules(Demography, Concomitant Drug, Concomitant Non-drug Treatment), though they does not show homogeneity of variance, the P-values of nonparametric test-Wilcoxon test analysis also are all< 0.05.(2) To compare the quality of EDS created based on the two different templates by doing the reconciliation between them and the EDS applied in real clinical trial projects, the average match rates are calculated for eight core modules. Among them, the match rates of optimized EDS template with Demography (100.0%), Medical History (98.3%), Adverse Event (100.0%), Medication Error (100.0%), Concomitant Non-drug Treatment (100.0%) are all higher than or equal to OC EDS template with Demography (97.1%), Medical History (88.1%), Adverse Event (98.4%), Medication Error (97.8%), Concomitant Non-drug Treatment (100.0%); besides, match rates of OC EDS template with Vital Signs (98.1%), Subject Summary (97.3%), Concomitant Drug (97.0%) are higher than the optimized template with Vital Signs (95.2%), Subject Summary (96.1%), Concomitant Drug (92.3%). Overall, match rate of optimized EDS template is higher than that of OC EDS template.Conclusion:Under the current research conditions, the optimized new EDS template can decrease the time for creating EDS by 63% compared to the OC EDS template, which means it is more adapted to the DataLabs system. Besides, after comparing with the EDS applied in the 9 selected real clinical trial projects, the quality of the optimized EDS template is closer to the needs of real clinical trial than that of OC EDS template.