CAB/VP and OpenCDISC Rules Reconciliation Project
The purpose of this project is to continue the effort of improving SDTM validation rules, ensuring that all rules are both clear and effective for use by industry. An industry sub-group known as CAB/VP recently consolidated SDTM rules from five different vendors after deliberating on discrepancies and inconsistencies.
The goal of this CAB/VP and OpenCDISC Rules Reconciliation Project is to compare the CAB/VP group results with the most current list of the OpenCDISC SDTM rules to determine if there are rules in the CAB/VP list that are better or more complete than the latest version of the SDTM rules implemented by OpenCDISC, which contains the latest input from the FDA.
Rules Under Review and Reconciliation
The CAB/VP was established during 2Q2010 as the result of an expressed CDISC community need. The goal of the group was to contribute to the overall quality of clinical data by providing an agreed-upon mechanism for assessing conformance to the CDISC/SDTM Standard and corresponding Implementation Guide.
A team initially composed of 20+ members from CDISC, Industry and Vendors, for the most part CAB representatives or delegates, with participation from the FDA, met normally on a bi-weekly basis. The team members all possessed strong technical expertise and knowledge of the SDTM standard. Over time, the team composition varied up and down and the CAB/VP as original established officially ended as of April 2012. A few of those original team members continued the effort and became part of the newly formed [WG1P4] team.
The scope of the CAB/VP included the evaluation of compliance rules for Data and Metadata (SDTM data and associated metadata in the define.xml document) and the initial focus was Regulatory submissions to the FDA.
The initial phase was to develop a uniform set of checks independent of implementation, where rules are concise and unambiguous, include both data and metadata, are not ‘value’ specific and include an agreed severity of failure.
A second phase was intended to develop golden set(s) of test data and associated check’s results that included positive and negative tests, with the potential to enable CDISC/SDTM Compliance Checks Certification. This phase was later on deemed not to be the responsibility of CDISC.
The approach followed during the initial phase was to review the SDTM Standard Compliance checks implemented by vendors to assess differences in meaning (independent of style and granularity), potentially adding missing checks, assign categories to each of the checks (structure; content/terminology; metadata: consistency of self-contained in SAS XPT files with define.xml plus not contained in SAS XPT files), determine the nature of the rule (compliance with the SDTM standard or other quality assessment) and establish a consensus for severity of failures and consistency of user messages.
The validation rules evaluated were the ones implemented in the following software tools: WebSDM / Phase Forward, OpenCDISC Validator / OpenCDISC, SAS Clinical Standards Toolkit / SAS, CheckPoint / Octagon and SAS/DI / Business & Decisions.
The CAB/VP team faced many challenges that included work coordination among team members as well as a more suitable infra-structure for collaborative work. The work completed included the review and deliberation of close to 1200 checks that varied in form/style or granularity however not in content to get to a consolidated list of 200+ checks. Among the next steps under consideration by the team were: 1) the reconciliation with the latest version of checks that over time got implemented by OpenCDISC since it has been the list of checks most closely reviewed by the FDA; 2) the addition of validation checks/rules to cover the SDTM/IG 3.1.2 Addendum; 3) the completion of standard validation rule messages.