Challenges implementing CDISC standards on real-world clinical trial data and addressing anomalies

Clinical trials are thoughtfully crafted endeavors, meticulously designed and executed with stringent protocols. To ensure the quality, consistency, and interoperability of the data collected throughout the trial FDA recommends adhering to CDISC (Clinical Data Interchange Standards Consortium) standards, in addition to the procedures outlined in the study protocol. These standards provide a comprehensive framework of guidelines and best practices for the collection, organization, and submission of clinical trial data. However, real-world scenarios often deviate from expected norms.

Implementing CDISC standards into real-world clinical trial data poses challenges due to the inherent complexity and heterogeneity of such data. There may be unexpected or unusual data values that do not conform to standard data definitions proposed by the SDTM (Study Data Tabulation Model) and ADaM (Analysis Data Model) standards. Moreover, addressing anomalies can be daunting, given the presence of unexpected or atypical data values that defy conventional definitions. While some occurrences and data may remain beyond control, proactive planning and management can mitigate certain anomalies.

What are the Challenges in Implementing CDISC standards?

Some challenges of implementing CDISC standards on real-world clinical trial data and accounting for anomalies are:

• Data Heterogeneity: This refers to the variability or diversity of data across different study participants or research sites. Clinical trial data can be collected from various sources, including different sites, countries, and systems, leading to data heterogeneity. This heterogeneity can make it challenging to apply standardized data definitions and formats for analysis. A typical example of this would be the laboratory data collected and analyzed in a study. Safety assessments often include laboratory tests to monitor parameters such as liver function, renal function, or hematological profiles. Using more than one laboratory may introduce variations in testing methods, reference ranges, units of measurement, or reporting formats. Inconsistent data collection practices can make it difficult to integrate data across different sites or studies.
• Non-standardized Data: Real-world clinical trial data may not always conform to CDISC standards, making it difficult to map the data to standard formats. This can be due to differences in data collection methods, data dictionaries, and data formats such as dates or events terminology. For example, non-standardized data collection can affect the reporting of concomitant medications used by participants during the clinical trial. Variations in medication names, dosages, frequencies, or duration can make it challenging to accurately convert these non-standard vocabularies to CIDSC terminology using dictionaries such as WHO DD, SNOMED, or MedDRA. 
• Missing/Incomplete Data: CDISC standards define specific variables or data elements that need to be collected and presented for various domains (e.g., demographics, adverse events, vital signs, etc.). Incomplete or missing data can result in missing variables or incomplete datasets, as data may be insufficient to fully populate required data elements. This can occur due to missing or incomplete data from study participants, accidental loss of data, or other data quality issues. For instance, incomplete or missing dates for certain participants or time points, make it challenging to derive the study day required in the SDTM domains and adhere to CDISC standards during data analysis or regulatory submissions. Imputation of partial dates can turn out to be quite challenging when presented with inconsistent date formats from various sites.
• Data Anomalies: Anomalies can occur in clinical trial data, such as outliers or unexpected values, which may not conform to standard data definitions. These anomalies can make it challenging to map the data to standard formats, leading to the need for data transformation and mapping. Examples of data anomalies could be adverse events occurring after the end of participation has been recorded for a subject but before the subject leaves the clinical site. Another example is having to re-screen and re-enroll screen failed subjects resulting in duplication of certain demographic data such as age and leading to inconsistencies if not handled appropriately.

Steps to Mitigate CDISC Challenges 

To address these challenges, the following steps can be considered:

1. Develop a Comprehensive Data Mapping Plan: A comprehensive data mapping plan should be developed as early as possible in the trial to map the heterogeneous data to CDISC standards. The data mapping plan developed at BioPharma Services includes guidelines for handling non-standardized data, incomplete data, and data anomalies. Data that is collected per the specific standards of the country such as date formats are converted to standardized units per trial analysis requirements before combining it. Following this standardized data collection practices recommended by CDASH can reduce data heterogeneity and help mitigate the impact of non-standardized data collection.
2. Collaborate with external Data sources and all internal stakeholders: Collaboration with data providers can be useful to understand the nuances of the data and to identify and address any anomalies or non-standard data definitions. At BioPharma Services, in addition to maintaining open dialogue with external data providers, we also involve other stakeholders including our clinical investigators, medical monitors, clinical research associates, CRCs, CDMs, statistical programmers, biostatisticians tasked with the responsibility to collect, clean, and ensure the integrity of clinical trial data right from the inception of the trial. This will ensure a smooth beginning-to-end transformation and compliance of all data.
3. Quality Control: Regular quality control checks are performed to ensure that the data collected within all the clinical trials conducted at BioPharma Services conforms to CDISC standards. This includes data completeness checks, data harmonization checks, and data validation through manual review, database edit checks, and programming back-end checks. Implementing these at regular intervals throughout the trial progress instead of waiting till trial completion enables us efficient and quick turnaround while implementing and developing CDISC-compliant datasets and submission packages. 
4. Data Transformation and Mapping: It is vital to have experienced and well-trained staff that have an end-to-end understanding of the clinical trial process and reporting requirements from regulatory agencies to be involved in data transformation and mapping. Our team involved in the mapping and transformation of data has decades of experience with CDISC requirements and regulatory guidelines for data submission.  Certain tools and techniques can be used to convert the data to the required format. This may include data normalization, cleaning, and imputation techniques to handle incomplete and missing data. 

Why Choose BioPharma Services for your Next Drug Development Project?

In summary, while clinical trials strive to adhere to predefined protocols and timelines, the reality is that they often encounter unexpected instances or events that require flexibility, adaptability, and proactive management to ensure the integrity of the research and the safety of the participants. At BioPharma Services we embrace these principles and implement the above mitigation practices along with the CDISC standards from the inception of each clinical trial. We design source data collection forms and CRFs with consideration of the CDASH guidelines. This approach enables us to streamline data collection processes, elevate data quality standards, bolster interoperability, and expedite regulatory submissions, ultimately enhancing trial outcomes for the benefit of all stakeholders involved.

Written by: Sushma Meka, Associate Director, Biometrics Operations.