Role of Statistics in Improving Data Quality: ICH E6 R2
Clinical trials are conducted to collect pharmacokinetics, pharmacodynamics, safety and efficacy data for getting the marketing approval of a drug. Multiple trial sites where the clinical trial is being conducted generate these data which are then collected by the sponsor. The data are then submitted to the regulators which review the data and grant approval for the drug if the data are found adequate. To design, conduct, record and report the clinical trials, sponsors must have to follow FDA Guidance for Good Clinical Practice (GCP) which is accepted internationally. Adherence to these guidelines helps to protect the rights of participants and ensures the well-being and safety of the trial participants. The first ICH-GCP guideline document ICH-E6-R1 was published in 1996 after public consultation and editorial corrections. The purpose of this document was mutual acceptance of clinical study data among regulatory agencies of different countries. An addendum ICH-E6-R2 to this original document ICH-E6-R1 has been published on 1st March 2018 which has implemented a number of changes throughout the document. There is also new info for the use of a statistical approach which has defined the role of statistics in improving the quality of data in a clinical study.
Statistics is a field specializing in analysis, tabulation, and interpretation of data. The addendum ICH-E6-R2 stresses the need for a good statistical approach in a clinical trial to produce quality data. To monitor clinical trials, a sponsor needs to develop the prioritized, systematic and risk-based statistical approach. There are various approaches which can improve the monitoring efficiency and effectiveness of a trial. Sponsor has different choices including On-site Monitoring, a combination of Centralized and On-site Monitoring, or Centralized Monitoring. On-site Monitoring is conducted at the trial site(s), and Centralized Monitoring is a remote assessment of extracted data performed periodically by trained and qualified persons. Centralized Monitoring is risk-based monitoring and used for the evaluation of risks in site data in real time from a single location (off-site). With this approach, the sponsor can decide when they need to send the Clinical Research Associates to a particular site. This helps in avoiding the review of risks on each investigational site. Addendum ICH-E6-R2 favors Centralized Monitoring that helps in differentiating reliable and potentially unreliable data. Centralized Monitoring may help in:
- Identifying inconsistencies in the data, missing data, and data outliers
- Identifying variations in the data across multiple clinical sites
- Identifying protocol deviations
- Analyzing individual site performance and characteristics
- Identifying sites which require more monitoring
- Identifying errors in data collection and reporting
- Examining consistency, the variability of data across multiple clinical sites.
- Identifying manipulations in data
- Protecting data integrity
To enhance the data quality in clinical trials, there is another combination approach, i.e., Centralized Statistical Monitoring which is a combination of Centralized Monitoring and Statistical Monitoring. Statistical Monitoring is the complex statistical algorithms to find out data anomalies and outliers, which results in getting information about different monitoring, communication or escalation actions in accordance with the Trial Master Plan (TMP) and communication plan.
Therefore, Centralized Statistical Monitoring helps to remove discrepancies in data and increases the overall quality of data. By applying this approach, a sponsor can identify the sites that are providing low-quality data and can make better management for these sites. Overall, Centralized Statistical Monitoring plays an important role in the generation of quality data by removing data inconsistencies and manipulations.