Key Data Concepts
Case Report Form (CRF) Development
Once a research idea has been developed and formally outlined in a protocol, it is important to consider the data you need to collect in order to answer your research question. Clinical trials Case Report Forms (CRFs) are designed to capture all study specific data points in accordance with the protocol. CRFs need to be carefully constructed to:
- Target the required information. You do not want to collect redundant, unnecessary data points.
- Ensure valid responses are provided. You want to limit data collection error.
Following these points will ensure increased reliability when the data is compiled and analysed.
Database Development and Data Dictionaries
A database is used to collect, store, update and retrieve clinical trial data. The database design should reflect the CRFs being used in a clinical trial and thought should be given as to how you wish to utilize and analyse your trial data when considering the functionality of your database.
Each field in your database should be defined in a data dictionary. Data definitions can change from organisation to organisation making interpretations and integrations difficult and deceptive. Text data must be converted to numeric data for it to be analysed in most statistical software. A data dictionary allows you to define each of your fields and the values that go in it. This also facilitates the interpretation of any statistical analysis.
Data Management and Data Quality Checking
It is not enough to simply collect data for a clinical trial. It is imperative to the outcome of the trial that this data be accurate and reliable. Therefore trial data should be entered, tracked and reviewed to ensure that there are no missing data points and that the data received makes sense. For example, that a value falls within an expected range or that if a particular data point is of great significance (e.g. this could be the date of disease progression), this value had been reported accurately and matches the source documentation (e.g. an imaging report). A specific process should be set up and followed with regards to handling data to ensure that all trials are being held to the same standard. Depending on the exact nature of the trial additional quality assurance may be considered. This could include reviewing data entry at a particular time point to assess the data entry error rate, or having an external expert review data pertaining a particular trial endpoint.
Data Monitoring and Safety
To support the data management processes performed in-house by the sponsor, monitoring at the trial site (hospital, clinic or institution) also occurs whereby the clinical data collected is reviewed against the source medical documentation. Of particular importance is the review of safety data. When testing a new treatment in patients it is critical to monitor any adverse outcomes to ensure the experimental treatment is not impacting upon the patient’s well being in a way that is unmanageable or unacceptable in comparison with current available treatments. Data regarding adverse events (AEs) and serious adverse events (SAEs) should be collected and regularly reviewed by the appropriate persons (Study operations team, medical monitor, clinician etc.), with SAEs being reported to Ethics where appropriate.
There will also be specific time points in a trial whereby the efficacy and safety data is reviewed by people independent of the trial, forming a Data Safety Monitoring Committee (DSMC). The DSMC reviews relevant data sets and provides a report outlining any issues detected, stating if the trial is still considered relevant (i.e. the data does not indicate that one treatment has any significant benefit/detrimental impact in comparison to the other) and safe.