Developments in Digital Data
By Exploristics Chief Data Scientific Officer – Kimberley Hacquoil
When I first started working in the pharmaceutical industry, most clinical trials used paper case report forms (CRF) to collect information from patients. I remember the very long, often cumbersome, daily diaries and questionnaires about patients’ disease and quality of life. There were also included handwritten notes from the investigators and other assessments carried out when the patients visited the sites. All this information had to be collated and turned into a useable format so the biostatistics team could analyse the data and provide insights into whether the study was a success or not. A lot of time, effort and resource was spent transcribing the data into an electronic form, along with quality checking (QC) to make sure there were no mistakes in this manual process prone to human error. This often meant that the cost was high (both for sponsor and patient burden), to introduce new exploratory assessments to advance clinical design.
I remember the first time I worked on a study that had an electronic diary to capture information. Patients had to use a telephone to call up a centre to log their results each day. These data were more readily available to statisticians and had built in checks to provide ongoing data validation. At the time, this was a massive step forward in improving the efficiencies and site and patient burden in clinical trials. It allowed teams to consider this type of data earlier in clinical development and progress exploratory or surrogate endpoints.
The changing data landscape
In recent years, the clinical data landscape has been transformed with an explosion in the types of digital data now available from a range of sources. Digital data in clinical trials is a broad term and includes a variety of data types including electronic consent, electronic patient reported outcomes (ePROs), wearable sensors and virtual patient visits. Some of these are not new to clinical trials. However, nowadays data is both instant and constant in everyday life. Therefore, real world data in the form of digital health and technology data is becoming more standard and more complex in the clinical trial setting and support of regulatory submissions. Part of this shift is due to COVID which has accelerated changes to clinical trials through decentralisation and the need for technology to provide solutions. In fact, it’s estimated that by 2025, 70% of clinical trials will utilise digital sensors. This is a stark difference to when I started working 15 years ago.
The benefits for patients
The current and evolving COVID-era data landscape offers many potential benefits to patients:
- Virtual site visits mean that patients can enrol into a clinical trial which they might not have been able to before due to travel or location restrictions. This opens opportunities for patients to access different and more relevant investigative treatments to manage their disease.
- Real-world digital data can be used to better understand disease progression and the patient journey. This will enable a more targeted approach to patient selection and population enrichment for clinical trials. Clinical trials will therefore be able to correctly identify and treat the right subset of patients.
- Being in a clinical trial no longer necessitates lots of additional assessments – wearable sensors can use the data that is already being collected as part of everyday life. This reduced burden on patients will improve the experience. It will also enable investigators to quickly access data and information they might not have been able to access before, which will help treatment and management of the disease.
- Collecting information in real time through apps makes it not only convenient for patients but also improves the accuracy as they don’t need to remember what they experienced days or hours ago. This better-quality data means improved likelihood of success for trials when the drug is effective.
- By using more real-life data in clinical trials such as digital data and technologies, it will support regulatory agencies’ understanding of the applicability of a treatment in this setting. Hence improving the ability to develop drugs which will be beneficial to patients in everyday life.
The challenges for developers
Whilst the emerging data landscape undoubtedly brings dividends, it also raises new challenges for developers of therapeutics. The use of digital data in clinical trials requires additional consideration by the sponsors to ensure the practicality of running a study. Undue reliance on its use could lead to excluding certain patient groups or demographics/countries by requiring technologies not readily available in those settings. It will be important to balance this as we need to design studies with a global mindset. Making sure that the right tool is used for the job will be important as there are a lot of potential options in this space. Data collection methods and safeguarding against technical risks will be important to plan for, in addition to scrutinising and understanding data quality issues and reliability and validity of endpoints. Compliance related to digital data, for example sensor wearables, will be vital to manage and mitigate where possible. Data security and privacy will require careful consideration in this highly regulated industry. A sponsor will also need to consider analysis methods and advanced analytics opportunities within these types of trials.
A more central role for statisticians
As a statistician, having more data available provides an exciting and stimulating prospect, and we should be embracing the technological advances we have seen in recent years. However, it can be a little overwhelming to have so much data and information at our fingertips. The challenge now is to make sure that we are getting the most out of this vast evolving data to benefit patients and that we are collecting the right data to make meaningful decisions. Therefore, it is now time to bring statisticians into the heart of the development process to ensure that studies can answer the clinical questions posed and to navigate developers through the choppy waters of the digital data deluge.
Read more:
Disparate and desperate, curse of the diminishing n
Are external control arms worth the extra effort?
Birth of the Data Science team
The liberation of real-world data
Going synthetic with real-world data
Watch more: