How to manage data from wearables in clinical trials — from overcoming regulatory issues to handling challenges associated with dirty data.

Digital technologies are challenging the traditional models of clinical trials, reimagining study designs and the creation and collection of data. In feasibility studies, digital technologies, such as wearables and sensors, have demonstrated value in deepening insights into the patient experience and the impact of investigational products on everyday life. More recently, pharma and biotech companies have initiated harnessing wearables and the data generated from them to inform primary and secondary endpoints.  

Yet, with these advances come burdens and challenges, such as understanding how to best manage, store and interpret large data sets, as well as determine what will pass regulatory submissions. It also includes recognising and handling the challenges associated with dirty data, which can manifest as duplicate data/information, inaccuracies and unused data. Not having proper data hygiene can cause timeline delays, wasted efforts and misinformed decisions, leading to poor clinical trial outcomes and even failure. 

Additional data challenges to applying wearables to clinical trial conduct include patient acceptance, and privacy and security issues. To successfully implement wearables and other digital health technologies in clinical trials given all these limitations, sponsors will need to not only deal head-on with regulatory issues and managing huge datasets, but also have a data management strategy in place. 

Overcoming regulatory concerns

The lack of clear regulatory guidelines and standardisation across countries and regions has led to cautionary usage of digital health technologies for endpoints in clinical research. Therefore, it is no surprise that in a recent webinar hosted by ICON, 40 percent of participants reported that regulatory authorities are the biggest barrier to the adoption of digital endpoints.1

Under the circumstances, sponsors should follow best practices and will need to ensure they have the necessary change control processes in place and that the devices, themselves, are capturing the data needed in the real world. All data submitted to regulators need to meet minimum standards in terms of validity, reliability, sensitivity and robustness. Moreover, regulatory agencies will require similar standards to support the use of data from a specific device in any given study.2

Regulatory survey graphic

Managing huge datasets 

Data management is vital to the success of using digital technologies in a clinical study, and is a key consideration to regulators. When devices are selected and used in a study without a data strategy in place, sponsors will often find challenges in cleaning dirty data, which can cause escalating costs. In fact, a report found that source data verification costs for Phase 1 trials account for about 15 percent of study costs, totalling upwards of $326,000. The study also found that site monitoring costs — which included costs for collecting and checking case report forms, source data verification, and the review and maintenance of drug accountability and query resolution — totalled more than one million dollars per study in phase 2 and 3.3 Additionally, it’s important to note that dirty data can not only increase development costs, but also prevent a new drug application, as it would require enough data to support a complete and high-quality application, leading to a longer time to market. 

Consulting a statistician early on to develop a data strategy, such as discussing what data to include or exclude, can save a considerable amount of money and time. For example, data processing and cleaning, in addition to identifying errors, can be set up to occur in near real-time, increasing the accuracy of data and processing speed. Also, using machine learning and other artificial intelligence approaches can help to manage high volumes of data. 

Developing a data strategy

As previously discussed, data is abundant, vast, and complex, making its collection, management and analysis challenging. The volume and diversity of data being collected contribute to concerns about high levels of delays and inefficiencies. To mitigate issues when analysing data and developing insight, sponsors should create a data management strategy. Requirements that sponsors will need to discuss with their statistician team include: 

  • Data cleaning: understanding how data can be cleaned to generate an analysis-ready dataset
  • Data validity: defining what constitutes a valid data set for a specific study
  • Data continuity: ensuring data continuity, and outlining how to handle missing data
  • Data compliance: determining compliance, including data stream compliance with 21 CFR Part 11 and patient compliance with device use 
  • Data veracity: establishing data veracity, which includes planning how a device’s algorithm will generate the analysable variables, in addition to measuring the degree to which the data is accurate, precise, interpretable and trusted
  • Data Ingestion: deciding how data will be ingested, such as streaming in real time or sending in batches
  • Data standardisation: providing a framework for data standardisation and quality 

Advancing data in clinical research

While data can be messy, implementing a strategy and controlling the process of data collection allows sponsors to prevent or mitigate dirty data as it arises in clinical trials. Despite the challenges that wearables and other digital health technologies present to clinical research, taking advantage of the rich data they provide can bring novel insights to ultimately enhance patient safety. 

To learn more about data challenges when using wearables in clinical trials, read our white paper — Advancing digital endpoints: An end-to-end approach to managing wearable devices through clinical development.

Read the whitepaper


  1. McCarthy, M., Ballinger, R., and Lewis, H. (2020). Webinar: Digital Endpoint Strategy and Validation. ICON, plc.
  2. Walton, M.K., et al. (2020) Considerations for development of an evidence dossier to support the use of Mobile sensor technology for clinical outcome assessments in clinical trials. Contemporary Clinical Trials 91:105962
  3. Sertkaya A, Wong HH, Jessup A, Beleche T. Key cost drivers of pharmaceutical clinical trials in the United States. Clin Trials. 2016 Apr;13(2):117-26. doi: 10.1177/1740774515625964. Epub 2016 Feb 8. PMID: 26908540.