Skip to main content

Table 2 Barriers in the use of routinely collected health data for randomized controlled trials and options for improvement

From: Routinely collected data for randomized trials: promises, barriers, and implications

General barriers or issues Pressing questions Possible solutions, actions and additional comments
Data
▪ Availability ▪ Management ▪ Linkage ▪ Accuracy ▪ Validity ▪ Is the desired outcome variable or RCD source available? ▪ Will it be possible to achieve the same data quality and accuracy with RCD as in traditional trials? ▪ Is the data linkage and management feasible in institutions with limited IT infrastructure? ▪ A central register of databases available for clinical trial research would be helpful, ideally with details about data quality. ▪ Establish core outcomes and structured outcome assessments in routine care ▪ Create RCD trial guidelines and RCD source validation guidelines to help standardize their use and reduce sources of bias or uncertainty ▪ Increase IT presence (particularly data analysts) to health research teams ▪ The more RCD is sought out and used in research, the greater is its availability and differentiation.
Regulatory and ethics
▪ Collecting and obtaining the data ▪ Using and sharing the data ▪ What type of release must be given by the patients before their data can be collected or shared? ▪ Is it ethical to use RCD without asking for their permission, even if their data are anonymized? ▪ Can this data be considered of value and morally be sold? ▪ How are concerns about privacy and informed consent approached (particularly in the context of population-wide trials or Zelen designs)? ▪ Are data safety standards applied to RCD just as strictly as they are to traditional actively collected data? ▪ Who is responsible for the safety of the data? ▪ Ethical guidelines specifically regarding the collection and dissemination of RCD should be developed. ▪ Ethics and approval committees should deepen their knowledge of these novel ethical challenges. ▪ Whereas personal data are collected daily from many sources (e.g., phone use), collection, storage, and dissemination of data related to health require more definite ethical oversight and greater transparency to the general public. ▪ After safety issues are defined, researchers and stakeholders must ensure that data are safely handled, with full transparency of access.
Costs
▪ Obtaining the data ▪ Managing the data ▪ Will data collectors (e.g. health insurers) share their data? Freely or at a cost? ▪ Is a constant increase in the generation of routine data really reducing the overall trial costs if the same institution collected the data in the first place? ▪ When is the use of RCD cost-effective? ▪ The financial worth of health data is not defined or explored; empirical data are necessary to determine the cost of both producing and maintaining health data ▪ Health data are already legally sold to many industries, and regulations/legislation must catch up with this aspect.
Novelty
▪ Bureaucratic obstacles ▪ Unawareness ▪ Training to generate, collect, prepare, manage and analyze RCD for trials ▪ Will approval committees understand the implications of using RCD sources for clinical trials? ▪ What are the challenges that can be expected bureaucratically because most submission templates do not assume the use of RCD and absence of patient contact? ▪ Are data anonymization techniques clear? ▪ What training is required to qualify individuals who generate, collect, prepare, and manage RCD for clinical trial research? ▪ Develop, in collaboration with approval committees, RCD-specific templates and submission forms, especially in such studies where no patient contact is foreseen and therefore speedy approval is desired. ▪ Educate regarding data anonymization and confidentiality risks ▪ Include the concept of using RCD for RCT in clinical research education and teaching ▪ Create and use reporting guidance specifically for RCD-RCTs
  1. IT Information technology, RCD Routinely collected health data, RCT Randomized controlled trial