New requirements for clinical trial transparency provide new opportunities for informatics research

This editorial continues a series focused on improving the credibility, transparency, reproducibility, and replicability of biomedical informatics research.^1,2 It addresses the new responsibilities and opportunities that arise from the recent clarification and expansion of requirements for clinical trials registration and deposit of summary results data and trial documents in ClinicalTrials.gov.

NEW REQUIREMENTS

It should not be news that in 2016 the Final Rule for Clinical Trials Registration and Results Information Submission (42 CFR Part 11)³ clarified and expanded the requirements and procedures for registering and submitting summary results for certain clinical trials of drugs, biologics, or devices to ClinicalTrials.gov, in accordance with Section 801 of the Food and Drug Administration FDA Amendments Act of 2007. The expanded requirements include deposit of the full protocol document and statistical analysis plan (if not included in the protocol) at the time of results submission. Submitters have the option of depositing them sooner. Simultaneously in 2016, the U.S. National Institutes of Health (NIH) stated that the ClinicalTrials.gov submission requirements in the Final Rule would apply to all clinical trials funded in full or in part by the NIH, regardless of study phase, type of intervention, or whether they are subject to laws and regulations governing trials of FDA-regulated products.⁴

The Final Rule and the new NIH policy apply to clinical trials initiated, completed, or funded after early 2017. Other biomedical research funders have issued similar policies that cover all clinical trials they support, some prior to 2017. As a result, the range of biomedical informatics studies for which submissions are required has increased substantially. For example, depending on the funder, trials of decision support tools, sensing devices, and data and information display techniques for patients, clinicians, and scientists are now subject to ClinicalTrials.gov registration and results submission.

In addition, the recently revised “Common Rule” for Protection of Human Subjects (45 CFR 46),⁵ effective July 19, 2018, requires posting of informed consent forms for clinical trials supported by U.S. government agencies. Depositing the form in ClinicalTrials.gov meets this requirement.

The clarification and expansion of mandates for clinical trial transparency are major steps forward in making evidence from human research studies available to advance science and practice, improve clinical research safety and accountability, and meet ethical obligations to study participants. They also impose new requirements on many researchers, expand the data available for consequential informatics research, and provide new opportunities for innovation in clinical research informatics.

NECESSARY ACTION BY THE BIOMEDICAL INFORMATICS COMMUNITY

Meeting the new requirements, using the expanding body of information about clinical studies, and improving the informatics infrastructure that supports clinical research will require action from a broad range of biomedical informatics researchers, including those who submit manuscripts to Journal of the American Health Informatics Association (JAMIA).

• Informatics researchers who conduct studies with human participants:

  • Follow standard guidelines for developing protocols^6,7 when defining a study to ensure a full description of what is planned. This will make clinical trial registration easier.

  • Register your research in ClinicalTrials.gov at the start of the study, whether mandated or not. Many important clinical journals already require early registration as a condition for later publication of study results.⁸JAMIA is likely to impose such a requirement in the future. Based on searches of PubMed or PubMed Central on February 7, 2019, only 20 JAMIA papers (published between 2002 and 2019) describe a clinical study that was registered in ClinicalTrials.gov. There is substantial room for improvement.

  • Include resources and mechanisms for generating the summary results tables required by ClinicalTrials.gov in the plan for any study with human participants. The 2016 Final Rule has made these requirements much clearer, and there are supporting information and education materials on the ClinicalTrials.gov site.⁹ Producing summary results tables is a significant scientific task, not a clerical one. It can also be a step toward publishing the results in a journal. By design, ClinicalTrials.gov requirements are aligned with good reporting practices as specified in the CONSORT (Consolidated Standards for Reporting Trials) guidelines.¹⁰ These guidelines are required or recommended for submissions to many journals, including JAMIA.

• Informatics researchers who use clinical trials data in their work:

  • Use ClinicalTrials.gov as a data source for systematic reviews and meta-analyses. As of February 7, 2019, only 1 systematic review published in JAMIA stated that ClinicalTrials.gov was 1 of the sources searched for evidence. This should be a standard practice. Summary results often appear in ClinicalTrials.gov before they are published elsewhere and in more complete form, and the database includes many trials with potential informatics relevance. To mention 2 examples, more than 4000 ClinicalTrials.gov records describe studies involving “mobile” interventions and more than 675 records describe studies of the impact of “decision support.”

  • Remember that the volume, completeness, and types of information available in ClinicalTrials.gov are increasing due to requirements effective in 2017 and 2018. ClinicalTrials.gov’s utility as a data source for particular lines of research is changing as its content expands.¹¹ To date, JAMIA authors have made creative use of ClinicalTrials.gov data for natural language processing research on topics such as eligibility criteria, patient–trial matching, plant-based medications, potential drug target discovery, and identification of molecular effects of drugs, as well as to monitor compliance with trial registration policies and assess the readability of ClinicalTrials.gov records themselves. All of these topics deserve more research attention, as do many others. The recently expanded requirements and related system changes have resulted in a growing body of protocols, statistical analysis plans, and informed consent forms in ClinicalTrials.gov. Although in PDF format, these documents are a potential source of useful information for those designing new studies or for direct analysis. For example, the more than 750 informed consent forms already available in ClinicalTrials.gov constitute a novel resource for research on health literacy, risk communication, and clinical research ethics.¹²

• The Clinical Research Informatics Community:

  • Focus research and development on reducing the burden associated with generating summary results tables for ClinicalTrials.gov.The expanded requirements for submission of summary results tables that went into effect in 2017 apply to thousands of clinical researchers, including many in the informatics field. The Final Rule eliminated much of the previous ambiguity about the who, what, and when of ClinicalTrials.gov results submissions. The characteristics of the summary results tables are now known in advance, although details obviously vary depending on the arms, interventions, outcomes, etc. in each specific trial. Greater certainty about what is required provides a firmer foundation for progress on providing useful automated support for generation of summary results tables from details captured in the many different systems used to manage the conduct of clinical trials. This is a major challenge and a worthy priority for clinical research informatics. Even small steps and partial successes could yield major benefits in reducing burden, improving data quality, and making research evidence available more quickly.

To state the obvious, greater transparency about informatics research will provide better evidence for rigorous reviews and a better foundation for design, funding, and conduct of more consequential future research. We expect JAMIA authors to meet requirements to register trials and submit summary results data and trial documents to ClinicalTrials.gov and to use ClinicalTrials.gov routinely as a data source for reviews and meta-analyses. We look forward to receiving many more manuscripts that report on innovative research uses of the growing body of information in ClinicalTrials.gov. We especially invite reports on informatics interventions designed to assist clinical investigators by improving the quality and reducing the burden of submitting data to ClinicalTrials.gov.