Priorities for modernizing ClinicalTrials.gov: Five ideas to facilitate transparency in observational research
Ilker Oztop, Ph.D.
ClinicalTrials.gov (CT.gov) is the largest public clinical registry and results database in the world, providing knowledge of clinical research to the public to foster trust amongst the public and inform decision-making.
Recently, the National Library of Medicine (NLM) announced plans to modernize the technical infrastructure of the site, and on December 30, 2019, it released a Request for Information (RFI) on CT.gov Modernization, soliciting “public input to guide the NLM in planning infrastructure enhancements aimed at users and submitters of CT.gov”.
Aetion had a chance to respond to this RFI and share the priorities of observational researchers and regulatory scientists working with real-world data (RWD) and real-world evidence (RWE).
Here, we share our key recommendations for the NLM as they work to modernize CT.gov, in advance of a public hearing on April 30, 2020, where the NLM will deliberate on the responses.
- To improve website functionality, integrate CT.gov entries with RxClass identifiers at the drug, drug class, and disease level for FDA approved prescription medications
RxClass, a web application also maintained by the NLM, is a medical/clinical ontology navigator that links drug classes from several sources (e.g., ATC, FDB, MeSH, SNOMED-CT among others) to their RxNorm drug members (i.e., prescription medications approved for human use in the U.S.). Of all these hierarchical drug class and disease identifiers, CT.gov currently recommends, but does not require, the use of MeSH and SNOMED-CT vocabularies when registering the clinical trial “Condition or Disease” of interest. A standard registration requirement to use RxClass identifiers and link to RxClass from within CT.gov entries upon regulatory approval would be of immense value to observational researchers and regulatory scientists, who currently have to rely on independently curated ontologies, cross-walking tables, or other resource-intensive research workflows; for example, when studying pre- versus post-approval drug-related safety events at the pharmacologic class level. ENCePP/EU-PAS Register is a good reference, where the search functionality allows filtering for “Study drug> Substance Class (ATC Index).”
- Tag completed trials where individual participant data (IPD) is made available besides aggregated clinical study results
Vivli.org is an independent, non-profit organization that offers a global data sharing and analytics platform with a focus on sharing IPD from completed clinical trials, study protocols, statistical analysis plans, and clinical study reports. Each vivli.org entry maps to a unique CT.gov identifier (NCT), but registrants are currently not required to reference this pairing in CT.gov. Requiring CT.gov registrants to enter this information whenever available would facilitate locating the IPD, and expedite secondary research. For example, historical IPD from completed clinical trials may serve as external controls to augment clinical trials, and inform related clinical development programs. Eventually, we recommend CT.gov itself hosts IPD under respective NCT entries, if available.
- Enable “embargo” function until planned public release date when registering retrospective (i.e., secondary data) studies
As previously discussed at ISPOR 2019, there is concern around protecting intellectual property (IP) when registering retrospective observational studies and detailed protocols, since a publicly available retrospective database study protocol can easily be executed by other parties—assuming underlying data availability. An agreed-upon “embargo date”, until which a registered retrospective database study protocol should not be publicly released, would help protect any IP rights of the registrants, and thus serve to incentivize consistent reporting and transparency. A similar regulation to learn from pertains to how the FDA is required under FDAAA § 916 to publish the Action Package of Approval of a New Molecular Entity (including the FDA’s multi-discipline review of the application) not later than 30 days after the approval date. This provision defines a time window when the FDA and the applicant can consult and agree on, for example, which parts of the multi-discipline review should be censored before public release to prevent potential IP infringement.
- Incentivize standardized registration of observational studies using RWD such as cohort studies
Under current legislation, observational research is not subject to the registration and results submission requirements, since they do not meet the criteria for “applicable clinical trials” as defined in federal law. While lack of such legislation led to inconsistent and not comprehensive reporting of observational research to date, multiple stakeholders ”did not let perfect be the enemy of good” and have developed standards and best practices for internal alignment and wider adoption.As recently discussed in detail at the ISPOR 2019 Summit, and then documented in the ensuing white paper, “standardized observational study registration—particularly for studies leveraging RWD sources—has been proposed as an important mechanism for improving transparency and trust into RWE generation”. While we encourage readers to peruse the white paper for further reference, an example NCT entry for standardized observational study registration is the RCT DUPLICATE study, “Replication of the CANVAS Diabetes Trial in Healthcare Claims,” where core registration elements included: full code and algorithm definitions; detailed and time-stamped study protocol; including eligible study population definition, eligible cohort entry dates, inclusion/exclusion criteria, primary and secondary outcome measures (including time frame of observation); a Statistical Analysis Plan; and results tables.
- Align retrospective cohort study elements to be registered in CT.gov with the FDA RWE submission cover letter (once finalized)
Aetion’s response to the FDA’s Request For Comments on a uniform cover letter to identify RWE submitted as part of a regulatory submission discussed the value in capturing detailed real-world study elements to facilitate transparent, auditable, and reproducible RWE generations. Aligning retrospective cohort study elements registered in CT.gov with the final FDA RWE submission cover letter would not only enable streamlined transparency, but also reduce administrative burden for the registrants. We encourage readers to peruse Aetion’s full response to the RFC for further reference.
As the NLM explores how best to modernize CT.gov, Aetion is committed to sharing the perspectives of observational and regulatory scientists to inform its decisions. By facilitating consistent and efficient registration of RWD analyses on CT.gov, NLM can help CT.gov become the standard bearing global registry of choice for all clinical trials and RWD studies alike, in line with the evolving regulatory frameworks around RWE.