2021 CTOACommunityBasedClinicalTrial

From GM-RKB
(Redirected from Lin, Gebel et al., 2021)
Jump to navigation Jump to search

Subject Headings: Clinical Research Ontology.

Notes

Cited By

2020

2020

Quotes

Abstract

Driven by the use cases of PubChemRDF and SCAIView, we have developed a first community-based clinical trial ontology (CTO) by following the OBO Foundry principles. CTO uses the Basic Formal Ontology (BFO) as the top level ontology and reuses many terms from existing ontologies. CTO has also defined many clinical trial-specific terms. The general CTO design pattern is based on the PICO framework together with two applications. First, the PubChemRDF use case demonstrates how a drug Gleevec is linked to multiple clinical trials investigating Gleevec’s related chemical compounds. Second, the SCAIView text mining engine shows how the use of CTO terms in its search algorithm can identify publications referring to COVID-19-related clinical trials. Future opportunities and challenges are discussed.

Introduction

...

Discussion

We presented our initial development of a community-based Clinical Trial Ontology (CTO) using an ontological realism approach. This initiative was driven by the need for greater interoperability between major clinical trial registries, and by two real-world use cases relating to PubChemRDF and SCAIView. CTO aims to provide a small set of ontologically engineered clinical trial specific terms for the ontology community. Many clinical trial-related standards, terminologies and vocabularies have been adopted in different circles and used in healthcare practice. These include CDISC, NCIt, and SNOMED CT.

Among the OBO Foundry ontologies, OBI, OPMI, OCRe, and ERO have defined many terms and relations relevant to clinical trial and clinical study research. Yet, no clinical trial focused and formally defined ontology has hitherto been available to the OBO Foundry community. Other previous non-BFO-compliant work include the PHUSE community’s Mini Study Ontology[19,20], CTO-NDD[10], and the Cochrane PICO Ontology (https://linkeddata.cochrane.org/pico-ontology). The PHUSE community focuses on representing submission data conformant to the CDISC standard in the RDF format, with no attention to semantics. None of these ontologies fits the needs of the use cases presented in this paper. In fact, the developers of OPMI and CTO-NDD themselves formed the CTO development team precisely to address their deficiencies. The CTO-NDD has therefore been refined to form part of the current CTO, and OPMI has donated multiple clinical trial-specific terms to CTO.

Although CTO has provided the basis for a simplified strategy for linking clinical trials to investigational drug and disease data in the PubChemRDF use case, and also provided synonyms and definitions for CTO-based text mining in SCAIView, challenges still remain for those who are developing this ontology. For example, what is a clinical trial? Traditionally, clinical trials are randomized, double-blind interventional studies in which both investigators and patients are unaware of which treatment is being administered. The ClinicalTrials.gov includes both interventional studies and observational studies because observational studies have been used for regulatory decision making. Furthermore, in the Framework for FDA’s Real-World Evidence Program (https://www.fda.gov/media/120060/download), FDA considers the traditional clinical trial as a type of clinical study. It is unclear if observational studies will in the future be considered as clinical trials given that they are non-interventional. In the CTO development team, how to ontologically represent and make distinctions between clinical study, clinical trial, and interventional and observational study is currently under active discussion.

The meanings of ‘study design’ and ‘study type’ are often ambiguous. CTO provides a distinct ontological treatment of study design as a plan specification (Figure 2). CTO contains many study design terms, such as allocation (e.g., randomized), intervention model (parallel assignment), primary purpose or masking (double blinded), cross-over design, factorial design, sequential, single arm, as well as study types including health services research, diagnostic test, basic science, prevention, prognosis study, screening, treatment study, epidemiological research, interventional clinical trial of medicinal product. These terms and relations among them have not as yet however been properly defined, and their hierarchies are under development in the current CTO.

Another challenge lies in the BFO-realism based approach, which has its roots in Aristotelian realism, and requires that the universals represented by the terms asserted in an ontology exist in space and time in their respective instances[20]. The definitions for terms in realism-based ontologies must adhere to very strict criteria. Each term must be provided with an Aristotelian definition which states the individually necessary and jointly sufficient conditions which must be satisfied be instances of the corresponding class[21]. These conditions are however not satisfiable in every case. Some difficult terms from this perspective might include the class of all those things capable of being investigated in a clinical trial (e.g. investigational drugs, devices, vaccines, dietary supplement, behavioral treatment, and more) and the ‘status’ of a clinical trial. In clinical trials, besides ‘condition’ and ‘disease’, clinical trials might investigate also for instance ‘conditions’ or ‘quality of life’ or ‘health risk’. The OBO foundry has established that the universe ‘disease’ is a subclass of ‘disposition’; however, there is no adequate ontological agreement regarding use of terms such as ‘condition’, ‘quality of life’ and ‘health risk’. The status of a clinical trial, including ‘completed’, ‘recruiting’, ‘terminated’, ‘withdrawal’ and the like, is something that changes over time. By treating ‘status’ as a subclass of ‘realizable entity’ this issue can be addressed in a BFO conformant matter. However, a definition has still to be formulated that will reach a consensus among the CTO development team.

The further development of CTO will continue with its BFO-based approach and evolve as a community effort, thereby supporting WHO’s goal of further standardizing registration and reporting of clinical trials. Further applications of the ontology will also be explored, including application to study design, to results comparison across clinical trials, linking out to other data resources, and to the improvement in the development of specific text mining algorithms to identify relevant publications on specific clinical trials.

References

;

 AuthorvolumeDate ValuetitletypejournaltitleUrldoinoteyear
2021 CTOACommunityBasedClinicalTrialBarry Smith
Asiyah Yu Lin
Stephan Gebel
Qingliang Leon Li
Sumit Madan
Johannes Darms
Evan Bolton
Martin Hofmann-Apitius
Yongqun Oliver He
Alpha Tom Kodamullil
CTO: A Community-Based Clinical Trial Ontology and Its Applications in PubChemRDF and SCAIViewH2021