Automatic classification of registered clinical trials towards the Global Burden of Diseases taxonomy of diseases and injuries

Ontology-based categorization of clinical studies by their conditions

OBJECTIVE

The free-text Condition data field in the ClinicalTrials.gov is not amenable to computational processes for retrieving, aggregating and visualizing clinical studies by condition categories. This paper contributes a method for automated ontology-based categorization of clinical studies by their conditions.

MATERIALS AND METHODS

Our method first maps text entries in ClinicalTrials.gov's Condition field to standard condition concepts in the OMOP Common Data Model by using SNOMED CT as a reference ontology and using Usagi for concept normalization, followed by hierarchical traversal of the SNOMED ontology for concept expansion, ontology-driven condition categorization, and visualization. We compared the accuracy of this method to that of the MeSH-based method.

RESULTS

We reviewed the 4,506 studies on Vivli.org categorized by our method. Condition terms of 4,501 (99.89%) studies were successfully mapped to SNOMED CT concepts, and with a minimum concept mapping score threshold, 4,428 (98.27%) studies were categorized into 31 predefined categories. When validating with manual categorization results on a random sample of 300 studies, our method achieved an estimated categorization accuracy of 95.7%, while the MeSH-based method had an accuracy of 85.0%.

CONCLUSION

We showed that categorizing clinical studies using their Condition terms with referencing to SNOMED CT achieved a better accuracy and coverage than using MeSH terms. The proposed ontology-driven condition categorization was useful to create accurate clinical study categorization that enables clinical researchers to aggregate evidence from a large number of clinical studies.

Medical Concept Normalization in Clinical Trials with Drug and Disease Representation Learning

Motivation

Clinical trials are the essential stage of every drug development program for the treatment to become available to patients. Despite the importance of well-structured clinical trial databases and their tremendous value for drug discovery and development such instances are very rare. Presently large-scale information on clinical trials is stored in clinical trial registers which are relatively structured, but the mappings to external databases of drugs and diseases are increasingly lacking. The precise production of such links would enable us to interrogate richer harmonized datasets for invaluable insights.

Results

We present a neural approach for medical concept normalization of diseases and drugs. Our two-stage approach is based on Bidirectional Encoder Representations from Transformers (BERT). In the training stage, we optimize the relative similarity of mentions and concept names from a terminology via triplet loss. In the inference stage, we obtain the closest concept name representation in a common embedding space to a given mention representation. We performed a set of experiments on a dataset of abstracts and a real-world dataset of trial records with interventions and conditions mapped to drug and disease terminologies. The latter includes mentions associated with one or more concepts (in-KB) or zero (out-of-KB, nil prediction). Experiments show that our approach significantly outperforms baseline and state-of-the-art architectures. Moreover, we demonstrate that our approach is effective in knowledge transfer from the scientific literature to clinical trial data.

Availability and implementation

We make code and data freely available at https://github.com/insilicomedicine/DILBERT.

An Interoperable UMLS Terminology Service Using FHIR

The Unified Medical Language System (UMLS) is an internationally recognized medical vocabulary that enables semantic interoperability across various biomedical terminologies. To use its knowledge, the users must understand its complex knowledge structure, a structure that is not interoperable or is not compliant with any known biomedical and healthcare standard. Further, the users also need to have good technical skills to understand its inner working and interact with UMLS in general. These barriers might cause UMLS usage concerns among inter-disciplinary users in biomedical and healthcare informatics. Currently, there exists no terminology service that normalizes UMLS’s complex knowledge structure to a widely accepted interoperable healthcare standard and allows easy access to its knowledge, thus hiding its workings. The objective of this research is to design and implement a light-weight terminology service that allows easy access to UMLS knowledge structured using the fast health interoperability resources (FHIR) standard, a widely accepted interoperability healthcare standard. The developed terminology service, named UMLS FHIR, leverages FHIR resources and features, and can easily be integrated into any application to consume UMLS knowledge in the FHIR format without the need to understand UMLS’s native knowledge structure and its internal working.

A scoping review describes methods used to identify, prioritize and display gaps in health research

BACKGROUND AND OBJECTIVES

Different methods to examine research gaps have been described, but there are still no standard methods for identifying, prioritizing, or reporting research gaps. This study aimed to describe the methods used to identify, prioritize, and display gaps in health research.

METHODS

A scoping review using the Arksey and O'Malley methodological framework was carried out. We included all study types describing or reporting on methods to identify, prioritize, and display gaps or priorities in health research. Data synthesis is both quantitative and qualitative.

RESULTS

Among 1,938 identified documents, 139 articles were selected for analysis; 90 (65%) aimed to identify gaps, 23 (17%) aimed to determine research priorities, and 26 (19%) had both aims. The most frequent methods in the review were aimed at gap identification and involved secondary research, which included knowledge synthesis (80/116 articles, 69%), specifically systematic reviews and scoping reviews (58/80, 73%). Among 49 studies aimed at research prioritization, the most frequent methods were both primary and secondary research, accounting for 24 (49%) reports. Finally, 52 (37%) articles described methods for displaying gaps and/or priorities in health research.

CONCLUSION

This study provides a mapping of different methods used to identify, prioritize, and display gaps or priorities in health research.

A mapping of 115,000 randomized trials revealed a mismatch between research effort and health needs in non-high-income regions

BACKGROUND

Concerns exist as to whether the allocation of resources in clinical research is aligned with public health needs. We evaluated the alignment between the effort of clinical research through the conduct of randomized controlled trials (RCTs) and health needs measured as the burden of diseases for all regions and a broad range of diseases.

METHODS

We grouped countries into seven regions and diseases into 27 groups. We mapped all RCTs initiated between 2006 and 2015 that were registered at the WHO International Clinical Trials Registry Platform to regions and diseases. The burden of diseases in 2005 was mapped as disability-adjusted life years (DALYs), based on the 2010 Global Burden of Diseases study. Within regions, we defined a research gap when the proportion of RCTs concerning a disease in the region was less than half the relative burden of the disease.

RESULTS

We mapped 117,180 RCTs planning to enroll 42.6 million patients and 2,220 million DALYs. In high- versus non-high-income countries, 130.9 versus 6.9 RCTs per million DALYs were conducted. We did not identify any research gap in high-income countries. We identified research gaps for all other regions. In particular, for Sub-Saharan Africa, we identified research gaps for common infectious diseases (CID) and neonatal disorders (ND): 5.8% (95% uncertainty interval 4.7-6.9) and 2.0% (0.9-4.5) of RCTs in Sub-Saharan Africa concerned CID and ND, although these diseases represented 22.9% and 11.6% of the burden in the region, respectively. For South Asia, we identified research gaps for the same two groups of diseases.

CONCLUSIONS

In non-high-income regions, the conduct of RCTs was misaligned with the distribution of major causes of burden, in particular infectious diseases and neonatal disorders in Sub-Saharan Africa and South Asia.

Making Sense of Big Textual Data for Health Care: Findings from the Section on Clinical Natural Language Processing

Objectives: To summarize recent research and present a selection of the best papers published in 2016 in the field of clinical Natural Language Processing (NLP). Method: A survey of the literature was performed by the two section editors of the IMIA Yearbook NLP section. Bibliographic databases were searched for papers with a focus on NLP efforts applied to clinical texts or aimed at a clinical outcome. Papers were automatically ranked and then manually reviewed based on titles and abstracts. A shortlist of candidate best papers was first selected by the section editors before being peer-reviewed by independent external reviewers. Results: The five clinical NLP best papers provide a contribution that ranges from emerging original foundational methods to transitioning solid established research results to a practical clinical setting. They offer a framework for abbreviation disambiguation and coreference resolution, a classification method to identify clinically useful sentences, an analysis of counseling conversations to improve support to patients with mental disorder and grounding of gradable adjectives. Conclusions: Clinical NLP continued to thrive in 2016, with an increasing number of contributions towards applications compared to fundamental methods. Fundamental work addresses increasingly complex problems such as lexical semantics, coreference resolution, and discourse analysis. Research results translate into freely available tools, mainly for English.