Randomized clinical trials are valuable in determining the effectiveness of health treatments. But problems with design, execution or reporting of the trial process can lead to unreliable findings, excessive costs, and, potentially, harm for patients.
Associate Professor Halil Kilicoglu and his colleagues seek to address this problem with the help of a $1,328,502 grant from the National Institutes of Health. The funding for this project, "Computational Methods, Resources, and Tools to Assess Transparency and Rigor of Randomized Clinical Trials," will be awarded over the next four years.
"I am leading a team of computer/information scientists and clinical research methodologists in developing datasets, natural language processing (NLP) methods, and ultimately software tools that will help various stakeholders of biomedical communication assess and improve the reporting quality in randomized clinical trial protocols and result publications," Kilicoglu said.
NLP, he explained, refers to the use of algorithms to model natural human language so that computers can process and understand what humans have written or said.
Although often costly, randomized clinical trials provide the most robust evidence for determining how well therapeutic interventions such as drugs can work. However, according to Kilicoglu, clinical trials often suffer from poor methodological and reporting quality (also known as rigor and transparency, respectively). This can render the trials' findings questionable, and waste dollars meant to find treatments.
Kilicoglu's group will help address the problem by applying NLP methods to both the assessment of clinical trial protocols (generally published before the trial is launched) and their result publications (published after the study is completed). The methods will look at whether the trials' procedures and results are reported in appropriate detail. Kilicoglu and his colleagues will also collaborate with journals to pilot test the tools that are developed.
The work of Kilicoglu and his team will be helpful to all involved in clinical research, including scientists, journal editors, peer reviewers, and funders. His team will develop a set of models, resources, and tools that will assist these stakeholders in maintaining high reporting standards, synthesizing evidence, and promoting open science practices. The work is intended to contribute to improvements throughout the scientific ecosystem, leading to better clinical care and health policy.
Kilicoglu's group will work with the National Center for Supercomputing Applications (NCSA) on the Urbana campus, as well as clinical research methodologists from the University of North Carolina, University of Arkansas, and Indiana University.
Kilicoglu earned his PhD in computer science from Concordia University in 2012. Prior to joining the iSchool faculty, he worked as a research scientist at the U.S. National Library of Medicine, National Institutes of Health.