Clinical Trials are the mandatory path for developing and bringing a new drug or vaccine to the market. Unfortunately, according to a study conducted by MIT, 86 percent of the drugs will fail during this process. This very high failure rate not only has consequences on the Pharmaceutical companies' bottom line, but it precludes potentially safe and efficacious drugs from reaching patients that could benefit from them. Recruitment is one of the main bottlenecks, is time-consuming, and very expensive. According to Chunhua Weng from Columbia University (New York), "Recruitment is the number one barrier to clinical research."
FDA has been promoting enrollment practices that could enhance the diversity of clinical trial populations, through broadening eligibility criteria. However, how to broaden eligibility remains a significant challenge. We propose an AI approach to Cohort Optimization (AICO) through transformer-based natural language processing of the eligibility criteria and evaluation of the criteria using real-world data. The method can extract common eligibility criteria variables from a large set of relevant trials and measure the generalizability of trial designs to real-world patients. It overcomes the scalability limits of existing manual methods and enables rapid simulation of eligibility criteria design for a disease of interest. A case study on breast cancer trial design demonstrates the utility of the method in improving trial generalizability.
Clinical trials play important roles in drug development but often suffer from expensive, inaccurate and insufficient patient recruitment. The availability of massive electronic health records (EHR) data and trial eligibility criteria (EC) bring a new opportunity to data driven patient recruitment. One key task named patient-trial matching is to find qualified patients for clinical trials given structured EHR and unstructured EC text (both inclusion and exclusion criteria). How to match complex EC text with longitudinal patient EHRs? How to embed many-to-many relationships between patients and trials? How to explicitly handle the difference between inclusion and exclusion criteria? In this paper, we proposed CrOss-Modal PseudO-SiamEse network (COMPOSE) to address these challenges for patient-trial matching. One path of the network encodes EC using convolutional highway network. The other path processes EHR with multi-granularity memory network that encodes structured patient records into multiple levels based on medical ontology. Using the EC embedding as query, COMPOSE performs attentional record alignment and thus enables dynamic patient-trial matching. COMPOSE also introduces a composite loss term to maximize the similarity between patient records and inclusion criteria while minimize the similarity to the exclusion criteria. Experiment results show COMPOSE can reach 98.0% AUC on patient-criteria matching and 83.7% accuracy on patient-trial matching, which leads 24.3% improvement over the best baseline on real-world patient-trial matching tasks.
Clinical trials represent the forefront of medicine bringing new treatment options to patients and caregivers. Unfortunately, close to 90% of clinical trials fail to meet recruitment goals. This results in expensive delays that often result in early trial termination, or simply the inability to gather sufficient data to draw efficacy conclusions. These clinical trials failures slow down research, delay patient access to life-saving treatments, and contribute to rising drug costs. Enrollment Challenges · Most patients are unaware of which clinical trials are being conducted and if they qualify to participate.
To kick off the celebration of PLOS Medicine's 15th Anniversary, Specialty Consulting Editor Sanjay Basu discusses the journal's contributions to scientific communication and his favorite article from the past 15 years. It's fitting that one of PLOS Medicine's most viewed and cited articles remains the cult classic, Why Most Published Research Findings Are False (2005). The article codifies the challenge taken up as a mantle by the contributors and editors of PLOS Medicine for the last 15 years: to make science more transparent, reproducible, and trustworthy. Thirteen years after first highlighting the problem of research that was nonsensical or misleading– often due to p-hacking and other practices that slice and dice results to find a "significant" outcome– PLOS Medicine also published a new framework for maximizing the value of clinical research (Academic response to improving value and reducing waste: A comprehensive framework for INcreasing QUality In patient-oriented academic clinical REsearch (INQUIRE), 2018). The framework flew under the radar by comparison to articles concerning the problem of low-value clinical research, but the article remains one of my favorite PLOS Medicine articles of the last 5 years.