In an era where over 90% of experimental therapies in human trials fail to make it to market, QuantHealth emerges as a beacon of innovation to predict clinical trials, endpoints, and outcomes. Coupling advanced AI with an unparalleled database, this trailblazer aims to reshape the clinical trial landscape by predicting clinical trial outcomes through advanced clinical trial outcomes modeling. We sat down with Orr Inbar, the driving force behind QuantHealth, to delve into how they’re harnessing AI technology to address the pressing challenges of today’s clinical research.
Moe Alsumidaie: Orr, the QuantHealth team has been pioneering some innovative approaches to model validation. Could you delve into the intricacies of how logical, mathematical, and clinical validation contribute to the robustness of your model to predict clinical trial outcomes?
Orr Inbar: Our model validation strategy is three-pronged, each addressing different aspects of the model’s reliability.
First, we have logical validation. It’s about introspecting the model and ensuring it behaves logically. For instance, if we simulate an older population, we’d anticipate their mortality rates would increase. It’s these fundamental checks that ensure the model is grounded in reality.
Next is mathematical validation, which is an industry-standard in machine learning. Here, we employ a series of quantitative metrics, such as accuracy, sensitivity, and specificity. Balancing these metrics is crucial. For example, a model that correctly predicts the presence of a common medical outcome might be accurate, but it also needs to be able to predict its absence, to be precise. We’re always refining our models to find that perfect balance, ensuring they’re usable and effective.
Lastly, clinical validation is where things get truly interesting. While the previous validations are at the patient level, clinical validation takes looks at the clinical trial level, aggregating multiple patient predictions to the cohort level. We back-test our models against high-profile phase three trials from recent years, essentially trying to recreate their results. Using this technique, we’ve reproduced over 30 clinical trials with 86% accuracy for the primary endpoints.
Moe Alsumidaie: That’s enlightening. With such a comprehensive approach, how does QuantHealth address real-world challenges beyond the ability to predict clinical trial outcomes, like patient dropouts during trials?
Orr Inbar: Great question. Patient dropouts are indeed a significant concern in clinical trials. Our platform goes beyond the ability to predict clinical trials outcomes. We simulate various scenarios, such as the likelihood of a patient dropping out due to the burdens of frequent site visits. We can also predict patients’ adherence to the trial protocol by assessing their medical history. This supports patient retention and the trial’s overall success.
Moe Alsumidaie: Personalized medicine is gaining traction. How does your platform cater to this paradigm shift, ensuring patients are matched to the right trials?
Orr Inbar: Personalized medicine is at the heart of what we do. Every patient is unique, and our platform reflects that. When patients are placed on trials via our platform, there’s a higher assurance of the experimental therapy’s efficacy. This personal touch ensures that physicians and patients are better informed about how apt a trial is for a patient. For instance, while one drug might work wonders for a particular demographic, it might not be as effective for another. Our platform provides these insights, paving the way for genuinely personalized treatment strategies.
Moe Alsumidaie: Orr, with the increasing importance of data in today’s clinical trials, ensuring data integrity and security is paramount. How does QuantHealth ensure that the vast data collected remains transparent and secure?
Orr Inbar: You’ve touched upon a critical aspect of our work, Moe. At QuantHealth, while we’re committed to offering transparency, we never compromise on security. Our platform is built on state-of-the-art encryption methods and complies with global data protection standards. We have stringent access controls, ensuring that stakeholders can only access data they’re authorized to view. Beyond the technology, our team undergoes regular training to stay updated on best practices in data protection. It’s a holistic approach – combining technology with human oversight to ensure data integrity.
Moe Alsumidaie: As we look to the future, how do you envision technology, especially platforms like QuantHealth’s, reshaping clinical trials, especially in patient recruitment and retention?
Orr Inbar: The future of clinical trials is deeply intertwined with technological advancements. Platforms like ours are set to make trials more precise, efficient, and patient-centric. By ensuring that we only enroll patients who respond to the drug safely and efficiently, we will not only require fewer patients to be enrolled, but we will also prevent many dropouts. Doing trials this way will drive system-wide efficiency, increase the velocity at which drugs are approved, and help reduce costs.
Moe Alsumidaie is Chief Editor of The Clinical Trial Vanguard. Moe holds decades of experience in the clinical trials industry. Moe also serves as Head of Research at CliniBiz and Chief Data Scientist at Annex Clinical Corporation.