The Impact Artificial Intelligence Is Already Having on Real-World Evidence

Artificial intelligence (AI) is driving significant improvements in healthcare, especially in the realm of real-world evidence (RWE). RWE – clinical evidence derived from real-world data (RWD) – is vital to helping life sciences companies better understand the patient journey and therapies used in the real world.

AI in medical research has the ability to remove the guesswork from the patient journey by analyzing and organizing a vast amount of RWD across formats, such as electronic health records (EHRs), claims, imaging, and genomics, where valuable information about patient care is documented. 

By combining AI and human expertise, we can revolutionize how we generate, interpret, and apply RWE to inform decision-making in therapy development.

By combining AI and human expertise, we can revolutionize how we generate, interpret, and apply RWE to inform decision-making in therapy development. The faster we can produce high-quality insights with real-world evidence and AI, the faster we might see treatments that will improve the lives of patients.

Enhancing Real-World Data Collection and Processing with AI

Traditional methods of collecting and processing healthcare data to generate RWE can be cumbersome and inefficient. However, AI algorithms in medical research – particularly machine learning (ML) and natural language processing (NLP) approaches – have streamlined the process by swiftly aggregating relevant information from unstructured data sources, such as clinical notes within EHRs. 

This is where clinicians document the “why” of patient care – “why” a particular diagnosis was determined or “why” certain medications were prescribed. Approximately 80% of clinical data is unstructured and untapped, yet is essential to generating comprehensive insights on patient outcomes and determining longitudinal trends in care.

Since clinical notes are typed in manually by clinicians, robust AI processing capabilities are necessary to extract and curate the information such that it becomes structured, standardized and reliable in order to enable more streamlined querying of the data.

Verana Health, with its advanced VeraQ^® population health data engine, is unique in its ability to utilize clinician-guided ML and NLP techniques to surface relevant clinical notes from some of the largest EHR-based specialty databases in medicine. Its team of expert physicians and data scientists possess deep data analytics and RWE expertise to create robust AI models that can search keywords and analyze patterns within clinical notes. 

Advancing Real-World Data Analysis and Insights

AI-powered analytics enable precision medicine by helping to identify trends and generate insights at a granular level, which can help tailor treatments.

AI-powered analytics enable precision medicine by helping to identify trends and generate insights at a granular level, which can help tailor treatments. Verana Health’s AI algorithms are high-caliber in that they venture far beyond identifying standard diagnostic codes (i.e., International Classification of Diseases) to understand disease prevalence and progression. They go a step further by extracting key variables from unstructured clinical notes to evaluate additional details about the disease and the effectiveness of treatments at scale.

Many of the key variables used to identify geographic atrophy (GA), for example, come from data derived from unstructured fields in clinical notes. The key to unlocking details is fine-tuning NLP models to flag keywords and patterns of language consistent with clinically relevant context, such as “vision worse,” “visual acuity deteriorating,” or “subfoveal involvement.” ML algorithms can identify correlations between patient characteristics and treatment outcomes, predict disease progression, and evaluate the long-term effects of interventions.

Other key variables involved in identifying and tracking GA progression can be found in imaging data. AI can analyze medical images (e.g., X-rays, MRIs, ophthalmic images and CT scans) to detect key variables, such as lesion size, location, and growth rate. This can be done with not only high accuracy and efficiency, but also with objectivity by removing variability that might be attributable to a more subjective manual review workflow.

By combining unstructured and structured data, AI models can be trained to identify GA prevalence and progression at scale. The results are cohesive and comprehensive datasets for analysis. In Verana Health’s case, Qdata^® modules are research-ready, fit-for-purpose datasets that can help unlock quality research insights along the entire drug and medical device development lifecycle. 

Accelerating Drug Development and Approval

Real-world evidence and AI are helping to accelerate the drug development process by providing more accurate, real-time assessments of treatment effectiveness and safety. Drug development is a lengthy and costly process, but AI and RWE can help expedite this in a number of ways. AI can help to quickly surface potential candidates for clinical trials from expansive specialty care databases. It can also assist with generating rapid insights that reflect more current patient populations and practice patterns that can be used to optimize trial design while also serving as a tool that can enable monitoring post-market safety using RWD. 

For rare diseases, where clinical trial data is often scarce, RWE and AI provide invaluable insights. By analyzing medical registries and other RWD sources, AI can process clinical notes to identify a larger number of these patients that can be connected to effective treatments. This is particularly important when developing therapies for rare conditions, where traditional research methods are often impractical. 

Additionally, regulatory bodies, such as the Food and Drug Administration, are increasingly accepting RWE-based approaches in support of drug approvals. AI is facilitating the generation of this evidence more robustly and efficiently than ever before.

Impact of AI on Medical Research

Overall, secure and advanced AI is transforming medical and scientific research by enhancing the speed, accuracy, and personalization of healthcare. To ensure data privacy and security, Verana Health enforces strict policies to protect patient data, including ensuring all data is de-identified in compliance with HIPAA guidelines. 

As it continues to evolve, the impact of AI on real-world evidence will only grow, leading to more informed healthcare decisions and improved patient outcomes. The future of medicine is not just in treating patients, but understanding them deeply, and individually, through the lens of AI-driven RWE.