The Hol Picture

The landscape of mental health research is evolving, and tools like Natural Language Processing (NLP) hold immense potential to bridge the gap between clinical trials and real-world evidence. As we strive to enhance our understanding of mental health outcomes and treatment effectiveness, a strategic implementation of NLP tools becomes pivotal.

This blog explores a high-level overview of how we developed NeuroBlu NLP - Holmusk’s NLP models specifically tailored for extracting disease-specific clinical features from unstructured clinical text. These NLP-derived clinical features are then integrated into our structured data and made available in NeuroBlu, Holmusk’s powerful data analytics software for behavioral health. The incorporation of NLP-derived features into structured data provides a more comprehensive, easily accessible estimate of patient phenotypes in the real-world, ultimately benefiting patient care.

Real-world evidence has been increasingly recognized as an important complement to clinical research. In clinical research, randomized controlled trials often have stringent inclusion criteria which may exclude certain patient groups, such as those with comorbidities. When clinical trial populations diverge from patient populations seen in a real-world care setting, insights gained might not be widely applicable to all people. Analyzing real-world data, such as clinical information recorded in electronic health records (EHR) as part of routine clinical care, can provide insights that are generalizable to real-world populations.

In the realm of behavioral and mental health, electronic health records (EHR) data is a goldmine of information. However, the challenge lies in the fact that these data often need more density for comprehensive analysis.

Innovation in behavioral health has long been limited by a lack of available data to support evidence-based decision-making. This lack of data can be attributed to many factors. Some of these factors are impossible to change, such as the heterogeneity of how the same diagnosis presents differently from patient to patient. Others could be changed but not without cooperation across behavioral health stakeholders; for example, a lack of consensus regarding how to define patient improvement, in addition to which tools to use to measure improvement, has resulted in a lack of standardization unseen in other areas of health.

As biopharmaceutical companies and regulatory bodies look toward adopting EHR-derived real-world data to complete studies more efficiently and affordably, one major concern that often arises is what has come to be known as “data missingness.”

The treatment of behavioral health conditions has historically been challenging. Despite the prevalent use of the DSM-V in clinical settings, clinicians lack a detailed and standardized vocabulary to discuss these conditions. This is due to a range of factors, from wide differences in disease presentation to stigma surrounding mental health conditions. The lack of standardized vocabulary has led to a subjective approach in treating these conditions, with each clinician relying on his or her own experience.

However, as more and more patients who are treated for behavioral health conditions are documented within an electronic health record (EHR) system, researchers now have a valuable tool for studying and improving the treatment of behavioral health conditions. By bringing together vast quantities of real-world data to understand how care and treatment are delivered in clinical practice, we can start building a set of standard definitions and objective measures for mental health conditions.

This is why we have created the NeuroBlu Database, in which we have extracted and organized EHR data from behavioral health clinics across the U.S. Our NeuroBlu data has thus far been leveraged by 5 of the top 15 biopharmaceutical companies with a behavioral health pipeline. These companies can benefit greatly from EHR-derived real-world data, particularly in the areas of research and development, medical affairs, and health economics and outcomes research (HEOR).

Previously on this blog, my colleague Alex has called for a paradigm shift to clinical data derived from the EHR as front-line evidence. He also alluded to some of the limitations of claims data, which have traditionally been used as a primary form of real-world data.

This is an excerpt from a longer post that originally appeared on Going Digital: Behavioral Health Tech.

Real-world data (RWD) are becoming increasingly critical to clinical research. The FDA has put forth definitions surrounding RWD, as well as issued guidance around its use in research emphasizing the principle of data being “fit-for-purpose”—selecting the data needed to answer the question at hand. Meanwhile, stakeholders engaged in clinical development have increasingly recognized that RWD will enable them to conduct studies faster, at a lower cost, and often, with a more representative and diverse population.

However, not all RWD is fit-for-purpose—that is, captured and stored in such a way that the data is ready to address the question at hand. In order to move forward with using RWD in a way that is efficient and effective, we need to build a shared understanding of the different types of data within the broad umbrella that is RWD and make clear which type of RWD is fit-for-purpose for a specific question.

I recently led a team of researchers in a large-scale observational study, which was accepted and published by The Lancet Psychiatry. To our knowledge, our study was the first to study the impact of early clinical trajectory across multiple psychiatric diagnoses.

I recently had the pleasure of joining Real-World Wednesday, a conversation hosted weekly on Clubhouse covering various topics about real-world data. First off, I’d like to thank the hosts of the RWD-RWE Club, Matt Veatch and Aaron Kamauu, for the invitation and the engaging discussion.