During the prescribing process, a physician’s ability to accurately and reliably match the most appropriate therapy option to the patient in question is limited by the quantity and quality of available clinical information. Although randomized clinical trials are considered the gold standard for establishing the safety and efficacy of prescription drugs, and verified clinical findings from the trial report the label of approved therapy, the overall utility of ‘these trials are limited by a variety of factors.
A particular limitation is the use of strict inclusion and exclusion criteria during essay registration. Although the resulting homogeneous patient population offers some benefit to researchers during the initial trial, the clinical findings are not generally applicable to subgroups of patients who were not represented in the trial or to patients with one or more chronic conditions. or comorbidities. Therefore, additional studies are needed to further refine the complete prescription information of the drug and expand physicians ’understanding of the therapy over time once it has been marketed.
In recent decades, the gap between randomized clinical trial findings and real-world prescribing conditions has widened, in part due to the fact that most current chronic diseases have multiple approved therapies available, and many of them Current adults are managing one or more chronic conditions. or comorbidities simultaneously.
Studies that use advanced techniques to develop real world evidence (RWE) based on routinely collected data, both in electronic medical records and in insurance claims databases, can help bridge this gap. But these studies must use state-of-the-art data modeling and analysis techniques if they are to produce data-based knowledge that is accurate and reliable enough to inform treatment.
In search of evidence-based medicine
To support evidence-based medicine, ongoing studies are routinely conducted after the drug has entered the market and is routinely prescribed to patients. These studies involve a complex analysis of real-world data and aim to help pharmaceutical and healthcare stakeholders to develop broader and deeper clinical findings related to the performance of a particular therapy in specific subpopulations of patients in the over time.
Such studies are critical, as the findings allow the medical community to better understand the clinical and safety profile of the drug in a much broader and more diverse patient population (compared to those enrolled in the trial). Given the limitations of time and resources, it is simply not possible or practical for any randomized clinical trial to study all possible combinations of therapy, in each subpopulation of patients, while taking into account an infinite combination of comorbidities. Instead, physicians should rely on their best judgment when prescribing therapies to patients who do not meet the criteria of those enrolled in the actual trial. These studies help physicians select the most appropriate therapy option (from a crowded therapeutic class) for specific patients who may be managing multiple chronic conditions or comorbidities. But well-designed RWE studies can help bridge information gaps and provide relevant clinical knowledge to inform prescribing in specific disease states, therapy classes, and patient populations.
Make sure the RWD is accurate for secondary use
Fortunately, ongoing advances in two data analysis techniques: subgroup analysis i comparative efficiency – enable in-depth clinical design to be designed and conducted using real-world data routinely collected from EHR systems and insurance claims databases and using the results of the study to create a sufficiently valid and accurate RWE to report treatment. The ability to accurately and reliably leverage the large amount of data routinely collected in the healthcare field and use it to amplify the clinical outcomes of the randomized clinical trial creates enormous opportunities to improve prescribing and health outcomes. of patients. The resulting knowledge generates direct benefits for the patient, as well as financial dividends for the healthcare system and the payer community improving care.
The greater the validity of an RWE study, the more appropriate it will be to modify the care and adapt the therapy according to its results. Achieving high validity requires explicit effort and experience in designing RWE studies that provide the required accuracy and generalization, and in devising ways to address some of the shortcomings inherent in current available real-world data.
While mountains of data are collected every day in the EHR claims systems and databases are a ubiquitous part of the current healthcare experience, much of the information available is not necessarily accurate enough for secondary use. For example, the quality and accuracy of EHR data is sufficient to capture the details of a patient’s meeting with their physician, and claim data is sufficient to determine whether the health plan should reimburse the patient. meeting. However, the information may not be accurate or clear enough for secondary use in a clinical study unless additional effort is made at the beginning of the study.
Using a study design paradigm that combines deep phenotyping with linked outcomes (discussed below), stakeholders can develop advanced RWE-based studies that provide clinically relevant findings to inform prescribing decisions. Data related to a specific patient phenotype can be found in the EHR system, as this source is rich in both structured data in the patient chart and unstructured narrative information (such as notes and comments that the physician has added to the chart of the patient) related to the patient’s ongoing health journey. However, EHR data are often incomplete and rarely detailed in a sufficiently consistent manner for study purposes. For example, when a patient sees a doctor or specialist who is in a different practice who is not affiliated with the patient’s primary EHR system or has an emergency room or hospital meeting that produces clinical or treatment data that does not they are married to the patient’s EHR. system.
To help bridge this gap, additional information about health outcomes can be found in the patient insurance claims / billing record. These databases provide a rich source of information for all incidents and interventions related to healthcare that the patient has experienced and that have been sent to the health plan for reimbursement.
By linking the outcome data found in the claims databases to the deep phenotyping information found in the EHR data, current advanced RWE studies are able to provide highly valid clinical findings and high precision.
What’s at stake
Advanced RWE-based studies allow for better clinical care and allow physicians to prescribe the most appropriate therapy among several competing options that may be available in the same therapy class, based on improved knowledge of how this therapy works in different subgroups. of patients who may not have done so. adequately represented in the underlying essay. The ability to reliably connect the rich phenotyping information that can be found in EHR data, with additional knowledge related to the results that can be found in the claims database, creates many opportunities to offer a more prescriptive personalized and improve overall health outcomes.
But this design and execution process involves a variety of data analysis challenges in terms of processing, determining accuracy, alignment, privacy, and data security. To address some of these needs, artificial intelligence (AI) can be applied to extract the relevant information needed for the clinical study and create the most appropriate subset of patient records that meets the study criteria. RWE. The model can then run small-scale testing to confirm the accuracy of the protocol standards before conducting the larger-scale study.
This process allows studio designers to access the most relevant and accurate data, specifically gathering data “appropriate for the purpose.” For example, if a given study protocol requires 80% accuracy, the use of deep phenotyping to collect real-world data from the most appropriate patients (those who meet the study criteria) would improve overall accuracy. Combining this more robust source data with linked database data from the claims database, and then using an augmented AI model with natural language processing techniques, allows for a strong feedback loop that verifies the level precision required.
When advanced RWE studies can show that a specific therapy is more effective in a given subgroup of patients, patients in that subgroup have access to this option as a standard of care. The approach described here: the wider use of advanced RWE studies that are explicitly designed to include deep phenotyping and linked results using TO THE to enable comparative effectiveness: it offers a way forward in healthcare.
This approach not only creates an incremental improvement in healthcare, but also a disruptive shift to better inform the standard of care of patients and physicians to improve prescribing. Doing so guarantees the best possible treatment, minimizes the progression of the disease and promotes optimal health outcomes. Patients deserve nothing less, and the approach also has broader financial and business implications, as deeper clinical insights based on real treatment data support help inform drug prices, formulary placement and reimbursement decisions.
About Dan Riskin
Dan Riskin is the founder and CEO of Verantos, the market leader in high-precision real-world (RWE) evidence generation. Dr. Riskin, recognized worldwide as an expert in artificial intelligence for health, has developed products that influence the care of millions of patients annually.
Dan is an expert in artificial health intelligence and a successful series entrepreneur. The products it has developed and marketed influence the care of millions of patients each year. He has spoken on medical technology at the FDA, CMS, NASA, DARPA, NSF and NIH. His contributions to data-driven healthcare have been featured in Forbes, The Wall Street Journal, and other prominent media outlets. Dan served on the Obama Health Policy Committee and testified before Congress on the 21st Century Care Initiative.