March 31, 2021
Each of the more than 5 billion prescriptions filled every year holds a promise of improved or sustained health or quality of life. For a significant number of patients, however, that promise is a broken one.
Suboptimal prescribing results in about 275,000 lives lost and $528B wasted every year in the United States. Many of the problems stem from a practice of trial and error or one-size-fits-all prescribing that results in medication therapy failure.
For patients, therapy failure can result in prolonged illness, decreased quality of life and even dangerous adverse events. The most common types of medication therapy failure are:
- 57% Inadequate therapy
- 15% Non-adherence
- 15% Adverse reactions
- 7% Unnecessary therapy
- 7% Dose too high
For healthcare organizations, this broken promise results in expenses that quickly add up to a high and unnecessary cost:
- $271.6B Long-term care admissions
- $7.8B Additional prescriptions
- $37.8B Provider visits
- $37.2B ED visits
- $174B Hospitalizations
How do we solve this problem?
We start ensuring patients get the right prescription the first time by replacing current models with precision health.
Precision health, as defined by UCLA Health, “takes into account differences in people’s genes, environments, and lifestyles and formulates treatment and prevention strategies based on patients’ unique backgrounds and conditions.”
That sounds simple enough. Take the patient’s health history, current condition, current treatments, lab test results, pharmacogenetic testing and compare all of that information to the recommendations for each drug choice. So why aren’t providers and health systems rushing to adopt precision health? In many cases, each piece of information must be gleaned from a different source and then manually compared. For a healthcare provider who has limited time, the comparison alone is overwhelming before even taking into consideration the required data mining.
The good news is that with the ability to build Precision Health Insight Networks (PHINs), healthcare systems are now able to tear down data silos and bring all of the information needed for precision prescribing into a single platform. With all of the data in one place, machine learning can be used to derive insights that identify otherwise hidden risks and inform prescribing so that the patient gets a medication plan designed specifically to be the safest and most effective for them.
A new solution: hc1 Precision Rx Advisor
When hc1 developed the award-winning hc1 PGx Advisor®, the goal was exactly that—use PHINs to normalize and organize the patient’s data in the hc1 Precision Health Platform™, and then produce a detailed Rx Scoring Report using a real-time database of drug interactions, FDA warnings and pharmacogenomic markers combined with prescription and medical claims, formulary and demographic data.
Since 52% of the 50 most common drugs have genetic implications for efficacy and patient safety, the potential to impact patient outcomes and reduce costs was huge using PGx testing alone. One hc1 evaluation of a health plan with 43,000 covered lives for pharmacogenomic implications identified $15 million in avoidable costs from medication therapy failures impacting 3,000 patients.
In the time since launching PGx Advisor, we’ve come to recognize that while, yes, using PGx testing to guide precision prescribing does make a big impact, there is even more potential to reduce risk for patients and reduce costs by taking into account other lab testing data.
Recognizing the potential impact of taking into account other laboratory and patient data, hc1 has chosen to expand on PGx Advisor and is now bringing to market PrecisionRx Advisor™, the only solution that combines Medication Therapy Management (MTM), PharmacoGenomics (PGx), clinical lab data and integrated engagement and monitoring workflows in a single platform. By integrating lab and patient data essential to precision prescribing and uncovering otherwise hidden risks, PrecisionRx Advisor enables an optimal medication plan designed just for the patient and lowers the overall cost of care.
Atul Butte, M.D., Ph.D., offered a clear example of using lab data at the University of California to inform precision prescribing and improve diabetes care and reduce cost in a February 2018 Regenstrief Institute “Work In Progress” session (advance to 1:07:00 minutes for the example). Dr. Butte is the Chief Data Scientist, University of California Health System, Priscilla Chan and Mark Zuckerberg Distinguished Professor of Pediatrics, Bioengineering and Therapeutic Sciences, and Epidemiology and Biostatistics, UCSF, and also presented at the hc1 and Becker’s Healthcare 2020 Precision Health Virtual Summit.
In his 2018 presentation, Dr. Butte begins with the Standards in Medical Care in Diabetes published by the American Diabetes Association. The standard practice first-line therapy is to prescribe Metformin and Comprehensive Lifestyle including weight management and physical activity. When that therapy is evaluated after 90 days using Hemoglobin A1C testing (HbA1C) and found to not be sufficient, then a medication change or addition may be necessary.
The University of California evaluated their medication strategies for 71,702 Type 2 Diabetes patients across all campuses for a year following each change in dose and medication over 90-day evaluation cycles. During the evaluation, they identified 6,543 different medication trajectories.
Using machine learning, they compared the model for predicting medication class increase within the first 90 days when the patient was prescribed Metformin only and determined that if the patient’s largest HbA1c was ever measured greater than 8.8 or their fasting glucose was ever higher than 2.10, Metformin alone was unlikely to be an effective treatment.
“We know with our data it’s not going to work,” Dr. Butte said. “What’s the next move you’re going to make? Just make that move now; save 90 days here. In our patients’ data-driven protocol, we know it’s not going to work. Skip ahead a move. That’s the whole point here.”
Real-time insights at the point-of-care
The hc1 Precision Health Cloud is also able to use machine learning principles as the University of California did but can do so in real-time at the point of care.
PrecisionRx Advisor provides the ability to not only optimize medication therapy for patients, thereby reducing healthcare costs, but also offers a tool for pharmacists to audit in real-time whether a drug is the right fit for the patient and recommend a change to the physician before filling the prescription. This enables providers and pharmacists to collaborate and make precision prescribing decisions with just a click.
If the amount of savings estimated for a single health plan based on PGx implications alone for just 3,000 patients is $15 million, the savings from optimizing medication plans across the entire population is bound to be even more significant.
If you are interested in learning more about PHINs, the hc1 Precision Health Platform, PrecisionRx Advisor or just want to talk about the potential precision prescribing has to improve healthcare, we’d love to talk to you. Contact us here.