Electronic Health Reporter

By Shivrat Chhabra, CEO, Dosis.

Now more than ever, the healthcare industry is leveraging new technologies to provide patients with improved, innovative care. The innovation attracting the most buzz in the healthcare industry today is artificial intelligence (AI). However, despite the ongoing hype of robots and algorithms as industry game-changers, results to date from early applications of AI in healthcare have fallen short of realizing dreams of sweeping improvements.

IBM’s Watson is an excellent example of how these improvements “in healthcare” will require a more step-by-step approach and may take longer to achieve than initially thought. In 2011, Watson garnered worldwide attention by winning a game of Jeopardy against two of the show’s greatest champions. Within healthcare, Watson’s win gave rise to hope that AI was on the precipice of full-scale deployment that would transform the industry and dramatically improve patient outcomes.

For several reasons, that hasn’t quite happened yet, and Watson has found it challenging to deliver improved patient outcomes. While those critical of AI have been quick to jump on these struggles, it’s crucial to acknowledge that Watson suffers from several common obstacles faced by AI in healthcare. These include the lack of high-quality data that can be used to train an algorithm, the low number of available training cases, implicit bias, and the differences in guidelines between the U.S. and other countries.

However, as the industry collectively works to address these issues, I envision three major areas where AI will soon transform personalized medicine.

Individualizing the patient-clinician relationship

Clinicians are already equipping themselves to better serve their patients with the predictive and organizational benefits of AI. This technology will move the field away from a “one-size fits all” approach and make the clinician-patient relationship more individualized, fostering trust.

This would be no small feat for improving the patient-clinician relationship, especially for those suffering from chronic conditions. A study by West Corporation in 2018 found that only 12 percent of chronic condition patients feel strongly that their provider is doing a good job of delivering information specific to their needs and condition.

When a clinician provides patients with unique, individualized solutions, patients feel empowered and are more comfortable speaking up throughout the treatment process. When a patient is comfortable enough to report symptoms, no matter how trivial they may seem, personalized medicine thrives.

With the help of AI, clinicians can search extensive amounts of information to find the causes of patient-reported symptoms and alter patient care accordingly. These improvements can be referenced by other clinicians and lead to large-scale medical breakthroughs.

Personalizing drug dosing

Perhaps one of the most useful medical advancements aided by AI is the creation of personalized drug dosing. With this, clinicians input patient data and receive individual medication plans based on the stage of the disease or illness, biology and medication type.

Researchers using AI platforms have already demonstrated how AI can enable dynamic dosing to meet a patient’s individual needs. Within the clinical study at UCLA, healthcare providers used an AI platform to identify how a prostate-specific antigen (PSA) in a prostate cancer patient’s blood evolved as their drug combination and dosages changed. These biomarkers were used by the AI platform to update prescribed medicine and dosage levels in real time. As a result, the patient was able to resume a completely normal and active lifestyle.

Many illnesses and medications require exact dosing to maintain treatment efficacy and patient health. However, with most chronic illnesses, doctors still prescribe dosing based on what is optimal for the general population versus on an individual level.

Additionally, most treatment plans are fixed for a period of time versus being dynamically adjusted in real-time. With personalized dosing, clinicians can easily track progress and ensure that patients receive the most effective dose possible. These platforms also allow clinicians to use current patient data to refine and improve treatments for future patients.

Personalized cost meets personalized care

Advancements in AI can vitally bolster healthcare’s newfound focus on value-based care. With value-based care, payments are often based on individual patient risk and anticipated cost. Using AI to personalize care, providers can develop an individualized treatment to minimize individualized cost.

Simply put, clinicians will be able to provide the treatment options that optimize patient outcomes and minimize cost. Both AI and personalized medicine make innovation easier for clinicians by providing them with the necessary tools to practice efficient and effective value-based care.

As AI starts to make research, dosing and treatment decisions easier and more precise, clinicians will be able to keep their focus on providing patients with effective treatments. The focus on the best individual treatment plan will reduce trial and error and allow clinicians to monitor patient status more closely, reducing preventable, costly events, such as side effects and hospitalizations.

With quality data and cooperative infrastructure, AI and personalized medicine will work hand-in-hand to provide a level of care never before thought to be possible.