Home / Resources / Articles / Innovative Artificial Intelligence (AI) Tool Optimizes Care for T2D Patients  

Innovative Artificial Intelligence (AI) Tool Optimizes Care for T2D Patients  

Oct 6, 2020
 

Author: Steve Freed, R.PH., CDE, et. al.


A new clinical software using artificial intelligence is being used for those with type 2 diabetes in Kentucky. It shows promising results in improving the Quality of Life for its residents and lowering treatment costs. This feature will be a 3-part series. 

In Part 1:  Adherence-related issues are hemorrhaging the costs of diabetes-related treatment in America. To offset this tragedy, GlucosePATH is utilized by a team of healthcare professionals dedicated to improving the quality of life while decreasing avoidable costs.  

 

Over one-in-three Americans are promised to be diagnosed with diabetes by 2050, according to the Centers for Disease Control [1].  The 2019 Kentucky Diabetes Report reveals daunting numbers of adults with diabetes: more than doubled between 2000 and 2017, to 442,480 [2]. Kentucky’s Medicaid program spends over $117 million on diabetes-related treatments, making it the third-most-expensive chronic disease behind the cerebrovascular disease and chronic obstructive pulmonary disease. 

Meeting compliance with Kentucky’s health challenges, the state runs a State-University Partnership program (SUP), which pairs the government agencies with in-state university research departments. For the quality improvement project reviewed here, Kentucky’s Department of Medicaid Services (DMS) partnered with Northern Kentucky University (NKU), and the hospital and physicians of Saint Elizabeth Healthcare in Covington, Kentucky. The healthcare team at St. Elizabeth is known for its holistic approach to assisting patients through multilevel collaboration. 

Kentucky’s two-year project, which ended June 30th, 2020, focused on adult Medicaid patients with an A1C level of 8 or higher. Outcomes measured included the change in patient hemoglobin A1C, patient and system medication cost, and hospitalization claims for unplanned hospitalization visits.  The project goals included improving medication adherence by involving patients in the decision-making process, reducing clinical inertia, aiding physicians in clinical decision-making, and, ultimately, reducing the cost of treating diabetes. 

Cost is an essential factor when it comes to managing diabetes. It is vital to reduce the cost to patients so that patients will take their medications. An increase in adherence should ultimately lead to a decrease in hospitalization visits due to uncontrolled diabetes. It is also important to ask patients about their medication preferences, such as oral agents, since patients with injection phobia may be reluctant to take their medicine. 

A drug regimen’s success depends heavily on the patient’s adherence to the treatment. Pharmacists’ consensus is that the two primary obstacles to patients’ treatment adherence are preference and cost. Kentucky’s project began by having pharmacists speak by telephone with individual patients about their diabetes treatment concerns before an initial office visit. The extra care to reach out to their patients led to more communication and shared decision-making between physicians, pharmacists, and patients. The pharmacists entered each patient’s responses into the computer software, which used that information and the patient’s insurance formulary to identify promising treatment regimens.  While the patient’s physician made the final decision regarding the patient’s treatment regimen, the project empowered pharmacists and patients to identify promising options using the software before presenting those options to the physician.  

Because neither the software nor the pharmacist had made the previous treatment recommendation, it might have helped avoid clinical inertia – approximately 94 of the 156 patients enrolled in the project changed regimens to include all or some of the softwares recommendations.  The A1C of these patients was reduced by approximately 2.15 points (about 20%), a statistically significant difference versus the projects control group.  Taken together, the benefits of team-based care may be the future for managing life-long illnesses such as diabetes.  

Besides incorporating patient preferences, healthcare providers are faced with the constant introduction of new medicines and dosages.  More than 160 combinations of diabetes medications and doses are available, with different levels of efficacy, side effects, and cost.  With that many individual choices, there are millions of ways to build multi-drug regimens to treat diabetes.  Because of this, it often takes years to reach therapeutic goals, while toxic sugar levels in the blood are degenerating the patients micro-and-macro vascular system. 

The pharmacists in Kentuckys project used the GlucosePATH AI software to avoid these obstacles. 

The software goes through every possible combination of medicines that could be prescribed.  For each combination, the software uses current treatment standards to evaluate efficacy and safety.  It also identifies barriers to treatment compliance and generates personalized strategies to facilitate treatment compliance. For example, GlucosePATH provides medication costs and pre-authorization information to mitigate potential cost problems. This transparency may build patient trust and improve health outcomes.  

The software suggested up to ten regimen options for helping the patient reach therapeutic goals.  This AI technology expands on the physicians typical prescribing habits and provides better alternatives that can be implemented in the patient treatment plan to control the A1C levels. Physicians who participated in the project concluded that AI is an excellent tool for expanding their knowledge of treatment options. Additionally, more than 80% of physicians found that the AI improved their patient outcomes in their practice setting [3]. 

GlucosePATH is designed to reach therapeutic goals within three months. In the Kentucky project, patients who wholly followed the regimen proposed by the software (and agreed to by the patient, pharmacist, and provider) saw an average A1c decrease of approximately two points (about 20%) in 90 days. Also, those patients’ A1c did not substantially increase for the duration of the project.  The chart below, from the project report, shows the A1c change.  

(Full indicates patients who followed the full regimen recommendation of the pharmacist/patient/provider/software; Partial represents patients who followed some of the regimen recommendations; Alternate represents patients who followed a regimen different than their current regimen, and one not recommended by the team or software; and No change are the patients who didnt change their regimen.) 

By experiencing dramatic positive change, patients may be more motivated to stay committed to improving their lifestyle with exercise and diet. 

Artificial intelligence is the key to healthcare breakthroughs. AI tools assist rather than replace healthcare professionals by providing shared decision making and improving personalized, patient-centered care. GlucosePATH equips physicians with various treatment options for patients with type 2 diabetes by curating their particular needs and integrating the medication cost into the treatment decision-making process. This project demonstrates how to reach therapeutic goals by integrating GlucosePATH software into your practice. An opportunity to act will be provided at the end of this series. 

 

Part 1:   Adherence-related issues are hemorrhaging the costs of diabetes-related treatment in America

Part 2: Collaborative Analysis Concludes that the AI Tool Improves Outcomes for Type 2 Diabetes Patients

An examination of the project architecture and the research methods are reported with the results supporting GlucosePATH’s positive impact on patient outcomes, ultimately decreasing total costs.  

Part 3: Optimize Clinical Diabetes Management with AI Technology

User testimonials about the experience and exciting findings using the software. A more in-depth examination of the benefits of applying decision-support software in clinical practice.  

 

Authors:

Steve Freed, R.PH., CDE, with:

Peter Jay Won Pharm.D. Candidate, University of South Florida, Taneja College of Pharmacy 

Aleksandra Kusic, PharmD Candidate, Florida A&M University, College of Pharmacy and Pharmaceutical Sciences  

Joan Prifti, PharmD Candidate, Lake Erie College of Osteopathic Medicine, LECOM School of Pharmacy 

See more about type 2 diabetes in our condition center.