Automated Prescription Adherence: Get Timely Medical and Appointment Alerts

The use of AI is increasingly evident in various aspects of the healthcare industry including surgeries, automation, precision therapeutics, and remote clinical trials. It has also expanded its scope to medical adherence. If the patient does not adhere to the prescriptions, the efficiency of the whole system would be at stake. But, in today’s challenging lifestyle which demands activity round the clock, following the prescription schedules is better said than done. Further, the complexity of the schedule increases with the patient’s age or health condition. The elaborate medication at segregated hours is difficult to cling to which often delays recovery and, in some cases, worsens the situation. Thus, AI tools are increasingly being used in the healthcare industry to help patients adhere to drug regimes. 

Current AI technologies such as mobile phone applications, reminder systems, tools for patient empowerment, instruments that can be used in integrated care, and machine learning help the patient to intimate required medical actions at the right time. Also, AI-assisted interventions aim to improve communication between patients and physicians, monitor drug consumption, and ultimately, increase adherence levels leading to better clinical outcomes and an increase in the quality of life of the patients.

The intended clinical outcomes make medical adherence critical as non-adherence might lead to multiple negative outcomes including mortality and hospital admissions. Data from US Medicare members with one or more of the diseases including diabetes, hypertension, and/or high cholesterol demonstrated high rates of non-adherence to medications for these conditions. 76% were non-adherent to one of the three medicines, whereas 32% were non-adherent to more than one target medication class.

The factors that primarily contribute to poor medical adherence can primarily be categorized into patient-related factors such as health literacy, multimorbidity, and lack of involvement in the treatment decision-making process and physician-contributed factors such as communication barriers, or having multiple physicians providing care. Further, healthcare system-related issues include limited access to care and a lack of health information technologies.

The urgency to establish medical adherence among patients has given rise to the need for AI–intervention in the field of medical adherence.  AI has been increasingly used to foster coordination within and between healthcare organizations and healthcare professionals through the development of care systems known as integrated care programs, which aim to improve clinical outcomes and the experience of patients.

Technologies Used to Increase Medical Adherence

Mobile Phone Applications

Mobile applications have been increasingly used in assessing and encouraging medical adherence. Apps have been developed for stroke patients that measure adherence by taking direct oral anticoagulant therapy. Apps have also been developed for patients with schizophrenia and cognitive impairment.  

Reminder Systems

This technology uses health communication to improve medical adherence. Automated systems are used to provide reminders to take medicines. These could be SMS-based refill reminder solutions or patient-sourced Observation of Daily Living (ODLs). The information collected from these systems was made available to the healthcare providers to improve decision-making. 

Chatbots and Robot Assistants

These have indirect medication adherence benefits. For instance, “Vik” a chatbot has been used to empower patients with breast cancer by sending personalized text messages. Additionally, it provided diverse and relevant quality-checked information about breast cancer including but not limited to its epidemiology, treatments, and side effects, as well as information about lifestyles, fertility, reimbursement, and patient's rights. Robot assistants have been used for self-management of diabetes and insulin control in children. 

Benefits of AI in Medical Adherence

AI has been used to assess the extent to which patients adhere to a prescription

 There are numerous direct and indirect and objective and subjective methods for measuring patient adherence including patient-self report, electronic measures, and pharmacy refill and claims data. Still, a combination of these methods is anticipated to be a useful approach to increase the validity and reliability of adherence management which is possible by the utilization of machine learning methods. 

AI has been used to increase patient adherence to a prescription

AI-assisted technologies have resulted in increased adherence of patients. This has improved the clinical outcomes. Adherence to cardioprotective medications in patients with diabetes and ischemic heart disease, statins, and beta blockers in patients with acute myocardial infarction has resulted in lower all-cause mortality. A meta-analysis demonstrated a 21% reduction in long-term mortality risk with good medication adherence and a 17% higher risk of all-cause hospitalization associated with non-adherence in older adults. Increased medical adherence leads to a better knowledge of medications and improvement in the quality of life, physical function, and symptoms. 

Saving Time for Healthcare Providers by Accurate and Remote Monitoring of Patients 

AI-assisted technologies enable a more accurate understanding of patients’ medication adherence lending physicians more time to develop personalized strategies to optimize adherence. Further, it could monitor disease management in times when face-to-face access to healthcare is restricted. Potential access to adherence via smartphones and tablets helps in the remote monitoring of patients and increasing communication with healthcare practitioners.

Challenges to Increasing the Scope of AI in Patient Adherence

Heterogeneity in the Characteristics of the Patients

The characteristics of the user may impact the effectiveness of an AI tool in a number of ways such as the engagement with AI solution can potentially depend on the race, ethnicity, language, age, and social determinants of health. Additionally, technological, ethical, trust factors, communication barriers, and regulatory concerns contribute to the perceived risk. These factors can also contribute to the trustworthiness of chatbots among patients.  

Depersonalization of Care Relationships Affecting the Patient-Centered Model of Healthcare

Patients, relatives, and healthcare professionals fear that the use of AI will lead to depersonalized medication. There has been a concern that automation may lead to discrimination against minority groups affecting the patient-centered model of care. Further, a systematic review of adherence interventions reported that face-to-face interventions were more effective than interventions delivered by a computer, telephone, and/or text message.  

Development and Recruitment of AI- Ready Workforce

Training the employees is critical to the success of any AI technologies. However, as per the insider reports of the healthcare industry, only 50% of healthcare providers render training to their employees.  Additionally building an AI-ready workforce would require changes in the approaches of training and talent acquisition. The challenge increases in building the critical capability of how AI can solve big complex problems, among the employees. 

Exploitation of Healthcare Systems by Using Basic Algorithms

AI-washing occurs when the developers add an AI label to a software platform with basic algorithms to excite and exploit buyers including healthcare systems and individual practitioners. This problem can be addressed by training the employees of the healthcare systems at multiple levels. 

Economic Constraints to the Development of Novel Technologies

The development and initiation of AI intervention may be expensive and thus a relevant barrier to the utilization of these technologies in healthcare research and practice. Studies have indicated that medical non-adherence is associated with a substantial increase in patient cost hence, care should be taken that Ai assisted interventions are cost-effective. 

The use of AI may be key to understanding the complex interplay of factors that underlie medication non-adherence in patients. AI-assisted interventions increase adherence levels leading to better clinical outcomes and increasing the quality of life of patients. However, research on this topic is still sparse, especially within the development and evaluation of AI solutions regarding their effectiveness in different patient groups and the barriers to widespread adoption. The development and use of AI in healthcare research and practice requires careful consideration of ethical principles, risks and benefits, privacy, access and usability, and data management. Effective adherence support will require a combination of interventions to optimize adherence at multiple levels, including the patient, the healthcare professional, and the healthcare systems.