Available online at ajdhs.com

Asian Journal of Dental and Health Sciences

Open Access to Dental and Medical Research

Copyright  © 2024 The   Author(s): This is an open-access article distributed under the terms of the CC BY-NC 4.0 which permits unrestricted use, distribution, and reproduction in any medium for non-commercial use provided the original author and source are credited

Using AI to increase medication adherence

Pallav Dave *

Regulatory Compliance Analyst, Louisville, KY,40223, USA

 

 

Introduction

Medication adherence remains a significant challenge despite numerous measures taken to improve adherence and years of advocacy on the importance of the same. Globally, non-adherence remains as high as 40% to 50%, particularly for patients grappling with chronic diseases.^1,2 Medication adherence is imperative because non-adherence has a direct impact on patient outcomes.³ In fact, medication adherence has a more direct impact on patient outcomes than the treatment itself. In the United States, non-adherence accounts for up to 50% of treatment failure, contributes to about 25% of hospitalizations and leads to up to 125,000 deaths.⁴ Despite the measures that have been taken to increase medication adherence, it remains a significant challenge mainly because the responsibility of medication adherence largely falls on patients. As a result, it because extremely challenging to attain significant success. Different factors contribute to the high rates of medication non-adherence. Patient-related barriers account for the highest cases of non-adherence. They include forgetting to take medications when expected, forgetting to go for refills, poor understanding of dosage and schedules, patient attitudes and beliefs about medication, side effects, impaired cognition, and lack of patient engagement in the treatment decisions among others.⁵ 

Although the factors that contribute to non-adherence are known, addressing them is still a significant challenge. Different traditional measures of ensuring adherence such as pill counts, examining the rates of prescription refills, and electronic medication monitors have been taken over the years but despite this effort, the problem persists.⁶ This necessitates new measures to be adopted to increase medication adherence. The increasing adoption of technology in healthcare shows a lot of promise in different areas including medication adherence. According to research by Treskes et al., smart technology has been shown to increase medication adherence in patients with cardiovascular disease.⁷ The research investigated the potential of different types of technologies including phone interventions, short message service (SMS), smart pill boxes, web-based interventions, and mobile apps and showed that they had a lot of potential to increase adherence.⁷ Technology-based methods of monitoring medication adherence show a lot of promise as they have been shown to have accuracy when it comes to monitoring adherence rates. However, successful implementation of these methods is still a challenge with results indicating mixed results when it comes to their efficacy in monitoring adherence.^8,9 The mixed results on the effectiveness of these technologies require further research in this area. 

AI-powered technologies are coming up as better alternatives to improve medication adherence. Research has shown that these technologies improve medication adherence for different conditions.^10,11,12,13 To aid in medication adherence, AI-powered tools can perform a number of functions. These functions include identifying non-adherent individuals, sending reminders, educating patients on the importance of medication adherence, and monitoring patients’ adherence.¹¹ Identifying non-adherent individuals is important because it will lead to dedicated and personalized efforts to increase adherence. It is also important to educate patients on the importance of adherence and send reminders. All these measures can help to increase adherence. 

This review seeks to explore how AI can be used to increase medication adherence. It bases its findings on past review papers that have documented the use of AI to increase medication adherence. As such, it explores both the positives and limitations of the same. 

Importance of Medication Adherence 

The World Health Organization recognizes the importance of medication adherence on disease outcomes by noting that non-adherence leads to poor patient outcomes and compounds the challenge of improving health.¹⁴ Medication non-adherence also results in wastage of the already limited healthcare resources. The problem of non-adherence is a significant challenge with research showing that up to 50% of people living with chronic illness do not take their medications as prescribed or expected.^2,15,16

Poor adherence has a negative impact on health. In addition to increasing morbidity and mortality rates, non-adherence leads to wastage, disease progression, increased hospital visits and admissions, and lower quality of life.^6,17,18 In addition to the negative impact on health, medication non-adherence has significant cost implications. According to Cutler et al., the disease-specific economic cost of medication non-adherence averages $949 to $44 190 per person.¹⁹ When it comes to all causes, the cost rises significantly to an average of $5271 to $52 341.¹⁹ These cost implications pose a significant burden to healthcare systems which are already burdened by high costs of care which impacts people’s ability to get medications for prescribed conditions.²⁰

Addressing non-compliance remains a significant challenge because it is attributed to different causes. The causes can be patient-related, treatment-related or even provider-related. Patient-related non-adherence results from patient views and beliefs, social contexts, access or service issues.¹⁴ Patients' views and beliefs play a crucial when it comes to non-adherence. Beliefs and views about the illness can inform their decision to adhere to medication or not.¹⁴ For instance, if patients believe they need medication for their disease, then they are likely to adhere to it. Patients who believe the illness can go away without the medication are likely to be non-adherent. Non-adherence can also be informed by patients' cultural or religious beliefs. Non-adherence may also be informed by the patients’ motivation to manage illness. If they are motivated to manage the illness, then they are likely to adhere to medication. Other factors such as forgetfulness, psychological stress, inadequate knowledge and skills, anxiety about medication adverse events, and lack of perceived need for treatment also inform patients’ decision to not adhere to medication. Inadequate knowledge and skills can affect patients’ ability to read and understand medication instructions which can inform decisions not to adhere to medication.⁶ Research has documented that patients who have low health literacy are likely to have negative disease outcomes.^21,22 This is informed by factors such as the inability to understand medication instructions hence making them not to take medications or take them in a way that is not prescribed. Such can lead to negative outcomes. Fear of side effects is another commonly reported factor that contributes to medication non-adherence.²³ Fear of side effects may be informed by factors such as negative reactions towards similar medication in the past, unpleasant withdrawal effects, and patient perceptions of the drugs.²⁴ 

Other patient-related factors that are attributed to poor medication adherence are complex regimens that involve taking many medications with varying dosage schedules, long-term drug regimes, inadequate access, cost, inconvenience of taking medications, unpleasant taste, and busy work schedule.^6,23

Provider-related barriers can also contribute to poor medication adherence. Poor provider-patient communication is one of the provider-related barriers that is attributed to medication non-adherence.²⁵ Effective communication is important to improve adherence because it enables patients to understand their illness better, the benefits of adhering to the medication regimen, and the risks that may arise if medication is not adhered to.²⁶ Patient-provider communication is identified as instrumental in the patient-provider relationship which has been identified as a factor that informs medication non-adherence.²⁷ For example, a poor-patient provider relationship can make it difficult for patients to ask relevant questions regarding their medication. Failing to clarify issues of concern can lead to confusion, misunderstanding, and subsequent non-adherence. Poor-patient-provider relationships may also make patients feel that doctors are not telling their patients the whole truth. Such feelings may inform non-adherence to the prescribed medications. Good communication makes patients feel more involved and are likely to take more control over their health including taking medications as prescribed. Provider-related barriers can also result from inadequate knowledge.^28,29,30 Providers who lack adequate knowledge about a disease or the assigned treatment are less likely to provide patients with the needed information. This results in non-adherence. 

Patient-physician discordance has also been attributed to non-adherence. Patient-physician discordance is defined as the difference between patient and physician evaluations of health-related information.⁶ Such discordance leads to poor health outcomes. Research suggests patient-physician discordance as one of the factors that lead to non-adherence.⁶ If patients and physicians have different views of health-related information, conflict is likely to emerge affecting medication adherence. 

Limitations of Traditional Methods of Ensuring Medication Adherence

Traditional measures of ensuring medication adherence have shown mixed results with some showing high success rates in increasing adherence while others showing very little success when it comes to increasing adherence. Some of the most commonly used measures to measure adherence are pill counts, self-reports, measures involving clinician assessment, patient diaries, and patient interviews.^2,31 Some of these methods have had more success than others. For example, although pill count is a simple and low-cost method of assessing medication adherence, it has several limitations. Pill count is not feasible when it comes to assessing non-discrete dosages. Underestimation of adherence also occurs regularly because pill count does not consider the likelihood of having surplus medication. Other limitations that are associated with pill count are the inability to determine whether the dosage units removed are really taken and the inability to determine whether the patient is following the regimen even after the removal of the right number of units.³² 

Clinician assessment and patient self-report have low success rates when it comes to measuring adherence rates. It is difficult to determine whether the patient is correctly reporting adherence by simply relying on an assessment of the same or a self-report.² Patients may also have difficulty communicating or answering the questions as expected. However, despite being highly unreliable, they are still preferred because of low cost, simplicity, and real-time feedback. Patient diaries are also a simple method to measure adherence but they are unreliable, limited by overestimation, and false reporting.³³ 

Medication events monitoring systems (MEMS) have shown more accuracy when it comes to medication adherence compared to methods such as pill counts, patient diaries, and self-reports. Research has shown that MEMS is able to determine non-adherence because it records both the time and date when medication is taken from a container.² As such, it is able to establish whether there is an abnormal medication-taking pattern or the number of doses missed. Research has shown that the ability to note the exact time and date the medication is taken from the bottle makes it more reliable that patient diaries or self-report. Besides, research has shown that MEMS helps to increase adherence.^34,35 

The limitations of traditional measures of medication adherence are what inform the need to explore other measures to increase adherence. AI-powered measures are being considered valuable tools to increase adherence by both predicting adherence and providing a comprehensive assessment of patients’ adherence behavior.³⁶ AI has been explored for different conditions including chronic diseases. In some instances, it has been shown to increase adherence by up to 70 to 80%.³⁷ With such high adherence rates, AI tools are likely to be a good alternative to traditional methods of ensuring medication adherence. 

The Role of AI in Increasing Medication Adherence 

Capitalizing on AI and machine learning-based technologies to increase medication adherence can have a positive impact on patient outcomes. AI-based tools and technologies together with traditional tools of ensuring medication adherence can make a significant impact in this area and lead to positive outcomes. There is research documenting the use of different AI-powered technologies in ensuring medication adherence.¹¹ The technology has a lot of potential when it comes to increasing adherence because it leverages data to offer personalized and intelligent support to patients that can aid in behavioral change and lead to increased adherence. AI-powered technologies and tools have documented varied levels of success in different areas of medication adherence including prediction, reminders, and monitoring of patient adherence. 

One of the areas where AI-based tools and technologies are being used is to predict medication adherence and non-adherence. Predicting adherence or non-adherence is important in increasing medication adherence. It helps to determine the patient’s likelihood of adhering to the prescribed medication regimen. According to Koesmahargyo et al., the prediction of medication adherence traditionally relies on the assessment of factors such as tolerability, treatment length, and demographics.³⁸ However, with the introduction of technologies with the ability to automatically measure medication dosing in real-time, it becomes easier to measure adherence. Such technologies rely on real-time data. As such, it becomes much easier to predict whether the patient will adhere to medication or not. AI-based technologies provide accuracy when it comes to making such predictions. Gu et al. suggest that predicting medication adherence can be done using ensemble learning and deep learning models.³⁹These models provide an opportunity to identify patients who have a high risk of non-adherence within a specific time frame. Wu et al. found that using machine learning algorithms to predict non-adherence in type 2 diabetes patients helped to identify risk factors that increased the likelihood of non-adherence.⁴⁰ Identifying risk factors made it easier to provide individualized diabetes care and education hence increasing the likelihood of adherence. Similarly, Li et al. established that machine learning could be used to predict the risk of non-adherence in patients with type 2 diabetes.⁴¹ Predicting medication adherence and non-adherence is imperative because it helps to identify factors that make it difficult for patients to adhere to medications and design personalized solutions to the problem. For instance, it becomes easier to tailor patient education and care if factors that make it difficult to adhere to medication are known. 

In addition to predicting adherence levels, AI-powered tools can be used to increase adherence. Research has documented the success of AI when it comes to improving medication adherence.^11,12 Labovitz et al. showed that AI improved adherence in patients on anticoagulation therapy by up to 50%.¹² The study showed that even patients who had limited experience with using smartphones were able to use the technology effectively to increase their adherence. The study attributed the improvement in adherence to a number of factors including real-time monitoring, ability to accurately monitor medication ingestion, and change in patient behavior. AI-based technologies can also be used to monitor patient compliance. A novel artificial intelligence platform on mobile devices increased dosing compliance by 17.9% in patients with schizophrenia.⁴² For patients who were monitored using the AI platform, the compliance was 89.7% which was higher than the 71.9% that was recorded for patients who were not monitored using AI.⁴² The AI platform collected data on the date and time stamps the medication was taken for each pill. It also collected data on missed doses, skipped doses, and doses taken in the clinic. Other data that was collected was visual confirmation of the patient ingesting the drug which was done using the tool. Visual confirmation increased the accuracy of the platform when compared to traditional methods of monitoring medication adherence such as pill counts or patients’ self-reports.

AI-powered technologies can also be used to increase medication adherence in conditions that require long-term therapies such as tuberculosis (TB) and chronic diseases. Long-term drug regimens are recognized as a barrier to medication adherence.⁶ Diseases such as TB are managed through long-term therapies with some therapies lasting up to 6 months. These long-term therapies pose a significant challenge to patients which explains why non-adherence is commonly reported in TB patients.⁴³ Considering the importance of adhering to medication when it comes to the management of TB, it is important to use technology to increase adherence. One study has documented the significance of using AI algorithms to increase medication adherence in patients living with TB.⁴⁴ The study documented success with AI algorithms resulting in effective management of TB therapies. Besides, research has documented that digital adherence technologies have high success rates in increasing adherence to TB therapy particularly when they are integrated with clinical strategies.^45,46 Some of these technologies are powered by AI. Other areas where AI can be used to increase medication adherence are to increase patients’ rates of refill, monitor patients’ adherence, and monitor adverse events.

AI has a lot of potential when it comes to increasing medication adherence. Leveraging this potential can lead to positive outcomes for both patients and healthcare systems. AI-powered technologies can be used to address patient-related barriers that affect medication adherence such as forgetfulness and poor reporting. Leveraging AI can also help to reduce costs associated with medication non-adherence. Some studies have documented that AI-powered tools significantly increase adherence which means that they can help to fill the gaps left by traditional methods of measuring adherence such as pill counts, patient self-reports, and clinical assessment. However, to gain the full benefits of AI, there is a need to address the challenges posed by AI use in monitoring medication adherence. 

Challenges of Using AI to Increase Medication Adherence

Although AI shows a lot of potential in addressing medication adherence challenges, it presents unique barriers that need to be addressed to optimize its potential. One of the unique barriers presented by AI use in medication adherence is ethical issues.¹¹ The use of AI in healthcare has always presented a challenge when it comes to patient data, privacy, and confidentiality.⁴⁷ AI and machine learning deal with large datasets most of which contain patient-sensitive information. As a result, securing patient privacy and confidentiality when it comes to the use of this data becomes challenging. Securing patient privacy and confidentiality when it comes to the use of AI in ensuring medication adherence becomes particularly challenging considering such technologies require the use of patient data to be effective. For instance, to ensure patients adhere to medication, it is important to have personal data on these patients. Securing this data and ensuring patient anonymity can be challenging limiting the use of AI to increase medication adherence. 

An additional issue that emerges from the use of AI to increase medication adherence is the lack of knowledge and skills on how to effectively use these technologies. Research has documented limited skill sets as one of the challenges that limit the use of AI in healthcare. A lack of knowledge and skills on how to use these technologies to monitor adherence can limit their use and implementation.³⁶ Limitations of knowledge on the available AI-based technologies also limit their use. Although significant advances have been made in the development of AI-based solutions for different areas of healthcare, most of these technologies are not known to healthcare providers limiting their use. Addressing the limitations in knowledge and skillsets can increase AI use in monitoring medication adherence. 

Conclusion

Increasing medication adherence is vital to improve patient outcomes. However, despite the implementation of different measures to improve adherence, it still remains a significant challenge with these measures producing mixed results. Most of the traditional measures of ensuring medication adherence are limited by the fact that they do not verify medication administration. Without verifying medication administration, it becomes really challenging to measure adherence. Increasing medication adherence is recognized as instrumental in improving patient outcomes. Non-adherence also compounds the challenge of improving health. Other problems attributed to non-adherence are increased morbidity, increased mortality, increased hospital visits and admissions, significant disease progression, and poor quality of life. Non-adherence also costs healthcare systems billions of dollars each year. Leveraging technologies such as AI can help to address the limitations of traditional approaches to ensuring medication adherence. AI-powered technologies are being used in different areas of healthcare and medication adherence is not an exemption. AI can be used to both predict adherence and improve adherence. However, to gain the full benefits offered by AI, it is important to address the challenges these technologies present. For instance, ethical issues remain a prevalent problem when it comes to AI use in healthcare. Maintaining patient privacy and confidentiality on sensitive issues is a significant problem. The use of AI to increase medication adherence is also limited by limited knowledge and skills on how to use this technology effectively. 

References

1. Lee EK, Poon P, Yip BH, et al. Global burden, regional differences, trends, and health consequences of medication nonadherence for hypertension during 2010 to 2020: a meta‐analysis involving 27 million patients. Journal of the American Heart Association. 2022;11(17):e026582.  https://doi.org/10.1161/JAHA.122.026582 PMid:36056737 PMCid:PMC9496433

2. Lam WY, Fresco P. Medication adherence measures: an overview. Biomed Research International. 2015;2015(1):217047. https://doi.org/10.1155/2015/217047 PMid:26539470 PMCid:PMC4619779

3. Brown MT, Bussell JK. Medication adherence: WHO cares?. In Mayo Clinic Proceedings 2011 Apr 1 (Vol. 86, No. 4, pp. 304-314). Elsevier. https://doi.org/10.4065/mcp.2010.0575 PMid:21389250 PMCid:PMC3068890

4. Kim J, Combs K, Downs J, Tillman F. Medication adherence: The elephant in the room. Us Pharm. 2018;43(1):30-4.

5. Neiman AB. CDC grand rounds: improving medication adherence for chronic disease management-innovations and opportunities. MMWR. Morbidity and Mortality Weekly Report. 2017;66. https://doi.org/10.15585/mmwr.mm6645a2 PMid:29145353 PMCid:PMC5726246

6. Jimmy B, Jose J. Patient medication adherence: measures in daily practice. Oman Medical Journal. 2011;26(3):155. https://doi.org/10.5001/omj.2011.38 PMid:22043406 PMCid:PMC3191684

7. Treskes RW, Van der Velde ET, Schoones JW, Schalij MJ. Implementation of smart technology to improve medication adherence in patients with cardiovascular disease: is it effective?. Expert Review of Medical Devices. 2018;15(2):119-26. https://doi.org/10.1080/17434440.2018.1421456 PMid:29271661

8. Dave P, How Digital Health is Revolutionizing Healthcare and Contributing to Positive Health Outcomes, Journal of Drug Delivery and Therapeutics. 2024; 14(6):287-293 https://doi.org/10.22270/jddt.v14i6.6640

9. Mason M, Cho Y, Rayo J, Gong Y, Harris M, Jiang Y. Technologies for medication adherence monitoring and technology assessment criteria: narrative review. JMIR Mhealth and Uhealth. 2022;10(3):e35157. https://doi.org/10.2196/35157 PMid:35266873 PMCid:PMC8949687

10. Bohlmann A, Mostafa J, Kumar M. Machine learning and medication adherence: scoping review. JMIRx Med. 2021;2(4):e26993. https://doi.org/10.2196/26993 PMid:37725549 PMCid:PMC10414315

11. Babel A, Taneja R, Mondello Malvestiti F, Monaco A, Donde S. Artificial intelligence solutions to increase medication adherence in patients with non-communicable diseases. Frontiers in Digital Health. 2021;3:669869. https://doi.org/10.3389/fdgth.2021.669869 PMid:34713142 PMCid:PMC8521858

12. Labovitz DL, Shafner L, Reyes Gil M, Virmani D, Hanina A. Using artificial intelligence to reduce the risk of nonadherence in patients on anticoagulation therapy. Stroke. 2017;48(5):1416-9. https://doi.org/10.1161/STROKEAHA.116.016281 PMid:28386037 PMCid:PMC5432369

13. Pinho S, Cruz M, Ferreira F, Ramalho A, Sampaio R. Improving medication adherence in hypertensive patients: A scoping review. Preventive Medicine. 2021;146:106467. https://doi.org/10.1016/j.ypmed.2021.106467 PMid:33636195

14. World Health Organization. Adherence to long-term therapies: evidence for action. World Health Organization. 2003. Available from: https://iris.who.int/handle/10665/42682 

15. Glombiewski JA, Nestoriuc Y, Rief W, Glaesmer H, Braehler E. Medication adherence in the general population. PLoS One. 2012;7(12):e50537. https://doi.org/10.1371/journal.pone.0050537 PMid:23272064 PMCid:PMC3525591

16. Yeaw J, Benner JS, Walt JG, Sian S, Smith DB. Comparing adherence and persistence across 6 chronic medication classes. Journal of Managed Care Pharmacy. 2009;15(9):728-40. https://doi.org/10.18553/jmcp.2009.15.9.728 PMid:19954264 PMCid:PMC10441195

17. Mongkhon P, Ashcroft DM, Scholfield CN, Kongkaew C. Hospital admissions associated with medication non-adherence: a systematic review of prospective observational studies. BMJ Quality & Safety. 2018;27(11):902-14. https://doi.org/10.1136/bmjqs-2017-007453 PMid:29666309

18. Dave P, How AI Can Revolutionize the Pharmaceutical Industry. Journal of Drug Delivery and Therapeutics. 2024; 14(6):179-183. https://doi.org/10.22270/jddt.v14i6.6657

19. Cutler RL, Fernandez-Llimos F, Frommer M, Benrimoj C, Garcia-Cardenas V. Economic impact of medication non-adherence by disease groups: a systematic review. BMJ Open. 2018;8(1):e016982. https://doi.org/10.1136/bmjopen-2017-016982 PMid:29358417 PMCid:PMC5780689

20. Van Alsten SC. Cost-related nonadherence and mortality in patients with chronic disease: a multiyear investigation, National Health Interview Survey, 2000-2014. Preventing Chronic Disease. 2020;17. https://doi.org/10.5888/pcd17.200244 PMid:33274701 PMCid:PMC7735485

21. Berkman ND, Sheridan SL, Donahue KE, Halpern DJ, Crotty K. Low health literacy and health outcomes: an updated systematic review. Annals of Internal Medicine. 2011;155(2):97-107. https://doi.org/10.7326/0003-4819-155-2-201107190-00005 PMid:21768583

22. Shahid R, Shoker M, Chu LM, Frehlick R, Ward H, Pahwa P. Impact of low health literacy on patients' health outcomes: a multicenter cohort study. BMC Health Services Research. 2022;22(1):1148. https://doi.org/10.1186/s12913-022-08527-9 PMid:36096793 PMCid:PMC9465902

23. Khan MU, Shah S, Hameed T. Barriers to and determinants of medication adherence among hypertensive patients attended National Health Service Hospital, Sunderland. Journal of Pharmacy and Bioallied Sciences. 2014;6(2):104-8. https://doi.org/10.4103/0975-7406.129175 PMid:24741278 PMCid:PMC3983739

24. Smith LE, Webster RK, Rubin GJ. A systematic review of factors associated with side‐effect expectations from medical interventions. Health Expectations. 2020;23(4):731-58. https://doi.org/10.1111/hex.13059 PMid:32282119 PMCid:PMC7495066

25. Ghosh P, Balasundaram S, Sankaran A, Chandrasekaran V, Sarkar S, Choudhury S. Factors associated with medication non-adherence among patients with severe mental disorder-a cross sectional study in a tertiary care centre. Exploratory Research in Clinical and Social Pharmacy. 2022;7:100178. https://doi.org/10.1016/j.rcsop.2022.100178 PMid:36161207 PMCid:PMC9493377

26. Świątoniowska-Lonc N, Polański J, Tański W, Jankowska-Polańska B. Impact of satisfaction with physician-patient communication on self-care and adherence in patients with hypertension: cross-sectional study. BMC Health Services Research. 2020;20:1-9. https://doi.org/10.1186/s12913-020-05912-0 PMid:33198739 PMCid:PMC7670590

27. Stavropoulou C. Non-adherence to medication and doctor-patient relationship: Evidence from a European survey. Patient Education and Counseling. 2011;83(1):7-13. https://doi.org/10.1016/j.pec.2010.04.039 PMid:20541884

28. Dave P, The Vital Role of Pharmacists in Diabetes Self-care, Journal of Drug Delivery and Therapeutics. 2024; 14(5):229-233 https://doi.org/10.22270/jddt.v14i5.6582

29. Krishnamoorthy Y, Rajaa S, Rehman T, Thulasingam M. Patient and provider's perspective on barriers and facilitators for medication adherence among adult patients with cardiovascular diseases and diabetes mellitus in India: a qualitative evidence synthesis. BMJ Open. 2022;12(3):e055226. https://doi.org/10.1136/bmjopen-2021-055226 PMid:35332041 PMCid:PMC8948385

30. Brown MT, Bussell J, Dutta S, Davis K, Strong S, Mathew S. Medication adherence: truth and consequences. The American Journal of the Medical Sciences. 2016;351(4):387-99. https://doi.org/10.1016/j.amjms.2016.01.010 PMid:27079345

31. Ho PM, Bryson CL, Rumsfeld JS. Medication adherence: its importance in cardiovascular outcomes. Circulation. 2009;119(23):3028-35. https://doi.org/10.1161/CIRCULATIONAHA.108.768986 PMid:19528344

32. Van Onzenoort HA, Verberk WJ, Kessels AG, et al. Assessing medication adherence simultaneously by electronic monitoring and pill count in patients with mild-to-moderate hypertension. American Journal of Hypertension. 2010;23(2):149-54. https://doi.org/10.1038/ajh.2009.207 PMid:19927136

33. Nunes V, Neilson J, O'flynn N, et al. Medicines adherence: involving patients in decisions about prescribed medicines and supporting adherence. Other. NICE: National Institute for Health and Clinical Excellence, London, UK. Available from: https://www.ncbi.nlm.nih.gov/books/NBK55447/ 

34. van Heuckelum M, van den Ende CH, Houterman AE, Heemskerk CP, van Dulmen S, van den Bemt BJ. The effect of electronic monitoring feedback on medication adherence and clinical outcomes: A systematic review. Plos One. 2017;12(10):e0185453. https://doi.org/10.1371/journal.pone.0185453 PMid:28991903 PMCid:PMC5633170

35. Demonceau J, Ruppar T, Kristanto P, et al. Identification and assessment of adherence-enhancing interventions in studies assessing medication adherence through electronically compiled drug dosing histories: a systematic literature review and meta-analysis. Drugs. 2013;73:545-62. https://doi.org/10.1007/s40265-013-0041-3 PMid:23588595 PMCid:PMC3647098

36. Pulice E, Coustasse A. AI-Driven solutions promote medication adherence. Pharmacy Times. Available from: https://www.pharmacytimes.com/view/ai-driven-solutions-promote-medication-adherence 

37. Warren D, Marashi A, Siddiqui A, et al. Using machine learning to study the effect of medication adherence in Opioid Use Disorder. Plos One. 2022;17(12):e0278988. https://doi.org/10.1371/journal.pone.0278988 PMid:36520864 PMCid:PMC9754174

38. Koesmahargyo V, Abbas A, Zhang L, et al. Accuracy of machine learning-based prediction of medication adherence in clinical research. Psychiatry Research. 2020;294:113558. https://doi.org/10.1016/j.psychres.2020.113558 PMid:33242836

39. Gu Y, Zalkikar A, Liu M, et al. Predicting medication adherence using ensemble learning and deep learning models with large scale healthcare data. Scientific Reports. 2021;11(1):18961. https://doi.org/10.1038/s41598-021-98387-w PMid:34556746 PMCid:PMC8460813

40. Wu XW, Yang HB, Yuan R, Long EW, Tong RS. Predictive models of medication non-adherence risks of patients with T2D based on multiple machine learning algorithms. BMJ Open Diabetes Research and Care. 2020;8(1):e001055. https://doi.org/10.1136/bmjdrc-2019-001055 PMid:32156739 PMCid:PMC7064141

41. Li M, Lu X, Yang H, et al. Development and assessment of novel machine learning models to predict medication non-adherence risks in type 2 diabetics. Frontiers in Public Health. 2022;10:1000622. https://doi.org/10.3389/fpubh.2022.1000622 PMid:36466490 PMCid:PMC9714465

42. Bain EE, Shafner L, Walling DP, Othman AA, Chuang-Stein C, Hinkle J, Hanina A. Use of a novel artificial intelligence platform on mobile devices to assess dosing compliance in a phase 2 clinical trial in subjects with schizophrenia. JMIR mHealth and uHealth. 2017;5(2):e7030. https://doi.org/10.2196/mhealth.7030 PMid:28223265 PMCid:PMC5340925

43. Sazali MF, Rahim SS, Mohammad AH, et al. Improving tuberculosis medication adherence: the potential of integrating digital technology and health belief model. Tuberculosis and Respiratory Diseases. 2023;86(2):82. https://doi.org/10.4046/trd.2022.0148 PMid:36597583 PMCid:PMC10073608

44. Kim K, Kim B, Chung AJ, Kwon K, Choi E, Nah JW. Algorithm and System for improving the medication adherence of tuberculosis patients. In2018 International Conference on Information and Communication Technology Convergence (ICTC). 2018 (pp. 914-916). IEEE. https://doi.org/10.1109/ICTC.2018.8539402

45. Liu X, Thompson J, Dong H, Sweeney S, Li X, Yuan Y, Wang X, He W, Thomas B, Xu C, Hu D. Digital adherence technologies to improve tuberculosis treatment outcomes in China: a cluster-randomised superiority trial. The Lancet Global Health. 2023;11(5):e693-703. https://doi.org/10.1016/S2214-109X(23)00068-2 PMid:37061308

46. Subbaraman R, de Mondesert L, Musiimenta A, Pai M, Mayer KH, Thomas BE, Haberer J. Digital adherence technologies for the management of tuberculosis therapy: mapping the landscape and research priorities. BMJ Global Health. 2018;3(5):e001018. https://doi.org/10.1136/bmjgh-2018-001018 PMid:30364330 PMCid:PMC6195152

47. Lee D, Yoon SN. Application of artificial intelligence-based technologies in the healthcare industry: Opportunities and challenges. International Journal of Environmental Research and Public Health. 2021;18(1):271. https://doi.org/10.3390/ijerph18010271 PMid:33401373 PMCid:PMC7795119