Editorial: Mathematical Modeling of Medication Nonadherence

Sean D Lawley^1*Teresa Gibson²Jiao Zheng³

• ¹The University of Utah, Salt Lake City, United States
• ²Rochester Institute of Technology, Rochester, United States
• ³Shanghai Jiao Tong University, Shanghai, China

Medication adherence is the process by which patients take their medications as prescribed [1]. Medication nonadherence is an age-old problem, as even Hippocrates warned physicians to "keep a watch also on the faults of the patients, which often make them lie about the taking of things prescribed" [2]. Today, medication nonadherence results in over 100,000 preventable deaths and more than $100 billion in healthcare costs each year in the United States alone [3,4].In fact, the World Health Organization has claimed: [5] "increasing the effectiveness of adherence interventions may have a far greater impact on the health of the population than any improvement in specific medical treatments."A former United States surgeon general famously observed, "drugs don't work in patients who don't take them." [6].Medication nonadherence is challenging to understand and mitigate for many reasons. For one, nonadherence can be erratic, as patients do not miss doses on neat, predictable schedules. Indeed, adherence data for an individual patient over time shows doses taken on-time, doses taken late, doses skipped, double doses, etc. [7]. Furthermore, these seemingly stochastic, temporal fluctuations in adherence vary considerably between patients [8], with perhaps at least six qualitatively distinct patterns seen in different patients [9]. In addition to temporal and patient heterogeneity, the physiological consequences of nonadherence can vary considerably between drugs. For example, missing a morning dose of one medication may go largely undetected by the patient or might entail a lethargic afternoon, but a missing a dose of an antiepileptic drug might cause a seizure [10]. Moreover, even within a specific drug class, some drugs maintain efficacy despite lapses in adherence (so-called "forgiving" drugs), whereas other drugs require nearly perfect adherence to be effective [11].Due to this complexity, mathematical and computational approaches are emerging as powerful tools to combat medication nonadherence. For instance, stochastic analysis can leverage the science of pharmacometrics to investigate remedial dosing protocols and design regimens to mitigate nonadherence [12][13][14][15][16][17][18][19][20][21]. Furthermore, "drug forgiveness" has been quantified with a number of approaches [22][23][24][25][26][27][28][29][30][31][32][33][34][35], which enables the identification and design of drugs which are robust to nonadherence. In effect, the mathematical equations are a laboratory to conduct experiments which would not be feasible in human trials, such as quantifying how therapeutic efficacy depends on adherence rates and patterns. In addition, methods employing artificial intelligence (AI) and sophisticated statistical approaches are being applied to swaths of adherence data to predict, detect, and ameliorate nonadherence [36][37][38][39][40].This Research Topic presents five articles in this burgeoning field. Three