Mucherino, Sara (2022) Adherence patterns across multiple medications in chronic diseases: an innovative drug utilization model. [Tesi di dottorato]

	Testo Mucherino_Sara_COMPLETA_35.pdf Visibile a [TBR] Amministratori dell'archivio Download (7MB) | Richiedi una copia
Anteprima	Testo Mucherino_Sara_OSCURATA_35.pdf Download (5MB) | Anteprima

Abstract

Medication-taking behavior is extremely complex and individual, requiring numerous multifactorial strategies to improve medication adherence (MA). Hence, MA is a key factor associated with the effectiveness of all pharmacological therapies but is particularly critical for medications used for chronic conditions. The treatment of chronic illnesses often includes the long-term use of pharmacotherapy, but although these medications are effective in treating chronic diseases, their full benefits are often not realized because ~50% of patients do not take their medications as prescribed. Therefore, is widely recognized that suffering from one or multiple chronic conditions with a corresponding increase in medication utilization are at an increased risk of medication nonadherence. Despite the central importance of medication adherence in clinical practice and policy, medication adherence is difficult to define and measure. One of the possible reasons for the difficulty in uniquely assessing, predicting, and measuring adherence to drug therapies is the lack of a harmonized process for measuring adherence and the use of routine measures of adherence in clinical practice. Hence, indicators to measure adherence through pharmacy claims databases generally return a static and dichotomous measure of MA (Adherent/Not-Adherent). This problem stems from an underlying misconception about the nature of adherence, as the idea that adherence is a single stable behavior, instead of the reality that adherence encompasses a set of different and dynamic behaviors. While definitions have evolved over time (e.g. from compliance to adherence and persistence), the more recent developments on the EMERGE Guidelines have moved towards defining separate elements of adherence (initiation, implementation, and persistence) that are thought to describe the processes involved in medication taking, treating the term “adherence” as an overarching term. Therefore, in addition to the definition, the measurement of adherence through the use of both direct and indirect methods is also reaching a new frontier: Medication adherence is a process divided into three operational and quantifiable phases. In this scenario, this dissertation has explored and faced challenges in medication adherence research and its relation with patient complexity in terms of multimorbidity and polypharmacy by implementing an innovative drug-utilization (DU) models based on longitudinal calculation of medication adherence by exploiting the crasis between: DU research, ML/AI models (Data science applications) and medication adherence to major chronic diseases. Main findings of this PhD thesis address all the developments and discoveries observed to date regarding the measurement of medication adherence through indirect methods, namely the use of Big Data. Such joining tract between disciplines has enabled the implementation of a recently developed algorithm by Dima A. and colleagues, allowing measurement and visualization of all adherence profiles of patients treated with specific drug therapies throughout the entire pharmacological treatment period. The algorithm was implemented by characterizing patients with similar medication adherence estimates and evaluating their baseline and clinical characteristics as potential determinants of non-adherence. Thus, findings address that medication nonadherence is a complex problem rooted in a multitude of interconnected factors some of them modifiable and predictable upstream. Future studies are needed to understand the underlying complexity and guide future interventions in real clinical practice.

Downloads

Actions (login required)

Modifica documento