In this thesis I will focus on the use of multilevel models on dyadic data. This topic was driven by a randomized clinical trial (RCT) on the efficacy of a motivational interview (MI) in improving self-care in patients with heart failure (HF). Longitudinal mixed models were used to evaluate the effect of MI on a battery of primary and secondary outcomes in patients and caregivers, including self-care, caregiver contribution to self-care, HF symptoms, generic and specific quality of life, anxiety, depression, sleep quality, caregiver preparedness, use of emergency services and mortality. Secondary outcomes included also the mutuality between patient and caregiver, which evaluates the positive quality of the care relationship by separate questionnaires to the patients and to the caregivers. Using the dyad as the unit of analysis, rather than the individuals, multilevel modelling is an appropriate and effective methodology for considering the dyadic context. In this framework it is possible to investigate the impact of health and behaviour changes on both the patient and caregiver's perceived relationship quality. The problem of interdependence in dyadic data is further complicated when measuring dyadic outcomes repeatedly over time. Then, it is necessary not only to account for the non-independence of the members within a dyad, but also for the dependence of the longitudinal measures within one member. Therefore, this thesis studies the application of multilevel models to longitudinal dyadic data in healthcare research, evaluating advantages and disadvantages of this approach and commenting on the implications of these models.
Questa tesi si focalizza sull'utilizzo di modelli multilevel su dati diadici. Questo argomento è stato applicato su un trial clinico randomizzato sull'efficacia di un colloquio motivazionale (MI) nel migliorare il self-care nei pazienti con scompenso cardiaco. Sono stati utilizzati modelli lineari ad effetti misti longitudinali per valutare l'effetto del MI su una batteria di end-points primari e secondari nei pazienti e nei caregiver, inclusi il self-care, il contributo del caregiver al self-care, i sintomi dello scompenso cardiaco, la qualità della vita generica e specifica, l'ansia, la depressione, qualità del sonno, preparazione del caregiver, uso dei servizi ospedalieri e mortalità. Gli end-points secondari includevano anche la mutualità tra paziente e caregiver, che valuta la qualità positiva della relazione di cura mediante questionari separati somministrati ai pazienti e ai caregiver. Utilizzando la diade come unità di analisi, piuttosto che i singoli individui, i modelli multilevel sono una metodologia appropriata ed efficace per considerare il contesto diadico. In questo contesto è possibile esaminare l'impatto della salute e dei cambiamenti comportamentali sulla qualità della relazione percepita sia dal paziente che dal caregiver. Il problema dell'interdipendenza nei dati diadici è ulteriormente complicato quando si misurano ripetutamente gli outcomes diadici nel tempo. Quindi, è necessario non solo tenere conto della non indipendenza dei membri all'interno di una diade, ma anche della dipendenza delle misure longitudinali all'interno di uno stesso membro. Pertanto, questa tesi studia l'applicazione di modelli multilevel applicati ai dati diadici longitudinali nella ricerca sanitaria, valutando vantaggi e svantaggi di questo approccio e commentando le implicazioni di questi modelli.
(2023). The efficacy of a motivational interview intervention in patients with heart failure and their caregivers: a dyadic analysis. (Tesi di dottorato, Università degli Studi di Milano-Bicocca, 2023).
The efficacy of a motivational interview intervention in patients with heart failure and their caregivers: a dyadic analysis
OCCHINO, GIUSEPPE
2023
Abstract
In this thesis I will focus on the use of multilevel models on dyadic data. This topic was driven by a randomized clinical trial (RCT) on the efficacy of a motivational interview (MI) in improving self-care in patients with heart failure (HF). Longitudinal mixed models were used to evaluate the effect of MI on a battery of primary and secondary outcomes in patients and caregivers, including self-care, caregiver contribution to self-care, HF symptoms, generic and specific quality of life, anxiety, depression, sleep quality, caregiver preparedness, use of emergency services and mortality. Secondary outcomes included also the mutuality between patient and caregiver, which evaluates the positive quality of the care relationship by separate questionnaires to the patients and to the caregivers. Using the dyad as the unit of analysis, rather than the individuals, multilevel modelling is an appropriate and effective methodology for considering the dyadic context. In this framework it is possible to investigate the impact of health and behaviour changes on both the patient and caregiver's perceived relationship quality. The problem of interdependence in dyadic data is further complicated when measuring dyadic outcomes repeatedly over time. Then, it is necessary not only to account for the non-independence of the members within a dyad, but also for the dependence of the longitudinal measures within one member. Therefore, this thesis studies the application of multilevel models to longitudinal dyadic data in healthcare research, evaluating advantages and disadvantages of this approach and commenting on the implications of these models.File | Dimensione | Formato | |
---|---|---|---|
phd_unimib_771380.pdf
accesso aperto
Descrizione: DOCUMENTO DI TESI
Tipologia di allegato:
Doctoral thesis
Dimensione
2.45 MB
Formato
Adobe PDF
|
2.45 MB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.