Analysis of psychological variables in hemodialysis patients using machine learning algorithms

Abstract

Introduction: Chronic kidney disease is a health problem that is increasing, highly related to psychological problems that often determine its evolution.

Objective: Describe the prevalence and association of psychological and demographic variables in relation to resilience in hemodialysis patients at the “Arnaldo Milián Castro” Hospital in Santa Clara.

Methods: Descriptive cross-sectional study, with non-probabilistic sampling, the 25-item resilience scales (CD-RISC) were used; The Hospital Anxiety and Depression Scale (HADS) and the Stress Symptom Scale (ESE). The processing was carried out with descriptive statistics and artificial intelligence algorithms.

Results: Resilience behavior was predominantly moderate in 54.5% of cases; Anxiety and depression did not indicate a case in 78.8% and 84.8% of the sample, while stress was classified as pathological in 57.6% of the patients. The variables that best associate and predict resilience were control under pressure (C.B.P), persistence, tenacity and self-efficacy (P.T.A) and adaptability and support networks (A.R.A), stress was the manifestation best related to resilience.

Conclusions: It is concluded that resilience is better predicted from the variations of the C.B.P, the P.T.A and A.R.A and that it is negatively linked to stress, with anxiety and depression not having a high predictive value in the machine learning model.