Trajectories of Chronic Multimorbidity Patterns in Older Patients

Overview

Following the MRisk-COVID project, MTOP (Multimorbidity Trajectories in Older Patients) study was developed. It is a retrospective observational study using Real World Data that aims to identify patterns of chronic multimorbidity in patients aged ≥65 years and their evolution and trajectories in the previous 10 years. The secondary objective is to identify the relationship between the trajectories of multimorbidity patterns in the previous 10 years and the severity of the infection by COVID-19.

Multimorbidity is associated with negative results and presents difficulties in clinical management. Recently, new methodologies are emerging based on the hypothesis that chronic conditions are associated in a non-random way forming multimorbidity patterns. However, there are few studies that study the temporal evolution and trajectories of these multimorbidity patterns, which could be associated with different prognoses and could allow better forecasting and planning. The primary objective of this analysis is to identify patterns of chronic multimorbidity in patients aged ≥65 years and their evolution and trajectories in the previous 10 years, using part of the MRisk-COVID project data. As a secondary objective, the investigators want to identify the relationship between the trajectories of multimorbidity patterns in the previous 10 years and the severity of the COVID-19 infection. This retrospective observational study has a historical cohort of 3958 patients ≥65 years of age suspected and confirmed of COVID-19 infection from February 1 to June 15, 2020 in the reference area of Parc Taulí University Hospital. The available data (real-world data) are socio-demographic and diagnostic variables provided by the Data Analytics Program for Research and Innovation in Health (PADRIS), which include sex, age and primary care diagnoses. To identify patterns of multimorbidity, the Clinical Classification Software, Chronic Condition Indicator, multiple correspondence analysis and cluster analysis using the fuzzy c-means algorithm will be used. Then, a clinical consensus process (Delphi-like) will be made of the clusters obtained. To identify the most probable trajectories along the three time points, each patient will be assigned to the cluster with the highest probability of membership. Descriptive and bivariate statistics will be performed.

Conditions

Eligibility

Locations (1)

Outcomes