Desenvolupament d'un model predictiu d'ingressos i reingressos hospitalaris no programats a Catalunya
Abstract
Objectives: the aim of this study was to analyze risk factors and probabilities of unexpected
hospitalizations and re-hospitalizations in the population attending primary care centers in an
area of Catalonia.
Methodology: a longitudinal retrospective study was carried out including data on PC and the Minimum Basic Data Set on Hospital Discharge of the Bellvitge and Viladecans hospitals, from
01/01/2006 to 12/31/2008 from the population attending PC centers of the sanitary subsector
Baix Llobregat (municipal areas of Castelldefels, Gavà, Viladecans, Begues and Sant Climent de Llobregat). Dependent variables were unexpected hospitalization and rehospitalization in 12 months and 6 months periods respectively, during the year 2008. Any unexpected admission between 01/01/2008 and 12/31/2008 was considered as the Index admission, and re-hospitalization was any unexpected readmission within the 6 months after the index admission. Sociodemographic factors (age, sex, area of residence), morbidity (selected diagnoses according to the ICD 10 Edition classification), and the use of services in the previous 2 years (visits to the health center, expended medication, previous hospital admissions, and cumulative days of stay at hospital) were antecedents included in the study. Logistic regression models of admissions and readmission were adjusted taking into account morbidity and use of services, stratified by sex and in the whole population.
Keywords
Predictive model; Hospital readjustments; Catalonia
Bibliographic citation
Rajmil L, López-Aguilà S. Desenvolupament d'un model predictiu d'ingressos i reingressos hospitalaris no programats a Catalunya. Barcelona: Agència d'Avaluació de Tecnologia i Recerca Mèdiques; 2010.
Audience
Professionals
Use this identifier for quote and/or link this document
https://hdl.handle.net/11351/3082This item appears in following collections
The following license files are associated with this item: