Predicción de dependencia en ancianos atendidos en atención primaria = Prediction of dependency in the elderly attended in primary health care

Puigvert Vilalta, Margarita
Martí Lluch, Ruth
Ponjoan, Anna
Objective: To identify the elderly population attended in primary health care with more risk of becoming dependent for activities of daily living. Methods: A longitudinal study, based on data recorded in the medical record of people from 75 years old. We constructed a Cox models to determine the risk of becoming dependent on the follow up. The discrimination capacity of the model was estimated with the concordance index. Results: The model showed as predictors of dependency: the mental impairment, according to the Pfeiffer test, severe (HR: 3,02; 95% CI 2,03-4,49) or moderate (HR: 1,35; 95% CI 1,08 to 1,68), homecare situation (HR: 1,82; 95% CI 1,44 to 2,31) and to have recorded history of falls (HR: 1,51; 95% CI 1,20 to 1,90). The C index of the derivation model was 0,72 (95% CI 0,68 to 0,76). Conclusions: The model accurately predicts patients most likely to be dependent ​
This document is licensed under a Creative Commons:Attribution - Non commercial (by-nc) Creative Commons by-nc3.0