2018 Conference Presentation
Background: Technological and scientific advances have produced a society with a high percentage of elderly people and patients with chronic diseases; these patients, who require long-term care, are frequent high users of health care services. The systematic use of stratification tools and predictive models for this group of patients can be useful for health professionals in decision-making processes.
Objective: The objective of this paper is to describe two new classifier systems to detect the risk of hospital admission of patients with chronic conditions, that can be used in health databases, and whose indicators facilitate the decision-making in the management of these patients from a primary health care starting point.
Methods: A set of variables, related to hospital admission of patients with chronic conditions, has been obtained by means of various empirical methods (focus groups, health database analysis, statistical processing). The pilot sample consisted of 1000 patients. To predict the probability of admission from the set of predictor variables, a logistic regression within the framework of Generalised Linear Models was used.
Results: Two classifier systems were designed. In the first one, the important variables were two: the total number of pathologies and the total number of emergency visits. While the second one included eight: the presence of respiratory diseases, heart diseases, diabetes, chronic pain, requiring palliative care, hemiplegia, and those who have already had a hospital emergency or home emergency attended by a primary health care unit. In order to measure the specificity and sensitivity of both classifier systems, reasonable values of 0.722 and 0.744 (first and second system respectively) of the area under the ROC curve were obtained.
Conclusions: The proposed classifier systems facilitate a change in the management model of patients requiring long-term care, which becomes proactive since the risk information is available for the physician’s reference in the computerised primary care system.