PDF(Sci-hub)
Abstract：
OBJECTIVE: Chronic diseases are the main cause of mortality worldwide. Study of the most prevalent diseases is essential, as well as the development of indicators of health services\' utilization and mortality in these patients. The objective of this study was to identify which comorbidity measure best predicts health services\' utilization and mortality in patients with diabetes mellitus in our environment.
METHODS: A longitudinal study was carried out in a cohort of diabetes mellitus patients diagnosed in 2006 in Zaragoza and followed up to 2010. Logistic regression predictive models were developed. The number of diagnosis, the number of ambulatory diagnostic groups (ADG), and the number of major ambulatory diagnostic groups (MADG) from the Ambulatory Care Groups system were used as comorbidity measures. The validity measure consisted of the improvement in the model\'s explanatory capacity (c-statistic).
RESULTS: The prevalence of diabetes mellitus was 8.8%. Both the number of diagnoses and comorbidity were associated with health services\' utilization and mortality. For mortality, the best indicator of comorbidity was the number of MADGs (c=0.763). The model adjusted by sex, age, number of MADGs, and number of hospitalizations had the highest explanatory capacity (c=0.818).
CONCLUSIONS: The ACG system allows resource consumption and mortality to be predicted in people with diabetes mellitus in our environment. This study confirms the substantial healthcare burden generated by patients with diabetes mellitus and the need to tackle this situation.
METHODS: A longitudinal study was carried out in a cohort of diabetes mellitus patients diagnosed in 2006 in Zaragoza and followed up to 2010. Logistic regression predictive models were developed. The number of diagnosis, the number of ambulatory diagnostic groups (ADG), and the number of major ambulatory diagnostic groups (MADG) from the Ambulatory Care Groups system were used as comorbidity measures. The validity measure consisted of the improvement in the model\'s explanatory capacity (c-statistic).
RESULTS: The prevalence of diabetes mellitus was 8.8%. Both the number of diagnoses and comorbidity were associated with health services\' utilization and mortality. For mortality, the best indicator of comorbidity was the number of MADGs (c=0.763). The model adjusted by sex, age, number of MADGs, and number of hospitalizations had the highest explanatory capacity (c=0.818).
CONCLUSIONS: The ACG system allows resource consumption and mortality to be predicted in people with diabetes mellitus in our environment. This study confirms the substantial healthcare burden generated by patients with diabetes mellitus and the need to tackle this situation.
摘要：
暂无翻译
