Agathe Guenego1,2, Guillaume Bouzillé2,3,4, Stéphanie Breitel5, Annabelle Esvant1,2, Jean-Yves Poirier1, Fabrice Bonnet1,2, Isabelle Guilhem1. 1. 1 Department of Endocrinology, Diabetes and Nutrition, Hôpital sud , CHU de Rennes, Rennes, France . 2. 2 Université de Rennes 1 , Rennes, France . 3. 3 INSERM , U1099, Rennes, France . 4. 4 CIC Inserm 1414, CHU Rennes , Rennes, France . 5. 5 AIR de Bretagne , Z.A.C Atalante Villejean, Rennes, France .
Abstract
BACKGROUND: Insulin pump failures had been assessed in our center by a prospective observational study from 2001 to 2007. The aim of this study was to update our data since 2008 and to determine whether there exist specific risk factors for insulin pump failures. METHODS: All insulin pump defects were prospectively collected between 2008 and 2013 in a monocentric cohort of 350 new pumps. Clinical consequences were recorded. Brand and model of pumps and type of defects and patients' characteristics (gender, type of diabetes, age at diabetes diagnosis, age at first pump, pump treatment duration, number of previous pumps, and number of previous pump failures) were tested for possible association with insulin pump failure. RESULTS: Malfunctions occurred in 239 (68%) pumps. The incidence rate was 33/100 pump-years. There were 28 (12%) complete failures, 17 (7%) alarms, 83 (35%) mechanical defects, and 105 (44%) minor defects. Survival curves did not differ according to pump brand and model. Hyperglycemia occurred in 2.9% of cases. In multivariate analysis, only patient age less than 40 years at the initiation of pump therapy was associated with higher risk of malfunction (hazard ratio 1.64; 95% confidence interval 1.19-2.24; P = 0.002). CONCLUSIONS: Pump malfunctions remain common with modern pumps. We report less complete failures than in our previous study. This could be because of improvement in quality of pumps or to our strategy of systematic screening and replacement in case of mechanical defects.
BACKGROUND:Insulin pump failures had been assessed in our center by a prospective observational study from 2001 to 2007. The aim of this study was to update our data since 2008 and to determine whether there exist specific risk factors for insulin pump failures. METHODS: All insulin pump defects were prospectively collected between 2008 and 2013 in a monocentric cohort of 350 new pumps. Clinical consequences were recorded. Brand and model of pumps and type of defects and patients' characteristics (gender, type of diabetes, age at diabetes diagnosis, age at first pump, pump treatment duration, number of previous pumps, and number of previous pump failures) were tested for possible association with insulin pump failure. RESULTS: Malfunctions occurred in 239 (68%) pumps. The incidence rate was 33/100 pump-years. There were 28 (12%) complete failures, 17 (7%) alarms, 83 (35%) mechanical defects, and 105 (44%) minor defects. Survival curves did not differ according to pump brand and model. Hyperglycemia occurred in 2.9% of cases. In multivariate analysis, only patient age less than 40 years at the initiation of pump therapy was associated with higher risk of malfunction (hazard ratio 1.64; 95% confidence interval 1.19-2.24; P = 0.002). CONCLUSIONS: Pump malfunctions remain common with modern pumps. We report less complete failures than in our previous study. This could be because of improvement in quality of pumps or to our strategy of systematic screening and replacement in case of mechanical defects.