Concordancia entre glicemia central y glucometría en pacientes con nutrición enteral o parenteral continuas en una unidad de cuidado intensivo de adultos
- Publication Year:
- Repository URL:
- Monitoreo fisiológico; Análisis químico de la sangre / métodos; Glucemia / análisis; Nutrición parenteral; Nutrición enteral; Cuidados críticos; 362.11; Análisis químico de la sangre; Glucemia; Monitoring; Physiologic; Blood Chemical Analysis / Methods; Blood Glucose / Analysis; Parenteral Nutrition; Enteral Nutrition; Critical care; Hyperglycemia; Concordance correlation coefficient
thesis / dissertation description
Introduction: Patients with artificial nutrition require monitoring of the plasma glucose levels in order to take the correct therapeutic decisions. These results may be obtained between 45 to 120 minutes by venous monitoring, and 15 to 30 seconds via capillaries. The difference between capillar and plasma glucose in a critically ill patient may fluctuate between 8% and 20% according to international parameters. Glucose readings by glucometry in a critical patient is used in Colombia as a reliable metric, although some authors recommend against its routine practice due to the high variability with respect to central glycemia. Currently, there are no studies in which that relationship is evaluated among patients receiving enteral nutrition or continuous parenteral in the context of critical care. Objective: To estimate the concordance between glucometry and central glycemia in critical patients with continuous nutritional support, and the relative difference in the absolute mean (MARD) <10%. The rapid measurement of blood glucose in critically ill patients must be accurate, to make appropriate therapeutic decisions. Glucometry in these patients is used in Colombia as a reliable measure. Some authors do not routinely recommend its use due to its high variability with central glycemia. Currently, there are no studies available in which this relationship was evaluated in patients receiving continuous enteral or parenteral nutrition. Methods: Study of prospective concordance using the Bland-Altman method and the concordance correlation coefficient (CCC). Fifty-six patients participated in the study, of which 23 (41.1%) were men. Fifty (88.9%) received enteral nutrition. We obtained a total of 196 pairs of samples of central glycemia and glucometry, of which 140 had been applied after starting nutrition. The median of the MARD was 8.55%. CCC was higher in 0.96 during fasting (IC95% 0.94-0.98) than in the postprandial 0.84 (IC95% 0.77-0.88). The mean and standard deviation of the glycemia-glucometry difference at time zero (in fasting) was similar to that of all the pairs of samples (7.7 and 7.3 mg / dL). The regression line (Passing & Bablok) for all paired samples had an interception of 3.186 (95% CI -8.000 and 12.267) with a systematic bias for glucometry of 3.186 mg/dL with glycemia values between 8 mg/dL and 12 mg/dL. This correlation was not observed during fasting, where the regression line showed an interception of -20.686 (95% CI -30.538 and -3.635). In the 140 paired samples of the glucometry and postprandial glycemia data, the regression line had a systematic bias of glucometry with respect to glycemia of 0.397 mg/dL, with values between 16 mg/dL and 11 mg/dL. Conclusion: There is a good but not sufficiently precise agreement between glycemia and glucometry in critical patients receiving an artificial nutrition. The routine use of glucometry as the only metric in ICU is therefore not recommended. Considering the stress load on patients with multiple punctures for readings of central glycemia, the implementation of continuous monitoring devices that allow the detection of episodes as well as duration of hypoglycemia is warranted.