Avaliação da acurácia diagnóstica entre glicemia capilar e glicemia venosa no diagnóstico de hiperglicemia hospitalar

Abstract

Hospital hyperglycemia is a very common clinical condition, including in patients requiring hospitalization, with stress hyperglycemia as one of the factors. This condition has negative repercussions for the patient, such as increased mortality rate and risk of cardiovascular events. The study was developed in the Fundação Santa Casa de Misericórdia do Pará - FSCMPA, in the medical clinic wards of the hospital, São Roque and São Paulo, and the glycemic values were collected in two ways, one by venipuncture and the other by capillary blood sampling. The venous puncture used a colorimetric method, with internal quality controls Level 1 and 2; the capillary glucose determination used electrochemical reactive strips (glucose dehydrogenase) by glucometer, using Level 1 and 2 control solutions. For the analysis of numeric variables, the Shapiro Wilk Normality test was used, so in case of normality the One Criterion ANOVA test will be calculated and in case of non-normality the Krulkal Wallis test. Categorical variables were evaluated using the Chi-square Adherence Tests for univariate tables and G and Chi-square Independence Tests for bivariate tables. BioEstat® 5.3 software was used for descriptive and analytical analyses, using the significance level α=0,05 or 5%. It was observed that 93.3% of patients had glycemic concentration higher than 140 mg/dL when performed via capillary blood glucose, while 90% had concentration higher than 140 mg/dL when performed venous blood glucose measurement. The capillary blood glucose was generally higher (140 to 450 mg/dL) compared to the venous blood glucose (118 to 394 mg/dL). However, the median variation between methods (9.8%) is within the confidence interval for the median variation, demonstrating that the values are within those recommended by the FDA, the ADA, and the ISO. As the result, capillary blood glucose can be safely used, as it is as accurate as venous blood glucose for the diagnosis of hospital-acquired hyperglycemia.