Blood glucose management in the critically ill : perception and clinical practice in the ICU

Abstract

Background Contradictory findings from multiple studies on blood glucose management in the ICU and the absence of clear international guidelines leads to varying perceptions and clinical practice among clinicians. Objectives To assess clinical practice and perceptions regarding blood glucose management in a Norwegian ICU. Methods We performed a self-performed survey for doctors working in the main and thoracic ICU at St. Olavs University Hospital regarding their blood glucose management perceptions. We also collected blood glucose management data from ICU patients from the last year from our electronic health record system. Data were analysed with descriptive statistics. Results The response rate for the survey was 72.5%. Blood glucose threshold for what the clinicians regarded as “hypoglycaemia” and “hyperglycaemia” were, respectively, 4.19 + 0.62 mmol/L and 9.60 + 1.68 mmol/L. The mean blood glucose value for patients was 7.66 + 2.23 mmol/L. 62.07% of surveyed clinicians believed that 11-40% of patients had blood glucose values above 10.0 mmol/L compared to 51.54% patients in patient data, 68.97% believed 1-30% of patients had blood glucose values below 5.0 mmol/L compared to 53.92% in patient data, and 68.97% believed that 1-10% of patients had blood glucose values below 2.2 mmol/L compared to 1.39% in patient data. 76% of clinicians believed that patients were in range (5.0-10.0 mmol/L) 61-90% of their stay in the ICU, while time in range for the median patient was 91% of monitoring time. Conclusions Clinician perception regarding definition of hyperglycaemia concedes well with blood glucose target protocol of the unit, whilst clinician definition of hypoglycaemia is lower than blood glucose target protocol. Clinicians in our survey underestimates the number of patients with blood glucose values above 10.0 mmol/L and below 5.0 mmol/L, but give good estimates for patients with values below 2.2 mmol/L. The findings from our study coincide with the findings of previous similar studies of clinician's perceptions, which might contradict the use of clinician's perceptions as the basis for further research and quality improvement measures. Further studies on this subject, including nurse's perceptions and multiple centres is needed.