Management of diabetes and hyperglycaemia in the hospital

Abstract

Hyperglycaemia in people with and without diabetes admitted to the hospital is associated with a substantial increase in morbidity, mortality, and health-care costs. Professional societies have recommended insulin therapy as the cornerstone of inpatient pharmacological management. Intravenous insulin therapy is the treatment of choice in the critical care setting. In non-intensive care settings, several insulin protocols have been proposed to manage patients with hyperglycaemia; however, meta-analyses comparing different treatment regimens have not clearly endorsed the benefits of any particular strategy. Clinical guidelines recommend stopping oral antidiabetes drugs during hospitalisation; however, in some countries continuation of oral antidiabetes drugs is commonplace in some patients with type 2 diabetes admitted to hospital, and findings from clinical trials have suggested that non-insulin drugs, alone or in combination with basal insulin, can be used to achieve appropriate glycaemic control in selected populations. Advances in diabetes technology are revolutionising day-to-day diabetes care and work is ongoing to implement these technologies (ie, continuous glucose monitoring, automated insulin delivery) for inpatient care. Additionally, transformations in care have occurred during the COVID-19 pandemic, including the use of remote inpatient diabetes management-research is needed to assess the effects of such adaptations.

Figure 1:. Individualised antihyperglycaemic therapy in hospitalised patients with diabetes

In critically ill patients, continuous insulin infusion is recommended followed by transition to subcutaneous insulin regimens once patients are stable and close to discharge from the intensive care unit.2 Subcutaneous insulin DKA protocols might be considered in patients with mild-to-moderate DKA (subcutaneous insulin protocol examples adapted for COVID-19 are available online). We discourage the widespread use of premixed insulin regimens in the hospital setting. BG=blood glucose. DKA=diabetic ketoacidosis. HHS=hyperosmolar hyperglycaemic state. OAD=oral antidiabetes drug. TDD=total daily dose. U=units. *Consider OAD if no contraindications (only DPP-4 inhibitors have been studied in randomised controlled trials); metformin is commonly used in the hospital setting but might be associated with lactic acidosis in high-risk patients (eg, sepsis, shock, renal or liver failure). †Antidiabetic agents include OADs and GLP-1 receptor agonists. ‡In patients with hypoglycaemia risk (frail, elderly, acute kidney injury), reduce starting dose to 0·15 U/kg per day (basal alone) or TDD 0·3 U/kg per day (basal–bolus).

Figure 2:. Remote glucose management during the COVID-19 pandemic

Real-time CGM or flash glucose monitoring data are transmitted via Bluetooth from the sensor to a receiver or smartphone. From a smartphone, sensor glucose data can be transferred to the cloud (via cellular signal or WiFi) and from there to real-time followers (health-care providers and telemetry) as well as to dashboard software (eg, LibreView or Dexcom CLARITY) for comprehensive assessment of multiple patients. Comprehensive glucose reports can be scanned and uploaded to the chart. Until more data on the reliability of inpatient CGM are available, a hybrid approach is recommended. Data can be documented and validated in the EHR (eg, sensor values are within 20% of POC glucose values for blood glucose concentrations >100 mg/dL). Steps for CGM implementation during COVID-19 were recently described by Galindo and colleagues.18 Steps for direct integration of CGM data into the EHR were recently described by Espinoza and colleagues.136 Additional information, scientific literature, and links are available online. CGM=continuous glucose monitoring. EHR=electronic health record. FGM=flash glucose monitoring. POC=point-of-care.