SAT138 “Zero”Ing On Hypoglycemia Caused By Hyperkalemia Treatments In The Inpatient Setting

Abstract

Abstract Disclosure: S. Kim: None. H. Tran: None. R.Y. Gianchandani: None. Background: Severe hyperkalemia can be life-threatening, therefore, is urgently treated in the hospital with intravenous (IV) insulin. An insulin dose of 10 units given with dextrose is considered standard of care. Hypoglycemia associated with hyperkalemia treatment was identified as a significant contributing factor to our overall hypoglycemia rate in the hospital. We designed a hyperkalemia quality improvement (QI) project based on recent studies demonstrating improvement in hypoglycemia rates with weight-based insulin dosing, without clinically impacting potassium-lowering effects. Methods: 5 months of retrospective data was collected before and after implementing a revised hyperkalemia orderset. Prior to implementation, regular insulin was dosed at 10 units intravenously along with 25 g of 50 percent dextrose (D50W). In the updated orderset, regular insulin was dosed at 0.1 units/kg of body weight up to a maximum of 10 units with 25 g of D50W. Additionally, the revised orderset included options for ordering alternative hyperkalemia treatment modalities along with monitoring parameters. A point of care BG was ordered at one hour post treatment. A utilization report was generated for pre- and post-implementation when IV insulin was ordered from hyperkalemia orderset. Additionally, patients who received IV insulin independent of the orderset during the post-implementation period were evaluated. The primary endpoint is the rate of hypoglycemia (BG <70 mg/dL) within 6 hours of IV insulin and D50W administration for hyperkalemia treatment. Results: In the pre-implementation period, 604 doses of insulin were administered from the original orderset with a hypoglycemia rate of 9.8% at 6 hours. Post-implementation, 322 doses were administered from the revised orderset which had a hypoglycemia rate of 5% at 6 hours. Partial adoption of the revised hyperkalemia orderset resulted in 281 doses of insulin treatment ordered independent of the new orderset and hypoglycemia rates remained high at 9.6% at 6 hours in this group. Most of these were in the emergency room where there was low orderset adoption.95% (21/22) of patients who experienced hypoglycemia despite using weight-based insulin dosing for hyperkalemia treatment had an initial BG ≤150 mg/dL and 59.1% (13/22) had a CrCl <30 mL/min or ESRD. Conclusion: A QI process utilizing weight-based insulin dosing for the treatment hyperkalemia reduced hypoglycemia rates by half. Patients who continued to have low BG post intervention were those with a pretreatment BG ≤150 mg/dL. We further revised our orderset to utilize a higher dose of dextrose (D50W 50 g) if pretreatment BG was less than 150 mg/dL and will be implementing this shortly. Additionally, we will work with the emergency room to assist in their hyperkalemia management. With follow-up data, we plan to reach negligible rates of iatrogenic hypoglycemia with our hyperkalemia interventions. Presentation: Saturday, June 17, 2023