2017 Resarch Forum

PHARM: M-1

Applicant & Principal Investigator: Alice Peng PharmD

Impact of education on insulin infusion protocol compliance Alice Peng PharmD, Jeffrey Jolliff PharmD, David Lash PharmD

INTRODUCTION : Intravenous insulin infusions serve a vital role in the hospital setting; significant harm can occur when used inappropriately. Kern Medical has two protocols that guide initiation, adjustment, and monitoring of insulin infusions: Hyperglycemic Crisis (DKA/HHS) and Standard Insulin Infusion (STD) protocols. PURPOSE: This study evaluates whether compliance, subsequent safety, and efficacy of Kern Medical’s protocols can be affected by staff education. METHOD: Staff education (16 one hour-long sessions) on the proper use of the protocols was completed (October 2016 - January 2017); 227 nurses and approximately 50 medical residents participated. Outcomes were assessed using a pre-post retrospective chart review study design. Patients in the pre-education (Pre-Edu) cohort (July 1 -September 30, 2016) and post-education (Post-Edu) cohort (January 1 - March 31, 2017) were compared. Compliance to the protocols was assessed (correct protocol initiation, timing, and adjustment of the insulin infusion). Safety and efficacy were evaluated (number of hypoglycemic events, time to blood glucose normalization, and anion gap closure). Nominal data was analyzed (Fisher’s exact tests); continuous data was analyzed with T-tests and Mann-Whitney U. RESULTS: A total of 23 cases were included in the Pre-Edu cohort; 15 in the Post-Edu cohort. In the Pre- and Post- Edu cohorts, the appropriate protocol was ordered 47.8% vs 100% (p=0.008). 457 and 213 blood glucose checks in the Pre- vs Post- Edu cohorts and 80.7% vs 85.0% (p=0.20) were within 15 minutes of the scheduled blood glucose check time; 86.9% vs 90.1% (p=0.25) responded according to the protocol. From initiation of insulin infusion, the anion gap closed in a median of 6.5 vs 9.75 hours (p=0.25); blood glucose normalized in a median of 3.5 vs 5 hours (p=0.34). Only one entry in the Pre-Edu cohort and none in the Post- Edu cohort involved blood glucose less than 70mg/dL (p=1.0). Five patients in the Pre-Edu cohort had one or more episodes of blood glucose less than 90mg/dL vs one patient in the Post-Edu cohort (p=0.37), leading to insulin infusion interruption and dextrose 50% administration per protocol; no resulting complications were documented. DISCUSSION: Pre-Edu, the incorrect insulin protocol was prescribed in 52.2% of the cases. Differences between protocols in how infusion rates are adjusted based on blood glucose, selecting the wrong insulin protocol could cause serious harm by over-correcting hyperglycemia or correcting too rapidly. Due to relatively small sample sizes and low incident rates, the study was underpowered to detect differences in safety and efficacy endpoints. However, the results indicate that education effectively clarified confusion regarding the two different protocols and significantly improved appropriate application of the protocols. Confounding factors (comorbidities, concurrent health problems) and other technical issues were not included in our analysis. CONCLUSIONS: Staff education significantly improved decisions involving which protocol to use.