From the 2024 HVPA National Conference
Niloofar Latifi MD (Hopkins), Nestoras Mathioudakis MD, MHS (Johns Hopkins Hospital), Mohammed Abusamaan M.B.Ch.B., M.P.H, Megan Tschudy MD, MPH, Sarah Olson MPH, Jianqiao Ma, MPH, Daniel Brotman MD
Background
Guidelines for inpatient diabetes management recommend insulin therapy for persistent hyperglycemia ≥ 180 mg/dL with a target glucose range of 140 – 180 mg/dL. Although guidelines recommend glucose monitoring for insulin administration, there is no standard monitoring approach for patients without evidence of persistent hyperglycemia. As such, hospitalized patients are frequently started on 3 – 4 times daily point of care glucose (POC-G) monitoring without further adjustment of monitoring intensity even absent proven hyperglycemia. This practice leads to waste of medical resources, nursing time, and patient discomfort from unnecessary finger sticks.
Objectives
Determine the prevalence and clinical characteristics of non-critically ill patients subjected to POC-G monitoring without hyperglycemia.
Methods:
We performed a secondary data analysis of a retrospective cohort derived from inpatients discharged between January 1, 2015, and May 31, 2019, from five academic and community hospitals.³ We limited the dataset to inpatients ≥ 18 years old with a length of stay ≥ 4 days and ≥ 2 POC-G values within 0-48 hours of admission. These criteria were used for the future goal of creating predictive models for euglycemia. We excluded serum glucose values and patients with ICU stays, Type 1 diabetes, admission glucose >500 mg/dL, inpatient use of non-insulin diabetes medications, and patients on long or intermediate-acting insulin at home or in the hospital. For patients with multiple hospitalizations, only the first admission was included. The final analytical cohort was divided into three subgroups – patients with all POC-G values ≤ 140 mg/dL (group 1), patients with maximum POC-G >140 and <180 mg/dL (group 2), and patients with max POC-G ≥180 mg/dL (group 3). All institutions initiated correctional insulin administration for glucose values ≥141 mg/dL.
The primary outcome was the prevalence of POC-G monitoring in patients with all POC-G values < 180 mg/dL during hospitalization. The secondary outcomes were the number of POC-G tests performed in this patient population and the frequency of insulin administration. ANOVA and Pearson’s chi-squared were used for continuous and categorical data, respectively.
Results:
A total of 9,228 unique patients were included in the study with a mean age of 65, 54% female, 55% white, and 34% black (Table 1). Inpatient recommended glucose values of < 180 mg/dL were achieved in 4,021 (43.6%) of patients, with 1,462 (15.8%) of all patients having all glucose values ≤ 140 mg/dL. Of patients who had an HgbA1C measurement during the hospitalization or within 90 days of admission, the mean values were 5.5%, 5.8%, and 6.6% for groups 1, 2, and 3, respectively (p <0.001). On average, group 1 patients had 14.3 POC-G tests during hospitalization without associated insulin administration. Group 2 patients had a mean of 22.3 POC-G tests with a median of 0 instances of insulin administration during hospitalization (IQR 0 – 1). Group 3 patients had 30.1 POC-G tests and a median of 5 instances of insulin administration during their stay (IQR 1 – 11).
Conclusions:
An average of 20,907 POC-G tests were obtained in patients who never received insulin, with an additional 57,066 POC-G tests in patients who did not reach the definition of inpatient hyperglycemia.
Clinical Implications:
The prevalence of low-value POC-G testing requires further study to identify appropriate patients for early de-escalation of monitoring with potential to reduce patient discomfort and healthcare costs from low-value use of staff time and POC-G testing.