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.1,2 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 (POC) glucose 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. In this study, we aimed to determine the prevalence and clinical characteristics of non-critically ill patients subjected to POC glucose monitoring with all glucose values ≤ 180 mg/dL during hospitalization.
Methods: We performed a secondary data analysis of a retrospective cohort dataset derived from five academic and community hospitals.3 We limited the dataset to inpatients ≥ 18 years old with minimum four glucose values who were discharged between January 1, 2015 and May 31, 2019. As our aim was to describe the prevalence of unnecessary POC glucose monitoring, we excluded serum glucose values and patients with Type 1 diabetes, admission glucose >500 mg/dL, and patients taking long or intermediate acting insulin at home or in the hospital. Patients with admission duration ≤ 24 hours, ICU stay, or inpatient use of non-insulin diabetes medications were also excluded. For patients with multiple hospitalizations, only the first admission was included. The final analytical cohort was divided in to three subgroups –patients with all POC glucose values ≤ 140 mg/dL (group 1), patients with maximum POC glucose 141 – 180 mg/dL (group 2), and patients with any POC glucose >180 mg/dL (group 3). All institutions initiated correctional insulin administration for glucose values ≥141 mg/dL. The primary outcome was the prevalence of patients undergoing POC glucose monitoring who have all POC glucose values ≤ 180 mg/dL during hospitalization. The secondary outcomes were the number of POC glucose tests performed in this patient population and the frequency of insulin administration. ANOVA and Pearson’s chi-squared were used for continuous and categorial data respectively.
Results: A total of 23,134 unique patients were included in the study with a mean age of 64, 52% female, 56% white, and 32% black (Table 1). Inpatient recommended glucose values of ≤ 180 mg/dL were achieved in 11,525 (49.8%) patients with 4,836 (20.9%) of patients having all glucose values ≤ 140 mg/dL. Of patients who had a HgbA1C measurement during the hospitalization or within 90 days of admission, the mean values were 5.5%, 5.9%, and 6.8% for groups 1, 2, and 3 respectively (p < 0.001). On average, group 1 patients had 9.5 POC glucose tests during hospitalization without associated insulin administration. Group 2 patients had a mean 15.8 POC glucose tests during hospitalization with a median 0 instances of insulin administration during their stay (IQR 0 – 1). Group 3 patients had 20.4 POC glucose tests and a median 3 instances of insulin administration (IQR 1 – 7) during their stay.
Conclusions: Unnecessary inpatient POC glucose monitoring without the need for insulin administration to achieve glycemic targets is prevalent. Our next step is to identify predictor variables that aid in classifying appropriate patients for early de-escalation of POC glucose tests.
