Diabetes
Kathleen Ruddiman, DO (she/her/hers)
Fellow
Wake Forest University School of Medicine
Winston Salem, North Carolina, United States
This study was approved by Institutional Review Board (Protocol #97795, Wake Forest University School of Medicine). We identified a cohort of participants based on electronic health record (EHR) search for insulin pump orders and included patients over 18 years old with diagnosis of diabetes managed with insulin pump admitted to AHWFB hospitals between 2016 and 2023. We conducted a retrospective analysis of EHR data routinely collected during hospital admissions. Independent variables included demographics, pregnancy status, level of service at admission, length of stay, surgery during admission, type of insulin pump, whether the pump was removed during admission, and glycemic data. Glycemic outcomes were defined as time within inpatient goal (140-180 mg/dL) and rates of hypoglycemia ( < 55 and < 70 mg/L). We compared differences in glycemic metrics before and after endocrinology consult. Results were analyzed using descriptive statistics and 2-tailed T-test.
Results:
Of 333 individuals with 505 hospital admissions included in our study, 64% were female, 84% were Caucasian, 78% were under age 65, 19% of females were pregnant, and 43% of patients underwent surgery during admission. Average A1c was 7.6%. Mean time from admission to endocrinology consult was 0.83 days; this did not differ significantly based on gender, race, or age >65 years. Length of stay was 1.48 days longer for patients with delayed endocrine consultation more than 1 day after admission (p= 0.198). Mean glucose prior to endocrine consult was 201.6mg/dL, and was 181.6mg/dL post-consult (p = 0.196). Prior to consult, rates of hypoglycemia < 70mg/dL and < 54mg/dL were 4.3% and 2.0%. After endocrine consult, these rates decreased to 3.8% (p = 0.847) and 1.4% (p=0.979), respectively.
Discussion/Conclusion:
In patients with diabetes managed by insulin pump therapy during hospital admission, we observed lower rates of hypoglycemia and lower mean glucose after endocrinology consult. Increased length of stay was observed in patients with delayed consult times. Though not statistically significant, these findings may have financial implications for total cost of hospitalization, and will inform further investigation of a best-practice hospital care pathway for insulin pump use.