TOPLINE:
In a single-center study, shifting the timing of nutritional insulin from pre- to post-prandial reduced the risk for hypoglycemia without affecting rates of severe hyperglycemia or hospital length of stay.
METHODOLOGY:
- The retrospective study performed at a community hospital evaluated adult inpatients receiving nutritional insulin.
- On June 28, 2019, hospital policy changed the timing of nutritional insulin from preprandial to administration after at least 50% of the meal was consumed.
- Analysis stratified by three time periods: (1) Preintervention, September 2018 to June 27, 2019; (2) immediate postintervention, July 2019 to February 2020; and (3) distant postintervention, September 2022 to August 2023).
- Outcomes included rates of hypoglycemia (glucose level, ≤ 70 mg/dL), moderate hypoglycemia (< 54 mg/dL), severe hypoglycemia (≤ 40 mg/dL), severe hyperglycemia (≥ 300 mg/dL), daily mean glucose, and length of stay (LOS).
- The number of patient-days analyzed for timepoints 1, 2, and 3 were 1948, 1751, and 3244 patient-days, respectively.
TAKEAWAY:
- Unadjusted rates of any hypoglycemia were significantly reduced across the three time periods, from 9.4% to 8.9% to 6.7% (P < .001), with a significant decrease between timepoints 1 and 3 (P < .001), but not between timepoints 1 and 2.
- No significant changes were seen in rates of moderate hypoglycemia (3.3% vs 3.2% vs 2.6%; P = .24), while there was a nonsignificant trend toward lower severe hypoglycemia rates (1.3% vs 1.1% vs 0.7%, P = .07).
- Rates of severe hyperglycemia didn't differ between the groups (27.4% vs 28.3% vs 29.5%, P = .27), although mean daily glucose did significantly increase over time (194.6 mg/dL vs 196.8 mg/dL vs 199.3 mg/dL, P = .003).
- Daily mean glucose significantly increased from timepoint 1 to 3 (P = .004), but not between 1 and 2.
- No changes were seen in LOS between the three time periods (7.0 vs 6.9 vs 7.1 days, P = .74).
- After adjustment for covariates and compared to baseline, timepoint 3 was significantly associated with a lower risk for severe hypoglycemia (odds ratio, 0.44; P = .009) and any hypoglycemia (0.71, P = .001),
- There was no association of study time period with development of moderate hypoglycemia or severe hyperglycemia.
IN PRACTICE:
"The most common cause of inpatient hypoglycemia revolves around the dosing of nutritional insulin, particularly in patients with decreased caloric intake. Although most rapid-acting insulin analogues are approved to be administered postprandially, traditionally they are administered before meal intake to allow sufficient time for absorption and rise in plasma insulin levels to counter postprandial blood sugar elevations. This practice has been shown to lower hemoglobin A1c in the outpatient setting. However, it can be potentially unsafe in an inpatient population. In particular, variability in oral intake and timing of inpatient meal delivery can be unpredictable, predisposing patients to a higher risk of hypoglycemia," the authors wrote.
SOURCE:
The study was conducted by Merit George, MD, of Johns Hopkins University School of Medicine, Baltimore, and colleagues, all of Johns Hopkins Medicine, and published in Diabetes Care and Clinical Practice.
LIMITATIONS:
Limitations include the retrospective nature of the study and a potential for unmeasured variables. Moreover, the study was conducted at a nonacademic community hospital and may not generalize to other settings. Glycemia was assessed intermittently, not with continuous glucose monitoring. Moreover, compliance couldn't be confirmed.
DISCLOSURES:
George has no disclosures. One coauthors received grant funding from Dexcom, and another reported consulting for Dexcom and EMD Serono.