Admission glycemic gap in the assessment of patients with intracerebral hemorrhage.

Background: Intracerebral hemorrhage (ICH) is the most common type of hemorrhagic stroke. Glycemic gap, determined by the difference between glucose and the HbA1c-derived average glucose, predicts poor outcomes in various clinical settings. Our main objective was to evaluate association of some admission factors and outcomes in relation to admission glycemic gap (AGG) in patients with ICH.

Methods: We retrospectively analyzed 506 adult patients with ICH between 2014 and 2019. AGG was defined as A1c-derived average glucose (28.7×HbA1c-46.7) subtracted from admission glucose. Admission factors and hospital outcomes indicative of poor outcome (i.e. death, gastrostomy tube, tracheostomy, and discharge status) were compared between patients with elevated (greater than 80 mg/dL) vs. non-elevated (less than or equal-to 80 mg/dL) AGG. Pearson chi-square test was used for independence, and multivariate analysis was used for association. SPSS and excel were used for all data analysis.

Results: We found that 67 of 506 (13%) ICH patients had elevated AGG with a mean of 137.3 mg/dL compared to 439 (87%) non-elevated AGG with a mean of 12.6 mg/dL. While mean and standard deviation values for age, weight,and body mass index were comparable between groups, the elevated AGG group had significantly higher admission glucose (286.1 ± 84.3 vs. 140.1 ± 42.5, p < 0.001), higher lactic acid (3.26 ± 2.04 mmol/L vs. 1.99 ± 1.33 mmol/L, p < 0.001), lower Glasgow Coma Scale (GCS) scores (7.70 ± 4.28 vs. 11.24 ± 4.14, p < 0.001), and higher ICH score (median 3, IQR 2-4 vs. median 1, IQR 0-3, p < 0.001). Higher AGG was associated with an increased likelihood of mechanical ventilation, and in-hospital mortality (74.6% vs. 38.3% and 47.8% vs. 15.0% respectively, p < 0.001). Placements of tracheostomy and gastrostomy were similar between the two groups (13.4% vs. 11.8%, p = 0.69% and 1.5% and 4.6%, p = 0.34 respectively). The higher AGG group had a more common poor discharge outcome to either long-term acute care, skilled nursing facility, and/or hospice (65.7% vs. 42.6%, p < 0.001). Hospital cost and length of hospitalization did not differ significantly. Although AGG was not an independent predictor of poor outcome, multivariate analysis showed it was significantly associated with poor outcome while admission glucose was not (p < 0.001 vs. p = 0.167).

Conclusion: Elevated AGG was associated with worse GCS and ICH scores on admission, as well as need for mechanical ventilation, in hospital mortality and poor discharge status. Elevated AGG has value in prediction of outcome, but existing understanding is limited.