Impact of pattern of admission on ICH outcomes.

BACKGROUND: Intracerebral hemorrhage (ICH) is associated with the highest mortality of all strokes. Admission to a Neurosciences Critical Care Unit (NCCU) compared to a general ICU has been associated with reduced mortality following ICH. Such association has led to several hospitals transferring ICH patients to Neuro-ICUs in tertiary care centers. However, delays in optimizing ICH management prior to and during transfer can lead to deleterious consequences. To compare functional outcomes in ICH patients admitted to our NCCU directly from the ED versus inter-hospital transfer admissions.
METHODS: Records of consecutive spontaneous supratentorial ICH patients admitted to The Johns Hopkins Hospital NCCU were reviewed. Patients with ICH related to trauma or underlying lesions (brain tumors, aneurysms, AVM) were excluded. We compared outcomes at discharge in patients admitted directly from the ED and inter-hospital transfers (IHT) using dichotomized modified Rankin Scale (Good outcomes: mRS 0-3). Other factors potentially impacting outcomes such as age, ICH volume, IVH volume, and admission GCS were included in the multiple logistic regression analysis.
RESULTS: 125 patients were included in the analysis (ED 61.6%; IHT 38.4%). There were no significant differences between the two groups in mean age (ED 63.4 +/- 13.1; IHT 63.4 +/- 15.2, P = 0.96), ICH volume (ED 31.4 +/- 37.6; IHT 33.5 +/- 42.8, P = 0.76), IVH volume (ED 6.0 +/- 11.2; IHT 8.0 +/- 14.5, P = 0.38), and GCS (ED 11.3 +/- 3.7, IHT 10.9 +/- 3.5; P = 0.44). 57.2% ED patients had good outcomes (mRS 0-3) at discharge compared to 37.5% IHT. This difference was statistically significant following univariate (P = 0.034, 95% CI .2151-.9416) and multivariate analysis (P = 0.028, 95% CI .1338-.8896). Odds (adjusted) of ED admissions having good outcomes was three times higher than IHT. Neurological deterioration (GCS decline 2 or more) was more common in IHT and, in subgroup analysis of IHT patients with warfarin-associated ICH, hematoma enlargement was significantly more likely than in direct ED admissions.
CONCLUSIONS: Patients with ICH brought directly to our ED had significantly better outcomes than IHT; we hypothesize this may be caused by delays in optimizing management prior to arrival at the facility with a dedicated Neuro-ICU. Nevertheless, other equally plausible hypotheses need to be prospectively tested.