Spontaneous intracerebral hemorrhage (ICH) is the most devastating type of stroke, and it is associated with high morbidity and mortality. Patients with a spontaneous ICH are routinely admitted to an intensive care unit (ICU). However, an ICU is a valuable and limited resource, and not all patients may require this level of care. The authors conducted a systematic review and meta-analysis evaluating the safety and outcome of admission to a step-down level of care or stroke unit (SU) compared to intensive care in adult patients with low-risk spontaneous ICH. PubMed, Embase, and the Cochrane Library were searched for randomized clinical trials and observational cohort studies. The Mantel-Haenszel method or inverse variance, as applicable, was applied to calculate an overall effect estimate for each outcome by combining the specific risk ratio (RR) or standardized mean difference. Risk of bias was analyzed using the Newcastle-Ottawa Scale. The protocol was registered in PROSPERO (CRD42023481915). The primary outcome examined was in-hospital mortality. Secondary outcomes were unfavorable short-term outcome, length of hospital stay, and (re)admission to the ICU. Five retrospective cohort studies involving 1347 patients were included in the qualitative analysis. Two of the studies had severity-matched groups. The definition of low-risk ICH was heterogeneous among the studies. Admission to an SU was associated with a similar rate of mortality compared to admission to an ICU (1.4% vs. 0.6%; RR 1.66; 95% confidence interval [CI] 0.24-11.41; P = 0.61), a similar rate of unfavorable short-term outcome (14.6% vs. 19.2%; RR 0.77; 95% CI 0.43-1.36; P = 0.36), and a significantly shorter mean length of stay (standardized mean difference - 0.87days; 95% CI - 1.15 to - 0.60; P < 0.01). Risk of bias was low to moderate for each outcome. The available literature suggests that a select subgroup of patients with ICH may be safely admitted to the SU without affecting short-term outcome, potentially saving in-hospital resources and reducing length of stay. Further studies are needed to identify specific and reliable characteristics of this subgroup of patients.
Read full abstract