Unplanned ICU admissions as signals of hospital performance: linking cost of poor quality and opportunity cost

Abstract Unplanned intensive care unit (ICU) admissions are consistently associated with poorer outcomes, longer ICU and hospital stays, and greater resource use. Beyond their clinical consequences, they may also indicate vulnerabilities in the recognition and escalation of clinical deterioration, adverse events, or care coordination. We conducted a structured narrative review of clinical, organisational, and economic evidence on unplanned ICU admissions, ICU cost measurement, and hospital performance. A documented iterative search and selection process was used, and the CHEERS 2022 statement was consulted solely to guide the extraction and reporting of economic information. The literature showed substantial heterogeneity in costing methods, with length of stay, staffing, and organ support consistently identified as major cost drivers. Unplanned ICU admissions were associated with higher mortality, longer stays, greater care intensity, and higher costs. In a matched cohort analysis, total admission costs were higher by US$13,424 in medical patients and US$21,448 in surgical patients, with 4.1 and 6.4 additional hospital days, respectively. Several studies also reported associations with delayed recognition of deterioration, suboptimal ward monitoring, and preventable adverse events. However, the literature rarely distinguished the unavoidable cost of clinical severity from potentially avoidable resource use or from the consequences of constrained ICU capacity. We therefore propose an analytical framework comprising three complementary dimensions: the necessary cost of treating critical illness, the potential cost of poor quality associated with potentially modifiable care process vulnerabilities, and the opportunity cost that may arise when scarce ICU resources are unavailable for other patients. This framework does not imply that all unplanned ICU admissions or their associated costs are avoidable, but it may support more explanatory cost analysis, targeted quality improvement, and evidence-informed resource allocation.