Objective To identify the latent class analysis patterns of risk factor for stress induced hyperglycemia (SIH) among ICU patients and explore their association with prognosis, aiming to provide reference for improving patient survival outcomes. MethodsThis retrospective study included patients with SIH admitted to the emergency ICU. Demographic, disease related data, in hospital mortality and 28 day mortality were collected. Latent class analysis was used to identify the risk factor patterns, and logistic regression was employed to analyze the association between different risk factor patterns and patient outcomes. ResultsA total of 621 patients were included, among whom 38 (6.1%) died in the hospital and 54 (8.7%) died within 28 days. Latent class analysis identified four risk factor patterns:stable glucose with low risk (26.9%), fluctuating glucose with insulin treatment (27.4%), older age with comorbidities (21.6%), and high risk infection (24.1%). Patients in the fluctuating glucose with insulin treatment, older age with comorbidities, and high risk infection patterns had significantly higher risks of in hospital mortality and 28 day mortality compared with those in the stable glucose with low risk pattern (all P<0.05). Conclusion There is heterogeneity in the risk factor patterns among ICU patients with SIH. Healthcare professionals should pay attention to patients in the high risk infection pattern, older age with comorbidities pattern and the fluctuating glucose with insulin treatment pattern, and provide targeted interventions by identifying potential risk factor pattern to improve their overall prognosis.