Development of a risk predication model for critical care needs in patients with intracerebral hemorrhage: a retrospective cohort

BackgroundIt is very important to provide the correct nursing care for patients with intracerebral hemorrhage (ICH), but the level of critical care needs in patients with intracerebral hemorrhage is not clear. The purpose of this study is to establish a risk model based on the epidemiological and clinical characteristics of ICH patients, to help identify the critical care needs of ICH patients.MethodsThe clinical data of ICH patients from January 2018 to September 2023 were analyzed retrospectively. The full cohort was used to derive the clinical prediction model and the model was internally validated with bootstrapping. Discrimination and calibration were assessed using the area under curve (AUC) and the Hosmer-Lemeshow tests, respectively.Results611 patients with ICH were included for model development. 61.21% (374/611) ICH patients had received critical care interventions. The influencing factors included in the model were Glasgow Coma Scale (GCS) score, intraventricular hemorrhage, past blood pressure control, systolic blood pressure on admission and bleeding volume. The model's goodness-of-fit was evaluated, which yielded a high area under the curve (AUC) value of 0.943, indicating a good fit. For the purpose of model validation, a cohort of 260 patients with ICH was utilized. The model demonstrated a Youden's index of 0.750, with a sensitivity of 90.56% and a specificity of 78.22%.ConclusionGCS, systolic blood pressure, intraventricular hemorrhage, bleeding volume and past blood pressure control are the main factors affecting the critical care needs of patients with ICH. This study has deduced a clinical predictive model with good discrimination and calibration to provide scoring criteria for clinical health care providers to accurately evaluate and identify the critical care needs of ICH patients, to improve the rational integration and allocation of medical resources.