Objective To explore the construction of a risk prediction model for long-term hospitalization care of respiratory emergency patients.

Methods A retrospective selection of 211 patients with respiratory emergencies was conducted, with 70%(148 cases) randomly selected as the training set and 30%(63 cases) as the test set.The length of hospitalization and nursing care for patients in the training set was statistically analyzed, including 46 patients in the long-term hospitalization group who exceeded the median length of stay (≥8 days) and 102 patients in the non-long-term hospitalization group.The general data and laboratory data of the two groups in the training set were compared, multivariate logistic risk was used to screen the relevant factors for long-term hospitalization and nursing risk of respiratory emergencies, a nomogram prediction model based on the training set was established, and a correction curve was drawn to verify the internal effectiveness of the prediction model, and the external effectiveness of the model in the test set was evaluated.

Results The proportion of patients receiving mechanical ventilation treatment in the long-term hospitalization group was higher than that in the non-long-term hospitalization group, and the comorbidity index (CCI), NRS 2002 score, and arterial partial pressure of carbon dioxide (PaCO₂) levels were higher than those in the non-long-term hospitalization group (P < 0.01).Multivariate logistic analysis showed that CCI, mechanical ventilation treatment, PaCO₂, and NRS2002 scores were independent risk factors for long-term hospitalization risk in patients with respiratory emergencies (P < 0.05 to P < 0.01).The consistency index of the above factors predicting long-term hospitalization risk in patients with respiratory emergencies was 0.951(95%CI: 0.919-0.982).In the test set, there were 19 long-term hospitalization patients and 44 non-long-term hospitalization patients.External validation of the model showed a sensitivity of 84.21%, a specificity of 88.64%, and an accuracy rate of 87.30 %.

Conclusions A risk prediction model for long-term hospitalization care of respiratory emergency patients has been established.Mechanical ventilation treatment, PaCO₂ level at admission, CCI, and NRS2002 scores are risk factors for long-term hospitalization care of respiratory emergency patients.