Prediction of Death after Multiple Wasp Stings with a Clinical Prognostic Model

Maohe Wang,Mei Qin,Amanda Y Wang,Jia Wei Zhao,Fei Deng,Guisen Li,Wei Wang
DOI: https://doi.org/10.21203/rs.3.rs-877961/v1
2021-01-01
Abstract:Abstract Background We aimed to assess the utility of the poisoning severity score (PSS) as early prognostic predictors in patients with wasp stings, and to explore a reliable and simple predictive tool for short-term outcomes. Methods From January 2016 to December 2018, 363 patients with wasp stings in Suining Central Hospital were taken as research subjects. In the first 24h of hospital admission, the PSS and Chinese expert consensus on standardized diagnosis and treatment of wasp stings (CECC) were used as the criterion for severity classification, and their correlation was analyzed. The patients were divided into survival and death groups according to the state of discharge. The factors that affect outcome were analyzed by logistic regression analysis. A clinical prognostic model of death was constructed according to the risk factors, and 1000 times repeated sampling was done to include the data to verify the model internally. Results The mortality of wasp sting patients was 3.9%. There was a correlation between PSS and CECC (r=0.435, P<0.001) for severity classification. Sex, age, number of stings, and PSS were independent risk factors for death. Based on the 4 independent risk factors screened by the above regression analysis, a nomogram model was constructed to predict the risk of death in wasp sting patients. The predicted value C-index was 0.962, and the internally verified AUC was 0.962(95%C.I. 0.936-0.988, P<0.001). Conclusions PSS is helpful in the early classification of the severity of wasp stings. Sex, age, number of stings, and PSS were independent risk factors for death in wasp sting patients. The nomogram model established in this study can accurately predict the occurrence of the risk of death.
What problem does this paper attempt to address?