In China, Nepal, Vietnam and other Asian countries, systemic organ damage or even death due to wasp venom is not rare1,5,12. This retrospective study found that the mortality rate in wasp sting patients was 3.9%, and death was only seen in the summer and autumn months from September through November among this cohort. Of the 14 deceased patients, all died between the 1st-3rd day after admission. The PSS in the death group was significantly higher than that of the survival group. Multivariate logistic regression analysis showed that female sex, increased age, higher number of stings and greater PSS grade were independent risk factors for death in wasp sting patients.
Wasp venom contains a variety of bioactive components, such as peptides, enzymes and amines3,13. Severe wasp stings can lead to systemic allergic reaction, rhabdomyolysis, shock, hemolysis, acute kidney injury (AKI), and even death2,14,15. Unfortunately, there is no specific antidote for wasp venom at present. It is especially important to classify the severity of wasp sting patients at an early stage, and to carry out the corresponding treatment. Patients with severe wasp stings should be promptly referred to a medical institution that can perform blood purification treatment, which can reduce mortality6,16,17. Chinese Society of Toxicology had prepared a consensus statement on the standardized diagnosis and treatment of wasp stings (CECC) in 20186. Nevertheless, a wider application of this consensus criteria is likely limited by the complex evaluation criteria. In Mong's literature, the use of poisoning severity score (PSS) for the assessment of the severity of the poisoning patients (including wasp sting patients) in the emergency department was reported10. In the present research, there was a correlation between PSS and CECC in assessing the severity of wasp stings. The severity of wasp stings was evaluated by PSS, including 59(16.2%) patients with grade 2 and 14(3.9%) patients with grade 3, where a severity poisoning ≥grade 2 (PSS) might lead to death. The severity of wasp stings was evaluated by CECC, including 56(15.4%) patients with severe wasp stings, where a severity poisoning ≥severe (CECC) might lead to death. PSS had higher accuracy and sensitivity value (89.0%, 92.9% respectively), whilst CECC had a higher specificity value (86.8%) when predicting the short-term clinical outcome in wasp sting patients. The results revealed that these two criteria determined the severity of poisoning and were able to predict the short‑term clinical outcomes due to wasp stings. As the PSS was more predictive and simpler, it is recommended to poisoning centers as effective criteria for classify the severity of wasp sting patients.
Wasp stings were the main cause of human death caused by animal injuries, and it was the main cause of community-acquired AKI in Asia16. In this study, the mortality of our wasp sting patients was 3.9%, which was consistent with Xie, etc.'s report1. When the severity was graded according to the PSS, the mortality was as high as 28.5% in the severe group and 15.2% in the moderate group. PSS was an independent risk factor for death in wasp sting patients. The severity was also related to the number of stings. In Xie and Liu, etc.'s literature, the overall incidence of severe complications was higher in the group with more than 10 stings1,18. In our patients, the mean number of stings in the death group was 30. Females accounted for 70% of patient deaths. Multivariate logistics regression showed that the number of stings and female sex were also independent risk factors for death in wasp sting patients. The average age in the death group was significantly higher than that in the survival group. As such, age was also one of the risk factors for death, which may be related to underlying diseases in elderly patients. The four indexes of female sex, age, the number of stings, and PSS were combined as prediction criteria, and predicting the death of wasp sting patients had high accuracy (AUC=0.962, 95%C.I. 0.936-0.988, P<0.001), which was more powerful than using PSS (AUC=0.890) or CECC (AUC=0.845) alone. Therefore, a nomogram prediction model was built on the base of independent risk factors determined through logistic analyses. The predicted and observed values were found to be similar, which indicates that this prediction model demonstrates a good degree of discrimination and calibration. Clinicians can identify the patients at high risk of death for patient management and reduce the fatality rate according to this nomogram. It is recommended to poisoning centers as effective criteria.
Limitations of this study
Our study has some limitations. First, as this was a single-center retrospective study, there may be selection bias in addition to possible confounding. Second, our prediction model only takes into account the relevant clinical data on admission, and does not consider the impact of previous diseases and treatment on the prognosis. Multi-center prospective studies will be needed to verify the accuracy of this model in future research.