Main Article Content

Classifying snakebite in South Africa: Validating a scoring system


D Wood
B Sartorius
R Hift

Abstract

Objective. To develop and validate a scoring system for managing snakebites in South Africa (SA).
Methods. We studied all snakebite admissions to a regional hospital in KwaZulu-Natal, SA. The primary outcome was an active treatment intervention (ATI) defined as antivenom treatment or any surgical procedure. The development cohort consisted of 879 patients with snakebite who presented to the Ngwelezane Hospital Emergency Department from December 2008 to December 2013. Factors predictive of ATI and the optimal cut-off score for predicting an ATI were identified. These factors were then used to develop a standard scoring system. The score was then tested prospectively for accuracy in a new validation cohort consisting of 100 patients admitted for snakebite to our unit from 1 December 2014 to 31 March 2015. Accuracy of the score was determined.
Results. Of 879 snakebite admissions, 146 in the development cohort and 40 of 100 in the development validation cohort reached the primary endpoint of an ATI. Six risk predictors for ATI were identified from the development cohort: age <14 years (odds ratio (OR) 2.13), delay to admission >7 hours (OR 4.63), white cell count >10 × 109/L (OR 3.15), platelets <92 × 109/L (OR 2.35), haemoglobin <7.1 g/dL (OR 5.68), international normalised ratio >1.2 (OR 2.25). Each risk predictor was assigned a score of 1; receiver operating characteristic curve analysis returned a value of >4 out of 6 as the optimal cut-off for prediction of an ATI (area under the curve 0.804; 95% confidence interval 0.758 - 0.84). Testing of the score on the validation cohort produced a specificity of 96.6% and a sensitivity of 22.5%. The positive predictive value and negative predictive value were 81.8% and 65.2%, respectively.
Conclusion. Our results show that the identified score is a useful adjunct to clinical assessment in managing snakebite. Its value is greatest when used in those patients who fall in the mild to moderate clinical category. Until our severity score has been validated (or modified) for use across SA, we propose to name it the Zululand Severity Score; a true SA Severity Score may follow.


Journal Identifiers


eISSN: 2078-5135
print ISSN: 0256-9574