Assessing the reliability and accuracy of nurse triage ratings when using the South African Triage Scale in the Emergency Department of District Headquarter Hospital of Timergara, Pakistan

Abstract

To assess inter and intra-rater reliability, as well as accuracy of nurse triage ratings when using the South African Triage Scale (SATS) in the Emergency Department (ED) of District Headquarter (DHQ) Hospital of Timergara, Pakistan. Methods: This is a cross-sectional study using 42 previously validated paper based vignettes. Fifteen ED nurses assigned triage ratings to each of the 42 vignettes under classroom conditions. Validation of the SATS was done using the vignettes as a reference standard. Graphical displays portrayed rating distribution and validation measures of sensitivity, specificity, overtriage and undertriage across different acuity levels. Results: The estimated Quadratically weighted Kappa (QWK) and Interclass Correlation was found to be substantial at 0.77(CI 0.69–0.85). Intra-rater reliability with exact agreement was shown to be 87% (CI 67–100) with one category discrepancy showing 100% agreement. An average sensitivity, 70%; specificity, 97%; overtriage, 14.7% and undertriage, 21.6% was shown. The Graphical displayed showed that Very Urgent, Urgent and routine acuity levels had acceptable levels of overtriage and undertriage. Conclusion: The SATS has been shown to be a reliable triage scale for a developing country such as Pakistan. With accuracy being acceptable in the context of Timergara, we would suggest further validation studies looking at simple ways of validating the triage scale bearing in mind the challenges facing a developing country ED.