Background: Typhoid fever is one of the most common infectious diseases in the world with up to 17 million cases and 600,000 deaths annually worldwide. With the growing body of evidence testing newer serologic tests developed to circumvent the problems of culture (time-consuming, requiring specialized personnel), this meta-analysis was done to evaluate and compare the diagnostic accuracy of these tes Methods: All studies that were original articles withhumansubjects testing serologic tests wherein the gold standard utilized is isolation of the organism from the blood of host of medium to high quality were included. Random effects meta-analysis and hierarchical summary receiver operating characteristics curve were used to evaluate diagnostic accuracy Results: A total of 12 studies evaluated the sensitivity and specificity of a single or a combination of the serologic tests. The pooled sensitivity and specificity of all the serologic tests are fair at 72% (0.70-0.74, 95% CI) and 78% (0.76-0.80, 95% CI) respectively. The pooled positive likelihood ratios for all the serologic tests are 3.52 (X2 = 648, p = 0.0) and the pooled negative likelihood ratio is 0.360 (X2 = 314.96, p = 0.0). The positive and negative likelihood ratios for dipstick (LR+ = 8.600, LR- = 0.248), Tubex (LR+ = 4.661, LR- = 0.256), Typhidot (LR+ = 5.753, LR- = 0.321), Widal test (LR+ = 2.034, LR- = 0.556) were calculated. Index tests with single studies (Typhidot- M, Mega Salmonella, Cromotest, ELISA and PCR) were not included in the calculation of likelihood ratios and ROC curves. The area under the curve for applicable index tests showed considerable differences with Tubex (AUC = 0.903) showing the greatest diagnostic accuracy than Typhidot (AUC = 0.836), dipstick (AUC = 0.806) and the least wasWidal (AUC = 0.660). Conclusion: The serologic tests studied in general have promising diagnostic accuracy in confirming typhoid fever; however this varies among the different index tests. Tubex was shown to have the overall greatest diagnostic accuracy compared to blood culture and should be used, and Widal was shown to have the least and should no longer be used.
Alejandria, M., Concepcion, A. O., Li, R. J., & Gutierrez, J. (2012). The sensitivity and specificity of serologic tests in the diagnosis of typhoid fever in adults: a meta-analysis. International Journal of Infectious Diseases, 16, e393. https://doi.org/10.1016/j.ijid.2012.05.521