The Optimum Cut-off Value of the Nelwan Score to Diagnose Typhoid Fever in Adults

Abstract IntroductionTyphoid fever can be challenging to diagnose since clinicians often depend merely on clinical presentation. Clinical scores are useful to provide more accurate diagnosis. Variables in Nelwan Score are derived from clinical signs and symptoms of suspected cases for typhoid. Diagnostic value of Nelwan Score based on a cut-off value has never been evaluated. MethodsA cross sectional study was conducted between July 2017 and January 2018 in five hospitals and two Primary Health Centers in Jakarta and Tangerang. The inclusion criteria were patients with 3-14 days of fever and gastrointestinal symptoms between July 2017 and January 2018. Diagnosis are confirmed by blood culture, rectal swab culture, or PCR. Cut-off analysis was performed by using Receiver Operating Characteristic (ROC) curve and diagnostic value was analyzed to generate sensitivity, specificity, predictive value and likelihood ratio. ResultFrom 233 subjects involved, 4.72% of them were confirmed to have typhoid fever. The optimal cut-off value of Nelwan Score is 10 with AUC 71.3%. This cut-off value has sensitivity 81.8%, specificity 60.8%, PPV 9.3%, NPV 98.5%, LR + 2.086, and LR – 0.299. ConclusionNelwan Score with cut-off value of 10 provides a good diagnostic performance as a screening tool for patients with suspected typhoid fever clinical presentation.