Repeated tourniquet testing as a diagnostic tool in dengue infection.

Prospective evaluation of repeated standard tourniquet testing as a diagnostic indicator of dengue infection was done. Included were patients admitted to a children's hospital in Kuala Lumpur on a clinical suspicion of dengue infection based on the World Health Organization (WHO) criteria. A standard method of tourniquet was performed on 79 patients on a daily basis following admission. subjects and negative in the remaining 14 subjects. Fifty-eight subjects were serologically confirmed cases, 4 indeterminate and the remaining 17 subjects had negative serology. For diagnostic classification, 13 had dengue fever, 49 with dengue haemmorhagic fever (DHF) while 17 had non-dengue infection. The sensitivity and specificity of the tourniquet test was 82.8% and 23.5% respectively. The positive predictive value (PPV) was 78.7% while the negative predictive value (NPV) was 28.6%. In addition, the tourniquet test aided in the diagnosis of one-fifth of patients with DHF, who presented with a positive tourniquet test as the only bleeding manifestation. It seems that in a hospital setting, the tourniquet test adds little to the diagnosis of dengue infection/DHF. A positive tourniquet test, repeatedly performed, was found clinically useful as a preliminary screening test in dengue infection as recommended by WHO. However, it was not very specific and had a high false positive rate.