Correlation between clinical and laboratory diagnosis of dengue in Sri Lanka.

Abstract

BACKGROUND: In Sri Lanka, diagnosis of dengue mainly depends on clinical signs and symptoms. Very few suspected patients from the state and private sector health institutions are tested by laboratory diagnostic assays compared to the number of dengue cases recorded all over the island. OBJECTIVES: To correlate clinical parameters with laboratory diagnosis in confirmation of dengue. RESEARCH DESIGN: Patients, clinically suspected of having dengue (n=201) were selected based on WHO criteria. Serum samples were tested using major 3 types of laboratory diagnostic assays; molecular, virus isolation and serology. Differences in clinical and laboratory data were analyzed on the basis of the final diagnosis assigned as dengue or non-dengue. Chi-square test was used for comparison of data. RESULTS: The proportion of laboratory diagnosed dengue patients were 80% (162/201). Mean platelet value and PCV in laboratory confirmed dengue patients were 92 247/mm3 (range 20 000-318 000) and 45% (range 31-59%) respectively. On comparison of the presence of clinical features that are used by the WHO for diagnosis of dengue, headache (129/162 vs 18/39, x2=23, p=0.00), limb pain (107/162 vs 18/39, x2=4.56, p=0.03) and external bleeding (67/162 vs 00/39, X2=27, p=0.00) showed significant association, with dengue infection. The infection was confirmed as definitive dengue in 75% (121 /162) and probable dengue in 25% (41/162). DISCUSSION: Surveillance based on clinical diagnosis may result in over estimation of the disease as clinical diagnosis is not specific enough. Laboratory confirmation of dengue suspected patients is important to measure the real incidence of the disease is needed in country like Sri Lanka.