Diagnostic Accuracy of World Health Organization Case Definitions for Acute Febrile Illness: A Tertiary Care Hospital-based Study.

Abstract

Background: The causes of acute febrile illness (AFI) in South India have been challenging to identify. As health care facilities in many endemic areas lack good laboratory support for the diagnosis of AFI, there is difficulty in ascertaining the causes of AFI, resulting in a delay in the initiation of treatment. This study aimed to evaluate the sensitivity and specificity of surveillance criteria and rapid diagnostic tests in the diagnosis of patients presenting with AFI and thrombocytopenia.

Materials and methods: Our study enrolled 225 participants presenting with AFI who fulfilled the predetermined inclusion and exclusion criteria. The study was conducted in the medical wards of YMCH from January 2018 to January 2019, after obtaining institutional ethics clearance. The patients were selected according to the inclusion criteria after obtaining informed consent. The collected data were analyzed using SPSS v23.0 operating on Windows 10 to determine the diagnostic characteristics, such as sensitivity and specificity of surveillance criteria and rapid diagnostic tests.

Results: Our study enrolled 225 participants, with a majority (71.3%) falling within the 20-40 age range, and a significant male predominance. Around 68% of patients had a fever lasting <5 days. The clinical diagnosis was made using surveillance criteria in all 225 participants, of which 44% of cases were dengue, 35% were malaria, 12% were leptospirosis, and 7% were scrub typhus. However, when a definitive diagnosis was made using more specific tests (ELISA for dengue, leptospirosis, and scrub typhus; MP smear for malaria), it was found that 32% had dengue, 26% had malaria, 7% had leptospirosis, 5% had scrub typhus, 27% had undifferentiated fever, and 1.8% had coinfection. The overall accuracy of the surveillance criteria in diagnosing dengue is 66.37%, malaria is 76.91%, leptospirosis is 84.07%, and scrub typhus is 89.46%.

Conclusion: This study validates the effectiveness of the surveillance case definition in identifying AFI with thrombocytopenia, demonstrating high sensitivity and moderate specificity. It should be followed by routine rapid diagnostic tests and more specific investigations to arrive at a definitive diagnosis of AFI.