Clinico-microbiological description and evaluation of rapid lateral flow immunoassay and PCR for detection of Burkholderia pseudomallei from patients hospitalized with sepsis and pneumonia: A twenty-one months study from Odisha, India

• The prevalence of melioidosis in patients presenting with sepsis without CAP, with sepsis and underlying CAP, with only CAP without sepsis was 31%, 22% and 3 % respectively. • All the LFI or PCR positive cases were culture positive from at least one or more samples (blood/sputum/urine/pus). • The sensitivity, specificity, positive predictive value and negative predictive value of TTS1 PCR was 78% (18/23 melioidosis patients), 100% (34/34 non-melioidosis patients), 100% (18/18 melioidosis patients) and 87% (34/39 non-melioidosis patients). • The sensitivity, specificity, positive predictive value and negative predictive value of Rapid LFI was 91% (21/23 melioidosis patients), 100% (22/22 non-melioidosis patients), 100% (21/21 melioidosis patients) and 91% (22/24 non-melioidosis patients). • On sample wise stratification of LFI and TTS1 with respect to culture, plasma/serum samples showed the highest discordance by PCR (9/55, 16%) and LFI (11/35, 31%), with respect to culture. The lowest discordance was noted in respiratory tract samples (2/32, 6%) by PCR and pus/body fluids samples (2/14, 14%) by LFI. • The clinical utility of PCR and LFI needs to be further validated in a large scale study for early diagnosis of septicaemic melioidosis. Establishing a diagnosis of melioidosis based on clinical grounds is difficult in hospitalized patients with sepsis or community acquired pneumonia (CAP). We aimed to ascertain the prevalence, clinico-epidemiological and laboratory profile of melioidosis in hospitalized patients with sepsis or CAP, and to evaluate the diagnostic utility of rapid lateral flow immunoassay (LFI) and PCR in comparison with culture. In all patients with sepsis or CAP, blood, sputum/throat swab, and urine sample were subjected to culture along with other samples based on clinical presentation. In addition, PCR assay targeting the type III secretion system 1 (TTS1) and LFI was performed. Thirty-three (33/196, 17%) out of the total 196 cases were diagnosed as melioidosis by culture. The prevalence of melioidosis in patients who had only sepsis without CAP, had both sepsis and CAP, had CAP without sepsis was 31% (26/84), 22 % (4/18) and 3%(3/94) respectively. All the LFI or PCR positive cases were culture positive from at least one or more samples (blood/sputum/urine/pus). The sensitivity, specificity, positive predictive value and negative predictive value of TTS1 PCR was 78% (18/23 melioidosis patients), 100% (34/34 non-melioidosis patients), 100% (18/18 melioidosis patients) and 87% (34/39 non-melioidosis patients). The sensitivity, specificity, positive predictive value and negative predictive value of Rapid LFI was 91% (21/23 melioidosis patients), 100% (22/22 non-melioidosis patients), 100% (21/21 melioidosis patients) and 91% (22/24 non-melioidosis patients). On sample wise stratification of LFI and TTS1 with respect to culture, plasma/serum samples showed the highest discordance by PCR (9/55, 16.3%) and LFI (11/35, 31.4%). The lowest discordance was noted in respiratory tract samples (2/32, 6.2%) by PCR and pus/body fluids samples (2/14, 14.2%) by LFI and these findings are in line with previous published literature. The clinical utility of PCR and LFI needs to be further validated in a large scale study for early diagnosis of septicaemic melioidosis.