Background:
Tuberculosis (TB) remains a major health challenge globally, including Pacific countries. Solomon Islands, with a population of ~700,000 reports approximately 460 new cases and 50 deaths every year3. Diagnosis of TB in Solomon Islands is based on clinical manifestation, chest x-rays and microscopy of sputum smear slides, with diagnosis more challenging in provincial settings where there is limited facilities, training, and resources. This study aims to compare the efficacy of microscopy with molecular testing (real-time PCR) as reference standard.
Materials and Methods:
Archived sputum smear slides (n= 449) were collected from three provincial TB laboratories from 2013 to 2021. Smear scrapings were extracted in Chelex, and an optimised real-time PCR assay targeting the IS61101 and mpt642 genes was applied to detect M. tuberculosis. Laboratory records of microscopy results were compared with real-time PCR results. Sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) for the microscopy was calculated. The level of agreement and any difference between microscopy and qPCR test results was determined. The p-value was calculated by the McNemar Test.
Results:
Of the 449 slides analysed, 79% had concordant results by microscopy and real-time PCR (kappa correlation: 0.40). No significant difference (p-value = 0.84) was observed between microscopy and qPCR test results. The sensitivity and specificity of microscopy, when compared to real-time PCR, were 54.5% (CI 44.2% to 64.4%) and 85.9% (CI 81.8% to 89.4%), respectively. The PPV and NPV of the microscopy method were 52.9% and 86.7% respectively.
Conclusions:
These data suggest that smear positive cases may be missed using microscopy considering only approximately half (54%) of slides that were positive by PCR were positive by microscopy. This low rate of detection is likely to have implications for patient health and transmission dynamics. Necessary actions need to be taken to improve TB diagnosis at provincial setting in SI. The rollout of the GeneXpert system may improve diagnosis.