Lysozyme levels for the diagnosis of tuberculous effusions in children.

Lysozyme level was measured in the fluid and serum of 42 tuberculous (25 pleural, 11 ascites and 6 pericardial) and 29 non-tuberculous (5 malignant, 9 empyema thoracis, 10 transudative ascites and 5 pyopericardium) effusions. The mean fluid lysozyme level was significantly raised in tuberculous pleural, ascites, and pericardial effusions in comparison to malignant pleural (p <0.001), transudative ascites (p < 0.001), and pyopericardium (p < 0.02) cases, respectively. The mean fluid/serum lysozyme ratio did not differ significantly between tuberculous and their corresponding non-tuberculous effusions. The confirmed tuberculous pleural effusion patients had significantly higher mean fluid lysozyme level and fluid/serum lysozyme ratio when compared with clinical cases (p < 0.05). The cut-off fluid lysozyme level of > or = 50/UI(-1) and fluid/serum lysozyme ratio of > or = 1.1 were considered for the diagnosis of tuberculous effusions; the sensitivity and specificity of fluid lysozyme and fluid/serum lysozyme ratios were 100, 100 per cent, and 97.6, 33.3 per cent, respectively, on excluding the patients with purulent effusions. A significant correlation was observed between the fluid and serum lysozyme levels in tuberculous effusions (r = 0.39,p < 0.01). Thus, fluid lysozyme was found to be a better and reliable test than fluid/serum lysozyme ratio for the diagnosis of tuberculous effusions in children.