Abstract:

Objective: To analyze the clinical features and risk factors associated with the prognosis of tuberculous meningitis (TBM) in children, and to provide evidence for early detection and intervention on high-risk TBM in pediatric patients. Methods: Clinical records of TBM children diagnosed in Shanghai Public Health Clinical Center (Fudan University) between 01/01/2013 and 12/31/2023 were reviewed retrospectively. The neurological outcomes of anti-TB treatment at month-12 were evaluated by the modified Rankin Scale (mRS) through outpatient visit and telephone interview. The clinical characteristics of children with good prognosis and poor prognosis were compared, univariate and multivariate analyses were used to determine factors associated with poor prognosis. Results: A total of 91 patients were eligible for the study. Among them, 62 had a good prognosis while the left 29 had a poor prognosis. The median (quartile) age was 3.00 (1.00, 8.00) years old, and 51 were boys. On admission, the poor prognosis group had higher proportion of abnormal muscle strength (12/29, 41.38%), confusion (16/29, 55.17%), and hydrocephalus (17/29, 58.62%) compared to the good prognosis group ((9/62, 14.52%), (13/62, 20.97%), and (22/62, 35.48%), respectively)(χ²=8.032, P=0.005; χ²=10.647, P<0.001; χ²=4.319, P=0.038). In contrast, the frequency of cough ≥2 weeks (4/29, 13.79%) and expectoration (1/29, 3.45%) were lower in the poor prognosis group than in the good prognosis group ((21/62, 33.87%) and (13/62, 20.97%), respectively)(χ²=3.997, P=0.046; χ²=4.659, P=0.031). The Glasgow Coma Scale scores of the poor prognosis group were 13, lower than the 15 in the good prognosis group (Z=-4.190, P<0.001), and the difference in Medical Research Council (MRC) disease stage was statistically significant (χ²=22.327, P<0.001). In addition, the median (quartile) cerebrospinal fluid (CSF) lactate dehydrogenase level of 60.00 (29.50, 128.00) IU/L, the use of external ventricular drainage (7/29, 24.14%), and the frequency of mechanical ventilation (19/29, 65.52%) were higher in the poor prognosis group than in the good prognosis group (39.50 (20.75, 64.00) IU/L, (2/62, 3.23%), and (25/62, 40.32%), respectively)(Z=-2.100, P=0.036; χ²=9.696, P=0.002; χ²=5.022, P=0.025). Multivariate logistic regression analysis showed that confusion on admission (OR=3.690; 95%CI: 1.045-13.034; P=0.043) and external ventricular drainage (OR=8.041; 95%CI: 1.064-60.760; P=0.043) were independent risk factors for poor prognosis in children with TBM. Conclusion: Children with poor prognosis of TBM were more likely to have severe neurologic symptoms (e.g., muscle strength abnormalities, low GCS scores), hydrocephalus, and elevated CSF lactate dehydrogenase, and require higher mechanical ventilation. In contrast, respiratory symptoms (e.g., cough, sputum) were more prominent in the good prognosis group, which may suggest early consultation or milder disease. Admission confusion and external ventricular drainage were independent risk factors for poor prognosis.

Key words: Children, Tuberculosis, meningeal, Prognosis, Risk factors