Abstract
Background:
Tuberculous meningitis (TBM) remains a leading cause of morbidity and mortality in children, particularly in endemic regions. Early diagnosis is crucial for optimal outcomes, yet remains challenging due to nonspecific clinical presentations and limited diagnostic sensitivity.
Objective:
To evaluate the diagnostic utility of various microbiological and neuroimaging modalities in clinically suspected cases of pediatric TBM and classify cases according to the established diagnostic criteria.
Materials and Methods:
A prospective observational study was conducted at the department of pediatrics of a rural tertiary care center in North India. Fifty children aged 6 months–15 years with suspected TBM were enrolled. All participants underwent comprehensive clinical evaluation, cerebrospinal fluid (CSF) analysis including cartridge-based nucleic acid amplification test (CBNAAT), adenosine deaminase (ADA) estimation, contrast-enhanced computed tomography imaging, and chest radiography. Cases were classified using the Marais diagnostic scoring system.
Results:
The study population had a median age of 3.2 years with female predominance (58%). All patients presented with fever, while vomiting (92%), irritability (82%), altered sensorium (78%), and seizures (72%) were other common manifestations. 68% of patients presented in advanced British Medical Research Council (BMRC) stages (II-III). CSF analysis revealed lymphocytic predominance in 52% and elevated protein (40–500 mg/dL) in 94% of cases. CBNAAT was positive in 18% of patients, with rifampicin resistance detected in 11% of positive cases. Elevated CSF-ADA (>10 U/L) was found in 42% of all cases and 89% of definite TBM cases. Neuroimaging abnormalities were present in 90% of patients, with hydrocephalus being the most common finding (50%), followed by basal exudates and meningeal enhancement (38%). Mortality was significantly associated with advanced BMRC staging, with 100% fatality in Stage III patients.
Conclusion:
Neuroimaging demonstrated the highest sensitivity for TBM diagnosis in children. While CBNAAT provided definitive diagnosis with high specificity, its sensitivity remained limited. A multimodal diagnostic approach incorporating clinical assessment and CSF analysis including ADA, CBNAAT, and neuroimaging is essential for optimal diagnostic accuracy in pediatric TBM.
Keywords: Adenosine deaminase, cartridge-based nucleic acid amplification test, children, diagnosis, neuroimaging, tuberculous meningitis
Résumé
Contexte:
La méningite tuberculeuse (MT) demeure une cause majeure de morbidité et de mortalité chez les enfants, particulièrement dans les régions endémiques. Le diagnostic précoce est crucial pour des résultats optimaux, mais reste difficile en raison de présentations cliniques non spécifiques et d’une sensibilité diagnostique limitée.
Objectif:
Évaluer l’utilité diagnostique de diverses modalités microbiologiques et de neuroimagerie dans les cas cliniquement suspects de MT pédiatrique et classifier les cas selon les critères diagnostiques établis.
Matériels et Méthodes:
Une étude observationnelle prospective a été menée au département de pédiatrie d’un centre de soins tertiaires rural du nord de l’Inde. Cinquante enfants âgés de 6 mois à 15 ans présentant une suspicion de MT ont été recrutés. Tous les participants ont bénéficié d’une évaluation clinique complète, d’une analyse du liquide céphalorachidien (LCR) incluant le test d’amplification des acides nucléiques par cartouche (CBNAAT), du dosage de l’adénosine désaminase (ADA), d’une tomodensitométrie avec injection de produit de contraste et d’une radiographie thoracique. Les cas ont été classifiés selon le système de score diagnostique de Marais.
Résultats:
La population étudiée avait un âge médian de 3,2 ans avec une prédominance féminine (58 %). Tous les patients présentaient de la fièvre, tandis que les vomissements (92 %), l’irritabilité (82 %), une altération de la conscience (78 %) et les convulsions (72 %) constituaient d’autres manifestations courantes. 68 % des patients se présentaient à des stades avancés du British Medical Research Council (BMRC) (II-III). L’analyse du LCR a révélé une prédominance lymphocytaire chez 52 % des cas et une protéinorachie élevée (40-500 mg/dL) chez 94 % des cas. Le CBNAAT était positif chez 18 % des patients, avec une résistance à la rifampicine détectée chez 11 % des cas positifs. Une élévation de l’ADA dans le LCR (>10 U/L) a été constatée chez 42 % de tous les cas et chez 89 % des cas de MT certaine. Des anomalies en neuroimagerie étaient présentes chez 90 % des patients, l’hydrocéphalie étant la découverte la plus fréquente (50 %), suivie par les exsudats de la base du crâne et le rehaussement méningé (38 %). La mortalité était significativement associée au stade avancé du BMRC, avec une létalité de 100 % chez les patients de stade III.
Conclusion:
La neuroimagerie a démontré la sensibilité la plus élevée pour le diagnostic de MT chez les enfants. Bien que le CBNAAT ait fourni un diagnostic définitif avec une spécificité élevée, sa sensibilité est restée limitée. Une approche diagnostique multimodale incorporant l’évaluation clinique et l’analyse du LCR incluant l’ADA, le CBNAAT et la neuroimagerie est essentielle pour une précision diagnostique optimale dans la MT pédiatrique.
Mots-clés: Adénosine désaminase, test d’amplification des acides nucléiques par cartouche, enfants, diagnostic, neuroimagerie, méningite tuberculeuse
INTRODUCTION
Tuberculosis (TB) remains a significant global health burden, with an estimated 3.4 lakh children affected annually in India alone, representing approximately 6% of total reported TB cases to the National Tuberculosis Elimination Program.[1] Tuberculous meningitis (TBM), the most severe manifestation of extrapulmonary TB, predominantly affects young children aged 6 months–4 years and carries substantial morbidity and mortality rates exceeding 50%.[2,3]
The pathophysiology of TBM involves hematogenous dissemination of Mycobacterium tuberculosis from primary pulmonary foci to the central nervous system, where rich foci develop in subpial or subependymal regions. Subsequent rupture of these foci into the subarachnoid space triggers the characteristic inflammatory cascade, leading to basal exudates, hydrocephalus, and vascular complications.[4,5]
Early diagnosis and treatment initiation are paramount for favorable outcomes, yet diagnosis remains challenging due to nonspecific clinical presentations and the paucibacillary nature of the disease.[6] Conventional diagnostic methods including acid-fast bacilli staining and culture have limited sensitivity, particularly in pediatric populations. Advanced molecular diagnostics such as cartridge-based nucleic acid amplification test (CBNAAT) and biomarkers like cerebrospinal fluid (CSF) adenosine deaminase (ADA) have shown promise in improving diagnostic accuracy.[7,8]
The Marais diagnostic scoring system provides a standardized framework for TBM classification, categorizing cases as definite, probable, or possible based on clinical, laboratory, and radiological criteria.[9] This study aimed to evaluate the diagnostic performance of various microbiological and neuroimaging modalities in pediatric TBM using this established classification system.
MATERIALS AND METHODS
Study design and setting
This prospective observational study was conducted at the author’s institution Uttar Pradesh University of Medical Sciences, Saifai, Etawah – a Rural tertiary care center that also receives patients from adjoining districts in western UP, India, from October 2022 to March 2024. The study protocol was approved by the institutional ethics committee.
Study population
Children aged 6 months–15 years presenting with clinical features suggestive of TBM were eligible for inclusion. Inclusion criteria comprised fever duration exceeding 2 weeks accompanied by headache, signs of meningeal irritation, altered consciousness, or focal neurological deficits. Exclusion criteria included prior antitubercular therapy exceeding 2 months, age below 6 months, and confirmed alternative causes of meningitis.
Comprehensive history taking included documentation of TB contact, immunization status, socioeconomic status,[10] and nutritional assessment. Physical examination focused on neurological evaluation, including Glasgow Coma Scale scoring, cranial nerve assessment, and evaluation for meningeal signs. Disease severity was classified according to the British Medical Research Council (BMRC) staging.
CSF was obtained through lumbar puncture under sterile conditions and analyzed for:
Routine examination (cell count, protein, and glucose)
CBNAAT using GeneXpert M. tuberculosis/resistance to rifampicin
ADA estimation
Conventional microscopy for acid-fast bacilli.
Blood investigations included complete blood count, erythrocyte sedimentation rate, and biochemical parameters. Tuberculin skin testing was performed using purified protein derivative.
All patients underwent contrast-enhanced computed tomography of the brain and chest radiography. Neuroimaging findings were systematically evaluated for hydrocephalus, basal exudates, infarctions, and tuberculomas.
Cases were classified according to the Marais diagnostic criteria:
Definite TBM: Microbiological confirmation through positive CSF microscopy, culture, or CBNAAT
Probable TBM: Total score ≥12 with neuroimaging or ≥10 without neuroimaging
Possible TBM: Total score 6–11 with neuroimaging or 6–9 without neuroimaging.
Statistical analysis
Descriptive statistics were calculated for all variables. Categorical variables were expressed as frequencies and percentages, whereas continuous variables were presented as means with standard deviations or medians with ranges. Statistical significance was set at P < 0.05.
RESULTS
Fifty children with suspected TBM were enrolled during the study period. The demographic and baseline characteristics are summarized in Table 1. The median age was 3.2 years, with 46% of patients under 5 years of age. Females constituted 58% of the study population. Most families (66%) belonged to lower socioeconomic status, with 94% residing in rural areas. Malnutrition was universal, with 38% having severe malnutrition and 62% showing mild-to-moderate malnutrition.
Table 1.
Demographic and baseline characteristics of study population (n=50)
| Characteristic | n (%) |
|---|---|
| Age distribution | |
| 6 months–2 years | 18 (36.0) |
| 2–5 years | 23 (46.0) |
| 5–10 years | 7 (14.0) |
| 10–15 years | 2 (4.0) |
| Gender | |
| Male | 21 (42.0) |
| Female | 29 (58.0) |
| Residence | |
| Rural | 47 (94.0) |
| Urban | 3 (6.0) |
| Socioeconomic status | |
| Lower | 33 (66.0) |
| Middle | 17 (34.0) |
| Nutritional status | |
| Mild-moderate malnutrition | 31 (62.0) |
| Severe malnutrition | 19 (38.0) |
| TB contact history | 23 (46.0) |
| BCG vaccination | 14 (28.0) |
| Tuberculin test positive | 18 (36.0) |
TB=Tuberculosis, BCG=Bacillus Calmette –Guerin
The clinical features and disease severity at presentation are detailed in Table 2. Fever was universal (100%), followed by vomiting (92%), irritability (82%), altered sensorium (78%), and seizures (72%). Signs of meningeal irritation were observed in 82% of cases. Advanced disease presentation was common, with 68% of patients classified as BMRC Stage II or III at admission.
Table 2.
Clinical features and disease severity at presentation (n=50)
| Clinical feature | n (%) |
|---|---|
| Symptoms | |
| Fever | 50 (100.0) |
| Vomiting | 46 (92.0) |
| Irritability | 41 (82.0) |
| Altered sensorium | 39 (78.0) |
| Seizures | 36 (72.0) |
| Headache | 33 (66.0) |
| Weakness | 15 (30.0) |
| Physical signs | |
| Meningeal signs | 41 (82.0) |
| Neck rigidity | 39 (78.0) |
| Kernig’s sign | 35 (70.0) |
| Brudzinski’s sign | 33 (66.0) |
| Neurological deficits | 20 (40.0) |
| Cranial nerve palsy | 14 (28.0) |
| Hemiparesis/hemiplegia | 6 (12.0) |
| BMRC stage | |
| Stage I | 16 (32.0) |
| Stage II | 24 (48.0) |
| Stage III | 10 (20.0) |
| Glasgow Coma Scale | |
| 13–15 (mild impairment) | 17 (34.0) |
| 9–12 (moderate impairment) | 23 (46.0) |
| <8 (severe impairment) | 10 (20.0) |
| Fundoscopic findings | |
| Papilledema | 29 (58.0) |
| Choroidal tubercles | 12 (24.0) |
| Optic atrophy | 4 (8.0) |
| Normal | 5 (10.0) |
BMRC=British Medical Research Council
Comprehensive laboratory results and diagnostic test performance are presented in Table 3. CSF analysis revealed lymphocytic predominance in 52% of cases, with elevated protein levels in 94% of patients. CBNAAT was positive in 18% of patients, while elevated CSF-ADA was found in 42% of all cases.
Table 3.
Laboratory findings and diagnostic test results (n=50)
| Parameter | Result, n (%) |
|---|---|
| CSF analysis | |
| Lymphocytic predominance (>50%) | 26 (52.0) |
| Protein elevation (40–500 mg/dL) | 47 (94.0) |
| Low glucose (<45 mg/dL) | 34 (68.0) |
| Microbiological tests | |
| CBNAAT positive | 9 (18.0) |
| Rifampicin resistance | 2/9 (22.2) |
| AFB smear positive | 3 (6.0) |
| CSF-ADA (>10 U/L) | |
| Overall positive | 21 (42.0) |
| Definite TBM cases | 8/9 (88.9) |
| Probable TBM cases | 11/28 (39.3) |
| Possible TBM cases | 2/13 (15.4) |
| Blood parameters | |
| Anemia (Hb <11 g/dL) | 38 (76.0) |
| Elevated ESR (>20 mm/h) | 26 (52.0) |
| Chest X-ray abnormal | 19 (38.0) |
| Hilar lymphadenopathy | 12 (24.0) |
| Pulmonary infiltrates | 5 (10.0) |
| Miliary pattern | 2 (4.0) |
CSF=Cerebrospinal fluid, ADA=Adenosine deaminase, TBM=Tuberculous meningitis, Hb=Hemoglobin, ESR=Erythrocyte sedimentation rate, CBNAAT=Cartridge-based nucleic acid amplification test, AFB=Acid-fast bacilli
Neuroimaging abnormalities and final case classification according to the Marais criteria are shown in Table 4. Abnormal neuroimaging was present in 90% of patients, with hydrocephalus being the most frequent finding (50%). Based on diagnostic criteria, 18% were classified as definite TBM, 56% as probable TBM, and 26% as possible TBM.
Table 4.
Neuroimaging findings, case classification and outcomes (n=50)
| Finding | n (%) | |
|---|---|---|
| Neuroimaging abnormalities | 45 (90.0) | |
| Hydrocephalus | 25 (50.0) | |
| Communicating | 18 (36.0) | |
| Obstructive | 7 (14.0) | |
| Basal exudates/meningeal enhancement | 19 (38.0) | |
| Cerebral infarctions | 9 (18.0) | |
| Tuberculomas | 3 (6.0) | |
| Case classification (marais criteria) | ||
| Definite TBM | 9 (18.0) | |
| Probable TBM | 28 (56.0) | |
| Possible TBM | 13 (26.0) | |
|
| ||
| BMRC stage | Outcome | n/N (%) |
|
| ||
| Stage I | Discharged without deficit | 16/16 (100.0) |
| Stage II | Discharged (with/without deficit) | 20/24 (83.3) |
| Stage II | Mortality | 4/24 (16.7) |
| Stage III | Mortality | 10/10 (100.0) |
| Overall mortality | 14 (28.0) | |
BMRC=British Medical Research Council, TBM=Tuberculous meningitis
Mortality significantly increased with advancing stage as shown in Table 4. Overall, mortality differed across stages (Chi-square test [χ2] = 33.47, P = 5.4 × 10−8). Pairwise Fisher’s exact tests showed no significant difference between Stage I and Stage II (P = 0.136), whereas mortality was significantly higher in Stage III compared with Stage II (P = 7.6 × 10−6) and Stage I (P = 1.88 × 10−7).
Figure 1 illustrates the diagnostic yield of various modalities used in the study. Neuroimaging demonstrated the highest sensitivity (90%), followed by CSF biochemical analysis (94% for protein elevation). The combination of clinical features, CSF analysis, and neuroimaging provided optimal diagnostic accuracy.
Figure 1.
Diagnostic yield of various modalities. CSF: Cerebrospinal fluid, ADA: Adenosine deaminase, CBNAAT: Cartridge-based nucleic acid amplification test
Figure 2 demonstrates the relationship between diagnostic test positivity and BMRC staging. CSF-ADA showed higher positivity rates in advanced stages, while neuroimaging abnormalities were consistently present across all stages. CBNAAT positivity was distributed across different stages without a clear pattern related to severity.
Figure 2.
Relationship between diagnostic test positivity and BMRC staging. BMRC: British Medical Research Council, CSF: Cerebrospinal fluid, ADA: Adenosine deaminase, CBNAAT: Cartridge-based nucleic acid amplification test
CSF-ADA >10 U/L: Higher in Stage II (10/21 cases), supporting its link to moderate-severe disease
Neuroimaging: Dominates all stages (90% overall), consistent with our data (hydrocephalus = 50%)
CBNAAT+: No stage-specific trend (low sensitivity).
DISCUSSION
This evaluation of diagnostic modalities in pediatric TBM reveals several important findings that contribute to our understanding of diagnostic approaches in this challenging condition. The demographic profile of our study population aligns with established epidemiological patterns, showing predominant involvement of young children from disadvantaged socioeconomic backgrounds with high rates of malnutrition.
The clinical presentation in our cohort was consistent with previously reported series, though the high proportion of patients presenting in advanced BMRC stages (68%) underscores the diagnostic challenges and delays commonly encountered in pediatric TBM.[11] The universal presence of fever, combined with the high frequency of vomiting and altered sensorium, reinforces the importance of maintaining high clinical suspicion in appropriate epidemiological contexts.
Our CBNAAT positivity rate of 18% falls within the range reported in pediatric studies, which typically show lower sensitivity compared to adult populations due to the paucibacillary nature of pediatric TBM.[12,13] The detection of rifampicin resistance in 22% of positive cases (2 out of 9) highlights the additional value of CBNAAT in guiding treatment decisions. While the sensitivity remains limited, the high specificity of CBNAAT makes it invaluable for definitive diagnosis when positive.
The diagnostic performance of CSF-ADA in our study, with 42% overall positivity and 89% positivity among definite TBM cases, demonstrates its utility as a supportive biomarker. The high specificity of elevated ADA levels among confirmed cases suggests its value in strengthening diagnostic confidence, particularly when used in conjunction with other clinical and laboratory parameters.[14,15]
The high frequency of neuroimaging abnormalities (90%) in our cohort, with hydrocephalus being the predominant finding, confirms the critical role of brain imaging in pediatric TBM diagnosis. The presence of characteristic features such as basal exudates and meningeal enhancement provides strong supportive evidence for TBM diagnosis, particularly in resource-limited settings where advanced microbiological tests may not be readily available.[16,17]
Our findings support the necessity of a multimodal diagnostic approach in pediatric TBM. No single test demonstrated sufficient sensitivity to reliably exclude the diagnosis, while the combination of clinical assessment, CSF analysis including ADA estimation, CBNAAT, and neuroimaging provided optimal diagnostic accuracy. This approach aligns with current international guidelines emphasizing comprehensive evaluation rather than reliance on individual tests.[18]
The strong association between BMRC staging and mortality observed in our study reinforces the critical importance of early diagnosis and treatment initiation. The 100% mortality rate among Stage III patients contrasts sharply with the absence of deaths among Stage I patients, highlighting the narrow therapeutic window for optimal outcomes in pediatric TBM.
Limitations
The relatively small sample size and single-center design may limit generalizability. In addition, follow-up was restricted to the hospitalization period, potentially underestimating long-term neurological sequelae. The paucibacillary nature of pediatric TBM may have affected the sensitivity of microbiological tests, and the lack of mycobacterial culture results limited our ability to fully assess diagnostic accuracy.
Clinical implications
These findings have important implications for clinical practice in endemic regions. The high sensitivity of neuroimaging suggests that brain imaging should be prioritized in suspected pediatric TBM cases, particularly when advanced microbiological tests are unavailable. The supportive role of CSF-ADA, combined with its relatively low cost and rapid turnaround time, makes it a valuable addition to the diagnostic workup. The detection of rifampicin resistance through CBNAAT emphasizes the importance of molecular testing in guiding appropriate treatment regimens.
CONCLUSION
This study demonstrates that pediatric TBM diagnosis requires a comprehensive, multimodal approach. Neuroimaging showed the highest diagnostic sensitivity and should be considered essential in the evaluation of suspected cases. While CBNAAT provides definitive diagnosis with high specificity, its limited sensitivity necessitates careful clinical judgment to avoid missed diagnoses. CSF-ADA estimation offers valuable supportive evidence, particularly when elevated. The strong association between disease stage at presentation and mortality underscores the critical importance of early recognition and treatment initiation. Healthcare providers in endemic regions should maintain high clinical suspicion and utilize available diagnostic modalities collectively to optimize outcomes in pediatric TBM.
Conflicts of interest
There are no conflicts of interest.
Acknowledgments
The authors acknowledge Prof. IK Sharma, UPUMS, who initiated the process, the nursing staff. and laboratory personnel who contributed to patient care and data collection.
Funding Statement
Nil.
REFERENCES
- 1.Central TB Division. India TB Report 2020. National Tuberculosis Elimination Program Annual Report. New Delhi: Ministry of Health and Family Welfare, Government of India; 2020. [Google Scholar]
- 2.Chiang SS, Khan FA, Milstein MB, Tolman AW, Benedetti A, Starke JR, et al. Treatment outcomes of childhood tuberculous meningitis: A systematic review and meta-analysis. Lancet Infect Dis. 2014;14:947–57. doi: 10.1016/S1473-3099(14)70852-7. [DOI] [PubMed] [Google Scholar]
- 3.Wang MS, Zhao M, Liu XJ. Risk factors for poor outcome in childhood tuberculous meningitis. Sci Rep. 2021;11:8654. doi: 10.1038/s41598-021-87082-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Davis AG, Rohlwink UK, Proust A, Figaji AA, Wilkinson RJ. The pathogenesis of tuberculous meningitis. J Leukoc Biol. 2019;105:267–80. doi: 10.1002/JLB.MR0318-102R. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Manyelo CM, Solomons RS, Walzl G, Chegou NN. Tuberculous meningitis: Pathogenesis, immune responses, diagnostic challenges, and the potential of biomarker-based approaches. J Clin Microbiol. 2021;59:e01771–20. doi: 10.1128/JCM.01771-20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Palacios CF, Saleeb PG. Challenges in the diagnosis of tuberculous meningitis. J Clin Tuberc Other Mycobact Dis. 2020;20:100164. doi: 10.1016/j.jctube.2020.100164. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Daniel BD, Grace GA, Natrajan M. Tuberculous meningitis in children: Clinical management and outcome. Indian J Med Res. 2019;150:117–30. doi: 10.4103/ijmr.IJMR_786_17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Nhu NT, Heemskerk D, Thu do DA, Chau TT, Mai NT, Nghia HD, et al. Evaluation of GeneXpert MTB/RIF for diagnosis of tuberculous meningitis. J Clin Microbiol. 2014;52:226–33. doi: 10.1128/JCM.01834-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Marais S, Thwaites G, Schoeman JF, Török ME, Misra UK, Prasad K, et al. Tuberculous meningitis: A uniform case definition for use in clinical research. Lancet Infect Dis. 2010;10:803–12. doi: 10.1016/S1473-3099(10)70138-9. [DOI] [PubMed] [Google Scholar]
- 10.Wani RT. Socioeconomic status scales-modified Kuppuswamy and Udai Pareekh's scale updated for 2019. J Family Med Prim Care. 2019;8:1846–9. doi: 10.4103/jfmpc.jfmpc_288_19. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.van Toorn R, Springer P, Laubscher JA, Schoeman JF. Value of different staging systems for predicting neurological outcome in childhood tuberculous meningitis. Int J Tuberc Lung Dis. 2012;16:628–32. doi: 10.5588/ijtld.11.0648. [DOI] [PubMed] [Google Scholar]
- 12.Bhatia R, Dayal R, Jindal S, Agarwal D, Goyal A. GeneXpert for diagnosis of tubercular meningitis. Indian J Pediatr. 2016;83:1353–5. doi: 10.1007/s12098-016-2096-0. [DOI] [PubMed] [Google Scholar]
- 13.Sharma V, Rajeshwari K, Kumar D, Gupta G. Clinicoepidemiological profile and prognostic factors in neurotuberculosis in children. Ann Child Neurol. 2023;31:103–12. [Google Scholar]
- 14.Gupta BK, Bharat A, Debapriya B, Baruah H. Adenosine deaminase levels in CSF of tuberculous meningitis patients. J Clin Med Res. 2010;2:220–4. doi: 10.4021/jocmr429w. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Rana SV, Singhal RK, Singh K, Kumar L. Adenosine deaminase levels in cerebrospinal fluid as a diagnostic test for tuberculous meningitis in children. Indian J Clin Biochem. 2004;19:5–9. doi: 10.1007/BF02894249. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Andronikou S, Smith B, Hatherhill M, Douis H, Wilmshurst J. Definitive neuroradiological diagnostic features of tuberculous meningitis in children. Pediatr Radiol. 2004;34:876–85. doi: 10.1007/s00247-004-1237-1. [DOI] [PubMed] [Google Scholar]
- 17.Garg RK, Malhotra HS, Jain A. Neuroimaging in tuberculous meningitis. Neurol India. 2016;64:219–27. doi: 10.4103/0028-3886.177608. [DOI] [PubMed] [Google Scholar]
- 18.Solomons RS, Visser DH, Marais BJ, Schoeman JF, van Furth AM. Diagnostic accuracy of a uniform research case definition for TBM in children: A prospective study. Int J Tuberc Lung Dis. 2016;20:903–8. doi: 10.5588/ijtld.15.0509. [DOI] [PubMed] [Google Scholar]