Tb Meningits

For personal use. Only reproduce with permission from The Lancet Publishing Group.

ARTICLES

THE LANCET Vol 360 October 26, 2002 www.thelancet.com 1287

Summary

Background The diagnosis of tuberculous meningitis isdifficult. Discrimination of cases from those of bacterialmeningitis by clinical features alone is often impossible, andcurrent laboratory methods remain inadequate or inaccessiblein developing countries. We aimed to create a simplediagnostic aid for tuberculous meningitis in adults on the basisof clinical and basic laboratory features.

Methods We compared the clinical and laboratory features onadmission of 251 adults at an infectious disease hospital inVietnam who satisfied diagnostic criteria for tuberculous(n=143) or bacterial (n=108) meningitis. Featuresindependently predictive of tuberculous meningitis weremodelled by multivariate logistic regression to create adiagnostic rule, and by a classification-tree method. Theperformance of both diagnostic aids was assessed byresubstitution and prospective test data methods.

Findings Five features were predictive of a diagnosis oftuberculous meningitis: age, length of history, white-blood-cellcount, total cerebrospinal fluid white-cell count, andcerebrospinal fluid neutrophil proportion. A diagnostic ruledeveloped from these features was 97% sensitive and 91%specific by resubstitution, and 86% sensitive and 79% specificwhen applied prospectively to a further 42 adults withtuberculous meningitis, and 33 with bacterial meningitis. Thecorresponding values for the classification tree were 99% and93% by resubstitution, and 88% and 70% with prospective testdata.

Interpretation This study suggests that simple clinical andlaboratory data can help in the diagnosis of adults withtuberculous meningitis. Although the usefulness of thediagnostic rule will vary depending on the prevalence oftuberculosis and HIV-1 infection, we suggest it be applied toadults with meningitis and a low cerebrospinal fluid glucose,particularly in settings with limited microbiological resources.

Lancet 2002; 360: 128792

IntroductionThe prompt diagnosis and treatment of tuberculousmeningitis saves lives.15 However, about 30% of patientswith tuberculous meningitis die despite anti-tuberculosischemotherapy.6 Delays in diagnosis and treatment areregarded as major contributing factors in the highmortality reported in many recent series.710 The diagnosisof tuberculous meningitis relies on isolation ofMycobacterium tuberculosis from the cerebrospinal fluid.Unfortunately, culture is too slow and insensitive to aidclinical decision-making. Direct Ziehl-Neelsen staining ofthe cerebrospinal fluid for acid-fast bacilli remains thecornerstone of rapid diagnosis, but this technique lackssensitivity.6 Newer diagnostic techniques, such as thosethat use PCR, have not been assessed completely,11 and arenot possible in most settings in the developing world wheremost cases of tuberculous meningitis are seen.12,13

Consequently, the decision to treat a patient fortuberculous meningitis is frequently empirical, irrespectiveof the diagnostic facilities available to clinicians. Until new,affordable, sensitive, and specific diagnostic assays becomeavailable, clinicians must depend on the discriminativeclinical and laboratory features of the disease for successfuldiagnosis and treatment.

The presenting clinical features of tuberculousmeningitis in adults are similar to those of many meningo-encephalitides, which result in frequent diagnosticconfusion. Delays in starting appropriate antibiotics fortuberculous meningitis or pyogenic meningitis worsensprognosis, yet physicians are often reluctant to startmonths of antituberculosis treatment without firmevidence. Diagnostic uncertainty arises commonly inpatients who present with a few days of headache, fever,and neck-stiffness; undefined treatment in the community;a low concentration of glucose in cerebrospinal fluid(


first scenario, patients are unnecessarily exposed to therisks of drug toxicity. In the second, many patients coulddie. Finally, a rule developed in children might not beapplicable to adults.

Nevertheless, there are strong arguments for developinga simple diagnostic algorithm for tuberculous meningitis inareas of high tuberculosis prevalence. First, tuberculousmeningitis tends to be commonest in areas with the leastclinical and laboratory resources. Second, a diagnostic ruledeveloped and used in high tuberculosis prevalence areas islikely to perform consistently. And third, early diagnosisand treatment improves outcome.15 We aimed to developsuch a diagnostic algorithm on the basis of clinical andlaboratory findings of adult patients in Vietnam.

MethodsPatientsThe adults described in this study were admitted to theClinical Research Unit at the Centre for TropicalDiseases, Ho Chi Minh City, Vietnam. The Centre forTropical Diseases is a 500-bed infectious diseases hospitalthat serves the local community, and acts as the tertiaryreferral centre for infectious disease for the whole ofsouthern Vietnam. The hospital treats about 30000inpatients and 52 000 outpatients each year. The ClinicalResearch Unit is a 15-bed ward for patients with central-nervous-system infection, severe malaria, sepsis, and renalfailure. Clinical data were recorded prospectively from alladults (>15 years old) admitted with suspected central-nervous-system infection to the ward between 1997 and2000.

ProceduresAfter providing informed consent, each patient underwentstandard history taking, examination, and baselineinvestigations (including lumbar puncture, chestradiography, and computed tomography of the head whenpossible). Patients judged to be at risk of HIV-1 infectionwere tested for antibodies to this virus. Each sample ofcerebrospinal fluid was centrifuged, and a portion ofdeposit was examined by microscopy with gram, Ziehl-Neelsen, and India ink stains. The remaining deposit wascultured on blood and chocolate agar, and Lwenstein-Jensen media.

After 48 h, a second lumbar puncture was done as partof routine hospital management in most patients to assessresponse to treatment. Those with culture-proven orsuspected bacterial meningitis received 10 days ofintravenous ceftriaxone (Roche, Basel, Switzerland; 2 gtwice per day). Those with suspected tuberculousmeningitis received four drugs (streptomycin, isoniazid,rifampicin, and pyrazinamide) for 3 months, followed bythree drugs (isoniazid, rifampicin, and pyrazinamide) for6 months. Clinical progress and response to treatmentwere recorded in individual study notes. The HospitalScientific and Ethical Committee approved the study.

ARTICLES

1288 THE LANCET Vol 360 October 26, 2002 www.thelancet.com

Tuberculous meningitis Bacterial meningitis p

Median (90% range) Number of patients Median (90% range) Number of patients

Age (years) 34 (1664) 143 41 (1770) 108 00076Male sex 91 (64%) 84 (78%) 00160Duration of illness (days) 12 (434) 142 3 (111) 107 00001Duration of fever (days) 10 (230) 139 3 (111) 106 00001Duration of headache (days) 10 (130) 136 3 (111) 106 00001Neck stiffness 120 (91%) 81 (84%) 00910Presence of coma before admission 49 (36%) 53 (50%) 00271Glasgow coma score (/15) 13 (715) 143 14 (615) 107 00807Pulse (/min) 90 (60120) 143 92 (72120) 108 00023Systolic BP (mm Hg) 120 (90150) 143 120 (90150) 108 07588Diastolic BP (mm Hg) 70 (5090) 143 70 (5090) 108 02267Hemiplegia 11 (8%) 4 (4%) 02820Cranial nerve palsies 32/143 (22%) 9/107 (8%) 00031Packed-cell volume (%) 40 (3048) 139 42 (3048) 104 01255WBC count (103/mL) 9800 (500016 200) 137 15250 (747031 500) 107 00001% neutrophils 80 (6089) 130 86 (7095) 107 00001Blood sodium (mmol/L) 135 (122143) 98 138 (125148) 102 00009CSF opening pressure (cm H20) 23 (944) 133 24 (7540) 60 09290Clear CSF appearance 81/141 (57%) 2/107 (2%) 00001CSF total WCC (103/mL) 300 (701090) 143 2583 (38220 000) 108 00001CSF % neutrophils 37 (184) 142 90 (6099) 108 00001CSF % lymphocytes 64 (1699) 142 10 (140) 108 00001CSF/blood glucose 028 (011052) 139 020 (003046) 101 00001CSF chloride (mmol/L) 85120 136 97121 67 00120CSF protein (g/dL) 191 (80490) 141 270 (89730) 107 00001CSF lactate (mmol/L) 54 (1598) 102 94 (21197) 92 00001

BP=blood pressure; WBC=white blood cell; CSF=cerebrospinal fluid; WCC=white-cell count.

Table 1: Univariate analysis comparing admission variables between patients with tuberculous and bacterial meningitis

Diagnostic criteria for tuberculous and bacterialmeningitis

Tuberculous meningitis Bacterial meningitisMycobacterium tuberculosis Pathogenic bacteria isolated isolated from cerebrospinal from cerebrospinal fluidfluid

Or Or

Clinical meningitis with Clinical meningitis with all of negative gram and India the following:ink stains, plus sterile Lymphocytes and neutrophilsbacterial and fungal cultures, in cerebrospinal fluidplus one or more of the Low concentration of glucose following: in cerebrospinal fluid (


Diagnostic criteriaDiagnostic criteria for tuberculous meningitis and bacterialmeningitis were applied to all patients admitted to theClinical Research Unit with meningitis. The criteria arepresented in the panel. There is no acceptable goldstandard diagnostic test for tuberculous meningitis.Culture and Ziehl-Neelsen staining of the cerebrospinalfluid are specific, but too insensitive to be used as the solecriteria for diagnosis. We therefore developed clinicalcriteria similar to those used by other researchers.14,15 Thecriteria for diagnosis of bacterial meningitis included thosefor partly treated bacterial meningitis (ie, sterilecerebrospinal fluid cultures), since this group of patients iscommonly misdiagnosed with tuberculous meningitis.Tuberculous meningitis was excluded in these patients ifthey fully recovered, without antituberculosis chemo-therapy, 3 months after admission. Untreated tuberculousmeningitis would almost always be fatal within this time.

Statistical analysisThe clinical and laboratory features of those fulfilling thediagnostic criteria for tuberculous meningitis and bacterialmeningitis were compared. These features comprised 26 clinical and laboratory variables at admission, includingcharacteristics of the cerebrospinal fluid sample.Additionally, data from a second cerebrospinal fluidsample were collected from all patients who received 48 hof intravenous ceftriaxone. Diagnostic uncertainty onadmission often leads clinicians in our hospital to use atrial of ceftriaxone, and reconsider the diagnosis after asecond cerebrospinal fluid examination. The purpose ofcollecting these additional data was to compare thecerebrospinal fluid variables in both groups, and todevelop a second diagnostic rule for patients managed by this approach. The relative change in eachcerebrospinal fluid variable was calculated by (a[CSF2]a[CSF 1])/a(CSF 1), where a is the specified variable,and CSF 1 and CSF 2 are the first and secondcerebrospinal fluid samples, respectively. All those givenimmediate antituberculosis chemotherapy were excludedfrom this analysis. The Kruskal-Wallis test was used to

compare continuous variables, and the 2 test (or Fishersexact test for small proportions) was used for categoricalvariables.

Three diagnostic aids were developed by means of twostatistical approaches: classification trees (classificationand regression tree, CART) and logistic regression. Theclassification trees were developed by consideration of allthe variables separately. The range of each variable wasdivided into two groups to obtain the best separationbetween patients with bacterial meningitis and those withtuberculous meningitis. The division corresponding to thebest separation was selected. The resulting subsets of caseswere then partitioned independently in turn. The processwas done recursively, until a stopping condition wassatisfied. Node deviance, which measures nodeheterogeneity, was set to 01 to stop the tree-growingprocess. Subsets smaller than 10 were not partitionedfurther.

Logistic regression was used to model the probability ofhaving tuberculous meningitis. A stepwise forward variableselection procedure was used to find independentpredictors of tuberculous meningitis with p-to-enter of005 or less, and p-to-remove of 0055 or more. Once thefinal model was constructed, the continuous variables inthe model were dichotomised by use of cutoffs from theunivariate classification trees, and the model was refitted.Rounded -coefficients from the model with dichotomisedvariables were used to define a diagnostic index for each ofthe clinical variables. A receiver operator characteristic(ROC) curve analysis was selected to find an optimumcutoff for the combined diagnostic indices. ROC analysiswas done on the original dataset, and the completed rule(diagnostic index with cutoff) was applied to the test data.

The three diagnostic aids were assessed byresubstitution and prospective test data methods. Thesensitivity and specificity were calculated and comparedwith the study diagnostic criteria. The resubstitutionmethod used the original data set. The test data methodused data recorded from a further 75 patients enrolled inthe same manner and subject to the same diagnosticcriteria. All analyses were done with STATA and Splus.

ARTICLES

THE LANCET Vol 360 October 26, 2002 www.thelancet.com 1289



CSF opening pressure (cm H2O) 27 (844) 43 14 (427) 58 00027Clear CSF appearance 21 (49%) 23 (23%) 00001CSF total WCC (103/mL) 470 (1732575) 51 760 (8012 000) 106 00484CSF % neutrophils 47 (487) 51 74 (897) 106 00001CSF % lymphocytes 53 (1396) 51 26 (390) 106 00001CSF/blood glucose 027 (01305) 47 043 (011069) 106 00001CSF chloride (mmol/L) 102 (81116) 48 115 (105126) 64 00001CSF protein (g/dL) 190 (98680) 49 110 (37580) 104 00001CSF lactate (mmol/L) 64 (2911) 36 48 (18104) 77 00105

CSF=cerebrospinal fluid; WCC=white-cell count.

Table 2: Second cerebrospinal fluid analysis taken after 4872 h of parenteral ceftriaxone



CSF opening pressure 8% (75 to 267) 36 41% (83 to 56) 51 00001CSF WCC 13% (68 to 964) 51 75% (97 to 918) 106 00001CSF % neutrophils 1% (45 to 500) 50 15% (88 to 23) 106 00001CSF % lymphocytes 5% (54 to 145) 50 144% (88 to 155) 106 00001CSF glucose/blood ratio 6% (66 to 92) 45 100% (61 to 833) 99 00001CSF lactate 0% (55 to 314) 27 48% (81 to 114) 70 00001CSF protein 2% (37 to 275) 49 54% (90 to 51) 103 00001CSF chloride 4% (14 to 8) 48 4% (8 to 21) 63 00001

CSF=cerebrospinal fluid; WCC=white-cell count.

Table 3: Change in cerebrospinal fluid variables after 4872 h of parenteral ceftriaxone


Role of the funding sourceThe study sponsor had no role in the study design; datacollection, analysis, or interpretation; or in the writing of thereport and the decision to submit the paper for publication.

Results357 adults were admitted to the Centre for TropicalDiseases with meningitis between 1997 and 2000. 251satisfied the diagnostic criteria for inclusion in this study:143 with tuberculous meningitis, and 108 with bacterialmeningitis. 66 of the 251 adults were tested for antibodiesto HIV-1: eight were positive (seven with tuberculousmeningitis, one with bacterial meningitis).

163 of 357 adults received antituberculosis chemotherapyfor suspected tuberculous meningitis: M tuberculosis wasisolated from the cerebrospinal fluid of 37 patients, and 106were defined as having clinical tuberculous meningitis.Supportive radiological evidence of tuberculous meningitiswas present in 85 of 106 adults with clinical tuberculousmeningitis. 20 patients treated for tuberculous meningitiswere excluded because they did not meet the study criteriafor the diagnosis of tuberculous meningitis, and all diedshortly after the start of antituberculosis chemotherapy. Nosignificant differences (p50% of that in blood), 15 died within 3 months, and 11 were lost to follow-up.

Admission variables are shown in table 1. Excluding the92 patients who received immediate antituberculosischemotherapy, the results from a second lumbar punctureafter 48 h of ceftriaxone were available in 157 of 251patients: 51 with tuberculous meningitis and 106 withbacterial meningitis (table 2). Two of 251 patients withculture-confirmed bacterial meningitis died before a secondsample could be taken. After 48 h of ceftriaxone, significantdifferences were apparent between the two groups (table 3).The cerebrospinal fluid variables changed little in those withtuberculous meningitis. However, in those with bacterialmeningitis, cerebrospinal fluid pressure, white-cell count,protein, and lactate fell, whereas the cerebrospinalfluid:blood glucose ratio rose.

Multivariate analyses were done to construct a diagnosticrule. Glasgow coma score, diastolic and systolic bloodpressure, presence of hemiplegia, and packed-cell volume atadmission were excluded due to the non-significant resultsby univariate analysis (table 1). Blood sodium,cerebrospinal fluid opening pressure, cerebrospinal fluidchloride, and cerebrospinal fluid lactate were excludedbecause of a large number of missing values. Only the totalduration of combined symptoms before admission wasincluded for analysis: length of fever, and headache, wereexcluded.

Stepwise logistic regression analysis found five variablesindependently associated with a diagnosis of tuberculousmeningitis on admission. The final logistic model with thesevariables is described in table 4. The formula for diagnosticindex was derived from the final model by dichotomisingthe variables and rounding the coefficients in the model.The index was adjusted to a positive scale for ease of use.The total diagnostic index (DI) was calculated for eachpatient according to the formula:

DI (age)+DI (blood white-cell count)+DI (history ofillness)+DI (cerebrospinal fluid white-cell count)+DI(cerebrospinal fluid % neutrophils)

The diagnostic index for each of the five variables is givenin table 5. The optimum cutoff for the total diagnostic index(by which to classify a patient as having tuberculousmeningitis) was found by use of an ROC curve (figure 1).The three points close to the top left-hand corner of thecurve correspond to cutoffs of 4, 3, or 2. The respectivediagnostic sensitivities and specificities for each point are97% and 91%, 96% and 91%, and 91% and 97%. Our

ARTICLES


Diagnostic index

Age (years)36 2


suggested diagnostic rule was therefore: if the patient has atotal diagnostic index score of 4 or less, he or she hastuberculous meningitis, and if the patient has a score ofmore than 4, he or she has bacterial meningitis. Noimprovement in the diagnostic rule was obtained by use ofdata from both admission and the second cerebrospinalfluid sample. Missing values, which reduced the sample sizeto 121 observations, might have accounted for this failure.

Figure 2 shows the diagnostic trees generated from theCART analysis. The performance of these diagnostic aids,and of that generated by logistic regression, are summarisedin table 6. Resubstitution of the original data set into theadmission diagnostic tree misclassified ten patients (twowith tuberculous meningitis and eight with bacterialmeningitis), giving 99% sensitivity and 93% specificity. Thesecond tree misclassified eight patients (three withtuberculous meningitis and five with bacterial meningitis),giving 93% sensitivity and 95% specificity.

The test data were recorded from a further 75 patientswho satisfied the study diagnostic criteria: 20 adults hadculture-confirmed tuberculous meningitis, 22 had clinicaltuberculous meningitis, 21 had culture-confirmed bacterialmeningitis, and 12 had clinical bacterial meningitis. Clinicaldata from this group were applied to each diagnostic aid.The sensitivities and specificities are presented in table 6.

DiscussionThe diagnosis of tuberculous meningitis in adults is difficultwhatever the resources available to the physician. This studyused diagnostic criteria derived from cerebrospinal fluidculture results and observed response to specific treatment.Such criteria suggest that even in the best settingscerebrospinal fluid culture is insensitive, and some cases oftuberculous meningitis or bacterial meningitis will never beproven microbiologically.

Because untreated tuberculous meningitis is always fatal,a clinical diagnostic aid or laboratory assay for tuberculousmeningitis must be sensitive. The potential toxic effects andduration of antituberculosis chemotherapy, and the limitedresources of many tuberculosis programmes, also mandatesdiagnostic specificity. At present, no rapid laboratorymethod for the diagnosis of tuberculous meningitis satisfiesthese requirements. A clinical diagnostic rule orclassification tree for tuberculous meningitis might improvethe sensitivity of current laboratory methods, and could beused in settings with limited microbiological diagnosticsupportie, where tuberculous meningitis is mostcommon.

We compared patients with tuberculous meningitis withthose having confirmed or probable bacterial meningitis fortwo reasons. First, both groups of patients requireimmediate decisions about chemotherapy, and second,those with bacterial meningitis, particularly partly treatedbacterial meningitis, are difficult to distinguish from thosewith tuberculous meningitis. Low cerebrospinal fluidglucose is usually present in both disorders, and forms animportant discriminating feature from other meningo-encephalitides. A low cerebrospinal fluid glucose (


laboratory features of the disease.19 However, HIV-1infection alters the differential diagnosis in meningiticadults: opportunistic infection with unusual pathogens mustbe considered, in particular Cryptococcus neoformans, whichcan present subacutely in a similar way to tuberculousmeningitis. The results of this study should therefore beapplied with caution in areas with a high prevalence of HIV-1 infection.

The univariate analysis of admission variables suggests aset of potentially discriminative clinical features. Patientswith tuberculous meningitis present with a longer history;they are more likely to have cranial nerve palsies; they willnot usually have a blood leucocytosis; and theircerebrospinal fluid will frequently be clear, with moderatenumbers of lymphocytes and neutrophils, in combinationwith an increased protein concentration and a low ratio ofcerebrospinal fluid:blood glucose. However, diagnosticuncertainty frequently persists despite the first cerebrospinalfluid analysis. In such cases, physicians in our hospital canassess response to treatment with a broad-spectrumantibiotic after 48 h. As expected, the cerebrospinal fluidvariables change little over these 48 h in those withtuberculous meningitis. The changes seen in patients withbacterial meningitis presumably reflect successfulantimicrobial effect. The dangers of delayedantituberculosis chemotherapy focus attention on adiagnostic aid that makes use of clinical features onadmission. Multivariate logistic regression analysis definedfive characteristics independently predictive of distinctionbetween tuberculous meningitis and bacterial meningitis:age, history of illness, white-cell count in blood, white-cellcount in cerebrospinal fluid, and percentage of neutrophilsin cerebrospinal fluid (table 4). We elected to use a cutoff of4 in the ROC analysis because it provides the greatestsensitivity (97%) with acceptable specificity (91%).Application of the test data revealed 86% sensitivity and79% specificity. These figures are similar to those of the bestavailable laboratory assays.6

Two diagnostic classification trees were developed: onefor admission, and a second to be used after a trial of broad-spectrum antibiotics (figure 2). When the test data wereapplied to the admission tree, sensitivity was similar to thatof the diagnostic rule (88%), but specificity was reduced(70%). The second tree performed less well, although thedata available to test this tree were small (14 patients with tuberculous meningitis, 33 with bacterialmeningitis). Nevertheless, a diagnostic tree incorporatingthe second set of cerebrospinal fluid variables is attractivegiven the striking differences documented between thosewith tuberculous meningitis and those with treated bacterialmeningitis, but such an aid requires further developmentand prospective assessment.

There are some important limitations to this study. First,the prevalence of both tuberculosis and HIV-1 infection willaffect the performance of the rule. We therefore suggest thatthis rule should not be used in areas of substantiallydifferent tuberculosis and HIV-1 prevalence to those ofsouthern Vietnam without prospective evaluation in thosesettings. Because this difference in prevalence will remain afundamental problem for all similar diagnostic methods,future research should also be directed at investigating theeffect of diagnostic algorithms on outcome.

This study suggests that simple clinical and laboratorydata can be used to help diagnose tuberculous meningitis inadults, and we propose an admission diagnostic rule with86% sensitivity and 79% specificity. We suggest that therule should be applied to adults in high tuberculosisprevalence settings with meningitis and a cerebrospinalfluid:blood glucose ratio of less than 50%. Since the

diagnosis and management of meningitis rest on clinical andcerebrospinal fluid assessment, efforts are necessary tosupport clinical and appropriate laboratory diagnosticservices in low-income countries. The search for acid-fastbacilli in the cerebrospinal fluid still represents the bestrapid laboratory diagnostic technique, but requires largevolumes of cerebrospinal fluid, and meticulous microscopyto achieve the best results.20 A clinical rule might focusattention on patients in whom such an approach iswarranted, and thereby optimise the use of often-limitedlaboratory resources.

ContributorsThe study was designed by G E Thwaites, C M Parry, and J J Farrar. Datawere collected by T T H Chau, N H Phu, L V Chuong, and D X Sinh, and analysed by G E Thwaites and K Stepniewska. G E Thwaites, C M Parry, J J Farrar, and N J White wrote the paper.

Conflict of interest statementNone declared.

AcknowledgmentsWe thank the doctors and nurses of the Clinical Research Unit, and the staffof the Microbiology Department, Centre for Tropical Diseases.The study was funded by the Wellcome Trust, UK.

References1 Fallon RJ, Kennedy DH. Treatment and prognosis in tuberculous

meningitis. J Infect 1981; 3: 3944.2 Humphries MJ, Teoh R, Lau J, Gabriel M. Factors of prognostic

significance in Chinese children with tuberculous meningitis. Tubercle Lung Dis 1990; 71: 16168.

3 Misra UK, Kalita J, Srivastava M, Mandal SK. Prognosis of tuberculousmeningitis: a multivariate analysis. J Neurol Sci 1996; 137: 5761.

4 Kalita J, Misra UK. Outcome of tuberculous meningitis at 6 and 12months: a multiple regression analysis. Int J Tuberc Lung Dis 1999; 3:26165.

5 Lincoln EM, Sordillo SVR, Davies PA. Tuberculous meningitis inchildren: a review of 167 untreated and 74 treated patients with specialreference to early diagnosis. J Paediatr 1960; 57: 80723.

6 Thwaites GE, Chau TT, Mai NT, et al. Tuberculous meningitis. J Neuro Neurosurg Psychiatry 2000; 68: 28999.

7 Girgis NI, Sultan Y, Farid Z, et al. Tuberculous meningitis, AbbassiaFever HospitalNaval Medical Research Unit. Am J Trop Med Hyg1998; 58: 2834.

8 Karstaedt AS, Valtchanova S, Barriere R, Crewe-Brown HH.Tuberculous meningitis in South African urban adults. Q J Med 1998;91: 74347.

9 Hosoglu S, Ayaz C, Geyik MF, Kokoglu OF, Ceviz A. Tuberculousmeningitis in adults: an eleven-year review. Int J Tuberc Lung Dis 1998;2: 55357.

10 Verdon R, Chevret S, Laissy JP, Wolff M. Tuberculous meningitis inadults: review of 48 cases. Clin Infect Dis 1996; 22: 98288.

11 Pfyffer GE. Nucleic acid amplification for Mycobacterial diagnosis. J Infect 1999; 39: 2126.

12 Wilson SM. Application of molecular methods to the study of diseasesprevalent in low income countries. Trans R Soc Trop Med 1998; 92:24144.

13 Foulds J, OBrien R. New tools for the diagnosis of tuberculosis: theperspective of developing countries. Int J Tuberc Lung Dis 1998; 2:77883.

14 Kumar R, Singh SN, Kohili N. A diagnostic rule for tuberculousmeningitis. Arch Dis Child 1999; 81: 22124.

15 Ahuja GK, Mohan KK, Prasad K, Behari M. Diagnostic criteria fortuberculous meningitis and their validation. Tubercle Lung Dis 1994; 75:14952.

16 Kay R, Cheng AF, Tse CY. Streptococcus suis infection in Hong Kong. Q J Med 1995; 88: 3947.

17 Nguyen TH, Hoang TL, Pham KC, et al. HIV monitoring in Vietnam:system, methodology, and results of sentinel surveillance. J Acquir Immune Defic Syndr 1999; 1: 33846.

18 Berenguer J, Moreno S, Laguna F, et al. Tuberculous meningitis inpatients with human immunodeficiency virus. N Engl J Med 1992; 326:66872.

19 Yechoor VK, Shandera WX, Rodriguez P, Cate TR. Tuberculousmeningitis among adults with and without HIV infection. Arch Intern Med 1996; 156: 171016.

20 Kennedy DH, Fallon RJ. Tuberculous meningitis. JAMA 1979; 241:26468.

ARTICLES


Documents

Tb Meningits