Indian Pediatrics 1999; 36:455-460

Shally A wasthi and Soofia Moin

From the Department of Pediatrics, King George's Medical College, Lucknow (UP), India.

Reprint requests: Dr. Shally Awasthi, C-4, Officer's Colony, Niralanagar, Lucknow 226 020, UP, India.
E-mail: [email protected]

Manuscript received: May 22, 1998; Initial review completed: July 27, 1998;
Revision accepted: November 19, 1998.

Objective: To assess the protective effectiveness of BCG vaccination against tuberculous meningitis, while controlling for age, nutrition and socio-economic status, in children 1 month to 12 years of age. Design: Case-control study. Setting: Secondary care referral and teaching hospital. Methods: Cases were those conforming to the definition of tuberculous meningitis and controls were patients admitted after every third consecutive case included in the study from September 1995 till the end of August 1997 and who did not suffer from any central nervous system disorder. Results: Among the 192 cases and 70 controls, BCG scar was present in 57.8% and 75.7%, respectively. The crude odd's ratio (OR) for tuberculosis meningitis with a BCG scar was 0.44 (95% CI, .24-0.81; p=0.008), while the adjusted OR was 0.53 (95% C/, 0.26-1.06; p value = 0.07) after controlling for weight, age, sex and place of residence. Higher weight for age and urban residence were associated with a decreased risk of tuberculous meningitis in the logistic model. Conclusion: BCG vaccination offers protection against tuberculous meningitis. Since improvement in weight for age was associated with a decreased risk of disease, further studies are needed to evaluate the association, if any, between nutritional status and vaccine efficacy.

Key words: BCG vaccine, Protective efficacy, Tuberculous meningitis.

TUBERCULOUS meningitis is an important cause of hospital admission in the developing countries(1). Studies have found BCG vaccination to be protective against tuberculous meningitis and disseminated tuberculosis(2-9). However, in most of the studies confounding by age, nutritional or socio-economic status could not be effectively analyzed, often due to overmatching of cases and control with reference to these variables. The current work was undertaken to study the protective efficacy of BCG vaccination, if any, controlling for age, sex, weight and other immunization and socio-economic status.

Subjects and Methods

The study was conducted among children between 1 month to 12 years of age admitted in the Pediatrics Department of King George's Medical College, Lucknow, from September 1995 till the end of August 1997. Among patients presenting with the symptom complex of fever, headache or irritability, with or without vomiting, and either altered sensorium or first episode of seizures or both, tuberculous meningitis was defined as: (i) the presence of all three of the following cerebrospinal fluid abnormalities (if lumbar puncture could be done): leukocytosis > 10 and <1000 X 10⁶ cells/L, predominantly lymphocytes (>80%), proteins ≥60 mg/L with or without CSF glucose > 1.66 mmol/L (>30 mg/dL) and cerebro-spinal fluid culture sterile for pyogenic bacteria or (ii) if lumbar puncture was not done, treatment as tuberculous basal meningitis with at least 4 antitubercular drugs with no other concurrent antibiotics. Every patient admitted after third consecutive cases was a control, if he/she did not suffer from any disorder of the central nervous system and whose parents gave consent for the study. While the tentative diagnosis of the control at the time of admission was noted, the diagnosis at the time of discharge or death was considered as final.

Sample size calculations were based on reported crude odd's ratio of 0.36 (95% CI 0.19- 0.70) for tuberculous meningitis given a BCG scar in children less than 5 years of age from India (1). With a 0.3 ratio of controls to cases an odd's ratio of 0.5 with 70% prevalence of BCG scar in the controls and an alpha of 0.05 and power of,0.8, the estimated number of cases and controls was 138 and 42, respectively, using Epi Info 6.0 statistical software(10). A 0.3 ratio of cases to controls was taken to maximize on power(11) and at the same time have efficient data collection.

Data was collected on the age, sex, immunization and socio-economic status, including the family income, as well as the number of doses of vitamin A taken. Weight and height of the children and their parents was also taken. The Epi Info 6.0 statistical software was used to calculate the z-scores for weight for age (10). Based on the WHO classification, children were classified as underweight when the z score for weight for age was <-2.00(10). Presence of BCG scar on the insertion of the left deltoid muscle was noted. Results of the following investigations, if done, were noted in all the cases: computerized axial tomographic scan (CTS) of the head, chest roentgenogram (postero-anterior view), fundoscopy and BCG diagnostic testing. The BCG diagnostic test was interpreted as positive if a papule with an induration of >5 mm appeared within 24 to 48 hours(12). Chest roentgenograms were performed within the first 72 hours and CT scans within the first ten days of hospitalization. Both of these investigations, when done, were ordered as a part of routine patient care by the physician in charge of the case.

Univariate analysis was done to study the distribution of various variables as well as to compare them among the cases and the controls. Univariate association between BCG scar and meningitis was done by calculating the crude odd's ratio with 95% confidence interval (CI). Logistic regression was done using STATA software(13) to find the association of tuberculous meningitis with BCG scar, while controlling for those variable that were found to have univariate association between the case-control status and were clinically meaningful.

Results

There were 192 cases and 70 controls. The diagnostic category of controls was as follows: diarrheal diseases (n=15, 21.4%), respi- ratory diseases (n=16, 22.9%), hematological diseases (n=7, 10%), renal diseases (n=9, 12.9%), hepatobiliary diseases (n=6, 8.6%), malnutrition (n=4, 5.7%), genetic/immune (n=6, 8.6%) and others (n=7, 10%). There was 16 deaths in the cases and none in the controls.

The diagnosis of tuberculous meningitis was based on cerebrospinal fluid examination in 179 (93.2%) cases. In the 13 cases (6.8%) where CSF examination could not be done, CT scan was done and there were abnormalities like basal exudates (n=8), ventricular dilatation (n=7) and meningeal enhancement (n=10). Since the CT scan findings were supportive of the diagnosis of tuberculous meningitis, these cases have been included in all analysis.

CT scan (brain) was done in 141 (80.1 %) of the 176 survivors. CT scan was normal in (n=7) 5% cases. The abnormalities found were basal exudates (n=87) in 61.7%, ventricular dilatation in (n=101) 71.6%, obstructive hydrocephalus in (n=56) 39.7%, cerebral edema in 3, and tuberculoma and hypodense areas in 1 case each. Fundus examination was done in all 192 cases and showed a normal optic disc and retina in 115 (59.9%), pallor in 31 (16.1%), papilledema in 29 (15.1%), hyperemia in 8 (4.2%) and optic atrophy in 9 (4.7%). Chest roentgenograms were taken for '170 cases and was interpreted as normal in 62 (36.5%) and showed abnormalities suggestive of pulmonary tuberculosis in 108 (63.5%) cases. The positive X-ray findings were primary complex (which included either presence of parenchymal or lymph node enlargement or both) in 84.3% (n=9l), bronchopneumonia in 9.3% (n=IO), consolidation in 3.7% (n=4) and miliary tuberculosis in 2.8% (n=3). BCG diagnostic test was done in 170 cases and was positive in 51 (30%).

The anthropometric, immunization and socio-economic characteristic of cases and controls has been compared in Table I. Cases were younger in age, had higher proportion of underweight and lower proportion of immunized children when compared to the controls (Table I). Two-thirds cases and 47.1 % controls were from the rural areas. Among the children coming from rural and urban areas 59% and 58%, respectively, were immunized with 3 doses of DPT/OPV and 61.5% and 65%, respectively, had a BCG scar. There was no association between immunization with 3 doses of DPT/OPV and residence in a rural area (Chi square = 0.44, P value = 0.6). The crude odds' ratio for tuberculous meningitis in children with a BCG scar was 0.44 (95% CI 0.24-0.81; P = 0.008). The prevented proportion of cases of tuberculous meningitis with a BCG scar was 56.04% (95% CI, 18.96%- 76.13% ).

The results of the logistic regression to assess the association between tuberculosis and BCG scar is given in Table ll. The adjusted odd's ratio for tuberculous meningitis given a BCG scar was 0.53 (95% CI, 0.26 to 1.06; p = 0.07), while controlling for age, weight, sex and place of residence (Table II).

Discussion

This was a hospital based case-control study. We observed that the association between the presence of BCG scar and tuberculous meningitis is confounded by age and weight, and place of residence (rural) is an in- dependent risk factor for it. On controlling for these, the risk reduction of tuberculous meningitis by BCG vaccination could not be proved beyond the accepted limits of chance.

On univariate analysis, we found a 56.04% (95% CI: 18.9%-76.1 %) protective efficacy of BCG against tuberculous meningitis. This is less than the 80.2% efficacy reported from Sao Paulo, with hospital controls(2) and 100% from Argentina(3) but in accordance with unclear protective efficacy reported from India in the past(14). We also observed that controlling for confounders resulted in a reduction in the protective efficacy of the vaccine. In fact, studies from Sao Paulo and Argentina have not controlled for these in their published analyses.

We did not match the controls with the age and sex of the cases. Hence the cases of tuberculous meningitis were younger and weighed less than the cases (Table l). The difference in
 

TABLE I

Characteristics of Cases and Controls

Variable	Case	Control	p value
Number	192	70
Male	115 (59.9%)	44 (62.9%)	0.7
Age (mo)	46.1 :t 35.8	66.9 :t 45.8	0.003
Weight (kg)	10.1 :t 4.6	15.5 :t 8.4	<0.0001
Height (cm)	90.7:t 22.1	103.6:t 27.3	0.001
Underweight	169 (88%)	40 (57.1 %)	<0.0001
Immunization status
DPT/OPV (3 doses)	101 (52.6%)	51 (72.9%)	0.01
Vitamin A (1 +)	100(52.1%)	48 (68.6%)	<0.01
BCG scar	III (57.8%)	53 (75.7%)	0.01
SE Status
# Rural	128 (66.7%)	33(47.1%)	0.004
Income (Rs.)	1765 :t 694	1991 :t 799	0.1
# I room house	153 (79.7%)	54 (77.1 %)	0.6
Height (Father)	158.4 :t 3.8	159.3 :t 4	0.09
Height (Mother)	153.1 :t 3.6	153.3 :t 3.4	0.5

TABLE II

Results of Logistic Regression Assessing the Association of Tuber cillo us Meninigitis with BCG Scar.

Variable	Adjsted OR	95% CI	P value	Coding
BCG Scar	0.53	0.26 to 1.06	0.07	I =Yes,O=No
Weight (Kg)	0.80	0.72 to 0.99	<0.0001	Continuous
Age (mths)	0.98	0.97 to 0.99	0.06	Continuous
Sex	0.95	0.49 to 1.84	0.9	1 = Male, 0 = Female
Rural	2.36	1.26 to 4.44	0.007	1 = Rural, 0 = Urban
Constant	18.54	4.31 to 79.04	<0.0001

Number of observations = 262; chi2(5) = 60.87, p value = <0.00001; Log likelihood =-121.64.

their weights may be an underestimation due to fluid retention in a few controls We could, therefore, study the effect of age, weight and sex on the protective efficacy of BCG scar against tuberculous meningitis and have con- trolled for this imbalance in the analysis. We found that for children with BCG scar, increased weight for age for both the sexes was associated with a lower risk of tuberculous meningitis.

It is known that the vaccine efficacy is altered by the nutritional, genetic and immuno- logical make up of the subjects(15,16). The in fluency of ethnicity on the protective efficacy of BCG vaccination against tuberculous meningitis needs to be studied further. The association of weight for age with tuberculous meningitis is feasible since children with lower weight-for-age are at increased risk of mortality due to infectious diseases( 17). In addition, we have found that children living in rural areas have an increased risk of tuberculous meningitis. Since there was no difference in the proportions who had a BCG scar on the basis of the place of residence, the reasons for the reduced protective efficacy of BCG scar in rural children need to be deter- mined. Possible explanations for this phenomenon may be an increased exposure to adult cases of tuberculosis in rural as compared to urban areas, poor ventilation and undetermined environmental factors.

The strengths of this study are in its case- control design and random selection of controls that allowed study of confounding with age, weight and sex. The weaknesses are in taking BCG scar as a surrogate marker of immune response to vaccination, whereas this may not always be true. Since making the diagnosis of tuberculous meningitis is not always foolproof, there may have been misclassification with other types of meningitis which may have biased the results towards null. This drawback of disease classification, however, has been present in most studies on this issue.

Since an association between BCG vaccination and tuberculous meningitis was found with and without controlling for known con- founders, we conclude that BCG vaccination is preventive against tuberculous meningitis. Also, since increased weight for age was found to be associated with a decreased risk of disease, further studies are needed to find the association between nutritional status and vaccine efficacy. Factors associated with an increased risk of tuberculous meningitis in the rural children in India have to be identified.

Acknowledgments

This work was partly funded by a grant from the UP State Council for Science and Technology (CST). The computing facilities used have been provided by the International Clinical Epidemiology Network (INCLEN),

USA and Clinical Trials Service Unit (CTSU), Oxford, UK.