INT J TUBERC LUNG DIS 14(11):13821387

2010 The Union

REVIEW ARTICLE

Diagnostic value of adenosine deaminase in cerebrospinal

uid for tuberculous meningitis: a meta-analysis
H-B. Xu,* R-H. Jiang, L. Li,* W. Sha, H-P. Xiao
* Department of Clinical Pharmacy, Shanghai Tenth Peoples Hospital, Tongji University School of Medicine, Shanghai,
Department of Tuberculosis, Shanghai Pulmonary Hospital, Tongji University School of Medicine, Shanghai, China
SUMMARY
O B J E C T I V E : To determine the accuracy of adenosine
deaminase (ADA) measurements in the diagnosis of tuberculous meningitis (TBM).
D E S I G N : After a systematic review of English language
studies, the sensitivity, specificity and accuracy of ADA
concentrations in the diagnosis of cerebrospinal fluid
(CSF) were evaluated using random effects models. Summary receiver operating characteristic curves were used
to summarise overall test performance.
R E S U LT S : Ten studies met our inclusion criteria. The

sensitivity of ADA in the diagnosis of TBM was 0.79

(95%CI 0.750.83), specificity 0.91 (95%CI 0.890.93),
positive likelihood ratio 6.85 (95%CI 4.1111.41), negative likelihood ratio 0.29 (95%CI 0.190.44) and diagnostic odds ratio 26.93 (95%CI 12.7356.97).
C O N C L U S I O N : Our data suggest that ADA in the CSF
can be a sensitive and specific target and a critical criteria for the diagnosis of TBM.
K E Y W O R D S : adenosine deaminase; cerebrospinal fluid;
tuberculosis

TUBERCULOSIS (TB) is a major cause of morbidity

and mortality, the global incidence of which is increasing by 0.4% per annum.1 Tuberculous meningitis
(TBM) is one of the most harmful infectious diseases,
and accounts for 1% of all forms of TB. About 30%
of TBM patients die despite anti-tuberculosis chemotherapy, and early institution of treatment is essential
for satisfactory improvement.2,3 The diagnosis of TBM
is difficult due to the low sensitivity of acid-fast bacilli (AFB) microscopy using Ziehl-Neelsen staining
of the cerebrospinal fluid (CSF), and defining the
growth of Mycobacterium tuberculosis bacilli is time
consuming.4,5 A quick, non-invasive test to confirm
TBM would therefore be helpful.
Adenosine deaminase (ADA) is an enzyme widely
distributed in tissues and body fluid used in the diagnosis of TB in pleural, meningeal and pericardial fluids.6 The measurement of ADA was initiated by Giusti
in 1981 and applied extensively in clinical practice.7
Although its use has become popular in many countries, there is no consensus regarding its current usefulness in clinical practice.
We performed a systematic review of the literature
and a meta-analysis to determine the usefulness of
ADA measurement in the diagnosis of TBM.

or the approval of institutional review boards. We

searched the following electronic databases: Medline (19662009), Embase (19802009), Web of Science (19902009), BIOSIS (19942009) and LILACS
(19802009). All searches were up to date as of May
2009. The search terms used included tuberculosis,
Mycobacterium tuberculosis, meningitic/meningitis, cerebrospinal effusion/cerebrospinal fluid, adenosine deaminase/ADA, sensitivity and specificity
and accuracy. We contacted experts in the speciality
and searched reference lists from primary and review
articles. Articles published in English were reviewed
and analysed, while conference abstracts and letters
to journal editors were excluded due to their limited
data content.
Studies were included in the present meta-analysis
if they provided both the sensitivity (true-positive rate)
and the specificity (false-positive rate) of ADA in the
diagnosis of TBM, or if they provided ADA values in a
dot-plot form, allowing test results to be extracted for
individual study subjects. As very small studies may
be vulnerable to selection bias, only studies including
at least 10 TBM specimens were selected for inclusion
in the study. Two reviewers (HBX and RHJ) independently judged study eligibility while screening the citations. Disagreements were resolved by consensus.

MATERIALS AND METHODS

Search strategy and study selection
As the present study was a meta-analysis based on
published articles, we did not require patient consent

Data extraction and quality assessment

The final set of English language articles was assessed
independently by two reviewers (HBX and RHJ). The
reviewers were blinded to publication details, and

Correspondence to: L Li, Department of Clinical Pharmacy, Shanghai Tenth Peoples Hospital, Tongji University School of
Medicine, Shanghai, China. Tel: (+86) 216 630 2761. Fax: (+86) 216 630 7668. e-mail: jiangruihua119@yahoo.cn

Diagnostic value of ADA for TBM

Table 1

1383

Summary of studies included

Study, year, reference

20068

Chotmongkol,
Kashyap, 20079
Lopez-Cortes, 199510
Corral, 200411
Kashyap, 200612
Mishra, 199513
Choi, 200214
Rohani, 199515
Ribera, 198716
Coovadia, 198617

Patients
n
177
153
180
62
281
35
182
119
205
78

Test results

Quality score

Assay method

Cut-off
IU/l

TP

FP

FN

TN

Giusti method
Giusti method
Kinetic determination
Spectrophotometric method
Giusti method
Giusti method
Giusti method
Giusti method
Giusti method
Giusti method

15.5
5
10
8.5
11.39
5
10
9
9
10

12
55
10
8
96
24
21
14
32
11

11
12
11
6
10
3
6
13
2
18

4
11
10
6
21
3
15
0
0
4

42
75
149
42
154
5
140
92
171
45

STARD QUADAS
9
14
12
15
13
11
15
13
9
13

7
9
7
10
10
9
11
8
8
11

IU = international unit; TP = true-positive; FP = false-positive; FN = false-negative; TN = true-negative; STARD = standards for reporting diagnostic accuracy;
QUADAS = quality assessment for studies of diagnostic accuracy.

disagreements were resolved by consensus. Data retrieved from the reports included participant characteristics, test methods, sensitivity and specificity data,
cut-off values, publication year and methodological
quality. The numbers of true-positive, false-positive,
false-negative and true-negative results are shown in
Table 1.
We assessed the methodological quality of the
studies using guidelines published by the standards
for reporting diagnostic accuracy (STARD) initiative
(maximum score 25) and the quality assessment for
studies of diagnostic accuracy (QUADAS) tool (maximum score 14).18,19 In addition, for each study the
following study design characteristics were also retrieved: 1) cross-sectional design (vs. case-control design), 2) consecutive or random sampling of patients,
3) blinded (single or double) interpretation of determination and reference standard results, and 4) prospective data collection. If no data on the above criteria were reported in the primary studies, we requested
the information from the authors. If the authors did
not respond to our letters, the unknown items were
treated as no.
Statistical analysis
The kappa () coefficient was calculated as a measure
of agreement between two reviewers. We used standard methods recommended for meta-analyses of diagnostic test evaluations.20 The following measures
of test accuracy for each study, e.g., sensitivity, specificity, positive likelihood ratio (PLR), negative likelihood ratio (NLR) and diagnostic odds ratio (DOR)
were computed using several statistical software programmes. The analysis was based on a summary receiver operating characteristic (SROC) curve.20,21 The
sensitivity and specificity for the single test threshold
identified for each study were used to plot an SROC
curve.21 A random effects model was used to calculate the average sensitivity, specificity and other
measures across studies.22 The term heterogeneity
when used in relation to meta-analyses refers to the
degree of variability. Fishers exact tests were used to
detect significant heterogeneity. To assess the effects of

STARD and QUADAS scores on the diagnostic values

of ADA, we included them as covariates in univariate
meta-regression analysis (inverse variance weighted).
We also analysed the effects of other covariates on
DOR (i.e., cross-sectional design, consecutive or random sampling of patients, single or double interpretation of determination and reference standard results, and prospective data collection). The relative
DOR (RDOR) was calculated to analyse the change
in diagnostic precision in the study per unit increase
in the covariate.23 As publication bias is a risk in
meta-analyses of diagnostic studies, we tested the potential presence of this bias using the Egger test.24

RESULTS
After independent review, 16 publications dealing
with ADA concentrations for the diagnosis of TBM
were considered to be eligible for inclusion in the analysis.814,2530 Studies were excluded if ADA concentration was determined only in TBM patients,30 they
recruited <10 patients with confirmed TBM,26 or
they did not calculate the sensitivity or specificity.25,27,29
Some of the data might reappear in the publication.13,28
A final 10 studies817 involving 375 patients with TBM
and 989 non-TBM patients were available for analysis; the clinical characteristics of these studies, along
with their QUADAS scores, are outlined in Table 1.
Quality of reporting and study characteristics
The average internal agreement between the two reviewers for items in the quality checklist was 0.9. The
average sample size of the included studies was 136
(range 35281). In all 10 studies,817 the diagnoses of
all patients with TBM were made based on positive
smears or cultures of M. tuberculosis in the CSF and/
or on clinical symptoms. There were also increased
proteins, decreased glucose and good response to antituberculosis drugs in the CSF. Our initial data were
affected by the poor quality of reporting in the primary studies. To overcome this problem, we contacted
all authors of the 10 studies by airmail as well as
e-mail where e-mail addresses were available. Six

1384

The International Journal of Tuberculosis and Lung Disease

Table 2

Characteristics of included studies

Study, year, reference

TB patients/
non-TB subjects

Reference
standard

Cross-sectional
design

Consecutive
or random

Blinded
design

Prospective

Chotmongkol, 20068
Kashyap, 20079
Lopez-Cortes, 199510
Corral, 200411
Kashyap, 200612
Mishra, 199513
Choi, 200214
Rohani, 199515
Ribera, 198716
Coovadia, 198617

15/53
66/87
20/160
14/48
117/164
27/84
36/146
14/105
32/173
15/63

Bac /clin
Bac /clin
Bac /clin
Bac
Bac /clin
Bac /clin
Bac /clin
Bac /clin
Bac
Bac /clin

Yes
Yes
Yes
No
Yes
No
Yes
Unknown
Unknown
Unknown

Yes
Yes
Yes
Yes
Yes
Unknown
Yes
Yes
Yes
Unknown

No
No
Yes
No
No
No
No
Unknown
No
Unknown

Yes
Yes
Yes
No
Yes
Yes
No
Yes
Yes
Yes

TB = tuberculosis; bac = bacteriology; clin = clinical course.

authors responded who could provide additional

data.812,14 As shown in Table 2, 5 of the 10 studies
(50%) were cross-sectional in design. In eight studies
(80%), the samples were collected from consecutive
patients. Eight studies (80%) were prospective.
Diagnostic accuracy
Figure 1 shows the forest plot of sensitivity and specificity of the 10 ADA assays in the diagnosis of TBM.
Sensitivity and specificity ranged from 0.50 to 1.00

(mean 0.79, 95% confidence interval [CI] 0.750.83)

and from 0.63 to 0.99 (mean 0.91, 95%CI 0.89
0.93), respectively. PLR was 6.85 (95%CI 4.1111.41),
NLR 0.29 (95%CI 0.190.44) and DOR 26.92
(95%CI 12.7256.97). 2 values of sensitivity, specificity, PLR, NLR and DOR were respectively 44.80
(P < 0.001), 63.60 (P < 0.001), 51.66 (P < 0.001),
34.82 (P < 0.001) and 29.75 (P < 0.001), indicating
significant heterogeneity for sensitivity, specificity,
PLR, NLR and DOR between the studies.
Unlike a traditional receiver operating characteristic (ROC) plot that explores the effect of varying
thresholds on sensitivity and specificity in a single
study, each data point in the summary ROC (SROC)
plot represents a separate study. The SROC curve
presents a global summary of test performance and
shows the trade-off between sensitivity and specificity.
A graph of the SROC curve for the ADA determination showing true-positive vs. false-positive rates from
individual studies is shown in Figure 2. As a global
measure of test efficacy we used the Q-value, the intersection point of the SROC curve with a diagonal
line from the left upper corner to the right lower

Figure 1 Forest plot of estimate of sensitivity (A) and specicity (B) for adenosine deaminase assays in the diagnosis of TBM.
Circles = point estimates from each study; error bars = 95%CIs.
Pooled estimate of sensitivity for adenosine deaminase assays
was 0.79 (95%CI 0.750.83); pooled estimate of specicity
was 0.91 (95% CI 0.890.93). TBM = tuberculous meningitis;
CI = condence interval.

Figure 2 SROC curves for adenosine deaminase assays.

Circles = each study in the meta-analysis (the size of each
study is indicated by the size of the solid circle); the maximum
joint sensitivity and specicity (i.e., the Q value) was 0.85; the
area under the curve was 0.9154. SROC curves summarise
overall diagnostic accuracy. SROC = summary receiver operating characteristic.

Diagnostic value of ADA for TBM

1385

Table 3 Weighted meta-regression of the effects of methodological quality and study design
on diagnostic precision of cerebrospinal adenosine deaminase in 10 assays
Covariates
STARD 13
QUADAS 10
Cross-sectional design
Consecutive or random
Blinded
Prospective

Studies
n

Coefcient

RDOR

P value

6
4
5
8
2
8

0.168
0.137
0.427
2.243
0.411
0.231

0.84 (0.441.63)
0.87 (0.391.97)
1.53 (0.337.10)
9.42 (0.61144.42)
0.66 (0.133.39)
1.26 (0.236.87)

0.5628
0.7041
0.5316
0.0932
0.5702
0.7565

RDOR = relative diagnostic odds ratio; STARD = standards for reporting diagnostic accuracy; QUADAS = quality
assessment for studies of diagnostic accuracy.

corner of the ROC space, which corresponds to the

highest common value of sensitivity and specificity
for the test. This point does not indicate the only or
even the best combination of sensitivity and specificity for a particular clinical setting, but represents an
overall measure of the discriminatory power of a test.
Our data showed that the SROC curve was positioned
near the desirable upper left corner of the SROC curve,
and the Q value was 0.85, while the area under the
curve (AUC) was 0.92, indicating a high level of overall accuracy.

no statistical significance in RDOR values between

studies of higher (STARD score 13, QUADAS score
10) and lower quality. There was no significant difference between studies that were or were not blinded,
cross-sectional, consecutive/random and prospective.
The evaluation of publication bias showed that the
Egger test was not significant (P = 0.171). The funnel
plots for publication bias (Figure 3) also did not show
any asymmetry. These results indicate a lack of potential for publication bias.

Multiple regression analysis and publication bias

Using STARD guidelines, a quality score for every
study was compiled on the basis of the title, introduction, methods, results and discussion (Table 1).18
Quality was scored using QUADAS, where 1, 0 or
1 were given if all criteria were fulfilled, unclear or
not achieved, respectively (Table 1).19 These scores
were used in the meta-regression analysis to assess
the effect of study quality on the RDOR of ADA in
the diagnosis of TBM. As shown in Table 3, there was

DISCUSSION

Figure 3 Funnel graph for the assessment of potential publication bias in adenosine deaminase assays. The funnel graph
plots the log of the OR against the SE of the log of the OR (an
indicator of sample size). Circles = each study in the metaanalysis. The result of the Egger test for publication bias was not
signicant (P = 0.171). OR = odds ratio; SE = standard error.

Making a differential diagnosis between TBM and

non-TBM is a critical clinical problem. Conventional
methods, such as the direct examination of CSF, are
positive in only 5~20% of cases. The rate of positivity
is about 40% in culture, which takes about 6 weeks.4,5
Cerebrospinal levels of some biomarkers have been
proposed to be helpful in the diagnosis of TBM, including ADA. The present meta-analysis shows that
the mean values of the sensitivity and specificity of the
ADA assays were respectively 0.79 and 0.91, and that
the maximum joint sensitivity and specificity (Q value)
was 0.85, while the AUC was 0.92, indicating a high
level of overall accuracy. We also noted that three
studies16,29,32 showed relatively low sensitivity (<0.70)
and six studies15,28,29,31 demonstrated low specificity
(<0.90) for the detection of ADA in diagnosing TBM.
DOR is a single indicator of test accuracy that
combines the data from sensitivity and specificity
into a single number and the ratio of the odds of positive test results in the patient with disease or without
disease.31 DOR values range from 0 to infinity, with
higher values indicating better discriminatory test performance (i.e., higher accuracy). In the present metaanalysis, we found that the mean DOR was 26.92,
also indicating a high level of overall accuracy, and
also presented both PLR and NLR as diagnostic accuracy, as the SROC curve and the DOR are not easy
to interpret and use in clinical practice, and ratios are
considered to be more clinically meaningful.32,33 A
PLR of 6.85 suggests that TBM patients have an
approximately 7-fold higher chance of being ADA
assay-positive as compared to patients without TBM.
However, if the ADA assay result is negative, the

1386

The International Journal of Tuberculosis and Lung Disease

probability that the patient has TBM is approximately

29%, which is not low enough to rule out TBM. These
data suggest that a negative ADA assay result should
not be used alone as a justification to exclude or discontinue anti-tuberculosis treatment. The choice of
therapeutic strategy should be based on the results of
microscopic examination of a smear or culture for
M. tuberculosis, as well as other clinical data, such as
response to anti-tuberculosis treatment.
An exploration of the reasons for heterogeneity
rather than the computation of a single summary measure is an important goal of meta-analysis.34 In our
meta-analysis, both STARD and QUADAS scores were
used to assess the effect of study quality on RDOR. We
found significant heterogeneity for sensitivity, specificity, PLR, NLR, and DOR among the studies analysed, although the exact mechanism responsible for
the significance was inexplicable. There were no differences between studies with or without blinded,
cross-sectional, consecutive/random and prospective
design, which may contribute to the quality of the
test performance.
Publication bias may also have been introduced by
the inflation of diagnostic accuracy estimates, as studies with positive results were more likely to be accepted for publication. TBM is not always diagnosed
by microbiological examination. The heterogeneity
detected in the present analysis may have limitations.
TBM was diagnosed in some patients with TBM only
on clinical course. This can cause non-random misclassification, leading to biased results.
In conclusion, current evidence suggests a potential role of ADA assays in confirming a diagnosis of
TBM. There is a great need to develop CSF biomarkers specific for TBM. The results of ADA assays
should be interpreted with clinical findings and other
examinations.
Reference
1 Dye C. Global epidemiology of tuberculosis. Lancet 2006;
367: 938940.
2 Nunn P, Williams B, Floyd K, et al. Tuberculosis control in the
era of HIV. Nat Rev Immunol 2005; 5: 819826.
3 Quan C, Lu C Z, Qiao J, et al. Comparative evaluation of early
diagnosis of tuberculous meningitis by different assays. J Clin
Microbiol 2006; 44: 31603166.
4 Thwaites G E, Caws M, Chau T T, et al. Comparison of conventional bacteriology with nucleic acid amplification (Amplified Mycobacterium Direct Test) for diagnosis of tuberculous
meningitis before and after inception of anti-tuberculosis chemotherapy. J Clin Microbiol 2004; 42: 9961002.
5 Thwaites G E, Chau T T, Mai N T, et al. Tuberculous meningitis. J Neuro Neurosurg Psychiatry 2000; 68: 288299.
6 Segura R M, Pascual C, Ocana I, et al. Adenosine deaminase
in body fluids: a useful diagnostic tool in tuberculosis. Clin
Biochem 1989; 22: 141148.
7 Giusti G. Adenosine deaminase. In: Bergmeyer H U, ed. Methods of enzymatic analysis. New York, NY, USA: Academic Press
1974: pp 10921096.
8 Chotmongkol V, Teerajetgul Y, Yodwut C. Cerebrospinal fluid
adenosine deaminase activity for the diagnosis of tuberculous

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

meningitis in adults. Southeast Asian J Trop Med Public Health

2006; 37: 948952.
Kashyap R S, Ramteke S P, Deshpande P S, et al. Comparison
of an adenosine deaminase assay with ELISA for the diagnosis
of tuberculous meningitis infection. Med Sci Monit 2007; 13:
BR200204.
Lopez-Cortes L F, Cruz-Ruiz M, Gomez-Mateos J, et al. Adenosine deaminase activity in the CSF of patients with aseptic
meningitis: utility in the diagnosis of tuberculous meningitis or
neurobrucellosis. Clin Infect Dis 1995; 20: 525530.
Corral I, Quereda C, Navas E, et al. Adenosine deaminase
activity in cerebrospinal fluid of HIV-infected patients: limited value for diagnosis of tuberculous meningitis. Eur J Clin
Microbiol Infect Dis 2004; 23: 471476.
Kashyap R S, Kainthla R P, Mudaliar A V, et al. Cerebrospinal
fluid adenosine deaminase activity: a complimentary tool in
the early diagnosis of tuberculous meningitis. Cerebrospinal
Fluid Res 2006; 3: 5.
Mishra O P, Loiwal V, Ali Z, et al. Cerebrospinal fluid adenosine deaminase activity and C-reactive protein in tuberculous
and partially treated bacterial meningitis. Indian Pediatr 1995;
32: 886889.
Choi S H, Kim Y S, Bae I G, et al. The possible role of cerebrospinal fluid adenosine deaminase activity in the diagnosis of
tuberculous meningitis in adults. Clin Neurol Neurosurg 2002;
104: 1015.
Rohani M Y, Cheong Y M, Rani J M. The use of adenosine
deaminase activity as a biochemical marker for the diagnosis
of tuberculous meningitis. Malays J Pathol 1995; 7: 6771.
Ribera E, Martinez-Vazquez J M, Ocana I, et al. Activity of
adenosine deaminase in cerebrospinal fluid for the diagnosis
and follow-up of tuberculous meningitis in adults. J Infect Dis
1987; 155: 603607.
Coovadia Y M, Dawood A, Ellis M E, et al. Evaluation of
adenosine deaminase activity and antibody to Mycobacterium
tuberculosis antigen 5 in cerebrospinal fluid and the radioactive bromide partition test for the early diagnosis of tuberculosis meningitis. Arch Dis Child 1986; 61: 428435.
Bossuyt P M, Reitsma J B, Bruns D E. Standards for reporting
of diagnostic accuracy: towards complete and accurate reporting of studies of diagnostic accuracy; the STARD initiative.
BMJ 2003; 326: 4144.
Whiting P, Rutjes A W, Reitsma J B, et al. The development of
QUADAS: a tool for the quality assessment of studies of diagnostic accuracy included in systematic reviews. BMC Med Res
Methodol 2003; 3: 25.
Deville W L, Buntinx F, Bouter L M, et al. Conducting systematic reviews of diagnostic studies: didactic guidelines. BMC Med
Res Methodol 2002; 2: 9.
Moses L E, Shapiro D, Littenberg B. Combining independent
studies of a diagnostic test into a summary ROC curve: dataanalytic approaches and some additional considerations. Stat
Med 1993; 12: 12931316.
Irwig L, Tosteson A N, Gatsonis C, et al. Guidelines for metaanalyses evaluating diagnostic tests. Ann Intern Med 1994; 120:
667676.
Suzuki S, Moro-oka T, Choudhry N K. The conditional relative
odds ratio provided less biased results for comparing diagnostic test accuracy in meta-analyses. J Clin Epidemiol 2004; 57:
461469.
Egger M, Davey Smith G, Schneider M, et al. Bias in metaanalysis detected by a simple, graphical test. BMJ 1997; 315:
629634.
Jakka S, Veena S, Rao A R, Eisenhut M. Cerebrospinal fluid
adenosine deaminase levels and adverse neurological outcome in
pediatric tuberculous meningitis. Infection 2005; 33: 264266.
Pettersson T, Klockars M, Weber T H, et al. Diagnostic value of
cerebrospinal fluid adenosine deaminase determination. Scand
J Infect Dis 1991; 23: 97100.

Diagnostic value of ADA for TBM

27 John M A, Coovadia Y M. Shortfalls in the use of adenosine

deaminase in tuberculous meningitis. Trop Doct 2001; 31: 138
139.
28 Mishra O P, Loiwal V, Ali Z, et al. Cerebrospinal fluid adenosine deaminase activity for the diagnosis of tuberculous meningitis in children. J Trop Pediatr 1996; 42: 129132.
29 Donald P R, Malan C, Schoeman J F. Adenosine deaminase
activity as a diagnostic aid in tuberculous meningitis. J Infect
Dis 1987; 156: 10401042.
30 Mann M D, Macfarlane C M, Verburg C J, et al. The bromide
partition test and CSF adenosine deaminase activity in the
diagnosis of tuberculosis meningitis in children. S Afr Med J
1982; 62: 431433.

1387

31 Glas A S, Lijmer J G, Prins M H, et al. The diagnostic odds

ratio: a single indicator of test performance. J Clin Epidemiol
2003; 56: 11291135.
32 Deeks J J. Systematic reviews of evaluations of diagnostic and
screening tests. In: Egger M, Smith G D, Altman D G, eds. Systematic reviews in health care: meta-analysis in context. London, UK: BMJ Publishing Group, 2001: pp 248282.
33 Jaeschke R, Guyatt G, Lijmer J. Diagnostic tests. In: Guyatt G,
Rennie D, eds. Users guides to the medical literature: a manual
for evidence-based clinical practice. Chicago, IL, USA: AMA
Press, 2002: pp 121140.
34 Petitti D B. Approaches to heterogeneity in meta-analysis. Stat
Med 2001; 20: 36253633.

RSUM
O B J E C T I F : Dterminer la prcision des mesures de ladenosine deaminase (ADA) dans le diagnostic de la mningite tuberculeuse (TBM).
S C H M A : Aprs une revue systmatique des tudes en
langue anglaise, la sensibilit, la spcificit et la prcision
des concentrations de lADA pour le diagnostic dans le
liquide crbrospinal (CSF) ont fait lobjet dune valuation par les modles deffets alatoires. On a utilis les
courbes sommaires dopration du receveur pour rsumer les performances globales du test.
R S U LTAT S : Dix tudes ont rpondu nos critres din-

clusion. La sensibilit de lADA pour le diagnostic de

la TBM a t de 0,79 (IC95% 0,750,83), sa spcificit
de 0,91 (IC95% 0,890,93), le ratio de vraisemblance
positive de 6,85 (IC95% 4,1111,41), le ratio de vraisemblance ngative de 0,29 (IC95% 0,190,44) et le
coefficient (rapport) de diagnostic de 26,93 (IC95%
12,7356,97).
C O N C L U S I O N : Nos donnes suggrent que lADA du
CSF peut tre une cible sensible et spcifique et un critre
critique pour le diagnostic de la TBM.

RESUMEN
O B J E T I V O : Determinar la exactitud de las mediciones
de desaminasa de adenosina (ADA) en el diagnstico de
la meningitis tuberculosa (TBM).
M T O D O : Despus de un anlisis sistemtico de las
publicaciones cientficas en ingls, se evalu la sensibilidad, la especificidad y la exactitud de las concentraciones del ADA del liquido cefalo-rachidiano (CSF) en el
diagnstico de la TBM mediante modelos de efectos
aleatorios. La eficacia global de la prueba se valid con
curvas de resumen de rendimiento diagnstico.
R E S U LTA D O S : Diez estudios cumplieron con los crite-

rios de inclusin. La sensibilidad de la concentracin del

ADA en el diagnstico de TBM fue 0,79 (IC95% 0,75
0,83), la especificidad fue 0,91 (IC95% 0,890,93), el
cociente de verosimilitud positivo fue 6,85 (IC95%
4,1111,41), el cociente de verosimilitud negativo fue
0,29 (IC95% 0,190,44) y el cociente de posibilidades
diagnsticas fue 26,93 (IC95% 12,7356,97).
C O N C L U S I N : Los datos obtenidos indican que la concentracin del ADA del CSF puede ser una variable sensible y especfica y un criterio esencial en el diagnstico
de la TBM.