Bjerrum et al. BMC Infectious Diseases  (2015) 15:407 
DOI 10.1186/s12879-015-1151-1
RESEARCH ARTICLE Open Access
Diagnostic accuracy of the rapid urine
lipoarabinomannan test for pulmonary
tuberculosis among HIV-infected adults in
Ghana–findings from the DETECT HIV-TB study
Stephanie Bjerrum1,2*, Ernest Kenu3, Margaret Lartey3,4, Mercy Jemina Newman5, Kennedy Kwasi Addo6,
Aase Bengaard Andersen1 and Isik Somuncu Johansen1
Abstract
Background: Rapid diagnostic tests are urgently needed to mitigate HIV-associated tuberculosis (TB) mortality. We
evaluated diagnostic accuracy of the rapid urine lipoarabinomannan (LAM) test for pulmonary TB and assessed the
effect of a two-sample strategy.
Methods: HIV-infected adults eligible for antiretroviral therapy were prospectively enrolled from Korle-Bu Teaching
Hospital in Ghana and followed for minimum 6 months. We applied the LAM test on urine collected as a spot and
early morning sample. Diagnostic accuracy was analysed for a microbiological TB reference standard based on
sputum culture and Gene Xpert MTB/RIF results and for a composite reference standard including clinical follow-up
data. Performance of sputum smear microscopy was included for comparison.
Results: Of 469 patients investigated for TB, the LAM test correctly identified 24/55 (44 %) of microbiologically
confirmed TB cases. Sensitivity of the LAM test was positively associated with hospitalisation (67 %), Modified Early
Warning Score > 4 (57 %) and subsequent death (71 %). LAM test specificity was 95 % increasing to 98 % for the
composite reference standard. A two-sample LAM test strategy did not improve test performance. Using concentrated
sputum for Ziehl-Neelsen and fluorescence microscopy in combination yielded a sensitivity of 31/55 (56 %) that
increased to 35/55 (64 %) when the LAM test was added. Surprisingly, nontuberculous mycobacteria were cultured in
34/469 (7 %) and associated with a positive LAM test (p = 0.008).
Conclusions: LAM test sensitivity was highest in patients with poor prognosis and subsequent death and did not
increase with a two-sample strategy. A rigorous sputum microscopy strategy had superior sensitivity, but the simplicity
of the LAM test holds operational possibilities as a TB screening method among severely sick patients.
Keywords: Tuberculosis, HIV, Lipoarabinomannan, Urine, Diagnosis, Africa, Ghana
Background highest TB diagnostic yield achieved by microbiological
Tuberculosis (TB) is the leading cause of death among screening of all HIV-infected individuals regardless of
HIV-infected individuals initiating antiretroviral therapy presenting symptoms [5]. While culture based diagnos-
(ART) in sub-Saharan Africa [1]. Undiagnosed TB re- tics remain the gold standard for TB diagnosis, it is
mains highly prevalent in this population as identified in time-consuming and unavailable in many TB endemic
studies of intensified TB screening pre-ART [2–4]; the settings. The fully automated GeneXpert®MTB/RIF
(“Xpert”) assay can provide diagnosis within 2 hours, but
* Correspondence: steph@medicinsk.dk
1 wide implementation of this assay is impeded by highDepartment of Infectious Diseases, Odense University Hospital, Odense,
Denmark costs and requirements for electricity and maintenance
2Institute of Clinical Research, University of Southern Denmark, Odense, [6, 7]. Often, TB diagnosis relies on symptom-based
Denmark screening, chest x-ray and sputum smear microscopy for
Full list of author information is available at the end of the article
© 2015 Bjerrum et al. Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Bjerrum et al. BMC Infectious Diseases  (2015) 15:407 Page 2 of 10
Acid Fast Bacilli (AFB) although these tools have before enrolment or who were unable to produce
demonstrated limited performance [3, 8–10]. Recently, sputum or urine samples were excluded. Presence of
the DETERMINE TB-LAM Ag test (“LAM test” Alere, TB related symptoms was not used as inclusion or
Waltham, MA, USA) was commercialised for point-of- exclusion criteria.
care TB diagnosis. It is a lateral flow immunochromato- We used a standardized questionnaire to record demo-
graphic strip test developed to detect the mycobacterial graphic and clinical details including vital measures,
lipoarabinomannan (LAM) antigen released from meta- height and weight, TB specific signs and symptoms in-
bolically active or degrading mycobacteria [11]. It can be cluding the WHO symptom screen with presence of more
applied directly on urine and provides a result within than one of the following symptoms: cough, fever, weight
30 minutes. Previous LAM performance studies have loss or night sweats [25] and previous history of TB. We
reported great variation in test sensitivity and specificity calculated the Modified Early Warning Score (MEWS) for
[12–19]. A consensus paper on the LAM test highlights each participant as an indicator for illness severity. MEWS
issues with its performance and appropriate target group is a scoring system based on the level of deranged physio-
and calls for further evaluations of the LAM test in logical parameters including systolic blood pressure, pulse
diverse settings [20]. rate, respiratory rate, body temperature and level of con-
We undertook a prospective study in Ghana to evaluate sciousness. A cut-off score > 4 has been considered pre-
the diagnostic accuracy of the LAM test to diagnose pul- dictive of poor outcomes [26]. Blood CD4 cell count was
monary TB among HIV-infected individuals eligible for obtained for participants at enrolment. Full blood count
ART. We assessed a two-sample strategy, similar to what and x-ray interpretation was obtained if available through
is conventionally applied for sputum microscopy, and ex- routine laboratory investigations.
plored the trade-off between the increase in sensitivity At enrolment, participants were requested to provide a
against reduction in specificity when using the LAM test spot urine specimen for LAM testing and one respira-
as an add-on test to sputum smear microscopy. tory specimen for sputum smear microscopy, mycobac-
terial culture and Xpert assay. Participants were further
Methods asked to deliver an early morning urine and sputum
Design sample within 7 days after enrolment. Medical records
This is a cross sectional diagnostic accuracy study with a were reviewed from time of enrolment to a minimum of
longitudinal follow-up of minimum 6 months. 6 months post-enrolment. If medical records were un-
available, an interview with the participant or relatives of
Study setting and population the participant was conducted by phone. The following
As part of the DETECT HIV-TB study, participants were were recorded: vital status, lost to follow-up, transfer
recruited prospectively between January 2013 and March out, start on antituberculous treatment.
2014 from the out- and inpatient departments, Fevers
Unit, Korle-Bu Teaching Hospital; Ghana’s largest public Ethics
hospital situated in the capital city Accra. The Fevers Informed consent was obtained in writing from each par-
Unit provides ART services including adherence coun- ticipant prior to enrolment in the study. The study proto-
selling, medical care and laboratory services for HIV- col was approved by the Ethical and Protocol Review
infected individuals. The overall HIV-epidemic in Ghana Committee, University of Ghana Medical School (MS-Et./
is moderate with a prevalence of 1.4 % in the general M.4–P 3.3/2012-13) and evaluated by the Developing
population [21]. In 2013, the TB Control Programme in Country Committee of the Danish National Committee
Ghana conducted a national TB prevalence survey that on Health Research Ethics (No. 1302133/1206169). Urine
found an overall adult TB prevalence of 356/100,000 LAM test results were not used for treatment decision-
(personal communication with the programme manager making, but sputum microscopy, culture and Xpert results
for the National TB Control Programme, Ghana, unpub- were communicated to the responsible clinicians.
lished data). This is much higher than the WHO estimated We followed the Standards for the Reporting of Diag-
prevalence of 71/100,000 and questions the estimated case nostic accuracy studies (STARD) criteria [27]. Additional
detection rates of 88 %, which is likely to be an overesti- file 1 displays the STARD checklist (For more on STARD
mation [22]. see http://www.stard-statement.org/).
HIV-infected adults were consecutively enrolled into
the study if ≥18 years and eligible for lifelong ART, i.e. Urine sample analysis
severe and advanced HIV clinical disease (WHO clinical Urine samples were transported to the Department of
stage 3 or 4 disease), a blood CD4 cell count ≤350 cells/ Medical Microbiology within Korle-Bu Teaching Hos-
μl or pregnant [23, 24]. Patients on antituberculous pital premises where study staff applied the LAM test on
treatment of more than 2 days within the last 3 months fresh urine in accordance with the manufacturer’s
Bjerrum et al. BMC Infectious Diseases  (2015) 15:407 Page 3 of 10
instructions. A 60 μL of unprocessed urine was applied immunocompromised HIV-infected individuals we used
to the sample pad at the bottom of the test strip. The the following composite TB case definition to categorize
test result was read between 25–35 minutes later and participants:
graded by comparing the test strip with a reference card.
We used the original 2012 reference scale card that con- “Confirmed TB” if M. tuberculosis culture positive or
sisted of 5 colour intensity grades. The test band was Xpert positive in any of the sputum samples.
graded as zero if no visual band appeared and graded 1 “Possible TB” if no positive culture or Xpert results
through 5 for a visualized band of equal intensity as for TB, but one of the following; sputum smear
those on the reference card. If a faint band was observed microscopy positive i.e. smears graded as scanty, 1+,
with intensity lower than the grade 1 cut-point it was re- 2+, and 3+; a clinical-radiological picture highly
corded as “faint”. Additional file 2 provides a figure of suggestive of TB and started on antituberculous
the reference card used and its interpretation. Each test treatment within two months; a clinical diagnosis of
was graded by two individual readers blinded to their active TB by a non-study clinician and started on
counterpart’s observations and to the patient identity treatment within two months; death within two
through the use of anonymous study ID’s. Reference months of enrolment reported to be due to TB per
standard results were not known at the time of LAM medical record.
testing. “Non–TB” if not meeting criteria for “Confirmed” or
“Possible” TB. Participants with growth of
Sputum sample analysis nontuberculous mycobacteria (NTM) and no positive
Sputum samples were analysed at the Reference TB cultures or Xpert results for M. tuberculosis were
Laboratory at Noguchi Memorial Institute for Medical assigned to this group.
Research or the TB laboratory at Korle-Bu Teaching
Hospital. Samples were processed according to standard- Statistical analysis
ized protocols for mycobacterial microscopy and culture Descriptive analysis was used to characterize the study
by trained laboratory technicians [28, 29]. Sputum sam- population and reported with interquartile range (IQR)
ples were decontaminated with N-acetyl-L-cysteine and and standard deviations (SD) as appropriate. Kappa
sodium hydroxide. Smears of centrifuged sputum sedi- statistics were used to determine inter-reader agreement
ment were examined microscopically and graded for between LAM test results and agreement between test
AFB using both Ziehl-Neelsen (ZN) staining method results reported with the standard error (SE). Accuracy
and fluorescence microscopy of auramine O stained measures (sensitivity, specificity, positive predictive values
smears. After re-suspension in phosphate buffer the spu- (PPV), negative predictive values (NPV) and likelihood ra-
tum sediment was cultured for mycobacteria using both tio (LR)) were calculated with 95 % Confidence Interval
solid Lowenstein-Jensen medium and the BACTEC (CI). In our primary analysis, we used a microbiological
mycobacteria growth indicator tube (MGIT) 960 system reference standard comparing “Confirmed TB” vs. partici-
(BD Diagnostics, Sparks, MD, USA) and incubated for pants with no positive cultures or positive Xpert results.
up to 8 and 6 weeks respectively. Positive cultures were In the secondary analysis we used a composite reference
re-assessed for AFBs using ZN smear microscopy and an standard for TB and combined “Confirmed TB” and “Pos-
anti-MPB64 antibody assay was used on AFB positive sible TB” for calculation of sensitivities versus “Non TB”
cultures to confirm presence of M. tuberculosis. The cases to calculate specificity. Figure 1 outlines the analysis
GeneXpert MTB-RIF assay was performed on either of groups. For subgroup analysis we stratified participants
fresh sputum sample or on sputum sediment according by: enrolment site (hospitalised patients vs. outpatients);
to the manufacturer’s specifications (Cepheid, CA, USA) CD4 cell count (CD4 < 100 cells/mm3 vs. CD4 ≥ 100
after it became available at the Chest Clinic TB labora- cells/mm3); MEWS (MEWS > 4 vs. MEWS ≤ 4); and
tory to also confirm TB. vital status at 2 months (dead vs. alive). Sensitivity and
specificity was compared across strata using chi-square
Diagnostic classification for analysis test or Fisher Exact test as appropriate. We determined
We defined a positive LAM test result as a test band with diagnostic accuracy for LAM test in combinations with
intensity equal to or greater than the grade 2 cut-point sputum smear microscopy and for the two-sample LAM
[14]. The first reading of the first sample was considered test strategy. When assessing performance of a combin-
the study result and used for all data analysis except for ation of tests, the result was considered positive if any of
analysis of inter-reader agreement and accuracy of a the tests were positive. The result was considered negative
two-sample strategy. if both tests were negative. McNemar’s test was used to
In the absence of a single suitable reference stand- compare two different test sensitivities and specificities.
ard for TB diagnosis in a population of severely The cumulative probabilities of death were estimated by
Bjerrum et al. BMC Infectious Diseases  (2015) 15:407 Page 4 of 10
Fig. 1 Flowchart of study participants and analysis. *Of 29 participants excluded; three (3) were on antituberculous treatment; twenty one
(21) had no sputum samples; and five (5) had no urine sample. The remaining 469 participants were eligible for analysis with at least 1
sputum and 1 urine sample available
means of the Kaplain-Meier method, compared according were defined as “Possible TB” (prevalence 9.6 %; 95 % CI
to LAM test results with the log-rank test. Statistical sig- 7.2-12.6). Sputum microscopy for AFB was positive in
nificance was defined as a two-sided p-value less than 0.05 31/55 (56 %) of “Confirmed TB” cases. Among the 45
and all analysis were conducted using STATA™ version “Possible TB” cases, 6/45 (13 %) were categorised as
13.1 software. “Possible TB” based on a positive sputum microscopy
for AFB, 25/45 (56 %) based on start-up of antitubercu-
Results lous treatment within 2 months, and 14/45 (31 %) died
Participants within 2 months of follow-up with TB stated as the
In the study period, 571 HIV-infected adults were cause of death.
screened and 469 were eligible according to our inclu-
sion criteria (Fig. 1). In total, 399 (85 %) were enrolled Urine LAM test performance
from the outpatient clinic and 70 (15 %) were hospita- All LAM tests provided a valid result with visible control
lised patients (Table 1). Participants produced a mean of bar. The distribution of LAM test results across band in-
1.8 (SD 0.36) urine samples with two urine samples tensity grades was: no band, 194 (41 %); faint band, 168
obtained from 396 (84 %) participants. Two sputum (36 %); grade 1, 62 (13 %); grade 2, 10 (2 %); grade 3, 10
samples were collected from 371 (79 %) participants (2 %); grade 4, 16 (4 %) and grade 5, 9 (2 %). Inter-rater
with a mean of 1.8 (SD 0.41) per participant. agreement between the readers as to presence versus ab-
TB was confirmed bacteriologically in 55 cases (preva- sence of a LAM test band with intensity of grade 2 cut-
lence 11.7 %; 95 % CI 9.1-15.0) and additional 45 cases point was 99.1 % (kappa 0.94; SE 0.05). Agreement as to
Bjerrum et al. BMC Infectious Diseases  (2015) 15:407 Page 5 of 10
Table 1 Characteristics of study population
All Confirmed TB Possible TB Non-TB
n = 469 n = 55 n = 45 n = 369
Enrolment site
Out patients 399 (85 %) 40 (73 %) 20 (44 %) 339 (92 %)
In patients 70 (15 %) 15 (27 %) 25 (56 %) 30 (8 %)
Median Age in years (IQR) 38 (31–45) 37 (29–44) 37 (33–41) 38 (32–45)
Female 301 (64 %) 24 (44 %) 29 (64 %) 241 (65 %)
Median CD4 (IQR)* 127 (35–256) 92 (41–176) 46 (12–176) 142 (41–282)
CD4 < 100 195 (43 %) 29 (55 %) 28 (65 %) 138 (39 %)
CD4 > =100 259 (57 %) 24 (45 %) 15 (35 %) 220 (61 %)
Mean MEWS (SD) 3 (1.95) 5 (2.20) 4 (2.37) 3 (1.59)
MEWS > 4 101 (22 %) 35 (63 %) 19 (42 %) 47 (13 %)
MEWS < = 4 368 (78 %) 20 (36 %) 26 (58 %) 322 (87 %)
Median HGB g/dl** (IQR) 9.7 (8.1-11.2) 8.2 (6.2-9.6) 8.0 (6–10) 10 (9–12)
Median BMI (IQR) 20 (18–22) 18 (16–19) 19 (16–21) 21 (18–23)
TB suspects, WHO symptom screen 426 (91 %) 53 (96 %) 43 (96 %) 330 (90 %)
Current Cough 222 (48) 42 (76 %) 28 (62 %) 152 (41 %)
Fever 238 (52 %) 45 (82 %) 28 (62 %) 165 (45 %)
Weight loss 390 (83 %) 51 (93 %) 39 (87 %) 300 (81 %)
Night Sweats 161 (34 %) 26 (49 %) 16 (36 %) 119 (32 %)
Previous history of TB 28 (6 %) 4 (7 %) 2 (4 %) 22 (6 %)
Overall mortality*** 81 (17 %) 18 (33 %) 23 (51 %) 40 (11 %)
*CD4 count missing for 15 participants
**HGB missing for 39 participants
***80 participants were lost to follow-up at 6 months and 11 transferred out
IQR = Interquartile range, SD = Standard Deviation
test band grade was 69.2 % (kappa 0.54; SE 0.03) mainly LAM test. In the subgroup analysis (Table 3), LAM test
due to inter-rater variability between recording of grade sensitivity increased significantly among hospitalised pa-
1 cut-point intensity vs. faint. Additional file 3a + b tients (p = 0.035), participants with MEWS > 4 (p = 0.008)
details the inter-rater variability. and in participants who died within 2 months of follow-up
(p = 0.013). The increase in sensitivity among participants
Analysis 1–Microbiological reference standard with low CD4 did not reach statistically significance. The
For grade 2 cut-point positivity threshold the urine LAM inverse pattern was seen for test specificity being signifi-
test had an overall diagnostic sensitivity of 44 % (95 % CI; cantly lower among strata with higher degree of diseases
30–58) and specificity of 95 % (95 % CI; 92–97) using the severity or death as outcome.
microbiological reference standard (Table 2). Additional
file 4 provides a table with the diagnostic accuracy of the Analysis 2–Composite reference standard for TB
LAM test by all grade cut-point. Additional file 5 displays When using the composite reference standard for
the Receiver Operator Characteristic (ROC) curve for the TB, the overall LAM test specificity increased to
Table 2 Overall performance of LAM test for diagnosis of tuberculosis
Total Sensitivity Specificity LR (+) LR (−) PPV NPV
N % (95 % CI) N % (95 % CI) (95 % CI) (95 % CI) % (95 % CI) % (95 % CI)
LAM test overall
Microbiological reference standard* 469 24/55 44 (30–58) 393/414 95 (92–97) 8.6 (5.1-14.4) 0.6 (0.5-0.8) 53 (38–68) 93 (90–95)
Composite reference standard** 469 36/100 36 (27–46) 360/369 98 (95–99) 14.8 (7.4-29.6) 0.7 (0.6-0.8) 80 (65–90) 85 (81–88)
*Primary analysis using a microbiological reference standard:” Confirmed TB” cases versus culture and Xpert negative cases
**Secondary analysis using a composite reference standard: “Confirmed TB” and “Possible TB” combined for calculations of sensitivity versus “Non TB”
LR (+) = Positive Likelihood Ratio, LR (−) = Negative Likelihood Ratio, PPV = Positive Predictive Value, NPV = Negative Predictive Value
Bjerrum et al. BMC Infectious Diseases  (2015) 15:407 Page 6 of 10
Table 3 The overall sensitivity and specificity of the LAM test and stratified by subpopulation
LAM test overall Total Microbiological reference standard Composite reference standard
Sensitivity Specificity Sensitivity Specificity
N % (95 % CI) p-value N % (95 % CI) p-value N % (95 % CI) p-value N % (95 % CI) p-value
469 24/55 44 (30–58) 393/414 95 (92–97) 36/100 36 (27–46) 360/369 98 (95–99)
By Department 469
Inpatient 70 10/15 67 (38–88) 47/55 86 (73–94) 17/40 43 (27–59) 29/30 97 (83–100)
Out patient 399 14/40 35 (21–52) 0.035 346/359 96 (94–98) 0.001 19/60 32 (20–45) 0.269 331/339 98 (95–99) 0.74
By CD4 strata* 454
CD4 < 100 195 14/29 48 (29–68) 148/166 89 (83–93) 24/57 42 (29–56) 130/138 94 (89–98)
CD4 > = 100 259 8/24 33 (16–55) 0.272 233/235 99 (97–100) <0.001 9/39 23 (11–39) 0.054 219/220 100 (98–100) 0.002
By MEWS 469
MEWS > 4 101 20/35 57 (39–74) 54/66 82 (70–90) 29/54 54 (40–67) 44/47 94 (83–99)
MEWS < = 4 368 4/20 20 (6–44) 0.008 339/348 97 (95–99) <0.001 7/46 15 (6–29) <0.001 316/322 98 (96–99) 0.061
By Vital status 397
at 60 days**
Dead 52 10/14 71 (42–92) 33/38 87 (72–96) 14/29 48 (29–68) 22/23 96 (78–100)
Alive 345 11/34 32 (17–51) 0.013 302/311 97 (95–99) <0.001 14/58 24 (14–37) 0.023 281/287 98 (96–99) 0.483
*15 participants with missing values for CD4 were excluded from analysis
**63 participants were lost to follow-up at 2 months and 7 transferred out
MEWS =Modified Early Warning Score
98 % (95 % CI; 95–99), but sensitivity reduced to Sputum smear microscopy performance alone and in
36 % (95 % CI; 27–46) (Table 2). Sensitivity increased for combination with LAM test using the microbiological
the strata with poor outcome, while specificity remained reference standard
above 94 % for all strata (Table 3). ZN microscopy detected 29/55 (53 %) of the confirmed
TB cases and ZN + fluorescence microscopy in combin-
LAM test performance with a two-sample strategy using ation detected 31/55 (56 %) (Table 4). Concordance be-
the microbiological reference standard tween LAM test and microscopy was moderate both for
Of all participants, 396/469 (84 %) was able to deliver ZN microscopy alone (91.4 %; kappa 0.45; SE 0.05) and
two urine samples for LAM test. Among these, the ZN + fluorescence microscopy (91.4 % kappa 0.47; SE
LAM test correctly identified 14/40 (35 %) of the 0.05). Similar to the LAM test accuracy, the sensitivity
“Confirmed TB” cases in the first sample tested and of sputum microscopy increased in subgroups with poor
additionally 2 “Confirmed TB” cases 16/40 (40 %) when prognosis (data not shown). Combining LAM test to ZN
including positive LAM test results from the second and fluorescence microscopy maximised sensitivity to
sample. The additional two cases both had a grade 1 35/55 (64 %) non-significantly higher than for the ZN
band intensity at the first LAM test. The increase in and fluorescence microscopy alone. There were four
sensitivity was not statistically significant (35 % vs. 40 %; additional TB cases detected by adding the LAM test; all
p = 0.5) and specificity dropped significantly from 96 % had CD4 < 100 cells/mm3 and two of them subsequently
to 92 % (p < 0.001) when the second LAM test was ap- died. The LAM false positive cases were 26 of whom 17
plied. The LAM test was positive in 40/458 (9 %) of the (65 %) were defined as “possible TB”.
spot samples and 45/407 (11 %) of the early morning
samples. Agreement between spot and morning LAM LAM test cross reactivity with nontuberculous
test results was 95.5 % (kappa 0.72; SE 0.05). Additional mycobacteria
file 6a + b detail the inter-variability between spot and NTM was cultured in sputum from 34 (prevalence
morning samples. Sensitivity of urine LAM test performed 7.2 %; 95 % CI 5.2-10.0) participants and additionally
on a spot sample was 40 % (21/52; 95 % CI: 27–55) two participants were co-infected with NTM and TB. The
slightly lower than when applied on an early morning LAM test was positive in 5/34 (15 %, 95 % CI; 5–31) par-
urine samples with a sensitivity of 42 % (18/43; 95 % ticipants with a positive LR+ of 1.6 (95 % CI; 0.7-3.8).
CI: 27–58). Specificity was 95 % (387/406; 95 % CI: Compared to those without any mycobacteria, a positive
93–97) and 93 % (337/364; 95 % CI: 89–95) for spot NTM culture was associated with a positive LAM test
and morning sample respectively. (4 % vs. 15 %; p = 0.008).
Bjerrum et al. BMC Infectious Diseases  (2015) 15:407 Page 7 of 10
Table 4 Sensitivity and specificity of the LAM test, sputum smear microscopy and combinations of the tests
Total Sensitivity Specificity LR (+) LR (−) PPV NPV
N % (95 % CI) N % (95 % CI) % (95 % CI) % (95 % CI) % (95 % CI) % (95 % CI)
LAM test 469 24/55 44 (30–58) 393/414 95 (92–97) 8.6 (5.1-14.4) 0.6 (0.5-0.8) 53 (38–68) 93 (90–95)
ZN microscopy 467 29/55 53 (39–66) 406/412a 99 (97–100) 36.2 (15.7-83.3) 0.5 (0.4-0.6) 83 (66–93) 94 (91–96)
ZN + Fluorescence microscopy 467 31/55 56 (42–70) 406/412a 99 (97–100) 38.7 (16.9-88.5) 0.4 (0.3-0.6) 84 (68–94) 94 (92–96)
LAM + ZN microscopy 469 34/55 62 (48–75)a 388/414 94 (90–96)a,b 9.8 (6.4-15.1) 0.4 (0.3-06) 57 (43–69) 95 (92–97)
LAM + ZN + Fluorescence Microscopy 469 35/55 64 (50–76)a 388/414 94 (91–96)a,c 10.1 (6.6-15.5) 0.4 (0.3-0.6) 57 (44–70) 95 (93–97)
For combinations of test a positive result was recorded if any of the tests were positive
ap < 0.05 for comparison with LAM test alone
bp < 0.05 for comparison with ZN microscopy alone
cp < 0.05 for comparison with ZN + Fluorescence microscopy
LR (+) = Positive Likelihood Ratio, LR (−) = Negative Likelihood Ratio, PPV = Positive Predictive Value, NPV = Negative Predictive Value, ZN = Ziehl Neelsen
Mortality reported as the most common AIDS defining event and
During the study 81 (17 %) of the participants died after cause of death [30]. We found that rapid urine LAM test
a median of 25 days from enrolment. The 6-month mor- could identify 44 % of confirmed TB cases with a specifi-
tality was 73 (16 %), 80 (17 %) participants were lost to city of 95 % and that a two-sample strategy for the LAM
follow-up (LTFU) and 11 (2 %) transferred out. Con- test did not improve sensitivity. The LAM test was easily
firmed TB patients had a 2-month mortality of 14/55 performed and had a high inter-reader reliability. Sensi-
(25 %) and at 6-month 18 (33 %) cases had died, 7 cases tivity increased to 67 % among hospitalised patients and
(13 %) were LTFU and 1 (4 %) transferred out. LAM was associated with death and high clinical illness score
positive participants had a significantly higher probability of (MEWS > 4); further, we found a tendency of increased
death compared to LAM negative in the overall population sensitivity among patients with lower CD4 cell counts.
(p < 0.001) (Fig. 2a) and among “confirmed TB” (p = 0.002) Our findings affirm that sensitivity of the LAM test is
(Fig. 2b). LTFU among LAM positive participants was not highest for the sickest patients and those holding the
significantly higher than among LAM negative (6/45; 18 % greatest risk of dying. This is consistent with findings by
vs. 74/424; 13 %, p = 0.5). In a sensitivity analysis assuming Lawn et al. who found increased LAM test sensitivity
that all participants LTFU had died, the probability of among patients with CD4 cell count < 100 cells/mm3,
death remained significantly higher for LAM-positive CRP > 200 mg/L, severe anaemia (<8.0 g/dL), advanced
than LAM-negative participants overall (p < 0.001) and symptoms and subsequent death [31, 32]. To our know-
among confirmed TB cases (p = 0.044). ledge, four other studies have evaluated urine LAM test
among HIV-infected individuals irrespective of the pres-
Discussion ence of TB symptoms [12–15] and four other studies
We evaluated the urine LAM strip test for TB diagnosis among HIV-infected TB suspects [16–19]. The lowest over-
among ART eligible adults in Ghana where TB is all LAM test sensitivity of 25 % as reported by Balcha et al.
Fig. 2 Cumulative probability of survival. a Shown for the full study population (n = 469) by LAM test status. b Shown for confirmed TB cases (n = 55)
by LAM test status
Bjerrum et al. BMC Infectious Diseases  (2015) 15:407 Page 8 of 10
in a study that included HIV-infected individuals with test results has been sparsely addressed. In a cohort of
CD4 < 350 regardless of TB symptoms [12]. The highest cystic fibrosis patients we previously reported that 2/23
sensitivity of 50 % was reported for grade 2 cut-point in a (8.7 %) of NTM infected patients were LAM-positive at
study among hospitalised TB suspects [18]. LAM test sen- a grade 2 cut-point increasing to 9/23 (39.1 %) for grade
sitivity was consistently the highest among hospitalised 1 cut-point [37]. In the context of HIV infected individ-
populations and strata with lower CD4 cell count, suggest- uals, 10 NTM culture positive cases have previously
ing that critically ill patients are the appropriate target been described to be LAM positive [15, 38–40]. We
group for LAM testing. No study has previously reported found a positive LAM test for 5/34 (15 %) of NTM cul-
a correlation with MEWS although this could be a simple ture positives. The prevalence of NTM culture positive
and objective alternative to identify the target group for was high in our study and associated with a positive
LAM testing. MEWS was originally developed to detect LAM test. While this raises concern that NTM among
critically ill patients at risk of catastrophic deterioration in HIV infected individuals affect LAM test specificity, its
high-income countries [26], but has since gained a wider importance needs to be evaluated in studies applying
application in medical wards and intensive care units to appropriate case definitions for NTM disease.
predict hospital admission and mortality also in resource A two-sample test strategy including an early morning
limited settings [33, 34]. A great advantage of MEWS is sample has not been evaluated to date for the LAM test.
that it is based on simple measures that are routinely col- We found that an additional sample did not increase
lected as part of most clinical consultations. Our finding overall performance of the LAM test and agreement be-
of a significantly higher risk of death for LAM-positive tween results for a spot and morning sample was high.
compared with LAM-negative TB patients further empha- A recent study explored a two-test strategy performed
sises that the LAM test identify TB patients with the on the same sample and did not show any added diag-
greatest clinical need. nostic value of the second test [13].
Specificity of the LAM test is paramount for utility of Previous studies found that the LAM test sensitivity
LAM as a screening test. We found considerably lower was superior to sputum smear microscopy and found
specificity among subgroups with signs of critical illness incremental diagnostic value of combining the two tests
or subsequent death than the target of 95 % for new [12, 16, 19]. We found higher sputum smear micros-
point-of-care tests for TB diagnosis [35]. Using the com- copy sensitivity that contrasts other studies comparing
posite reference standard, however, increased overall microscopy with the LAM test. This could be due to our
specificity to 98 % and to 94 % in all subgroups with use of sputum concentration and fluorescence microscopy
poor outcome. This probably reflects the well-known both known to increase diagnostic sensitivity [41, 42], and
limitations of sputum culture and Xpert to detect TB collection of two samples for the majority of participants.
among severely sick HIV-infected individuals. We sought A similar and rigorous sputum sampling and microscopy
to optimize our microbiological reference standard by per- methodology is not always achievable in a routine clinical
forming two sputum cultures on both solid and liquid setting and the LAM test could be preferred for reasons
media, and by adding Xpert results. Despite these efforts, other than a higher diagnostic accuracy alone.
the microbiological reference standard leads to underesti- The LAM test is simple to perform at the bedside and
mation of specificity. This is similar to findings by Peter et collection of urine is simple and poses minimal biohazard;
al. where LAM test specificity increased from 78 % to all attractive features in settings with overburdened TB la-
94 % when the composite reference standard was used boratories. Urine based diagnosis further holds a potential
instead of a microbiological reference standard [18]. An- to improve TB diagnosis in children with one study of
other key factor for test specificity is the positivity thresh- LAM test showing reasonable sensitivity in HIV-positive
old for the LAM test as shown in previous studies [13, 16, TB infected children [43]. A number of other biomarkers
18]. There is consensus of using grade 2 cut-point as posi- have been detected at increased levels in the urine from
tivity threshold and in 2014 the LAM test manufacturers patients with active TB [44, 45]. LAM remains among the
changed the reference scale card omitting the band corre- most promising so far, but is limited by its modest sensi-
sponding to grade 1. However, we report a large number tivity and use among HIV-infected individuals only.
of tests with band-intensity fainter than the original grade The strengths of this study are prospective data collec-
2 cut-point as did another prospective LAM study [16]. It tion, LAM testing on fresh urine and a minimum of
is important to acknowledge that such visible bands, al- 6 months’ follow-up. We chose to enrol HIV-infected indi-
though fainter than the positivity threshold, in a clinical viduals initiating ART regardless of clinical presentation as
setting could prompt clinicians to interpret the LAM test this target group has been identified as a particularly vul-
as positive and lead to over treatment. nerable group for prevalent and undiagnosed TB. We had
LAM is also a component of the NTM cell wall [36], several limitations with regard to our reference standard
but the possibility of NTM causing false positive LAM that did not include Xpert results for all participants or
Bjerrum et al. BMC Infectious Diseases  (2015) 15:407 Page 9 of 10
investigation of extrapulmonary samples that could have Acknowledgements
increased specificity further, especially among those with The authors wish to thank all who participated in the study as well as staff at
the Fevers Unit and designated laboratory for their contributions to realise it.
more advanced immunodeficiency. Moreover, we did not Financial support was received from the University of Southern Denmark
have capacity to perform sputum induction to improved and Odense University Hospital as well as a number of private funds in
sputum sample quality. Despite active follow-up through Denmark. ISJ was granted an associate research professor position by the
research council of Region Southern Denmark. Alere provided the Urine LAM
personal calls to participants and their relatives the LTFU tests at a reduced rate. None of the sponsors or Alere were involved in the
was high, but comparable to several other African HIV- design, implementation, analysis or write up of this study.
cohort studies [46]. Mortality is a frequent cause of LTFU
Author details
[47, 48] and mortality rates in our study could have been 1Department of Infectious Diseases, Odense University Hospital, Odense,
higher had we performed a more thorough follow-up with Denmark. 2Institute of Clinical Research, University of Southern Denmark,
e.g. house-visit. However, sensitivity analysis assuming that Odense, Denmark.
3Fevers Unit, Korle-Bu Teaching Hospital, Accra, Ghana.
4Department of Medicine, University of Ghana Medical and Dental School,
all participants LTFU had died did not change the associ- Accra, Ghana. 5Department of Medical Microbiology, School of Biomedical
ation between LAM-positivity and mortality. and Allied Sciences, College of Health Sciences, University of Ghana, Accra,
Ghana. 6Department of Bacteriology, Noguchi Memorial Institute for Medical
Research, University of Ghana, Accra, Ghana.
Conclusions
Despite modest sensitivity, we found that the LAM test Received: 27 May 2015 Accepted: 28 September 2015
has a potential to improve TB case detection when ap-
plied as a screening test among the sickest patients and
those at the greatest risk of dying, especially in settings References
without easy access to high quality microscopy, culture 1. Gupta A, Nadkarni G, Yang WT, Chandrasekhar A, Gupte N, Bisson GP, et al.
or more advanced diagnostic technologies. A two-sample Early mortality in adults initiating antiretroviral therapy (ART) in low- and
middle-income countries (LMIC): a systematic review and meta-analysis.
strategy did not improve test performance. PLoS One. 2011;6(12):e28691.
2. Lawn SD, Kranzer K, Edwards DJ, McNally M, Bekker LG, Wood R. Tuberculosis
Additional files during the first year of antiretroviral therapy in a South African cohort using an
intensive pretreatment screening strategy. AIDS. 2010;24(9):1323–8.
3. Bassett IV, Wang B, Chetty S, Giddy J, Losina E, Mazibuko M, et al. Intensive
Additional file 1: The STARD checklist. (PDF 302 kb) tuberculosis screening for HIV-infected patients starting antiretroviral
Additional file 2: The original 2012 LAM test reference scale card therapy in Durban, South Africa. Clin Infect Dis. 2010;51(7):823–9.
used in the study and interpretation. The integers under the reference 4. Wood R, Middelkoop K, Myer L, Grant AD, Whitelaw A, Lawn SD, et al.
card correspond to grade 1 cut-point (1), grade 2 cut-point (2), grade 3 Undiagnosed tuberculosis in a community with high HIV prevalence:
cut-point (3), grade 4 cut-point (4) and grade 5 cut-point (5). *Interpretation implications for tuberculosis control. Am J Respir Crit Care Med.
of positivity threshold is indicated by “Considered positive” (PDF 4536 kb) 2007;175(1):87–93.
5. Kranzer K, Houben RM, Glynn JR, Bekker LG, Wood R, Lawn SD. Yield of
Additional file 3: Inter-rater variability for reader 1 and 2. (a.) Shown
HIV-associated tuberculosis during intensified case finding in resource-limited
for presence versus absence of a test band with intensity grade 2 cut-point
settings: a systematic review and meta-analysis. Lancet Infect Dis.
or higher (b.) Shown by grade cut-points (PDF 57 kb)
2010;10(2):93–102.
Additional file 4: Accuracy of the LAM test by all grade cut-points. 6. Boehme CC, Nabeta P, Hillemann D, Nicol MP, Shenai S, Krapp F, et al.
(PDF 33 kb) Rapid molecular detection of tuberculosis and rifampin resistance. N Engl J
Additional file 5: Receiver operator characteristic curve for the LAM Med. 2010;363(11):1005–15.
test. The integers adjacent to points on the graph correspond to faint 7. Lawn SD, Brooks SV, Kranzer K, Nicol MP, Whitelaw A, Vogt M, et al.
(F), grade 1, grade 2, grade 3, grade 4 and grade 5 cut-point (PDF 42 kb) Screening for HIV-associated tuberculosis and rifampicin resistance before
Additional file 6: Inter-variability for spot and morning samples. antiretroviral therapy using the Xpert MTB/RIF assay: a prospective study.
(a.) Shown for presence versus absence of a test band with intensity grade PLoS Med. 2011;8(7):e1001067.
2 cut-point or higher (b.) Shown by test band grade cut-points (PDF 56 kb) 8. Getahun H, Harrington M, O’Brien R, Nunn P. Diagnosis of smear-negative
pulmonary tuberculosis in people with HIV infection or AIDS in resource-
constrained settings: informing urgent policy changes. Lancet.
Abbreviations 2007;369(9578):2042–9.
AFB: Acid fast bacilli; ART: Antiretroviral therapy; FM: Fluorescence microscopy; 9. Reid MJ, Shah NS. Approaches to tuberculosis screening and diagnosis in
IQR: Interquartile range; LAM: Lipoarabinomannan; LR: Likelihood ratio; people with HIV in resource-limited settings. Lancet Infect Dis.
MEWS: Modifies early warning score; MTB/RIF: Mycobacterium tuberculosis/ 2009;9(3):173–84.
resistance to rifampicin; NPV: Negative Predictive Value; NTM: Nontuberculous 10. Lawn SD, Wood R. Tuberculosis in antiretroviral treatment services in
mycobacteria; PPV: Positive Predictive Value; TB: Tuberculosis; WHO: World Health resource-limited settings: addressing the challenges of screening and
Organization; ZN: Ziehl-Neelsen; SD: Standard Deviation; SE: Standard Error. diagnosis. J Infect Dis. 2011;204 Suppl 4:S1159–67.
11. Lawn SD. Point-of-care detection of lipoarabinomannan (LAM) in urine for
Competing interests diagnosis of HIV-associated tuberculosis: a state of the art review. BMC
The authors declare that they have no competing interests. Infect Dis. 2012;12(1):103.
12. Balcha TT, Winqvist N, Sturegard E, Skogmar S, Reepalu A, Jemal ZH, et al.
Authors’ contributions Detection of lipoarabinomannan in urine for identification of active
SB, ISJ, MJN, KKA and ABA conceptualised and initiated the study in tuberculosis among HIV-positive adults in Ethiopian health centres. Trop
common. SB, ML and EK collected the clinical data. SB ran the LAM test with Med Int Health. 2014;19(6):734–42.
assistance from MJN. The microbiological analysis and laboratory work was 13. Drain PK, Losina E, Coleman SM, Giddy J, Ross D, Katz JN, et al. Value of
supervised by SB, MJN, and KKA. Data analysis and statistical analysis was urine lipoarabinomannan grade and second test for optimizing clinic-based
performed by SB and ISJ. SB wrote the paper with inputs from all authors screening for HIV-associated pulmonary tuberculosis. J Acquir Immune Defic
who also approved the final version of the manuscript before submission. Syndr. 2015;68(3):274–80.
Bjerrum et al. BMC Infectious Diseases  (2015) 15:407 Page 10 of 10
14. Drain PK, Losina E, Coleman SM, Giddy J, Ross D, Katz JN, et al. Diagnostic 36. Briken V, Porcelli SA, Besra GS, Kremer L. Mycobacterial lipoarabinomannan
accuracy of a point-of-care urine test for tuberculosis screening among and related lipoglycans: from biogenesis to modulation of the immune
newly-diagnosed HIV-infected adults: a prospective, clinic-based study. response. Mol Microbiol. 2004;53(2):391–403.
BMC Infect Dis. 2014;14:110. 37. Qvist T, Johansen IS, Pressler T, Hoiby N, Andersen AB, Katzenstein TL, et al.
15. Lawn SD, Kerkhoff AD, Vogt M, Wood R. Diagnostic accuracy of a low-cost, Urine lipoarabinomannan point-of-care testing in patients affected by
urine antigen, point-of-care screening assay for HIV-associated pulmonary pulmonary nontuberculous mycobacteria - experiences from the Danish
tuberculosis before antiretroviral therapy: a descriptive study. Lancet Infect Cystic Fibrosis cohort study. BMC Infect Dis. 2014;14:655.
Dis. 2012;12(3):201–9. 38. Gounder CR, Kufa T, Wada NI, Mngomezulu V, Charalambous S, Hanifa Y, et
16. Nakiyingi L, Moodley VM, Manabe YC, Nicol MP, Holshouser M, Armstrong al. Diagnostic accuracy of a urine lipoarabinomannan enzyme-linked
DT, et al. Diagnostic accuracy of a rapid urine lipoarabinomannan test for immunosorbent assay for screening ambulatory HIV-infected persons for
tuberculosis in HIV-infected adults. J Acquir Immune Defic Syndr. 2014. tuberculosis. J Acquir Immune Defic Syndr. 2011;58(2):219–23.
17. Peter JG, Theron G, Dheda K. Can point-of-care urine LAM strip testing for 39. Lawn SD, Edwards DJ, Kranzer K, Vogt M, Bekker LG, Wood R. Urine
tuberculosis add value to clinical decision making in hospitalised HIV-infected lipoarabinomannan assay for tuberculosis screening before antiretroviral
persons? PLoS One. 2013;8(2):e54875. therapy diagnostic yield and association with immune reconstitution
18. Peter JG, Theron G, Van Zyl-Smit R, Haripersad A, Mottay L, Kraus S, et al. disease. AIDS. 2009;23(14):1875–80.
Diagnostic accuracy of a urine LAM strip-test for TB detection in HIV-infected 40. Reither K, Saathoff E, Jung J, Minja LT, Kroidl I, Saad E, et al. Low sensitivity
hospitalised patients. Eur Respir J. 2012;40(5):1211–20. of a urine LAM-ELISA in the diagnosis of pulmonary tuberculosis. BMC Infect
19. Shah M, Ssengooba W, Armstrong D, Nakiyingi L, Holshouser M, Ellner JJ, et al. Dis. 2009;9:141.
Comparative performance of urinary lipoarabinomannan assays and Xpert 41. Steingart KR, Ng V, Henry M, Hopewell PC, Ramsay A, Cunningham J, et al.
MTB/RIF in HIV-infected individuals. AIDS. 2014;28(9):1307–14. Sputum processing methods to improve the sensitivity of smear
20. Lawn SD, Dheda K, Kerkhoff AD, Peter JG, Dorman S, Boehme CC, et al. microscopy for tuberculosis: a systematic review. Lancet Infect Dis.
Determine TB-LAM lateral flow urine antigen assay for HIV-associated 2006;6(10):664–74.
tuberculosis: recommendations on the design and reporting of clinical 42. Steingart KR, Henry M, Ng V, Hopewell PC, Ramsay A, Cunningham J, et al.
studies. BMC Infect Dis. 2013;13:407. Fluorescence versus conventional sputum smear microscopy for
21. Country AIDS response progres report, Ghana 2012–2013, Ghana AIDS tuberculosis: a systematic review. Lancet Infect Dis. 2006;6(9):570–81.
Commission, UNAIDS, 2014 [http://www.unaids.org/sites/default/files/country/ 43. Kroidl I, Clowes P, Reither K, Mtafya B, Rojas-Ponce G, Ntinginya EN, et al.
documents//GHA_narrative_report_2014.pdf] (Accessed on 01-10-15). Performance of urine lipoarabinomannan assays for paediatric tuberculosis
22. World Health Organization: Tuberculosis country profile, Ghana. 2013. in Tanzania. Eur Respir J. 2015;46(3):761–70.
Geneva: World Health Organization [https://extranet.who.int/sree/ 44. Wallis RS, Kim P, Cole S, Hanna D, Andrade BB, Maeurer M, et al.
Reports?op=Replet&name=%2FWHO_HQ_Reports%2FG2%2FPROD%2 Tuberculosis biomarkers discovery: developments, needs, and challenges.
FEXT%2FTBCountryProfile&ISO2=GH&LAN=EN&outtype=html] (Accessed on Lancet Infect Dis. 2013;13(4):362–72.
01-10-15). 45. Cannas A, Calvo L, Chiacchio T, Cuzzi G, Vanini V, Lauria FN, et al. IP-10
23. World Health Organization. Antiretroviral therapy for HIV infection in adults detection in urine is associated with lung diseases. BMC Infect Dis. 2010;10:333.
and adolescents: recommendations for a public health approach - 2010 46. Rosen S, Fox MP. Retention in HIV care between testing and treatment in
revision. Geneva: World Health Organization; 2010. p. 145. sub-Saharan Africa: a systematic review. PLoS Med. 2011;8(7):e1001056.
24. National HIV/AIDS/STI Control Programme. Guidelines for antiretroviral therapy 47. Brinkhof MW, Pujades-Rodriguez M, Egger M. Mortality of patients lost to
in Ghana, 2010. Ghana: Ghana Ministry of Health/Ghana Health Service; 2010. follow-up in antiretroviral treatment programmes in resource-limited
25. World Health Organization, Stop TB Dept. DoHA. Guidelines for intensified settings: systematic review and meta-analysis. PLoS One. 2009;4(6):e5790.
tuberculosis case-finding and isoniazid preventive therapy for people living with 48. Honge BL, Jespersen S, Nordentoft PB, Medina C, Da Silva D, Da Silva ZJ, et al.
HIV in resource-constrained settings. Geneva: World Health Organization; 2011. Loss to follow-up occurs at all stages in the diagnostic and follow-up period
26. Subbe CP, Kruger M, Rutherford P, Gemmel L. Validation of a modified Early among HIV-infected patients in Guinea-Bissau: a 7-year retrospective cohort
Warning Score in medical admissions. QJM. 2001;94(10):521–6. study. BMJ Open. 2013;3(10):e003499.
27. Bossuyt PM, Reitsma JB, Bruns DE, Gatsonis CA, Glasziou PP, Irwig LM, et al.
Towards complete and accurate reporting of studies of diagnostic accuracy:
the STARD initiative. Standards for Reporting of Diagnostic Accuracy. Clin
Chem. 2003;49(1):1–6.
28. Addo KK, Yeboah-Manu DK, Bonsu FA, Hanson Nortey NN, Dzata F, Chimzizi
R, et al. Tuberculosis culture/Drug susceptibility testing laboratory manual
for Ghana. Ghana: National Tuberculosis Control Programme, Ghana Health
Service; 2012.
29. Addo KK, Yeboah-Manu DK, Bonsu FA, Hanson Nortey NN, Dzata F, Chimzizi
R, et al. Tuberculosis Microscopy Laboratory Manual for Ghana. Ghana:
National Tuberculosis Control Programme, Ghana Health Service; 2012.
30. Sarfo FS, Sarfo MA, Norman B, Phillips R, Bedu-Addo G, Chadwick D. Risk of
deaths, AIDS-defining and non-AIDS defining events among Ghanaians on
long-term combination antiretroviral therapy. PLoS One. 2014;9(10):e111400.
31. Lawn SD, Kerkhoff AD, Vogt M, Wood R. HIV-associated tuberculosis:
relationship between disease severity and the sensitivity of new sputum-based Submit your next manuscript to BioMed Central
and urine-based diagnostic assays. BMC Med. 2013;11:231. and take full advantage of: 
32. Lawn SD, Kerkhoff AD, Vogt M, Wood R. Clinical significance of
lipoarabinomannan (LAM) detection in urine using a low-cost point-of-care
• Convenient online submission
diagnostic assay for HIV-associated tuberculosis. AIDS. 2012.
33. Asiimwe SB, Abdallah A, Ssekitoleko R. A simple prognostic index based on • Thorough peer review
admission vital signs data among patients with sepsis in a resource-limited • No space constraints or color figure charges
setting. Crit Care. 2015;19:86.
34. Burch VC, Tarr G, Morroni C. Modified early warning score predicts the need • Immediate publication on acceptance
for hospital admission and inhospital mortality. Emerg Med J. 2008;25(10):674–8. • Inclusion in PubMed, CAS, Scopus and Google Scholar
35. McNerney R, Daley P. Towards a point-of-care test for active tuberculosis: • Research which is freely available for redistribution
obstacles and opportunities. Nat Rev Microbiol. 2011;9(3):204–13.
Submit your manuscript at 
www.biomedcentral.com/submit