F inding A lgorithms for PLHIV Christina Yoon MD MPH MAS University of California San Francisco World TB Day Symposium March 23 2018 TB amp HIV TB leading cause of HIV deaths 1 in 3 worldwide ID: 806777
Download The PPT/PDF document "Yield and Efficiency of Novel TB Intensi..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Yield and Efficiency of Novel TB Intensified Case Finding Algorithms for PLHIV
Christina Yoon, MD, MPH, MASUniversity of California, San FranciscoWorld TB Day SymposiumMarch 23, 2018
Slide2TB & HIVTB - leading cause of HIV deaths (1 in 3) worldwide
Intensified case finding (ICF) recommended for all PLHIV presenting for care, regardless of symptomsICF = systematic TB screening, followed by confirmatory TB testing for all those who screen-positive2 main problems with current ICF algorithm: high false-positivity and inadequate sensitivityWHO. Global TB Report, 2015
WHO. 3I’s Guidelines, 2011
Slide3Screen for TB: WHO symptom screenCurrent cough, fever, night sweats, weight loss
Test for TB: Xpert MTB/RIF 1st line confirmatory test for PLHIV
ICF algorithm for PLHIVActive TB: Initiate TB treatment
Screen-positive
Xpert
-positive
Lawn, et al.
Plos
MED 2011
Abed Al-
Darraji
, et al. PLoS ONE 2013
LaCourse
, et al. JAIDS 2016
ICF yield: 40-60%
>80%
screen-positive
Misses 40-50% of all TB cases
Slide4Options for TB screeningC-reactive protein (CRP)I
nflammatory marker, levels rise with active PTB90% sensitivity and 70% specificity for culture-positive TB (cut-point 8 mg/L) Only test to meet the WHO targets for an effective TB screening test for PLHIV
Capillary blood
$2 per assay
R
esults in 3 min
Yoon, et al. Lancet ID 2018
+
=
Slide5Options for confirmatory TB testingIncreased speed: Determine TB-LAMLAM, an
MTb cell wall antigen Recommended as an “add-on test” for patients with advanced HIV and/or severe illnessIncreased sensitivity: Sputum liquid culture ($17 per MGIT)Gold standard for pulmonary TB diagnosis
Unprocessed urine
$4 per strip
Results in 25 min
WHO. Urine LAM 2015
+
=
Slide6Study Objective and DesignObjective:
To compare the diagnostic yield and efficiency of novel ICF algorithms to the current ICF algorithmDiagnostic yield - % of all TB cases detected (in reference to 2 MGIT cultures)Efficiency – number of confirmatory tests used to detect one case of active TB
Design:Prospective cross-sectional study of consecutive HIV-infected adults presenting for ART initiation
Slide7MethodsParticipants:
Adults with CD4 ≤350 cells/μL presenting to 2 HIV/AIDS clinics (Uganda) for ART initiation from 4/2014 to 12/2016Procedures: Symptom and POC CRP TB screening (8 mg/L cut-point)Sputum x1 for Xpert
and x2 for MGIT cultureUrine x1 for LAM (if CD4 ≤100 cells/μL; Grade 2 cut-point)Evaluated ICF algorithms combined…
Screening tests
: POC CRP or WHO symptom screen
Confirmatory testing strategies
: LAM,
Xpert
, and/or 1
rst
MGIT
Active TB definition:
≥1 positive sputum MGIT culture
Slide8Figure 1. Confirmatory testing strategiesAny CD4
CD4 ≤100Xpert
Xpert
+
Xpert
–
LAM
LAM
+
LAM
–
CD4
>100
Xpert
Xpert
+
Xpert
–
CD4 ≤100
CD4
>100
LAM
LAM
+
LAM
–
Xpert
Xpert
+
Xpert
–
MGIT
MGIT
+
MGIT -
1.
Xpert
only
2. LAM +
Xpert
3
. LAM +
Xpert
+ MGIT
Slide9Figure 2. Patient flow diagramParticipants enrolled
N=1511 N=1245 Excluded:19 Missing CD4 count3 Missing POC CRP15 Missing LAM
229 Contaminated MGITCD4 >100 cells/uLN=806 (65%)
CD4 ≤100 cells/
uLN=439 (35%)
No TB
N=705 (87%)
Active TB
N=102 (13%)
2,
Xpert
positive
0, 1
rst
MGIT positive
No TB
N=338 (77%)
Active TB
N=101 (23%)
56, Xpert positive83, 1
rst MGIT positive
6, LAM positive4, Xpert positive
0, 1rst MGIT positive26, LAM positive
65,
Xpert
positive
77, 1
rst
MGIT positive
Slide10Table 1. Demographics and clinical characteristics
Characteristic, N (%)
Total (N=1245)
Age (years)
†
33 (27-40)
Female
648
(
52%
)
CD4 count (cells/µL)
†
153 (67-252)
CD4 ≤100 cells/µL
439 (35%
)
Symptom screen
1100 (88%)
POC CRP ≥8 mg/L
498 (40%)
POC CRP (mg/L)
4.0 (2.5-24.4)
Culture-positive
TB
203 (16%)
Slide11Table 2. Diagnostic yield, incremental yield and # of false-positives for each ICF algorithm
Slide12Table 3. Number of confirmatory tests used and NNT to detect one case of active TB
Slide13Table 4. Individual test costs, ICF costs and costs per TB case detected
Slide14Conclusions
POC CRP-based ICF has similar diagnostic yield but is more than twice as efficient than symptom-based ICF Addition of LAM increases the speed of TB diagnosis among patients with advanced HIVAddition of a single liquid culture greatly improves ICF yield
Costs saved by using POC CRP-based ICF may be re-allocated to culture to improve ICF yield without increasing ICF cost
Slide15AcknowledgmentsUCSFA
dithya CattamanchiMakerere UniversityFred SemitalaMoses KamyaElly Atuhumuza
Alfred AndamaLucy AsegeSandra MwebeJane KatendeMartha
Nakaye
Johns Hopkins University
David Dowdy
Derek Armstrong
Funding
sources:
NIAID K23
UCSF-GIVI CFAR
NIAID/PEPFAR
UCSF Nina Ireland
Program in Lung Health