1. GOES fairly fast…with community to say stockouts and activism, hosp to say RAU, PHC to say ambulatory and CD4, then brief tools and HR, then summary.
Find place for… Yes you are right. It is the use of LAM that has led to a reduction of 2 days to 4 hours in the time from admission to initiation of TB treatment. However this may be specific to Kinshasa where all admitted have advanced HIV. In RAU type settings the CD4 is a necessary screening before doing LAM.
 The Kinshasa results is because we do systematic LAM in all admitted. However at PHC we don’t and it is clear that PIMA will shorten TAT as it will anywhere systematic LAM is not done. 
For LAM in Helena’s latest Plos MED paper:  
Of the 205 patients with laboratory-confirmed TB, 205 (100%) had a LAM result, 175 (81.4%)
a microscopy result, and 166 (81.0%) an Xpert result. The median time to result was 0 (IQR
0–0) days for LAM, 5 (IQR 3–8) days for microscopy, and 5 (IQR 2–9) days for Xpert. 
Moreover, the turnaround time was considerably longer for microscopy
and Xpert than it was for LAM.This is in AMBULATORY patients

2. Challenges in implementing LF-LAM in outpatient settings:
Experiences from MSF field programmes
Not much time so will give a brief overall perspective and then describe findings from a few experiences. Also don’t want to repeat presentation from this morning….but reminder that we are deeply concerned that success in pushing viral load and simple care for the 90…led to moving backwards on the 10s…ie CD4.
Reinvest in identifying risk and simplifying access at every level….and perhaps most importantly regenerating demand and outrage at gaps.
As others said ….relatively few and simple tools needed to identify those shifted down continuum from stable to critical.
Good evidence and policy on main killers and ways to prevent and treat but in reality….
RCTs and evidence don’t lead to policy (REMSTART great example)…especially when money is tight and those needing it are quiet. Low hanging fruits was all they wanted?
?reminder on changing nature of advanced hiv?...and debates. Context specific but all agree that unlike VL for stability.. there is urgency. Question is whether system can meet that need for urgency with centralised tools and referral or ensure quality with decentralised ones.
Multidisease on xpert?
Highlight here and in evening the remstart…trip showing that clinical trials ‘robustness’ is an illusion. Cd4 access, staff overload and willingness (trust), and MoH policy….not to mention all the tools..
Easy slides….what is stability?...abcd etc
Policy is result of institutions, interests, and ideas (including evidence). Look at profit to find interests….gendered…alcohol, smoking etc…
Tom Ellman MSF SA IAS 2019

3. No conflicts of interest to declare

4. What packages at which level of care?
Translating evidence into practice in real-world. Big gap between Remstart in study and in reality.
High hospital mortality directly relates to failure to take advantage of the narrow window of opportunity when seen at PHC level.
All as close to home as possible
Increasing complexity requires centralisation
But can be efficient in healthy health systems…ie where CD4 TAT is fast, patients access to f/u easy, and tracing manageable
Doesn’t work in CAR
Decentralisation requires simplification
But over-simplification can bring extra costs (POC tools and over/under treatment)…how to decide what is worth it? Definie cost-efficacy?

5. Realising the potential of POC tests in AHD Package of Care
Part of a bigger plan
Messages:
Where is it done by MSF in Africa .
First sites started in 2016
Without M&E systems action and evidence suffers
Khayelitsha

6. Use of LAM test for tuberculosis diagnosis
Helena Huerga
Epicentre Scientific Day
13th June 2019
Stolen from Helena to look at going beyond WHO guidelines and especially ambu

7. Methods
Observational study with prospective and cross-sectional components
Sites:
2 urban Health Centres, 1 referral HIV Centre, Maputo, Mozambique
3 rural Health Centres, Chiradzulu, Malawi
Out-patient and In-patient departments of Chiradzulu District Hospital, Malawi
Out-patients and In-patients Departments of Kabinda Hospital, Kinshasa, DRC (only feasibility)
We conducted an observational study with prospective and cross-sectional components in 8 health facilities in Malawi, Mozambique and DRC.

8. Study population: HIV adult patients
Group 1
Group 2
Group 3
Health Centres / OPD
TB symptoms
Medical wards
Irrespective of TB symptoms
HIV consultation
No TB symptoms reported
….and feasibility
CD4 <100
Any CD4
Any CD4
Current recommendations: only if both TB symptoms and CD4 <100 or seriously ill

9. Results – Patient characteristics
Group 1
Health Centre / OPD
Group 2
Medical Wards
Group 3
HIV clinic
TB symptoms
any CD4
N=485, n (%)
Irrespective symptoms
N=387, n (%)
No TB symptoms reported
CD4<100
N=360, n (%)
Age, median [IQR] 41
[34 – 49] 38
[32 – 45] 36
[31 – 43]
CD4, median [IQR]
341
[129 – 546]
173
[51 – 370] 31
[13 – 53]
On ART
438
(88)
300
(78)
172
(48)
Seriously ill* 35
(7)
126
(33) 13
(4)
Any TB symptom
485
(100)
349
(90)
162
(45)
Unable to produce sputum 40
(8)
133
(34) -
Unable to produce urine 5
(1.0) 3
(0.8)
(0.0)
This table shows the characteristics of the patients enrolled in the study.
I’d like to highlight some aspects:
- Among patients in group 1 who were ambulatory patients who self-reported TB symptoms, median CD4 was high and few patients were seriously ill.
*Seriously ill: temp >39°C, resp rate >30 /min, cardiac rate >120 /min, or unable to walk without help

10. Results – Patient characteristics
Group 1
Health Centre / OPD
Group 2
Medical Wards
Group 3
HIV clinic
TB symptoms
any CD4
N=485, n (%)
Irrespective symptoms
N=387, n (%)
No TB symptoms reported
CD4<100
N=360, n (%)
Age, median [IQR] 41
[34 – 49] 38
[32 – 45] 36
[31 – 43]
CD4, median [IQR]
341
[129 – 546]
173
[51 – 370] 31
[13 – 53]
On ART
438
(88)
300
(78)
172
(48)
Seriously ill* 35
(7)
126
(33) 13
(4)
Any TB symptom
485
(100)
349
(90)
162
(45)
Unable to produce sputum 40
(8)
133
(34) -
Unable to produce urine 5
(1.0) 3
(0.8)
(0.0)
Among hospitalized patients included irrespectively of their TB symptoms, almost all patients presented at least one TB symptom.
Among ambulatory patients coming for new or follow-up HIV consultation who did not self-reported TB symptoms, after asking actively, half of them present at least one TB symptom.
*Seriously ill: temp >39°C, resp rate >30 /min, cardiac rate >120 /min, or unable to walk without help

11. Results – Patient characteristics
Group 1
Health Centre / OPD
Group 2
Medical Wards
Group 3
HIV clinic
TB symptoms
any CD4
N=485, n (%)
Irrespective symptoms
N=387, n (%)
No TB symptoms reported
CD4<100
N=360, n (%)
Age, median [IQR] 41
[34 – 49] 38
[32 – 45] 36
[31 – 43]
CD4, median [IQR]
341
[129 – 546]
173
[51 – 370] 31
[13 – 53]
On ART
438
(88)
300
(78)
172
(48)
Seriously ill* 35
(7)
126
(33) 13
(4)
Any TB symptom
485
(100)
349
(90)
162
(45)
Unable to produce sputum 40
(8)
133
(34) -
Unable to produce urine 5
(1.0) 3
(0.8)
(0.0)
One third of the hospitalized patients could not produce sputum, while almost all patients could produce urine.
*Seriously ill: temp >39°C, resp rate >30 /min, cardiac rate >120 /min, or unable to walk without help

12. Diagnostic yield: proportion of positive among laboratory confirmed TB
All patients
Patients
CD4 < 100
CD4 100–199
LAM
82.4 (169/205)
86.4 (133/154)
70.6 (36/51)
Microscopy
33.7 (69/205)
19.5 (30/154)
43.1 (22/51)
Xpert
41.0 (84/205)
40.3 (62/154)
Of the 205 patients with laboratory-confirmed TB, 205 (100%) had a LAM result, 175 (81.4%) a microscopy result, and 166 (81.0%) an Xpert result. The median time to result was 0 (IQR 0–0) days for LAM, 5 (IQR 3–8) days for microscopy, and 5 (IQR 2–9) days for Xpert.
LAM sensitivity was 65% (63 LAM positive/97 Xpert or culture positive). LAM was able to diagnose a high proprortion of patients mainly because almost all patients had a LAM result, in contrast, almost a quarter of the patients (23%) with symptoms of TB did not have an Xpert result. In practice, patients without Xpert result, had to be diagnosed using clinical exam or microscopy.

13. Irrespective symptoms No TB symptoms reported
LAM positive results
Group 1
Health Centre/OPD
Group 2
Medical Wards
Group 3
HIV Clinic
TB symptoms
any CD4
N=485, n/N (%)
Irrespective symptoms
N=387, n/N (%)
No TB symptoms reported
CD4<100
N=360, n/N (%)
Overall
81/485
(17)
100/387
(26)
43/360
(12)
CD4 <100
30/96
(32)
49/141
(35) CD
12/75
(16)
13/61
(21) -
CD4 >=200
39/314
(13)
34/174
(20)
91/349
30/162
(19)
No TB symptoms
9/38
(10)
13/198
(7)
This table shows the proportion of patients with LAM positive results.
Among patients with self-reported TB symptoms irrespective of their CD4, 17% were LAM positive, and a relative high proportion of the patients with more than 100 CD4, in which LAM is not currently recommended had a positive LAM.

14. Irrespective symptoms No TB symptoms reported
LAM positive results
Group 1
Health Centre/OPD
Group 2
Medical Wards
Group 3
HIV Clinic
TB symptoms
any CD4
N=485, n/N (%)
Irrespective symptoms
N=387, n/N (%)
No TB symptoms reported
CD4<100
N=360, n/N (%)
Overall
81/485
(17)
100/387
(26)
43/360
(12)
CD4 <100
30/96
(32)
49/141
(35) CD
12/75
(16)
13/61
(21) -
CD4 >=200
39/314
(13)
34/174
(20)
91/349
30/162
(19)
No TB symptoms
9/38
(10)
13/198
(7)
In patients with no self-reported TB symptoms and low CD4, 12% were LAM positive and 7% of the patients with no TB symptoms had a positive LAM.

15. Mortality at 6 months Symptomatic out-patients <200 CD4 (group 1)
RR (95% CI)
p-value
aRR* (95% CI)
LAM positive and treated
12.8 1
LAM positive and not treated
36.8
2.9 (1.4–5.9)
<0.01
2.6 (1.3–5.2)
LAM negative and treated
2.9
0.2 (0.03–1.6)
0.14
0.3 (0.04–2.3)
0.26
Regarding the mortality, among patients with self-reported symptoms of TB and CD4<200, we found that the mortality of patients LAM positive not treated for TB was two times higher than those LAM positive treated and also higher than patients LAM negative treated. The higher risk of mortality persisted after adjusting for other factors such as seriously ill or CD4.
* Model adjusted for sex, age, BMI, heamoglobin, ART, seriously ill, CD4

16. TB Lam implementation Eswatini (PHC)
Overall in 2018
< 100 and TB symptoms
PHC facilities positivity = 7%
NHC positivity = 12%
Thanks to MSF and Eswatini MoH

17. MSF Malawi PHC: CD4 < 200 and TB symptoms
13% LAM positive
KDH OPD and IPD….The median TAT for LAM was 1:26 hours [IQR 0:50-3:51]. In comparison, the TAT for Gene Xpert was 2 days [IQR 2-3] and chest X-ray was available within 1-3 days. Among LAM-positive patients, a median 2:42 hours [IQR 1:30 – 8:01] elapsed from LAM request to first dose of TB drugs. More patients had a LAM test result compared to GeneXpert: 168/213 (78.9%) vs 77/213 (36.2%), p< LAM was perceived as easy to perform, and readily available.

18. FujiFilm LAM, Semi-Quant CrAg, HIV POC ARV levels, HIV POC Genotyping
PHC level: Advanced HIV Disease screening package ( CD4< 200 and/or symptomatic)
Urine Lam (25mins)
PIMA (25mins)
Eshowe south Africa….LAM for tomorrow. But its quite a package!
GeneXpert:
HIV VL, MTB/Rif
Visitect (40mins)
CrAg (10mins)
In the near future:
FujiFilm LAM, Semi-Quant CrAg, HIV POC ARV levels, HIV POC Genotyping

19. Case study : Immy CrAg test
A CD4 test must be done first to trigger CrAg test (if CD4<200)
For best results, this test requires that venous blood (EDTA) is first let to stand verticaly for a minimum of 10mins so that plasma can be collected after red-blood cells sediment naturally (as there are no centrifuges in many PHCs). Serum and CSF can be used. Whole blood can also sometimes be used.
Exactly 40ul of plasma is must be pipetted into a reaction tube and a 10min incubation before reading results
For a busy PHC, these steps could have a knock-on effect on the workload of HCW who also need to also multi-task
Allow an incubation time of 10 min before reading result
Add one drop of CrAg specimen diluent into the labeled cryovial tube
Add 40µl of plasma sample into the cryovial tube and mix.
Insert one CrAg test strip into the mixture.

20. Case Study 2: TB LAM (and FujiFilm TB LAM)
TB LAM: The simplest of the POCs but still needs watching
Fujifilm: More complex

21. Case study 3: Visitect Omega CD4 LFA
The test has 3 incubation steps which need accurate timings
This may prove difficult for clinicians who need to also multi-task
30 µL of sample
Results
1 drop Well A
3 drops Well B
90°
1 drop
90°
3 drops
Capillary
Wait 20 min
Wait 3 min
Wait 17 min
Venous (EDTA)
Results in ~40min

22. These test represent a significant burden
Risk of underuse of tests, unreliability of results, delayed responses
Will staffing increase?
Challenge of motivating and capacitating staff
Especially where guidelines and policies are lacking
Could breakdown the three things…ie underuse…..staff say ‘msf’ only, unreliable….staff don’t read on time, don’t use card,

23. The CQUIN Learning Network
TB LAM Poll
The CQUIN Learning Network

24. Challenges Reading card utilisation
Requirement for CD4 testing prior to LAM testing
Provision of samples
Children and menstruating women
Mistrust of LAM result
Heavy workload and low motivation
13% (56 minutes) of daily workload just for LAM in Malawi study
Underuse of LAM and mistimed reading of results
One case of ‘Invalid result’
(+ve LAM 4+ and no control line visible, even testing using different lot numbers. But other urine samples giving normal results).
Need better Post Market Surveillance and communication with company.
49 staff interviewed

25. LAM is not only for seriously ill patients and not only for IPD
LAM is feasible in our settings
Easy sample, easy test, fast TAT and high yield
Focus on ensuring access and ensuring HR capacity
LAM practice is far behind Guidelines and Policies
AHD Dashboard in development
Task sharing needed
Access to essential diagnostic, preventive and treatment tools
We need POC CD4
Funding and political commitment missing
Community awareness and activism needed
change

26. Some questions…. Who will look after the LAMs?
A new cadre?
Must we really confirm positive LAMs?
How many false + do we expect in high CD4?
Do they matter?
Planning with and without CD4?
Cant we stop worrying so much about IRIS?

27. Thank you https://samumsf.org/en/news/advanced-hiv-disease-toolkit
Acknowledgements: MSF and Epicentre colleagues (especially Helena Huerga), partners and patients