This Expert Panel is Archived.

While this Expert Panel is no longer active, we invite you to review and recommend past replies and resources. Membership for this Expert Panel is closed, but we hope you'll join us in one of the many communities on GHDonline.

Panelists of Overcoming Barriers to High-Quality TB Treatment During HIV Co-infection and GHDonline staff

Overcoming Barriers to High-Quality TB Treatment During HIV Co-infection

Posted: 10 May, 2016   Recommendations: 14   Replies: 52

While guidelines are clear about the need to treat tuberculosis (TB) in the setting of human immunodeficiency virus (HIV) infection, the reality is that many patients with HIV/TB co-infection do not receive high-quality TB treatment. The barriers to effective management of TB disease in the setting of HIV are present throughout the care cascade. Recognizing and diagnosing TB is difficult in the presence of HIV, leading to reluctance to initiate empiric treatment. Further, difficulties in drug susceptibility testing may impede the use of regimens for multi-drug resistance tuberculosis (MDR-TB). Then, if the decision to initiate TB treatment is taken, regimens must be appropriately designed: drug-drug interactions must be accounted for; drugs must be available. Finally, the treatment period itself is full of pitfalls, including immune reconstitution inflammatory syndrome (IRIS), heavy pill burdens and unpleasant side effects of combined HIV/TB treatment, all combining to leading to less than full adherence.
Strategies to overcome these barriers need to be discussed and shared among health care providers. This Expert Panel will seek to explore these solutions and disseminate them within the MDR-TB and HIV communities on GHDonline.

We are pleased to welcome the following panelists for this discussion:
- Junior Bazile, MD, MPH, Consultant in Health Systems and Institutional Capacity Development, The Bassiouni Group
- Riitta Dlodlo, MD, DMCH, MPH, Director, Department of Tuberculosis and HIV, the Union
- Serena Koenig, MD, MPH, Associate Physician, Brigham and Women’s Hospital (Division of Global Health Equity, and Division of Infectious Disease); Senior advisor, GHESKIO (Haitian Group for the Study of Kaposi’s Sarcoma and Opportunistic Infections); Assistant Professor, Harvard Medical School
- Dylan Tierney, MD, MPH, Associate Physician, Brigham and Women's Hospital
During our week-long discussion, panelists will address the following questions:
1. When is empiric TB treatment indicated in patients with HIV infection and how should possible drug resistance be accounted for in regimen design?
2. To what degree can diagnostic tests be expected to provide a microbiologic diagnosis of TB disease in the setting of HIV co-infection?
3. How should regimens for the co-treatment of HIV and TB be formulated to reduce pill burdens and avoid drug-drug interactions and compounded side effects?
4. What are best practices for managing IRIS?
5. What specific strategies can be used by community health workers to improve adherence to HIV and TB co-treatment?

We encourage you to join the conversation and share your comments and questions with the community.



Dr. Ivan Cortez Chalar Replied at 1:20 PM, 10 May 2016

Muy apenado por la confusión, pero para serle sincero mi ingles es cuasi nulo, peor aun para expresar algo escrito. Una vez mas me disculpo y lo lamento mucho

Enviado desde Correo para Windows 10

KWASI OFORI - ANTI Replied at 1:41 PM, 10 May 2016

The pill burden is in my opinion the biggest challenge in the treatment of HIV/TB coinfections. A significant number of patients default as a result of severe nausea and vomiting. I'll be grateful if our expert panel can adequately address this topic.

Lauren Alexanderson Replied at 2:39 PM, 10 May 2016

This expert panel looks fascinating! I'm excited to tune in next week!

Those following or speaking on the panel may be interested in a new database we launched last week of HIV/TB Co-Infection Guidelines from 24 countries around the world. We welcome submissions from those of you who might have updates or countries that we haven't yet included!

Explore it here!

Taye Balcha Replied at 4:08 PM, 10 May 2016

Dear Dylan,
Thank you for inviting me to this important discussion. I look forward to joining you as a panelist.

Kind Regards,

Elie NZIYOMAZE Replied at 4:30 PM, 10 May 2016

Thank for inviting me here Dylan ,
Looking forward gain more from our panelists~

Jane Carter Replied at 12:02 AM, 11 May 2016

Dear Lauren

Unfortunately your link doesn't seem to work. Kindly send a revised one.

Many thanks.

Jane Carter

HOMA MANSOOR Replied at 12:04 AM, 11 May 2016

Thanks for the invitation Dylan,looking forward to learn and exchange ideas with the panelist.

Mark Minnery Replied at 1:40 AM, 11 May 2016

A very interesting ad timely discussion. Im hoping that some panelists may be able to comment on structural health system problems (such as non-integration between tb and hiv programs) and different service delivery models which can enable higher quality care for coinfected patients while also providing a cost effective methods for addressing screening and promotional activities in high burden areas.

Getachew Desalegn Replied at 2:07 AM, 11 May 2016

Thank you. Will attend

*Getachew Wondimagegn Desalegn (MD, MPH)*

*KNCV Tuberculosis Foundation*
*P.O.Box 703 code 1110, Addis Ababa, Ethiopia*

*+251116189422 <%2B251116189422>*
*+251 911408602 <%2B251%20911408602>*

Tuberculosis is the second deadliest infection in the world. With early
detection and proper treatment, most people with tuberculosis can fully
recover. Our efforts and investment in tuberculosis control help to save
millions of lives worldwide. For more information, please visit <>).*tachew** Wondimagegn
Desalegn** (MD, MPH)*
*TB IC/HSS and PMDT Coordinator, TB CARE I*
*KNCV Tuberculosis Foundation*

*P.O.Box 703 code 1110, Addis AbabaEthiopia+251116189422; *

Ebrahim Hoosain Replied at 7:51 AM, 11 May 2016

I look forward to an interesting discussion of this topic by all participants.

Lauren Alexanderson Replied at 8:29 AM, 11 May 2016

Here is a new link. My apologies!

Jude Beauchamp Replied at 11:03 AM, 11 May 2016

Hi Dylan,
Thanks for the invitation.
Looking forward to this discussion about such an important issue.


*Jude BEAUCHAMP MD, MSc Candidate*

*Master of Medical Sciences in Global Health Delivery*

*Harvard Medical School*

<> *

Nelly Solomonia Replied at 11:15 AM, 11 May 2016

Dear colleagues,

Thank you very much for giving opportunity to participate in such interesting upcoming discussion. On the link shared by Dear Lauren you can not find information from Georgia. If someone will be interested of course I can share key recommendations of our National Guideline and experience of Georgian TB Physicians in management of TB/HIV co-infection.



Maya Habous Replied at 11:39 AM, 11 May 2016

It is my pleasure that I join with you .Many thanks Maya

Zacarias Mateus Replied at 12:39 PM, 11 May 2016

Thanks for the invitation

I Work with Patients with TB and HIV and its a realy Challenge dealing with some of this patients.

I hope learn more in this discussion...

Thanks you

Zacarias Mateus

Mutamuliza Janviere Replied at 3:12 PM, 11 May 2016

Jude Beauchamp
to a discussion
inOvercoming Barriers to High-Quality TB Treatment During HIV Co-infection

Dear Dylan,
Thanks for the invitation.
Looking forward to this discussion about such an important issue.

Jane Che Replied at 4:21 PM, 11 May 2016

Many thanks for the invitation. I look forward to a very interesting and useful discussion on this very important topic.

Kind regards,


Monique Germain Replied at 4:33 PM, 11 May 2016

Thank you for inviting me. I will be paying attention.

Allen Bain Replied at 4:47 PM, 11 May 2016

Dear Dylan,
I am hopeful to see some additional testing of Immunitor's V5 that has
appeared to be effective in several different clinical trials, including
with TB/HIV patients.

Virginia Lipke RN, MHA, ACRN, CIC Replied at 5:34 PM, 11 May 2016

Thank you for inviting me. I look forward to the sessions.

Madhuri Gandikota Replied at 6:38 PM, 11 May 2016

Thanks for the invite. Looking forward for an interesting session.

egh Eduardo Gotuzzo Replied at 7:15 PM, 11 May 2016

In my opinion we need to work together because each group defend their area
as propety and tha tis unfair with patients
For other part
Tbc need Communty DOT adn ARV need to b,e conducted for ID perosn
In tha tdorection we need to work together special becuase the aptient had
high risk to have tbmdr and also deceloped IRIS
Becias ethe patient received al least 7 drugs 4 for tbc amd 3 for HIV at
minumn becuase some need tmp.smx or other treatmemt for comorbilites
Also both need to sjare the information in tha tway obth know very well the
history and potential sdie effects
Also both nned to promote adhenrece the best of the treatment is adherence
to both
Finally the use of amsk until the patient is sputum negative is necesay
even the aptient feel is discriminatory .the main reason is neciase some
time tje hospital or health center are the place they adquired the tbcwhen
they shar ethe same lobby or waitng room to see doctor with other no tbc
petient but with HIV

El martes, 10 de mayo de 2016, Dylan Tierney via GHDonline <
> escribió:


*Dr. Eduardo Gotuzzo*, Director Instituto de Medicina Tropical "Alexander
von Humboldt"
*TELF:* (+51) 14823910, 4823903* SKYPE: *eduardo_gotuzzo * EMAIL: **e
< <


Av. Honorio Delgado 430.

San Martin de Porras,

Lima- Perú

* <>*
*IMT "AvH"* <>

<> [image:]




Este mensaje electrónico (e-mail) pertenece al remitente y puede contener
información confidencial legalmente protegida para uso del destinatario.
La divulgación de esta información, su copia o distribución para otro uso
esta terminantemente prohibida. Gracias.
This e-mail message belongs to the sender and may include confidential and
legally privileged information, for the exclusive use of the named
recipient. Any disclosure, copying or distribution of this information for
aims other than the intended purpose, is prohibited. Thank you.

Nataliya Moyseeva Replied at 8:35 AM, 12 May 2016

I think, main barieres are lack of TB and HIV drugs, postpones of ARV treatment. many settings don't provide appropriate managment of side effects. I have not find any survay targeted at comparison of quolity of brand drugs and generics (side effects)

minhaz rashid Replied at 11:36 AM, 13 May 2016

main difficulties in South Asia are-
1. more number of drop out cases
2. QUALITY drug shortage
3. duration of treatment
4. difficulty in diagnosis
5. lack of expertise
6. side effects ( specially streptomycin )
7. lack of tertiary health care facilities to diagnose complication.
8. decrease self motivation

Dylan Tierney Panelist Replied at 2:20 PM, 16 May 2016

I’d like to kick off this expert panel with a quick introduction. As many of you know, I’m one of the community moderators for the MDR-TB community. I’m an infectious disease physician by training but have been focused on implementation of diagnostics and treatments that can improve the quality of care for patients with TB in resource limited settings over the past five years.

I’m very excited to explore the day’s topic, which is focused on when to start empiric treatment for TB for patients with HIV. As we all know, pulmonary TB frequently exists in a paucibacillary state in the setting of HIV coinfection. Because there are fewer mycobacteria available for detection, arriving at a microbiologically confirmed diagnosis (either smear positive, NAAT positive or culture positive) is more difficult. So clinicians are often left having to decide whether patient with nonspecific clinical syndrome should be started empiric anti-TB treatment.

I would be very interested to hear how the other experts and community members manage this situation. What are the thresholds for starting therapy? Does the threshold depend on how sick the patient is? How are comorbidities accounted for?

Riitta Dlodlo Panelist Replied at 3:18 AM, 17 May 2016

Dear Dylan and Dear Colleagues,

Thank you for the kick off, Dylan, and posing a key question that many clinicians face daily in numerous settings around the world and particularly here in southern Africa which remains a TB-HIV hot spot.

Clinical manifestations of TB depend on a degree of immunosuppression in PLHIV: the more advanced the HIV disease, the more 'unclassical' the symptoms and signs may be. At the same time, with a quality sputum specimen and conscientious microscopy laboratory can pick up TB bacilli in a handsome proportion of PLHIV with TB (Munyati SS et al, CID 2005, 40:1818-27). XpertMTB/RIF adds onto diagnostic tools available to more and more patients, in addition to urine LAM - a simple bed-side test that is also fairly affordable (Peter JG et al, Lancet 2016, 1187-97).

Frequently, it is a race against losing one's patient: in high TB burden settings, the threshold to start anti-TB treatment should rather be lower than than high....

Let your views and experiences flow in & kind regards,

Riitta Dlodlo

ps. Let me try to match Dylan's brilliant introduction and add that I am a MD (Helsinki University, Finland) and MPH (KIT, the Netherlands) and have worked in both clinical and programmatic tuberculosis (TB), TB-HIV and related fields and overall health services management for over 25 years in Zimbabwe and have provided technical assistance in several sub-Saharan African countries. I am a strong proponent of patient-centred, decentralised and integrated services, for example, for persons living with HIV and TB.

I am also passionate about operational research to expand evidence base for health policy, practices and approaches in service provision and capacity building of health professionals. I have served as faculty member in several Union courses, including the International TB Course in Arusha, Tanzania, Zimbabwe and the programmatic TB-HIV course. I have published in TB-HIV and mortality.

Zacarias Mateus Replied at 7:00 AM, 17 May 2016

Dear All

Co infection HIV/TB is a big public health problem nowadays especially in patients with CD4 down. And if we experience make an active tuberculosis treatment study with the first line in all HIV patients with criteria to initiate antiretroviral therapy with no signs of active TB, then compare the proportion of them that after finishing treatment got active TB throughout life?
It would probably be very costly but the cost and benefit would depend on the results.

Best Regards

Zacarias Mateus Replied at 7:08 AM, 17 May 2016

In relation to strategies to improve adherence to HIV and co-TB treatment, I think we should provide inquiry to people who are on TB treatment and probably are doing
therapeutic bankruptcy and may have TB MDR.
In my experience despite being made the DOT, the monitoring of these patients in the community and defective, probably due to lack of staff with skills to identify these situations ....

Best Regards

Junior Bazile Panelist Replied at 10:26 AM, 17 May 2016

Thank you all,
As it has already been said, TB-HIV co-infection is a serious public health
problem in many developing countries. Very often it is hard for clinicians
to make the diagnosis of TB because of low level of mycobacteria and also
because those co-infected patients are usually so sick that they can't even
cough properly enough to get good quality sputum.
In my experience, when it is difficult to get good quality sputum for smear
or culture, the points that are considered for initiating TB treatment are:

Presence of constitutional symptoms (cough, weight lost, fever, chest pain)
Chest x-ray showing either cavitation, interstitial infiltrates or
para-hilar lymphadenopathy.

However since those signs and symptoms are not necessarily pathognomonic of
TB, other points have to be considered: the patient's past exposure to TB
infected persons. The burden of TB in the country or the community where
the patient lives has to also be taken into consideration.
I know that in many situations, clinicians would have divergent ideas on
whether to start TB treatment or not despite the presence of signs and
symptoms that can make one consider TB. In such situation there would be
consideration for potential pneumonia and treatment with macrolides would
be initiated. If after 7-10 of pneumonia treatment there is no improvement,
then TB treatment would be considered.

The use of Gene Xpert for TB diagnosis has been a great help however here
again the result of the test depend on the quality of the sputum.

Dylan Tierney Panelist Replied at 11:59 AM, 17 May 2016

I agree that there is a general awareness of the difficulty of diagnosing TB disease in HIV coinfection. The threshold for starting TB treatment in these patients should definitely be low.

My impression is that even though there is awareness about the issue, there are still many patients with a syndrome consistent with test-negative TB (e.g., smear negative, Xpert negative, urine LAM negative, culture negative) who don’t get empiric therapy. Why? Lack of awareness? Concerns about toxicity of empiric therapy? Some other reason?

I would be interested to hear from community members about why these patients don't get treated. Do you have access to local institutional documents that provide guidance on patients with HIV who you suspect have test-negative TB?

Serena Koenig Panelist Replied at 12:16 PM, 17 May 2016

I have a follow-on question to Dylan's question.

Now that the WHO says all HIV-infected patients should receive ART, ART staging procedures are no longer necessary, and we can start ART as early as the day of HIV testing. However, it is critical that we rule out TB first. We have been in discussion with policy makers about the importance of CXR. In Haiti, we find it very useful for the following reasons: (1) In non-coughing patients, it provides evidence that additional work up is required before starting ART. (2) in coughing patients, it allows us to provide same-day TB treatment - as we do not have same-day GeneXpert results (due to large number of tests, we have Day 2 results), and for those patients who have classic symptoms and classic TB CXR, we start same-day TB treatment. We believe that same-day treatment (either ART or TB as appropriate) is important.

1. I am interested in how valuable others have found CXR to be for patients with new HIV diagnoses, prior to ART initiation.

2. Also - though GeneXpert has been very helpful, we have found that an early morning sputum is more sensitive than a spot sputum. So we have the patient come back in the morning with early morning specimen. Are other sites doing this as well? It does complicate the diagnostic algorithm, but we worry that with spot specimens, we will have an increased number of false-negative GeneXpert tests.

Serena Koenig

Dylan Tierney Panelist Replied at 12:35 PM, 17 May 2016

Here's a study from Uganda by Nakiyina et al that describes how smear negative TB is managed there. (See link.) Out of 253 smear-negative outpatients with HIV coinfection, 10.7% ended up being mycobacterial culture positive but not started on empiric TB treatment. The biggest predictor for empiric TB treatment in the larger overall group was an abnormal chest x-ray.

So how should we consider the finding that 10% of smear negative TB went untreated? Is that a high or low percentage when thinking about delivering high quality care to these patients? What is your impression of the problem of where you work?

Attached resource:

Grigory Volchenkov, MD Replied at 6:46 AM, 18 May 2016

For benefit of not only patients but clinicians the holistic diagnostic approach towards at least most probable opprotunistic infections in HIV infected patients could be developed. I mean diagnostic algorythms, packages, cartridges, tools... We have to focus not only on TB... We should avoid by any means not only not diagnosing TB and not treating TB but wrong TB diagnosis for critically sick HIV infected patient.

NATALIA TAMAYO Replied at 7:13 AM, 18 May 2016

Thanks a lot Dr.Volchenkov to bring this up to the discussion. It's not the same scenario a HIV patient with CD4=500 cel/mm3, that one with CD4=50 cel/ml, when we are going for TB evaluation. Despite I'm sure we all use the TB diagnosis algorithm for both, should be in our mind a list of differential diagnosis, for sure much more longer for one with low CD4. Also the sensitivity of diagnostic method are different. We usually considered the "pooled" one, but we know that is not the same we can expect from a CXR of patient with severe immunodepression from one with CD4 > 350 cel/mm3; or for LAM that clearly is only recommended for diagnosis (not screening) in CD4<100 cel/mm3. I'm working in Myanmar, and for us Peniciolisis, Histoplasmosis and Cryptococcosis are still a concern and TB differential diagnosis (for PTB and EPTB). Luckily we have a lab that support us in diagnosis, but I can see how difficult could be in other context.
And I also agreed with Dr.Tierney, threshold for starting TB treatment in HIV coinfection, especially in patients severe immunodepressed, for whom the bacillary load is low and mortality high , should be low.

Florian Sauter Replied at 4:08 PM, 18 May 2016

Dear Dylan,

reasons why empiric TB treatment is not started often and/or early enough. Some experiences from NW Cameroon, where I used to work between 2012 and 2015:

Definitely unawareness and unavailability of the respective algorithms (e.g. PIH and MSF 2014) as well as fear of toxicity have their role.

But also:

Programmatic issues: In Cameroon I know about concerns from the side of the TB-program, that smear/test negative patients started on TB-treatment will dilute the cure rates down to below the aimed for edge of 85%. As a consequence, physicians are systematically encouraged to be very restrictive with diagnosing patients with smear/test negativ TB and empiric treatment for TB is still not appreciated.

The more diagnostic tools become available, the more ambitious physicians become, in the sense of “Now that we have the GeneXpert and culture, we can make a proper diagnosis and do not need empiric therapy any more…”, which can turn into a problem especially when it comes to TB diagnosis in HIV co-infected children, where sensitivity of all available confirmatory tests is still way too low. And even if sensitivity of TB-culture in adult patients is quite good, it takes too long for results to become available. In a small study we carried out in NW Cameroon out of about 180 patients with smear negative and culture positive TB, 16 (9%) were already dead at the time the liquid culture results became available.

Thanks a lot for initiating this important discussion!

Dylan Tierney Panelist Replied at 5:19 PM, 18 May 2016

Florian, I can imagine that the perverse incentive to avoid empiric treatment that you describe in Cameroon is also present in other places.

It seems like we have to find a way to adjust our monitoring and evaluation of TB program success to account for the possibility of empiric treatment. Does the community know of any countries that track empiric therapy in a way that doesn’t jeopardize the overall treatment success rate?

Dylan Tierney Panelist Replied at 5:19 PM, 18 May 2016

There are clearly important challenges to delivering high quality care related to the difficulty of diagnosing TB in HIV coinfection. But to quote the great Martin H. Fischer (1879–1962), diagnosis is not the end, but the beginning of practice.

As we all know, there are heavy pill burdens associated with concurrent HIV and TB treatment. This burden, compounded by drug-drug interactions and their side effects, often makes it difficult for patients to take all of their medications all of the time.

I’m very interested to learn from our other experts and community members about best practices to manage these issues.

Nicaise NSABIMANA Replied at 12:42 AM, 19 May 2016

Thanks Dylan for this discussion, how important it is! Co-infection TB and
HIV still to be really a serious public health problem by causing morbidity
and mortality in low income countries.
Most of hospital even referral or teaching hospital use to treat
empirically TB without evidence of acid fast bacilli. From my study "Role
of bronchoscopy in diagnosis suspected Pulmonary TB if sputum smear is two
time normal or with dry cough". From that study really bronchoscopy is a
good procedure to allow you getting a good specimen means material on which
you need to isolate acid fast bacilli. It should be LBA (lavage
bronchoalveolar, biopsy, etc).Sputum from HIV is often negative but LBA
showed to be efficient when sputum is negative or when patients have dry
cough. With these materials, the use of geneXpert on LBA or sputum after
bronchoscopy showed more important. Therefore it is used to detect also
resistance drug. First with rifampicin with geneXpert but with others drugs
when we cultured. Otherwise results showed that pneumonia, malignancy
tumor, fungus were treated with antiTB empirically by suspecting TB on
background of HIV. To be effective in diagnosis TB is the best way to avoid
MDR-TB in HIV patients. Healthy workers will be more effective by
motivation, a good number, and by a very good training. Hygiene is also an
Other important component in better management. Combined early diagnosis of
TB in HIV, good Hygiene, use of bronchoscopy, good training of all health
care provider and community health workers will help to improve burden
caused by couple TB-HIV.

migambi patrick Replied at 3:18 AM, 19 May 2016

Thx to all for the comment and sharing experience. As many of members said that majority of physician have fear to start empiric treatment is true but the issue is the that smear microscopy is less sensitive and specific. Based on the experience of TB prevalence survey from many countries, it shown that majority of patients with TB has not cough and they were diagnose because they using systematic chest x-ray.
In Rwanda we conducted active case finding among PLHIV by using symptoms and systematic chest x-ray as screening tool. TB presumptive was either PLHIV with one of the five clinical symptoms and/or abnormal chest x-ray and smear microscopy and genexpert were used for diagnostic to all presumptive. we screened 11091 PLHIV and 16% of them was TB presumptive(1824) but 65% of TB presumptive had abnormal chest ray only without symptom, we identified 92 TB cases and 65 TB cases were smear negative and xpert positive. The interesting finds was that among those 65 TB cases, 35 were presumptive by chest x-ray only and 30 had both abnormal chest x-ray and symptom.
Lesson learnt was for the proper diagnostic of TB among HIV we could use chest x-ray as screening tool to increase number of presumptive and sensitive diagnostic tool such genexpert, culture if possible etc.... now the question is non availability of chest x-ray in the low and middle countries which can be a challenge. Need for advocacy to avail sensitive tools to diagnose TB among PLHV

Naphtal Nyirimanzi Replied at 4:27 AM, 19 May 2016

Hello Members,

I think this is a good point from Rwanda.
However, I wonder if this can be systematically applied to PLHIV(we may overexpose them).
The other issue is for children: you know that it is difficult to diagnose TB in children. Do you think this method can apply to the children?
What do you think about the people living in overclouded areas like prisoners?

Thank you.

Andrew Black Replied at 7:16 AM, 19 May 2016

Thank you for this topic.
Where I work in South Africa we have a very low threshold to start empiric TB therapy in HIV infected persons especially when the CD4 cell count is low. Despite our low threshold (clinical symptoms suggestive of TB, no other diagnosis made that may explain the symptoms and failure to respond to course of amoxicillin) post mortem studies still reveal unsuspected TB to be responsible for a significant number of patient deaths.
There was initial resistance to empiric treatment from the TB programme but in our setting of high HIV and TB burden empiric treatment is no discouraged and rather than focusing on just cure more attention is paid to treatment completed (we do not distinguish between microbiologically confirmed and empiric treatment in our reporting).
The extent to which the diagnosis is pursued is often dependant on the level of care but a CXR prior to empiric treatment is encouraged. Where available abdominal sonars looking for splenic micro abscesses, free fluid, lymph nodes and caecal thickening is helpful.
The largest problem with a low threshold and acceptability of empiric therapy is it may be used as an easy way out and sufficient effort is not put into trying to make the diagnosis. Despite having access to rapid molecular diagnostics a large proportion of our patients notified as PTB and started on empiric therapy never have a sputum sent to the lab as collecting an early morning specimen after the patient has been encouraged to try deep cyclic breathing or inducing sputum is a hassle.
The importance of a confirmed diagnosis becomes apparent when the patient develops TB drug induced adverse events or fails to respord to the empiric treatment.
I firmly encourage all clinicians who start a person on empiric therapy to take responsibility for that patient and to follow them up closely to ensure they respond to the therapy and if they do not respond to investigate further. If the patient can not be followed up personally comprehensive notes on why treatment was started, base line symptoms and clinical findings (including CXR and blood results) must be sent to the person who will follow up the patient to allow for monitoring.
In our setting the majority of the patients started on empiric therapy probably do have tuberculosis but the small number who have other conditions such as fungal infections or lymphoma need to be identified early on in the course of their empiric TB treatment.
Once again thank you for the topic an the opportunity to share.
Andrew Black

Dylan Tierney Panelist Replied at 10:49 AM, 19 May 2016

The insightful comments from Nicaise, Patrick and Andrew have raised key points that I want to explore a bit further.

1. Bronchoscopy is an indispensable tool in the evaluation of the patient with HIV and a pulmonary syndrome. I definitely consider it a cornerstone of high quality care for these patients. When paired with airway sampling through lavage or biopsy, it can be used query a number of entities in the differential diagnosis, including infectious etiologies like TB, pneumocystis and other fungal pneumonias but can also noninfectious causes like pulmonary Kaposi’s sarcoma. The procedure, however, requires equipment and trained operators. I would be curious to know the percentage of patients with HIV and TB who have access around the world.

2. The chest x-ray is similar to bronchoscopy in that it can be used to evaluate a number of causes of pulmonary disease for patients with HIV. I agree that every patient with HIV should undergo a chest x-ray at the time of diagnosis. But also similar to bronchoscopy, the chest x-ray requires equipment and trained operators (i.e., radiologists). The advent of automated digital reading of the images has the potential to reduce the need for trained operators, thereby increasing the potential for each patient with HIV to be screened. (See link.)

3. A clinical (i.e., test-negative) diagnosis of TB should not be considered a soft diagnosis but rather a provisional diagnosis. I fully agree that every effort should be made to achieve microbiologic confirmation. (It should be noted that the microbiologic diagnosis becomes even more difficult once effective therapy is started because the anti-TB drugs begin decreasing the bacterial burden, making less bugs available to be detected.) A broad differential diagnosis should be kept in the fore of the clinicians mind as they move forward with treatment. If no other diagnosis can be made, the commitment to the TB diagnosis and treatment should be strong.

With any of these obstacles to providing high quality care, we can and should try to trace the problem back to a root cause. Why don’t patients at my hospital have access to bronchoscopy? Why are my patients with HIV not getting a screening chest x-ray? Why was my patient’s empiric TB treatment abandoned? Answers to these questions should stimulate follow up questions, each time going further towards the root. And to improve the quality of care for our patients with HIV and TB, some of the questions should stimulate us to undertake small and achievable actions to address the root causes we are uncovering.

Dylan Tierney Panelist Replied at 10:51 AM, 19 May 2016

Here's the link the article on automated chest x-ray.

Attached resource:

Nicaise NSABIMANA Replied at 12:25 PM, 19 May 2016

Sputum two times negative sorry for mistake by saying two times normal

Nicaise NSABIMANA Replied at 12:29 PM, 19 May 2016

Sputum two times negative or patients with dry cough need bronchoscopy.
Sorry for mistake by saying sputum two times normal

Riitta Dlodlo Panelist Replied at 3:36 PM, 19 May 2016

Thank you, colleagues, for this lively discussion.

Dylan has posed an interesting question about pill burden that PLHIV with TB face. Indeed, the number of medicines, including also cotrimoxazole, to be taken daily is high. At the same time, patients seem to be coping. Or am I missing something? Perhaps, I do not see or look for patients who do not turn up on their follow up dates - some may eventually become 'loss to follow up' because we, health workers, failed to address treatment interruption in time. Establishing good rapport with patients and their treatment buddies, providing adequate information about drug side effects and drug-drug interactions and giving concrete guidance what to do if a symptom develops are key in preventing treatment interruption.

What do you think? What is your experience? Who may have conducted operational and other research in this area?


Phumzile Mndzebele Replied at 4:24 AM, 20 May 2016

I am happy to be part of this discussion. I would appreciate if the pannel can further discuss the IPT to prevent the latent TB.

In Swaziland IPT coverage remains very low , HCWsare sceptical to prescribe INH due to the catastrophic side effects.

Please guide on issues of chemistry monitoring and EARLY WARNING INDICATORS.


Serena Koenig Panelist Replied at 11:06 PM, 20 May 2016

Hi Phumzile,
In Haiti, we just completed a study to evaluate same-day HIV testing and ART initiation. We included patients who did not have cough, and we conducted a CXR prior to enrollment. All patients received IPT. We have no toxicity from IPT, and only 1% of patients developed active TB within the subsequent 3 months, so we effectively ruled out TB. We also do this in our regular non-research clinics. We are happy to provide further advice on this if it is useful to you. Our goal is to have 100% compliance with IPT to reduce the incidence of TB.
Serena Koenig

Erica Lessem Replied at 9:33 AM, 22 May 2016

Serena, thanks so much for sharing your positive experience using IPT.

Phumzile, the risk of developing active TB far outweigh INH side effects.

The WHO's Guidelines on the Management of Latent Tuberculosis Infection
cover adverse event monitoring in Section 2.1 (page 22), available here:

Note that most adverse drug reactions are minor and occur rarely.

This is what they recommend for adverse events monitoring:
"Individuals who receive treatment for LTBI do not have active disease, and
therefore, it is mandatory to minimize risks during treatment.
Drug-specific adverse reactions can occur with isoniazid (asymptomatic
elevation of serum liver enzyme concentrations, peripheral neuropathy and
hepatotoxicity); rifampicin and rifapentine (cutaneous reactions,
hypersensitivity reactions, gastrointestinal intolerance and
hepatotoxicity). While most adverse drug reactions are minor and occur
rarely, the Panel noted that maximum attention should be paid to prevention
of drug-induced hepatotoxicity.

A systematic review was conducted to assess the best way to monitor and
manage hepatotoxicity and other adverse drug reactions, but no studies were
identified. A review of national guidelines (19–23) showed the presence of
consistent recommendations across the different guidelines based on expert
opinion, which were useful to inform the judgment of the Panel.

The Panel underlined the importance of routine regular clinical monitoring
of individuals receiving treatment for latent TB through a monthly visit to
health-care providers. The prescribing health-care provider should explain
the disease process and the rationale of the treatment and emphasize the
importance of completing the treatment. Those receiving treatment should be
educated to contact their health-care providers should they develop
symptoms, such as anorexia, nausea, vomiting, abdominal discomfort,
persistent fatigue or weakness, dark-coloured urine, pale stools or
jaundice. Whenever a health-care provider cannot be consulted at the onset
of these symptoms, treatment should be immediately stopped.

The Panel noted that there was insufficient evidence to support baseline
laboratory testing for measurements of serum aspartate aminotransferase,
alanine aminotransferase, and bilirubin. However, the Panel strongly
encourages baseline laboratory testing for individuals with the following
risk cofactors: history of liver disease; regular use of alcohol; chronic
liver disease; HIV infection; age more than 35 years; and pregnancy or the
immediate postpartum period (i.e., within three months of delivery). For
individuals with abnormal baseline test results, routine periodic
laboratory testing should be done."


Erica Lessem, MPH
Director, TB/HIV Project
Treatment Action Group
261 Fifth Avenue, Suite 2110
New York, NY 10016 USA
tel: +
mobile: +1.617.827.2461
fax: +

For the latest news in TB, check out

Please consider giving to TAG when you do your online shopping: make
purchases through Amazon via Click on the link to change
the charity and enter T A G Treatment Action Group, and select TAG; as long
as you enter via, this preference will be saved and 0.5%
of all your purchases will go to TAG.

maria zolfo Replied at 5:14 PM, 23 May 2016

many thanks for these important reflections, doubts, and recommendations.

I think we must also take into account the reality of the patients we serve. Unfortunately in many high burden countries the bulk of patients with undiagnosed TB are transmitting in the community, or attend overburdened nurse-run clinics.

Therefore I believe we need to
1) involve communities in TB diagnosis and follow-up (sputum collection, drug distribution points, reporting of outcomes...),
2) combine the low barrier for sputum collection with a performant diagnostic tool that can be used at decentralized level (Xpert if electricity is available and supply chains are functional)
and 3) assure referral to a clinician when a patient presents alarm signs (no response to AB and Xpert negative, dyspnoe, suspicion EPTB, suspicion DRTB...)

What are you experiences with combining strategies?

Greetings, Tom Decroo

Svetlana Kudryshova Replied at 1:28 PM, 31 May 2016

Целью лечения ВИЧ-инфекции является максимальное уменьшение количества вирусов в крови на максимально долгий срок. В настоящее время антивирусные препараты для лечения ВИЧ подразделяются 3 группы, и применяются в так называемых схемах высокоэффективной антиретровирусной терапии (ВЭАРТ), включающих комбинацию как минимум трех разных препаратов. Новейшие лекарства от ВИЧ позволяют повысить эффективность лечения, однако без тщательной заботы о собственном здоровье каждого отдельно взятого человека, даже самое лучшее лекарство от СПИДа не убережет от опасности. И, конечно, необходимо мощная поддержка иммунитета.

Masoud Dara, MD Replied at 3:00 PM, 31 May 2016

Thank you Svetlana for your input, indeed it is crucial to ensure HAART is available and patients receive a good care as you said.

In addition, with a wider access to ARVs across the world, clinicians shall be always conscious of the possibility of Immune Reconstitution Syndrome, mimicking worsening of the clinical conditions or lack of response to TB treatment.

This Expert Panel is Archived.

While this Expert Panel is no longer active, we invite you to review and recommend past replies and resources. Membership for this Expert Panel is closed, but we hope you'll join us in one of the many communities on GHDonline.

Panelists of Overcoming Barriers to High-Quality TB Treatment During HIV Co-infection and GHDonline staff