The utility of repeat Xpert MTB/RIF testing to diagnose tuberculosis in HIV-positive adults with initial negative result

Background: Amongst HIV-positive adults in South Africa with initial negative Xpert results, we compared the yield from repeating Xpert MTB/RIF (“Xpert”) on sputum to guideline-recommended investigation for tuberculosis (TB). Methods: A systematic sample of adults attending for HIV care were enrolled in a cohort exploring TB investigation pathways. This substudy was restricted to those at highest risk of TB (CD4<200 cells/mm 3 or unknown) who had a negative initial Xpert result. At attendance for the Xpert result, a repeat sputum sample was stored, and further investigations facilitated per national guidelines. Participants were reviewed monthly, with reinvestigation if indicated, for at least three months, when sputum and blood were cultured for mycobacteria, and the stored sputum tested using Xpert. We defined TB as “confirmed” if Xpert, line probe assay or Mycobacterium tuberculosis culture within six months of enrolment were positive, and “clinical” if TB treatment was started without microbiological confirmation. Results: Amongst 227 participants with an initial negative Xpert result (63% female, median age 37 years, median CD4 count 100 cells/mm 3), 28 (12%) participants had TB diagnosed during study follow-up (16 confirmed, 12 clinical); stored sputum tested positive on Xpert in 5/227 (2%). Amongst 27 participants who started TB treatment, the basis was bacteriological confirmation 11/27 (41%); compatible imaging 11/27 (41%); compatible symptoms 2/27 (7%); and unknown 3/27 (11%). Conclusions: Amongst HIV-positive individuals at high risk of active TB with a negative Xpert result, further investigation using appropriate diagnostic modalities is more likely to lead to TB treatment than immediately repeating sputum for Xpert. TB diagnostic tests with improved sensitivity are needed.


Introduction
Since 2011 the World Health Organization (WHO) has recommended Xpert MTB/RIF (Xpert; Cepheid, Sunnyvale, CA) as the initial diagnostic test for individuals being investigated for HIV-associated tuberculosis (TB) 1 .TB diagnosis in people living with HIV (PLHIV) is complicated by the high proportion who are smear-negative and/or have extrapulmonary disease 2 .Although Xpert has superior sensitivity to sputum microscopy, it is less sensitive than culture, with a pooled sensitivity of 61% for smear-negative, culture-positive TB among PLHIV 3 .
South Africa replaced smear microscopy with Xpert starting in 2011, for all individuals with symptoms suggesting TB 4 .Further evaluation of those who are HIV-positive and Xpertnegative comprises clinical reassessment, chest radiograph if available, sputum for mycobacterial culture, and treatment with antibiotic if clinically indicated 4 .In a South African study of 394 patients investigated for TB (irrespective of presence of symptoms) prior to antiretroviral therapy (ART) initiation, the sensitivity of Xpert for smear-negative, culture-positive TB increased from 43% to 62% when a second sample collected at the first visit was tested 5 .Mathematical modelling using a decision model from South Africa 6 suggested that replacing sputum culture with the cheaper option of a second Xpert would reduce loss to follow-up so 1% more patients would start TB treatment 7 , and save an estimated US$17.4 million per year 7 .This model assumed, based on limited data, the same sensitivity for the second Xpert test as for the first 7 , guidelines would be correctly followed 7 , and only 1% of those with TB symptoms start TB treatment based on a clinical diagnosis 6 .The strategy of sending a repeat Xpert for HIV-positive individuals whose initial Xpert result is negative has not been evaluated empirically.
The aim of our study was to describe the diagnostic yield from an immediate repeat Xpert on sputum, compared to sequential further investigation guided by South African recommendations, amongst HIV-positive adults being investigated for TB whose initial Xpert result is negative.

Methods
This "repeat Xpert" substudy was part of "Xpert for people attending HIV/AIDS care: test or review?"(XPHACTOR), a prospective cohort study evaluating a risk-based algorithm to prioritise Xpert testing amongst adults attending for routine HIV care in South Africa 8 .
XPHACTOR study population, recruitment and procedures XPHACTOR study flow, procedures and algorithm are described in detail in the supplementary material.In summary, we enrolled a systematic sample of adults (aged ≥18 years) attending four HIV clinics in Gauteng province, irrespective of presence of symptoms suggestive of TB, in the XPHACTOR study.Patients taking anti-tuberculosis treatment within the previous three months were excluded.Patients were enrolled into three groups: "on ART" (ART-experienced); "pre-ART" (in HIV care but not taking ART); and "HIV Testing and Counselling (HTC)" (newly-diagnosed HIV-positive).At the time of the study, ART eligibility comprised CD4 ≤350 cells/mm 3 or WHO clinical stage ≥3.Research staff screened participants for TB at monthly intervals to three months, using a standardised questionnaire which incorporated the WHO symptom screen (any one of current self-reported cough, fever, weight loss or night sweats, hereafter the WHO tool).A spot sputum sample was collected for Xpert for individuals at a priori highest risk of active TB according to the study algorithm, which prioritised testing for those with any of: current cough, fever ≥ 3 weeks, night sweats ≥ 4 weeks, BMI <18.5 kg/m 2 , CD4 <100 cells/mm 3 , or weight loss ≥10%; and at enrolment from all in HTC group or pre-ART with CD4 <200 cells/mm 3 (Supplementary material, Figure 1).At enrolment and follow-up visits, participants who submitted an Xpert sample were reviewed within one week, and if Xpert-positive, TB treatment was initiated.If Xpert was negative, research staff repeated WHO symptom screen and facilitated the Xpert-negative algorithm for all who were WHO tool positive, which comprised chest radiograph, spot sputum for TB culture, and/or antibiotic trial as clinically appropriate.The Xpert-negative algorithm was also facilitated, because of a priori high risk of active TB, for all pre-ART participants with CD4<200x10 6 /l who submitted sputum for immediate Xpert at enrolment to XPHACTOR.
At the three-month visit all participants had sputum and blood cultured for mycobacteria (Bactec MGIT 960 and 9240 systems).We allowed a broad window period around the three-month XPHACTOR main study final visit, until around six months, to maximise follow-up.
Study investigation results were returned to clinic staff, who were responsible for management decisions.Clinic records were reviewed at the end of the study to ascertain any additional relevant investigations (e.g.Xpert, mycobacterial culture and imaging requested as part of routine clinical care) and/or TB diagnoses.
Repeat Xpert substudy procedures XPHACTOR participants who were Xpert-negative with i) CD4 count<200 cells/mm 3 , or ii) new HIV diagnosis (HTC group) were eligible for this substudy, irrespective of presence of WHO

Amendments from Version 1
The manuscript has been revised in response to Reviewer comments regarding complicated presentation of the data and advice to exclude the LAM data.The major differences are 1) removal of the LAM data, 2) modification of Figure 1 to include information regarding the basis for TB treatment and additional yield of Xpert, 3) modification of the supplementary file to include a Figure 3 (new Figure 2 in supplementary material from the previous manuscript and clarification regarding the testing undertaken.
Any further responses from the reviewers can be found at the end of the article tool symptoms; these restrictions aimed to minimise unnecessary testing of individuals at lower risk of active TB.If a participant had more than one negative Xpert result during follow-up, only the first episode was included.
At attendance for Xpert result review, eligible participants were asked for an additional spot sputum sample for "repeat" Xpert, which was frozen at -80°C within 24 hours of collection.All stored samples were thawed and tested with Xpert at the end of the study to evaluate the diagnostic yield that could have been achieved if an immediate repeat Xpert had been sent at the Xpert result review visit.We decided a priori not to induce sputum for this substudy in order to reflect what would be achievable in routine practice.

Definitions
Repeat Xpert substudy entry and exit dates.Repeat Xpert substudy cohort entry date was defined as the date that the Xpert result review was conducted and sputum was collected for storage.Cohort exit date was defined as the last XPHACTOR study visit date.TB case definitions."Confirmed" TB was defined as a positive result on i) Xpert (on sputum sample) or ii) line probe assay (LPA) performed on smear-positive or cultured isolate (Geno-Type MTBDR plus, Hain Lifesciences) or iii) Mycobacterium tuberculosis (Mtb) culture, from any sample (including stored sputum and those requested by health care providers) collected within six months of XPHACTOR enrolment.Clinical TB was defined as TB treatment started within six months of enrolment ascertained from clinical records, self or family report, or reported in the context of a separate verbal autopsy sub-study, in the absence of microbiological confirmation.Six months was chosen because TB disease evolves gradually 9,10 ; data from Zimbabwe estimated the mean duration of smear-positivity prior to TB diagnosis amongst HIV-positive adults at 18-33 weeks 11 ."Not TB" was defined as absence of criteria for confirmed or clinical TB, and alive at least 3 months (the minimum follow-up period) after enrolment.Participants who did not fulfil the case definitions for TB or "not TB" were deemed to have unclassifiable outcome and excluded from analyses.
Pulmonary and extrapulmonary TB were classified in accordance with WHO definitions 12 .
We also summarise the basis on which treatment was initiated for those who started TB treatment.
Radiological definitions."Probable radiological TB" was defined as presence of any of cavitation, predominantly upper lobe infiltrates, pleural or pericardial effusion, or clear miliary picture on chest radiograph."Possible radiological TB" was defined as presence of any of lymphadenopathy (hilar or mediastinal), pulmonary nodules or other infiltrates.Participants with "probable" or "possible" radiological TB features, but without bacteriological confirmation, who started TB treatment within six months of substudy enrolment were assigned as having "clinical" TB.

Study outcomes
The co-primary outcomes were a) the proportion of participants with a positive repeat Xpert and b) the proportion of participants fulfilling study TB case definitions.

Statistical methods
Data were analysed using Stata 14 (Stata Corporation, College Station, TX, USA).
We did not undertake formal sample size calculation for this substudy as the sample size was all those eligible from the parent study.
We compared TB diagnoses made by Xpert using the sample stored at substudy enrolment, with all TB diagnoses fulfilling our case definitions during follow-up.We chose this pragmatic comparison because in real life, individuals with smear or Xpert-negative TB have sequential investigation, rather than all tests performed simultaneously.Our research staff facilitated the Xpert-negative algorithm when participants attended for Xpert result review, and therefore investigations are likely to have been initiated faster than in a routine setting.The proportion of TB diagnoses made by Xpert using the stored sputum was compared with TB diagnoses made during follow-up using McNemar's test.
In a sensitivity analysis restricted to participants who had at least one component of the Xpert-negative algorithm (chest radiograph, sputum for TB culture, or antibiotic trial) within a two-week window of providing the stored repeat Xpert sample, we compared the proportion of TB diagnoses made by Xpert using the stored sputum with TB diagnoses made by the Xpert-negative algorithm using McNemar's test.

Ethical approval
The study was approved by the ethics committees at the University of the Witwatersrand (approval # M120343), University of Cape Town (approval # 106/2012), and the London School of Hygiene & Tropical Medicine (approval # 6165).All consenting participants gave written consent or, witnessed verbal consent if unable to read or write.All ethics committees approved the consent form.Principles expressed in the Declaration of Helsinki were followed in the conduct of this research.

Results
Between September 2012 and March 2014, 235/410 (57.3%) potentially eligible participants were able to provide a sputum sample, stored for testing at study completion with Xpert (Figure 1).Eight participants with "unclassifiable" outcome were excluded, leaving 227 participants for analysis.

Participant characteristics
Characteristics of the 227 substudy participants and comparison with the 175 excluded because they were unable to produce sputum are presented in Table 1.The majority of participants were female (63%), median age was 37 years (interquartile range [IQR] 31,44), median CD4 count was 100 cells/mm 3 (IQR 51,147), and 26% had previously been treated for TB.78/227 (34%) of participants reported a TB symptom, most often cough  1 15/175 who were excluded because they did not produce sputum fulfilled case definitions for TB (clinical TB [9/15], confirmed TB [6/15]) and a further 3/175 had unclassifiable outcome 2 For 22/235 participants who provided more than one "repeat" sample (all Xpert-negative) only the result of the first sample and data from the associated review visit were used in the analysis 3 All had negative "repeat" Xpert. 4CXR features compatible with "Probable TB" (pleural effusion [4], cavitation and infiltrates [2], miliary TB [1] who had positive repeat Xpert); CXR features compatible with "Possible TB" (2); USS features compatible with TB (Pericardial effusion [1]; abdominal TB [1]). 5Xpert undertaken during study follow-up as part of routine clinical care or for research purposes in accordance with XPHACTOR study algorithm (Supplementary material, Figure 1). 6Samples collected during study follow-up as part of routine clinical care or for research purposes (sputum Mtb culture-positive [3] of whom one had positive repeat Xpert, blood Mtb culture-positive [1]). 7Two identified as having started TB treatment at verbal autopsy.One had M. xenopi identified in sputum culture prior to commencement of empiric TB treatment. 8Started on basis of compatible symptoms and positive sputum culture later identified as M. avium (1) and M. intracellulare (1); both had improvement in symptoms after treatment was initiated.
Tuberculosis diagnoses 12% (28/227) of substudy participants fulfilled case definitions for TB, of which 16 were confirmed and 12 were clinical (Table 2).One participant died before TB treatment could be commenced, and for one, the treatment start date was unknown.The remaining 26 started TB treatment at a median 49 days (IQR 0,108) after substudy entry.The range for time from substudy entry to earliest of positive TB investigation (including chest radiograph) or date TB treatment was started (amongst all fulfilling our case definitions for TB) was 0-118 days.24% (19/78) participants who were WHO tool positive when the sample for stored repeat Xpert was collected fulfilled TB case definitions (confirmed 11, clinical 8).

Basis for commencement of TB treatment (n=27).
Thirteen participants started treatment based on positive mycobacteriology.Amongst these, eleven started treatment because of a bacteriologically-confirmed TB result (Xpert [7, of which one was the stored repeat Xpert sample]; Mtb isolated from sputum [3] or blood [1]).A further two participants started treatment because of compatible symptoms and positive sputum mycobacterial culture (later identified as non-tuberculous mycobacteria [NTM]) with symptomatic improvement on standard TB treatment (Figure 1).
Eleven participants started TB treatment because of compatible imaging.Nine had compatible chest radiographs, of whom  Both pulmonary and extrapulmonary TB 3 2 (7%) Not recorded 3 (11%) LPA = line probe assay 1 Includes two participants for whom bacteriological confirmation was provided by stored sputum which was Xpert-positive; one of whom started treatment based on this result, and the other had already started TB treatment because of compatible chest radiograph (miliary TB). 2 Pleural effusion (3); positive mycobacterial blood culture (1); pericardial effusion (1) 3 Compatible abdominal ultrasound and sputum Mtb culture-positive (1), pleural effusion and sputum Xpert positive (1)   four were subsequently bacteriologically confirmed (Xpert 2, pleural fluid cultured isolate LPA-positive 1, positive Mtb sputum culture 1).Two participants started treatment based on ultrasound scans, one compatible with abdominal TB (subsequently confirmed by positive Mtb from sputum culture); and the other showing pericardial effusion.The basis for starting TB treatment was not clear for the remaining three participants (Figure 1).

Diagnoses made by repeat Xpert on stored sputum samples.
5/227 (2.2%) had a positive repeat Xpert.The sensitivity of repeat Xpert was 31.3% (5/16; 95% CI 11.0-58.7%)for the primary outcome of confirmed TB and 18% (5/28, 95% CI: 6.1%-36.9%)for the secondary outcome of confirmed / clinical TB combined) (Figure 2).In a matched analysis the odds of TB diagnosis was much greater by other modalities during follow-up than by the repeat Xpert, odds ratio 24.0 (95% CI: 3.9-986.9;p<0.0001,McNemar's test).Amongst the five participants with positive repeat Xpert, three were in the pre-ART group, and two in the on ART group.We were unable to undertake multivariable analysis to look at independent predictors of positive repeat Xpert on the stored sample because only five were positive.
In a sensitivity analysis restricted to 123 participants who had at least one component of the Xpert-negative algorithm within a two-week window of providing the stored repeat Xpert sample, 23 participants fulfilled our TB case definitions ( The participant who started TB treatment solely based on the stored repeat Xpert sample was in the pre-ART group with a CD4 cell count of 113 cells/mm 3 at substudy enrolment,  89 participants submitted sputum specimens for Xpert as part of routine care or because they fulfilled XPHACTOR algorithm criteria at monthly follow-up visits, for whom 6/89 (7%) were positive.An additional four participants had positive Xpert, of which three were stored sputum samples for repeat Xpert (bacterial confirmation provided solely by stored sample [2], also positive Mtb sputum culture [1]), and one was collected after the 3-month visit (Table 2).

Discussion
Among HIV-positive individuals at high risk of active TB, with a negative sputum Xpert result, very few TB diagnoses would have been made in this study by immediately repeating Xpert.We limited our study to HIV-positive individuals at highest risk of active TB, i.e. those who were WHO tool positive with CD4<200x10 6 /l, or pre-ART with CD4<200x10 6 /l, or newly diagnosed, of whom some were asymptomatic, to minimise unnecessary testing of individuals at lower risk of active TB.The low yield from repeat Xpert in those with negative initial Xpert is likely due to paucibacillary or subclinical disease which is more likely in asymptomatic individuals, or extrapulmonary disease.These TB diagnoses may be better identified by alternative diagnostic modalities, such as chest radiography.A South African study of patients with sputum screened for TB by Xpert and mycobacterial culture prior to ART initiation, using a gold standard of culture-confirmed TB (N=85), found that those who were Xpert-negative had higher CD4 cell counts and lower viral loads than those who were Xpert-positive 13 .We did not have enough positive stored Xpert results to undertake a similar analysis.
Our study illustrates the realities of implementing the test negative algorithm in HIV-positive individuals.Despite research staff facilitating the algorithm, less than half (100/227) of participants produced sputum for mycobacterial culture during follow-up (vs.100% assumed by Schnippel) 7 .We found sensitivity of the repeat Xpert was only 18% (5/28) for all TB or 31% (5/16) for bacteriologically-confirmed TB vs. 79% assumed by Schnippel 6,7 .Data from South Africa demonstrate poor adherence in routine care settings to TB diagnostic algorithms amongst HIV-positive individuals with initial negative Xpert test 14,15 .The aforementioned model 6,7 assumes 1% of patients with TB symptoms start TB treatment based on a clinical diagnosis, but we found this to be far greater; and the model does not consider extrapulmonary TB (one-fifth of our participants diagnosed with TB had only extrapulmonary disease).An economic evaluation of repeat sputum Xpert vs. the Xpertnegative algorithm for HIV-positive individuals using assumptions that are more realistic is needed.
Evaluation of the 2007 WHO algorithm for smear-negative TB (comprising chest radiograph, single sputum for mycobacterial culture, and antibiotic trial), in HIV-positive individuals being investigated for TB in Cambodia 16 , against a gold standard of culture-confirmed TB based on multiple specimens, demonstrated sensitivity of 60% 16 .Sensitivity of this algorithm is imperfect, and there is a risk of overtreatment when only clinical-radiological features are used to start TB treatment.40% (11/27) of our study participants who started TB treatment did so because of compatible imaging, of whom almost half were subsequently bacteriologically confirmed, highlighting its value to support rapid initiation of TB treatment.Our findings are in accord with data from the XTEND trial, which found that compatible chest radiograph was the main reason for initiating empiric TB treatment in a cohort of patients investigated for TB in primary care in South Africa, amongst whom microbiological confirmation was subsequently obtained for 13% 17 .South African national guidelines now recommend chest radiography for all individuals with symptoms suggestive of TB who cannot produce a sputum sample, but limited access to radiography facilities may limit implementation 4 .Amongst our study participants who provided sputum for mycobacterial culture prior to their 3-month visit there was a low yield of Mtb (3/100 [3%]), and the yield from further Xpert during follow-up was 7% (6/89), representing just over half (9/16) of all confirmed TB diagnoses.Our findings highlight the need for more sensitive diagnostic tests, and for repeating TB investigation using all available modalities, in HIV-positive individuals with initial negative sputum test result who remain symptomatic or have advanced immunosuppression.
Our study has some limitations.In the parent XPHACTOR study, Xpert testing was prioritised in people with BMI<18.5kg/m 2 or CD4<100, those newly diagnosed HIV-positive or pre-ART with CD4<200, as well as those with TB symptoms 8 .Thus the population in this substudy did not all have classic "TB symptoms" at the time of collection of either initial or repeat sputum samples for testing with Xpert.These individuals may have had paucibacillary or subclinical disease which would not have been detected by Xpert.However, TB prevalence in our substudy population was high, and we anticipate our results to be relevant at least to these high-risk groups.Our results are also relevant to efforts to improve TB case finding among asymptomatic HIV-positive people 18,19 .We froze all our raw sputum samples within 24 hours of collection, and all were thawed and tested within 6 months of collection, in line with other studies 20 .We assumed that all participants starting TB treatment or with a sample which was bacteriologically confirmed collected within six months of enrolment were likely to have had active TB at enrolment, regardless of whether it was diagnosable using sputum based tests at the time of enrolment.In fact, our study participants who started TB treatment commenced within a median of seven weeks from collection of the "repeat" Xpert sample.Some sputum samples for mycobacterial culture and chest radiographs were taken at an interval after participants returned for their initial Xpert test result, reflecting real-life investigation practice; we cannot be certain of the same result if they had been performed at the same time as sample collection for repeat Xpert.However, our findings suggest that following the Xpert-negative algorithm is more likely to lead to TB diagnosis than immediate repeat Xpert test.
Strengths of our study include systematic evaluation of participants and longitudinal follow-up which minimised the number of TB diagnoses missed, and the pragmatic nature of the study which reflected as far as possible real-life conditions, albeit with optimised implementation of TB diagnostic algorithms.

Conclusions
Amongst ambulatory HIV-positive individuals at high risk of active TB, if an initial Xpert is negative, the Xpert-negative pathway should be implemented and there should be a low threshold for investigating those who remain at high risk using all clinically appropriate diagnostic modalities.In addition, those for whom no TB diagnosis is made must be made aware of the importance of returning for review if symptoms persist or recur.Our findings do not support sending an immediate repeat Xpert and highlight the need for more sensitive diagnostic tests capable of detecting pulmonary and extrapulmonary TB.

Data availability
The XPHACTOR "Investigating TB if initial Xpert is negative" dataset, which includes data underlying this substudy, has been uploaded to the LSHTM Data Compass repository: https://doi.org/10.17037/DATA.284 21.
The reader will need to request the dataset from LSHTM (request access is provided within the data record) with a brief summary of how the dataset will be utilised.On request, a data sharing agreement will be made available which will first need to be signed, prior to provision of the dataset.This enables LSHTM to confirm that the reader is using the data for HIV or TB-related research, which is required because study participants consented to use of their data for HIV or TB-related research only.
The data is shared under a Data Sharing Agreement license (see above).
The study team wish to avoid unnecessary barriers to access and will seek to respond to data requests as quickly as possible.

Is the work clearly and accurately presented and does it cite the current literature? Yes
Is the study design appropriate and is the work technically sound?Partly

Are sufficient details of methods and analysis provided to allow replication by others? Partly
If applicable, is the statistical analysis and its interpretation appropriate?I cannot comment.A qualified statistician is required.

Are the conclusions drawn adequately supported by the results? Yes
Competing Interests: No competing interests were disclosed.

If applicable, is the statistical analysis and its interpretation appropriate? Yes
Are all the source data underlying the results available to ensure full reproducibility?Yes

Are the conclusions drawn adequately supported by the results? Yes
Competing Interests: No competing interests were disclosed.

Introduction:
Aim is awkwardly stated.Possibly reword with study population at the end, and try to make it more succinct.
○ LAM needs to be included as an aim from the start.Otherwise it feels like a very tangential add on.

○
Methods: I would like a formal definition of the primary outcome and any secondary outcomes.They are probably in there, but not explicitly stated enough.
○ I think a flow diagram of testing would be helpful to flesh out the methods.

○
It's not clear to me where LAM fits in.An Xpert negative pt is re-screened for symptoms, has a chest x-ray and sputum for culture, and/or antibiotic trial is given.But then suddenly under statistical methods, you mention LAM for the first time.LAM is not part of the SA algorithm.We need to know when this was done, and even why it is relevant to this study.

Results:
Not sure why 1 person who started TB treatment >6 months after enrollment would be excluded?This gets back to my previous point that you need to explicitly state the primary outcome, so we can understand why you would have excluded these 7 individuals.Although the authors note that their population was not limited to symptomatic individuals as a potential weakness, they do not explicitly state that an asymptomatic population is likely to have paucibacillary or subclinical disease, for which Xpert would not have good sensitivity.The SA algorithm for TB investigation starts with a positive symptom screen, not merely for being HIV positive with a low CD4 despite having no symptoms.So these findings are not necessarily generalizeable to actual practice.
○ I would also note that the prevalence of clinical diagnosis is not necessarily generalizeable.Its unclear the level of care that was given to these participants-was this routine care available in a primary public health clinic, or is this specialized care as part of a research study?These details could help the reader evaluate how generalizeable the findings are.

Conclusions:
The reference to the "Xpert negative pathway" is vague -authors should be specific about what kinds of followup are most useful.Authors should also include something about LAM.

Is the work clearly and accurately presented and does it cite the current literature? Yes
Is the study design appropriate and is the work technically sound?Yes

Are sufficient details of methods and analysis provided to allow replication by others? Partly
If applicable, is the statistical analysis and its interpretation appropriate?Yes Are all the source data underlying the results available to ensure full reproducibility?

Are the conclusions drawn adequately supported by the results? Yes
Competing Interests: No competing interests were disclosed.
I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.
I'm confused by the section on TB diagnosis.Those diagnosed by Xpert don't appear to be by the 2nd Xpert that was taken and stored, because lower down you detail those individuals.But a 2 nd Xpert is not part of the SA algorithm.Something needs to be done to make this clearer.
○ Response: Further Xpert investigation was performed during study follow-up both for research purposes in accordance with the XPHACTOR study algorithm (described in the methods and in the supplementary material) and also during routine clinical care.Therefore, participants may have submitted a number of sputum samples for Xpert during the course of the study, and a positive result from any sample was used to assign the TB case definition.This has been made clearer in the footnotes for new Figure 1.The main study methods in the supplementary material describe how we reviewed clinic records at the end of the study to ascertain additional relevant investigations (including TB investigation performed as part of routine clinical care) and/or TB diagnoses.
We have amended the methods section to clarify this.The text reads as follows: Investigation results were returned to clinic staff, who were responsible for management decisions.Clinic records were reviewed at the end of the study to ascertain any additional relevant investigations (e.g.Xpert, mycobacterial culture and imaging requested as part of routine clinical care) and/or TB diagnoses.

Response:
We have changed Figure 1 as per your recommendation.We think that the Venn diagram conveys the key study result at a glance and complements new Figure 1, particularly by showing graphically the overlap between individuals identified by repeat Xpert vs. the conventional algorithm.We would therefore prefer to retain it, if necessary in the supplementary material.
We have also amended the results section to clarify the basis for initiating TB treatment.The text reads as follows: Thirteen   Although the authors note that their population was not limited to symptomatic individuals as a potential weakness, they do not explicitly state that an asymptomatic population is likely to have paucibacillary or subclinical disease, for which Xpert would not have good sensitivity.The SA algorithm for TB investigation starts with a positive symptom screen, not merely for being HIV positive with a low CD4 despite having no symptoms.So, these findings are not necessarily generalizeable to actual practice.

○
Response: We do discuss the role of paucibacillary disease in the first paragraph of our discussion, and when describing study limitations but the text has been amended to include your recommendation to make a more explicit statement.We agree that current algorithms for TB start with symptom screening, but the need to identify active TB among asymptomatic HIV positive people is increasingly recognised, and our results have relevance to these efforts.We have added a sentence to the discussion to make this point.
The text reads as follows: We limited our study to HIV-positive individuals at highest risk of active TB, i.e. those who were WHO tool positive with CD4<200x10 6 /l, or pre-ART with CD4<200x10 6 /l, or newly diagnosed, of whom some were asymptomatic, to minimise unnecessary testing of individuals at lower risk of active TB.The low yield from repeat Xpert in those with negative initial Xpert is likely due to paucibacillary or subclinical disease which is more likely in asymptomatic individuals, or extrapulmonary disease.These TB diagnoses may be better identified by alternative diagnostic modalities, such as chest radiography.
Our study has some limitations.In the parent XPHACTOR study, Xpert testing was prioritised in people with BMI<18.5kg/m 2 or CD4<100, those newly diagnosed HIV-positive or pre-ART with CD4<200, as well as those with TB symptoms.Thus the population in this substudy did not all have classic "TB symptoms" at the time of collection of either initial or repeat sputum samples for testing with Xpert.These individuals may have had paucibacillary or subclinical disease which would not have been detected by Xpert.However, TB prevalence in our substudy population was high, and we anticipate our results to be relevant at least to these high-risk groups.Our results are also relevant to efforts to improve TB case finding among asymptomatic HIV-positive people.I would also note that the prevalence of clinical diagnosis is not necessarily generalizeable.Its unclear the level of care that was given to these participants-was this routine care available in a primary public health clinic, or is this specialized care as part of a research study?These details could help the reader evaluate how generalizeable the findings are.

○
Response: All decisions regarding initiation of TB treatment were made by the routine care provider, and patients attended their routine care appointments at which they may have had additional investigations (including Xpert and mycobacterial culture) arranged by the clinic.We reviewed clinic records at the end of the study to obtain information regarding additional TB investigations and TB diagnoses made in routine care.This has been made clearer in the methods section as detailed above.
As part of the research study additional Xpert testing was undertaken in accordance with the XPHACTOR study algorithm, and at the 3-month visit blood and sputum samples were taken for mycobacterial culture.As already described in the methods section our research staff facilitated the Xpert-negative algorithm when participants attended for Xpert result review, which should have been undertaken routinely for symptomatic individuals.Therefore these investigations are likely to have been initiated faster than in a routine setting.The research study is likely to identified both more confirmed TB and clinical TB diagnoses.We do refer to our optimised implementation of TB diagnostic algorithms when discussing the strengths of the study which should enable the reader to evaluate how generalizeable the findings are.Amongst ambulatory HIV-positive individuals at high risk of active TB, if an initial Xpert is negative, the Xpert-negative pathway (chest radiograph, mycobacterial culture on sputum, if appropriate antibiotic trial) should be implemented and there should be a low threshold for investigating those who remain at high risk using all clinically appropriate diagnostic modalities.

Figure 1 .
Figure 1.Flow chart of repeat Xpert substudy participants, basis for initiation of TB treatment, and yield from repeat Xpert MTB/RIF.C+ = culture-positive; CXR = chest radiograph; MTB = mycobacterium tuberculosis; NTM = non-tuberculous mycobacteria; USS = ultrasound scan; Xpert+ = sputum Xpert MTB/RIF positive.1 15/175 who were excluded because they did not produce sputum fulfilled case definitions for TB (clinical TB[9/15], confirmed TB[6/15]) and a further 3/175 had unclassifiable outcome 2 For 22/235 participants who provided more than one "repeat" sample (all Xpert-negative) only the result of the first sample and data from the associated review visit were used in the analysis3 All had negative "repeat" Xpert.4CXR features compatible with "Probable TB" (pleural effusion[4], cavitation and infiltrates[2], miliary TB[1] who had positive repeat Xpert); CXR features compatible with "Possible TB" (2); USS features compatible with TB (Pericardial effusion[1]; abdominal TB[1]).5 Xpert undertaken during study follow-up as part of routine clinical care or for research purposes in accordance with XPHACTOR study algorithm (Supplementary material, Figure1).6Samples collected during study follow-up as part of routine clinical care or for research purposes (sputum Mtb culture-positive[3] of whom one had positive repeat Xpert, blood Mtb culture-positive[1]).7 Two identified as having started TB treatment at verbal autopsy.One had M. xenopi identified in sputum culture prior to commencement of empiric TB treatment.8Started on basis of compatible symptoms and positive sputum culture later identified as M. avium (1) and M. intracellulare (1); both had improvement in symptoms after treatment was initiated.

Figure 2 .
Figure 2. Number of participants diagnosed with TB by "repeat" Xpert vs. number diagnosed during follow-up (N=28).CXR = chest radiograph; USS = ultrasound scan 1 CXR features compatible with "Probable TB" (pleural effusion[4], cavitation and infiltrates[2]); CXR features compatible with "Possible TB" (2); USS features compatible with TB (Pericardial effusion[1]; abdominal TB[1])2 One participant died before treatment commenced3 Two identified as having started TB treatment at verbal autopsy.One had M. xenopi identified in sputum culture prior to commencement of empiric TB treatment.4Started on basis of compatible symptoms and positive sputum culture later identified as M. avium (1) and M. intracellulare (1); both had improvement in symptoms after treatment was initiated.
participants started treatment based on positive mycobacteriology.Amongst these, eleven started treatment because of a bacteriologically-confirmed TB result (Xpert [7, of which one was the stored repeat Xpert sample]; Mtb isolated from sputum [3] or blood [1]).A further two participants started treatment because of compatible symptoms and positive sputum mycobacterial culture (later identified as nontuberculous mycobacteria [NTM]) with symptomatic improvement on standard TB treatment (Figure 1).

Figure 3
is totally confusing because people can be in more than one of the top boxes...I'm not sure what this adds, particularly how it speaks to your primary outcome.

○Response:
Although it does not speak to our primary outcome Figure3describes the volume and detail of investigation for TB that study participants underwent.As participants had multiple investigations they were included in more than one of the top boxes.This figure has been moved to the online supplement (Supplementary material, Figure2) and renamed "Investigations performed for participants from repeat Xpert substudy entry to 3month visit".******************** 5. Discussion: ******************** 6. Conclusions:The reference to the "Xpert negative pathway" is vague -authors should be specific about what kinds of followup are most useful.Authors should also include something about LAM.○Response:We have removed the LAM data from this manuscript.We have amended the text as per your recommendation.The text reads as follows:

Table 2 . Basis for TB diagnoses in repeat Xpert substudy (N=28).
What i'd like to know is what is the additional yield of Xpert, but I can't tell that from this figure.The way it looks is that all the people who had confirmed TB ○ were not diagnosed with clinical TB.However, I can't believe that is true.I would prefer to see who was diagnosed how up front (clinical, microbiological) and who wasn't ("not tb").Then from those boxes, show us how many in each were Xpert positive from the 2nd sputum.I'm confused by the section on TB diagnosis.Those diagnosed by Xpert don't appear to be by the 2nd Xpert that was taken and stored, because lower down you detail those individuals.But a 2nd Xpert is not part of the SA algorithm.Something needs to be done to make this clearer.Table 2 confuses me for the same reason that figure 1 confuses me. Figure 3 is totally confusing because people can be in more than one of the top boxes...I'm not sure what this adds, particularly how it speaks to your primary outcome.

Table 2
confuses me for the same reason that figure 1 confuses me.○Response:Table2has been simplified to report how case definitions were fulfilled and the site of TB; the summary of the basis upon which TB treatment was initiated has been moved to new Figure1.If you change figure1as suggested above, then you don't need a venn diagram.That diagram conveys very little information.○