Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

The Impact of Antiretroviral Therapy on Mortality in HIV Positive People during Tuberculosis Treatment: A Systematic Review and Meta-Analysis

  • Anna Odone ,

    Affiliations TB Modelling Group, Centre for the Mathematical Modelling of Infectious Diseases, London School of Hygiene and Tropical Medicine, London, United Kingdom, Department of Global Health and Social Medicine, Harvard Medical School, Boston, Massachusetts, United States of America, University of Parma, School of Medicine, Parma, Italy

  • Silvia Amadasi,

    Affiliation University Division of Infectious and Tropical Diseases, University of Brescia and Spedali Civili General Hospital, Brescia, Italy

  • Richard G. White,

    Affiliations TB Modelling Group, Centre for the Mathematical Modelling of Infectious Diseases, London School of Hygiene and Tropical Medicine, London, United Kingdom, TB Centre, London School of Hygiene and Tropical Medicine, London, United Kingdom

  • Theodore Cohen,

    Affiliations Center for Communicable Disease Dynamics, Department of Epidemiology, Harvard School of Public Health, Boston, Massachusetts, United States of America, Division of Global Health Equity, Brigham and Women's Hospital, Boston, Massachusetts, United States of America

  • Alison D. Grant,

    Affiliations TB Centre, London School of Hygiene and Tropical Medicine, London, United Kingdom, Department of Clinical Research, London School of Hygiene and Tropical Medicine, London, United Kingdom

  • Rein M. G. J. Houben

    Affiliations TB Modelling Group, Centre for the Mathematical Modelling of Infectious Diseases, London School of Hygiene and Tropical Medicine, London, United Kingdom, TB Centre, London School of Hygiene and Tropical Medicine, London, United Kingdom

The Impact of Antiretroviral Therapy on Mortality in HIV Positive People during Tuberculosis Treatment: A Systematic Review and Meta-Analysis

  • Anna Odone, 
  • Silvia Amadasi, 
  • Richard G. White, 
  • Theodore Cohen, 
  • Alison D. Grant, 
  • Rein M. G. J. Houben



To quantify the impact of antiretroviral therapy (ART) on mortality in HIV-positive people during tuberculosis (TB) treatment.


We conducted a systematic literature review and meta-analysis. Studies published from 1996 through February 15, 2013, were identified by searching electronic resources (Pubmed and Embase) and conference books, manual searches of references, and expert consultation. Pooled estimates for the outcome of interest were acquired using random effects meta-analysis.


The study population included individuals receiving ART before or during TB treatment.

Main Outcome Measures

Main outcome measures were: (i) TB-case fatality ratio (CFR), defined as the proportion of individuals dying during TB treatment and, if mortality in HIV-positive people not on ART was also reported, (ii) the relative risk of death during TB treatment by ART status.


Twenty-one studies were included in the systematic review. Random effects pooled meta-analysis estimated the CFR between 8% and 14% (pooled estimate 11%). Among HIV-positive TB cases, those receiving ART had a reduction in mortality during TB treatment of between 44% and 71% (RR = 0.42, 95%CI: 0.29–0.56).


Starting ART before or during TB therapy reduces the risk of death during TB treatment by around three-fifths in clinical settings. National programmes should continue to expand coverage of ART for HIV positive in order to control the dual epidemic.


Infection with HIV dramatically increases individuals' risk of developing active tuberculosis (TB) disease, as well as the risk of death during TB treatment [1]. While 13% of all people with TB are estimated to be HIV positive, they account for approximately a quarter of TB deaths [2]. The African region, where 250,000 deaths occurred among HIV-positive TB cases in 2012, accounts for 75% of HIV-positive TB cases [2].

Globally, in 2012, 46% of notified TB cases had a documented HIV test result and the coverage of Antiretroviral therapy (ART) among TB patients known to be HIV-positive was estimated to be 57% [2]. ART reduces the impact of HIV on incident TB, as illustrated by Suthar et al. who showed that the risk of TB is reduced by 65% [3]. Also, results from various trials and observational studies have consistently shown a benefit of ART on TB outcomes, and indicated that ART should be initiated as early as possible in the course of a TB episode [4], [5], [6], [7], [8], [9], [10], [11]. However, to date, there has been no systematic overview of the magnitude of benefit of ART on TB mortality. A recent systemic review by Straetemans et al. on TB mortality stratified subjects by HIV status but did not explore the effect of ART [12].

TB mortality in HIV-positive people is a key parameter to describe current state and progress in TB care and control [2]. Ideally, TB mortality is estimated through direct measurements using vital registration systems or mortality surveys, but these are often unavailable in resource limited settings, where TB incidence is highest [2]. Alternatively, TB mortality can be estimated indirectly by calculating the case fatality ratio (CFR), which is defined as the proportion of people with incident TB that die as a result of this episode of disease [13]. TB-CFR has been defined by Maher et al. as the proportion of TB cases that die within a specified time [14]. In particular, Mukadi et al. defined TB-CFR as the proportion of tuberculosis patients that die during TB treatment [15], which is the definition applied by Straetemans et al. in their review [12].

In this paper we systematically review the literature to estimate the mortality during TB treatment among HIV-positive TB patients receiving ART as well as its value relative to HIV-positive TB patients not receiving ART.


The review's methods were defined in advance following the Prepared Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines [16].

Criteria for considering studies

Study population.

We included studies that described mortality among HIV positive patients receiving TB treatment who initiated combination ART before or during TB treatment. All sex and age groups were included.

Our focus was on estimating the effect of ART as it applied to TB patients from the general population, and we excluded cohorts that were limited to recurrent, extra-pulmonary or known drug-resistant TB patients only. We also excluded studies focusing on populations of patients not representative of the general population such as miners [17], prisoners [18], healthcare workers [19] or injecting drug users [20].

Study design and sample size.

Eligible study designs included clinical trials, prospective cohort, retrospective cohort and case-control studies. Literature reviews were screened to retrieve relevant primary data. Studies which reported mortality results on fewer than 50 TB patients receiving ART were also excluded.

Outcome measures.

We focused on deaths occurring during TB treatment, the main operational parameter used in the field by most TB control programmes. This outcome has the benefit of a clearly defined at-risk period which facilitates between-study comparisons as well as programmme monitoring and evaluation. Papers were therefore excluded if they did not report mortality within six to eight months after start of TB treatment or otherwise restrict patient follow-up to the end of treatment.

We considered two primary outcome measures. Firstly, the TB-CFR defined as percentage of deaths among the study population occurring during TB treatment. Secondly, if the paper also reported TB mortality in HIV-positive individuals not on ART, we recorded the estimated relative risk (odds ratio (OR), relative risk (RR) or hazard ratio (HR)) of death during TB treatment by ART status.

Search methods for identification of studies

Studies were identified by searching the electronic databases Medline and Embase. The strategy was first developed in Medline and then adapted for use in the other databases (Appendix A). Studies published in English from 1996 (when combination ART first became available) through July2014 were included. In addition, the abstract books of the world conferences on lung health of the International Union against Tuberculosis and Lung Diseases (IUATLD) were manually scanned for the period 2004–2013 and further studies were retrieved from reference lists of relevant articles and consultation with experts in the field.

Data collection and analysis

Data extraction.

Identified studies were independently reviewed for eligibility by two authors in a two-step process; a first screen was performed based on title and abstract while full texts were retrieved for the second screen. At both stages disagreements between reviewers were resolved by discussion.

Data were extracted by one author supervised by a second author using a standardised data extraction spreadsheet. The data extraction spreadsheet was piloted on 10 randomly selected papers and modified accordingly. Data extraction included: authors' names, year and country of publication, study design, study setting, study period, age of study participants, information on TB type, TB diagnosis and drug resistance, information on time of ART initiation, follow-up time, information on analysis performed and outcomes of interest. We defined TB treatment to be standardized (following the definition used by Straetemans [12]) if it was described to be: i) in line with WHO recommendations, ii) following national or governmental guidelines, iii) direct observed treatment strategy (DOTS) or iv) defined as ‘standard’ [12].


We performed descriptive analysis to report the characteristics of the included studies. Midpoint values for the age of the included cohorts was based on the reported mean or median and were pooled as weighted averages. With regard to the pre-specified outcomes, we would expect variability between studies, e.g. based on average CD4 count in the study population or general quality of local health systems. We therefore applied random effects analyses to acquire an estimate of the average effect of ART on TB mortality, rather than assuming a single true value in a fixed effects approach [21]. With regard to the pre-specified outcomes, heterogeneity was assessed using the I2 statistic and visual inspection of forest plots [21]. Depending on data availability, we planned to conduct sub-group analyses (where relevant and possible) by WHO region, age-group, time interval between TB treatment start and ART start, laboratory-confirmed TB diagnosis and median CD4 count at start of study. If unadjusted and adjusted outcomes were available, we recorded the adjusted estimate to reduce the risk of confounding.

Quality assessment.

The same two authors who performed data extraction independently assessed the quality of the included studies using the Newcastle-Ottawa Assessment Scale (NOS) [22]. Level of quality was not set as a criterion for exclusion. Disagreements by reviewers were resolved by consensus.


Identified studies

We identified 2,129 records by searching the selected databases and listing references of relevant articles. After removing duplicates, 1,825 articles were retrieved. Papers were screened and selected as illustrated in Figure 1, resulting in 21 studies that were included in the systematic review and meta-analysis.

Characteristics of included studies

The characteristics of the included studies are reported in Table 1. The majority of the studies were conducted in the African (n = 11, 52%) [14], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32] and South-East Asian (n = 7, 33%) [33], [34], [35], [36], [37], [38], [39] regions; two studies were conducted in South America [40], [41] and one in Europe [42]. Four out of seven studies in the South-East Asian region were conducted in Thailand [33], [35], [37], [38] while one used data from eighteen sites in the Asia-Pacific region [39]. Both South American studies were conducted in Brazil.

More than half of included studies (n = 13, 62%) described a retrospective cohort [23], [24], [25], [27], [34], [37], [40], [42], seven (33.3%) described prospective cohort studies [14], [26], [33], [35], [38], [41] and one was a clinical trial [36]. We excluded seven papers [5], [6], [7], [8], [9], [43], [44] reporting results from large trials that focused on time of ART initiation during TB treatment as TB mortality data was only reported at time points that lay beyond our pre-specified ‘during treatment’ period [6], [7], [43], [44], the outcome was death combined with AIDS-defining illness and mortality data could not be extrapolated [7] or focused only on tuberculous meningitis [9]. Study enrollment periods ranged from 4 months [30] to 8 years [40]. The majority of the studies were conducted predominantly in urban settings (n = 14, 66.7%) [12], [23], [24], [25], [26], [29], [30], [31], [32], [36], [40], [41], [42]. Table 2 shows the levels of quality assigned to each study. We determined that the majority of the studies were susceptible to mild cohort selection bias because of their retrospective study design [23], [24], [25], [27], [34], [37], [40], [42]. We did not assess ten eligible studies for comparability as they were descriptive studies not reporting effect estimates [14], [23], [25], [27], [28], [31], [36], [37], [39], [42]. Among the rest, we judged that five had moderate comparability, because confounding factors were not fully adjusted or they were not specified [24], [29], [30], [34], [35]. For most studies, the assessment of outcomes was done by medical record review which we regarded as susceptible to mild outcome bias (Table 2).

Table 2. Source of data, method of death ascertainment or confirmation and quality assessment of the included studies.

Characteristics of the study populations

86% (n = 18) studies reported on source of data and method of data collection; of these, data were mostly collected from medical records (n = 10, 56%) and TB registers (n = 5, 28%). Three studies explicitly reported on how death was ascertained or confirmed (Table 2) [38], [39], [45].

For most studies (n = 19, 90.5%) only a fraction of the total patient population included TB cases that started ART before or during TB treatment. For example, in Agodokpessi et al., of the total patient population, 259/1086 (24%) were HIV positive and of these 259, 85 (33%) initiated ART before or during TB treatment [23]. On average, 50% of the studies' study population (range: 12–100%) fitted our criteria (Table 1). The number of included patients per study ranged from 75 [33] to 21,851 [29], with a median of 191 subjects. Studies included predominantly young adults (age midpoint: 34. years, SD: ±1.9).

All but two [14], [25] studies reported on TB type: among them, the percentage of individuals with pulmonary TB ranged between 19.4% and 100% (median: 66%). Where reported, most subjects were new TB cases (median percentage: 90%, range: 67%–100%). Drug susceptibility profiles were reported by nine studies. Three studies only included drug-sensitive TB cases and in the remaining studies, the proportion with multi-drug resistance was low and ranged from 0.1%–7% (median: 1.5%; included in the review). Three studies reported that some TB patients did not receive standardized therapy [35], [38], [42]. Details on TB treatment regimens are available in Table S3. Descriptions of TB type, TB category and drug susceptibility in the majority of studies referred to the original studies' entire cohorts and not to the subsets of the study population selected for this analysis.

Eleven studies reported that a proportion of their patients initiated ART before TB treatment (between 18 and 66%, Table 1), but the actual ART start dates were not reported [24], [26], [28], [29], [30], [33], [34], [38], [39], [41], [42]. Where all patients were initiated on ART during TB treatment, the median start times occurred in month 2 or 3 after starting TB treatment [14], [25], [27], [32], [36], [37]. The variability in reporting between studies limited our ability to examine the mortality impact of different ART initiation times with respect to the start of TB treatment.

Data on CD4 count were very heterogeneous and incomplete and did not allow quantitative assessment of effect modification (Table S2). In six (29%) studies, data on CD4 count were not reported at all [24], [27], [30], [32], [34], [40]; in four (19%) studies they were not available for the subgroup of the study population included in the meta-analysis [14], [33], [38], [41]; and in five (24%) studies the percentage of the study population included in the meta-analysis for which some measures of CD4 count were available did not exceed 55% [23], [26], [28], [37], [42]. Median CD4 count was available for >98% of subjects included in the meta-analysis in four (19%) studies [31], [35], [36], [39] and ranged from 48 to 152 cells/mm3.

Follow-up time corresponded to TB treatment duration in 18 (75%) studies. In Sinha et al., follow-up time was set at six months after TB treatment start or at TB treatment completion, and in Kendon et al., it was set at six months from ART start (ART was started within 56 days from TB treatment start) [36].

One study reported outcomes which were limited to bacteriologically-confirmed (positive sputum smear or culture result) TB cases [27], [35] and one study included only culture-confirmed TB cases [35].

Outcome measures

Sixteen studies reported data on CFR (Table 3) [14], [23], [25], [27], [28], [30], [31], [33], [34], [35], [36], [37], [38], [39], [41], [42]. Meta-analysis of the CFR showed high heterogeneity (I-squared  = 89.9%, p<0.001, Figure 2). The random-effects analysis suggested the CFR lay between 8% and 14% (point value  = 11%; Figure 2). The pooled CFR appeared slightly higher in the African region (11%–17%), and lower in the South-East Asian (7%–15%) region. The CFR was 4% (95%CI: 0%–7%) and 13% (95%CI: 6%–21%) in single studies set in the UK and Brazil respectively (Figure 2). Sanguanwongse et al. considered a subgroup of patients with very low CD4 counts at the start of TB treatment (CD4 <10 cells/mm3), which experienced a CFR of 21% (95%CI: 1%–32%) [35]. When restricted to the four studies where all patients were initiated on ART after starting TB treatment, the CFR was 12% (95% CI 8%–17%).

Figure 2. Forest plot of 16 studies reporting TB-CFR for HIV positive patients receiving ART (by Region).

Eleven out of 21 studies reported data on estimated relative risk of mortality comparing people taking vs. not taking ART [24], [26], [29], [30], [32], [33], [34], [35], [38], [40], [41]. Adjusted relative risks were available for all but one study [26]. A random effects meta-analysis showed that the relative risk of death during TB treatment by ART status was 0.42 with a 95%CI: 0.29–0.56; (I-squared  = 96.9%, p<0.001, Figure 3), which corresponds to a 44% to 71% reduction in mortality. When we restricted the analysis to studies considering only patients with smear positive TB (n = 2) the random effects relative risk was 0.11 (95%CI: 0.03–0.20) [35], [38].

Figure 3. Forest plot studies reporting the relative risk of death during TB treatment by ART status.


We estimate that mortality during TB treatment in HIV-positive individuals receiving ART under routine programmatic conditions lies between 8% and 14% and that ART reduces the mortality during TB treatment for HIV-positive TB cases by between 44 to 71%.

This is the first systematic assessment to quantify the impact of ART on TB mortality during TB treatment. In addition, we report not only pooled absolute mortality figures as done elsewhere [12] but also estimate the relative effect of ART on mortality, which has the advantage of being a less setting-specific and a more easily interpretable and generalizable parameter.

Our focus on deaths during TB treatment allowed us to make optimal use of the limited data, while providing a clear conceptual parameter that is widely used and reported by National TB Programmes, and which allows comparison between studies and other reviews. Mortality estimates over longer follow-up periods have been reported in recent trials [6], [7], [8], [43], but are not usually reported in surveillance systems, operational research and clinical settings. Previous reviews and reports have discussed the difficulties in trying to determine which deaths are directly attributable to TB, especially in low-resource settings [2], [12]. Different studies classified TB as either the primary or a contributory cause of death [12], [46]. In addition, according to the 10th revision of the International classification of diseases (ICD-10), deaths in HIV-positive TB patients are classified as HIV deaths [1]. In light of these considerations, we systematically assessed the effect of ART on death during TB treatment rather than deaths due to TB.

While the exclusion of several large RCTs reduced the number of studies and patients included, it was necessary to ensure between-study comparability regarding the time at risk of death. When reported, we note that these RCTs found relatively comparable TB-CFRs to our review [7]. We note that in the CAMELIA trial, the CFR was substantially higher than our estimate (18% and 27% for those initiating ART two or eight weeks into TB treatment), but this is likely to be due to the very low baseline CD4 count of patients included in that study (median  = 25 cells/mm3) and the longer follow-up time (median  = 25 months) [6]. As expected, other observational studies that considered follow-up periods greater than 3.5 years also find higher absolute mortality estimates [47].

The estimated range for CFR estimate of 8–14% in TB patients receiving ART is also below than the 19% point value reported by Straetemans et al. amongst all HIV-positive TB cases, which included some individuals receiving ART. This is consistent with expected effect of ART to reduce TB mortality [12], [27].

Sub-group analyses by geographical region suggested that the CFR was higher in the African region (14%, estimates based on 7 studies) and lower in the one study from a Western European setting (4%, in the UK), which mirror global TB mortality patterns [2]. The reported differences in CFR between regions might also be attributable to differences in immunosuppression at the time of clinical presentation which could affect the estimated benefits of ART. The protective effect of ART appeared stronger in people with smear-positive TB, reflecting the better treatment outcomes in this population across HIV/ART groups compared to patients without smear or culture confirmation. This could be due to some people who do not genuinely have TB being erroneously included in the smear negative group. However, given the small number of studies with bacteriological confirmation (n = 2), this result should be interpreted with caution.

A limitation of our work is that – due to a lack of data - we could not explore the effect of ART by CD4 count. The lack of data is due to the fact that the majority of the included studies reported data from TB registers, and surveillance systems in high HIV-TB burden and resource-limited settings where CD4 measurements are often unavailable [48]. Recent data from large RCTs have provided solid evidence that ART reduces mortality across a wide range of CD4 counts [5], [6], [7]. In addition, the 2013 WHO guidelines for antiretroviral therapy provide the evidence-based recommendation that in subjects with active tuberculosis ART should be initiated as soon as possible, irrespective of CD4 count [49]. However, we found that on average patients were started two to three months post TB treatment initiation, suggesting this may not have been policy or practice in the majority of observational studies included. If ART were provided to all HIV positive TB patients as per current guidelines, we would expect the CFR to decrease [8], [43], and to see an even greater reduction in relative mortality. The question of when to start ART in HIV-positive TB patients has been addressed by the above-cited RCTs [5], [6], [7], [8], [50] and - based on their findings - HIV and TB guidelines recommend that among HIV-positive TB patients with CD4 less than 50 cells/mm3, ART should be initiated within 2 weeks from TB treatment start and if CD4 above 50 cell/mm3, within 8 weeks [51], [52], [53], [54], [55]. Nonetheless, the issue remains controversial as results from a recent trial reported no difference in mortality between early and delayed ART for HIV-positive TB patients with CD4 counts of 220 cells/mm3 or more with authors arguing WHO guidelines should be updated accordingly [44] while, on the other hand, some researchers question the need of investing resources in other randomized, controlled trials on the same topic [56].

Nearly all data came from observational studies, which are more vulnerable to bias. For example, sicker patients might be more likely to be selected to start ART, thus introducing selection bias that would under-estimate the effect of ART. However, observational studies may better represent the patient population and clinical care provided in most settings. The majority of the studies considered single hospitals, local or district-level data and this might limit the ability to generalize results to national or regional level. Patients seen at referral hospitals may be sicker and may experience higher mortality, while at the same time, those treated in urban centres might have lower mortality than those treated in rural settings. Last but not least, most studies relied on TB notification registers and medical records whose degree of completeness and quality of reporting might be different in different settings, potentially introducing bias to pooled estimates. Similar to other reviews, we did not consider mortality amongst people lost to follow-up in our CFR estimates. Given that studies have suggested mortality among patients lost to follow up and transferred out averages 21% [57], our CFR is probably similarly underestimated [12]. As we did not consider studies focusing on special populations, our findings might not be generalizable to subpopulations with additional risk factors for mortality.

We found a high degree of heterogeneity between studies, which could be due to differences in clinical (e.g. CD4 count) or health system variables. We did not apply meta-regression because of the low number of studies, and the lack of information such as duration on ART and CD4 at start of TB treatment. By reporting results from random effects estimates, we acknowledge this heterogeneity. Due to the low number of studies contributing to the analysis, stratification by study quality was not possible. In addition, 95% of studies included in our systematic assessment have an observational study design which is subject to some risk of bias. However, since our aim was to describe the risk of death among HIV-infected TB patients under routine programmatic conditions, observational studies likely provide the most relevant data for this analysis.

In conclusion, we quantified the substantial impact of ART on reducing mortality during TB treatment. Collaborative tuberculosis-HIV activities are key components of the new Post-2015 Global Tuberculosis Strategy, approved by the World Health Assembly in May 2014 [58]. They include expanded collaboration between TB and HIV programmes and integrated tuberculosis and HIV service delivery in the field. For individuals newly diagnosed with TB, this means increased on-site HIV testing, and prompt referral for HIV care for all found to be HIV positive [59]. Improved harmonization relies on more sharing of clinical space and integration of medical records, staff, and training. These interventions promise to reduce delays to HIV diagnosis, facilitate early implementation of effective ART and reduce TB-related mortality in HIV positive patients.

Supporting Information

Table S2.

CD4 count at baseline and during follow up in the included studies.


Table S3.

TB treatment regimen in the included studies.


Author Contributions

Conceived and designed the experiments: AO RMGJH. Performed the experiments: AO SA RMGJH. Analyzed the data: AO SA RMGJH. Wrote the paper: AO SA RMGJH RGW TC ADG.


  1. 1. Corbett EL, Watt CJ, Walker N, Maher D, Williams BG, et al. (2003) The growing burden of tuberculosis: global trends and interactions with the HIV epidemic. Arch Intern Med 163: 1009–1021.
  2. 2. WHO (2013) Global Tuberculosis Report 2013. Geneva: World Health Organisation.
  3. 3. Suthar AB, Lawn SD, del Amo J, Getahun H, Dye C, et al. (2012) Antiretroviral therapy for prevention of tuberculosis in adults with HIV: a systematic review and meta-analysis. PLoS Med 9: e1001270.
  4. 4. Curran A, Falco V, Pahissa A, Ribera E (2012) Management of tuberculosis in HIV-infected patients. AIDS Rev 14: 231–246.
  5. 5. Abdool Karim SS, Naidoo K, Grobler A, Padayatchi N, Baxter C, et al. (2010) Timing of initiation of antiretroviral drugs during tuberculosis therapy. New England Journal of Medicine 362: 697–706.
  6. 6. Blanc FX, Sok T, Laureillard D, Borand L, Rekacewicz C, et al. (2011) Earlier versus later start of antiretroviral therapy in HIV-infected adults with tuberculosis. New England Journal of Medicine 365: 1471–1481.
  7. 7. Havlir DV, Kendall MA, Ive P, Kumwenda J, Swindells S, et al. (2011) Timing of antiretroviral therapy for HIV-1 infection and tuberculosis. N Engl J Med 365: 1482–1491.
  8. 8. Manosuthi W, Mankatitham W, Lueangniyomkul A, Thongyen S, Likanonsakul S, et al. (2012) Time to initiate antiretroviral therapy between 4 weeks and 12 weeks of tuberculosis treatment in HIV-infected patients: results from the TIME study. J Acquir Immune Defic Syndr 60: 377–383.
  9. 9. Torok ME, Yen NT, Chau TT, Mai NT, Phu NH, et al. (2011) Timing of initiation of antiretroviral therapy in human immunodeficiency virus (HIV)–associated tuberculous meningitis. Clin Infect Dis 52: 1374–1383.
  10. 10. Boulle A, Clayden P, Cohen K, Cohen T, Conradie F, et al. (2010) Prolonged deferral of antiretroviral therapy in the SAPIT trial: did we need a clinical trial to tell us that this would increase mortality? S Afr Med J 100: 566 –561.
  11. 11. Laureillard D, Marcy O, Madec Y, Chea S, Chan S, et al. (2013) Paradoxical tuberculosis-associated immune reconstitution inflammatory syndrome after early initiation of antiretroviral therapy in a randomized clinical trial. AIDS 27: 2577–2586.
  12. 12. Straetemans M, Glaziou P, Bierrenbach AL, Sismanidis C, van der Werf MJ (2011) Assessing tuberculosis case fatality ratio: A meta-analysis. PLoS One 6.
  13. 13. Glaziou P, Floyd K, Raviglione M (2009) Global burden and epidemiology of tuberculosis. Clin Chest Med 30: 621–636 vii.
  14. 14. Gandhi NR, Moll AP, Lalloo U, Pawinski R, Zeller K, et al. (2009) Successful integration of tuberculosis and HIV treatment in rural South Africa: the Sizonq'oba study. J Acquir Immune Defic Syndr 50: 37–43.
  15. 15. Mukadi YD, Maher D, Harries A (2001) Tuberculosis case fatality rates in high HIV prevalence populations in sub-Saharan Africa. AIDS 15: 143–152.
  16. 16. Moher D, Liberati A, Tetzlaff J, Altman DG (2010) Group P (2010) Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Int J Surg 8: 336–341.
  17. 17. Lim MS, Dowdeswell RJ, Murray J, Field N, Glynn JR, et al. (2012) The impact of HIV, an antiretroviral programme and tuberculosis on mortality in South African platinum miners, 1992–2010. PLoS One 7: e38598.
  18. 18. Davies NE, Karstaedt AS (2012) Antiretroviral outcomes in South African prisoners: a retrospective cohort analysis. PLoS One 7: e33309.
  19. 19. Casas EC, Decroo T, Mahoudo JA, Baltazar JM, Dores CD, et al. (2011) Burden and outcome of HIV infection and other morbidities in health care workers attending an Occupational Health Program at the Provincial Hospital of Tete, Mozambique. Trop Med Int Health 16: 1450–1456.
  20. 20. Altice FL, Kamarulzaman A, Soriano VV, Schechter M, Friedland GH (2010) Treatment of medical, psychiatric, and substance-use comorbidities in people infected with HIV who use drugs. Lancet 376: 367–387.
  21. 21. Higgins JP, Green S (2011) Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated March 2011], 5.1.0 ed. The Cochrane Collaboration.
  22. 22. Wells G (2011) The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. Available: Accessed 2014 Oct 23.
  23. 23. Agodokpessi G, Ade G, Ade S, Wachinou AP, Affolabi D, et al. (2012) Management of tuberculosis and HIV co-infection in Cotonou, Benin. Med Mal Infect 42: 561–566.
  24. 24. Henegar CE, Behets F, Vanden Driessche K, Tabala M, Bahati E, et al. (2012) Mortality among tuberculosis patients in the Democratic Republic of Congo. International Journal of Tuberculosis and Lung Disease 16: 1199–1204.
  25. 25. Kendon M, Knight SE, Ross A, Giddy J (2012) Timing of antiretroviral therapy initiation in adults with HIV-associated tuberculosis: Outcomes of therapy in an urban hospital in KwaZulu-Natal. South African Medical Journal 102: 931–935.
  26. 26. Nansera D, Bajunirwe F, Elyanu P, Asiimwe C, Amanyire G, et al. (2012) Mortality and loss to follow-up among tuberculosis and HIV co-infected patients in rural southwestern Uganda. International Journal of Tuberculosis and Lung Disease 16: 1371–1376.
  27. 27. Zachariah R, Fitzgerald M, Massaquoi M, Acabu A, Chilomo D, et al. (2007) Does antiretroviral treatment reduce case fatality among HIV-positive patients with tuberculosis in Malawi? Int J Tuberc Lung Dis 11: 848–853.
  28. 28. Ferroussier O, Dlodlo RA, Capo-Chichi D, Boillot F, Gninafon M, et al. (2013) Results of rapid and successful integration of HIV diagnosis and care into tuberculosis services in Benin. International Journal of Tuberculosis and Lung Disease 17 (11): 1405–1410.
  29. 29. Kaplan R, Caldwell J, Middelkoop K, Bekker LG, Wood R (2014) Impact of ART on TB Case Fatality Stratified by CD4 Count for HIV-Positive TB Patients in Cape Town, South Africa (2009–2011). J Acquir Immune Defic Syndr 66: 487–494.
  30. 30. Kayigamba FR, Bakker MI, Mugisha V, de Naeyer L, Gasana M, et al. (2013) Adherence to Tuberculosis Treatment, Sputum Smear Conversion and Mortality: A Retrospective Cohort Study in 48 Rwandan Clinics. PLoS ONE 8 (9).
  31. 31. Sileshi B, Deyessa N, Girma B, Melese M, Suarez P (2013) Predictors of mortality among TB-HIV Co-infected patients being treated for tuberculosis in Northwest Ethiopia: A retrospective cohort study. BMC Infectious Diseases 13 (1).
  32. 32. Tweya H, Feldacker C, Phiri S, Ben-Smith A, Fenner L, et al. (2013) Comparison of treatment outcomes of new smear-positive pulmonary tuberculosis patients by HIV and antiretroviral status in a TB/HIV clinic, Malawi. PLoS ONE 8 (2).
  33. 33. Akksilp S, Karnkawinpong O, Wattanaamornkiat W, Viriyakitja D, Monkongdee P, et al. (2007) Antiretroviral therapy during tuberculosis treatment and marked reduction in death rate of HIV-infected patients, Thailand. Emerg Infect Dis 13: 1001–1007.
  34. 34. Raizada N, Chauhan LS, Babu BS, Thakur R, Khera A, et al. (2009) Linking HIV-infected TB patients to cotrimoxazole prophylaxis and antiretroviral treatment in India. PLoS One 4.
  35. 35. Sanguanwongse N, Cain KP, Suriya P, Nateniyom S, Yamada N, et al. (2008) Antiretroviral therapy for HIV-infected tuberculosis patients saves lives but needs to be used more frequently in Thailand. J Acquir Immune Defic Syndr 48: 181–189.
  36. 36. Sinha S, Shekhar RC, Singh G, Shah N, Ahmad H, et al. (2012) Early versus delayed initiation of antiretroviral therapy for Indian HIV-Infected individuals with tuberculosis on antituberculosis treatment. BMC Infect Dis 12.
  37. 37. Tansuphasawadikul S, Saito W, Kim J, Phonrat B, Dhitavat J, et al. (2007) Outcomes in HIV-infected patients on antiretroviral therapy with tuberculosis. Southeast Asian J Trop Med Public Health 38: 1053–1060.
  38. 38. Varma JK, Nateniyom S, Akksilp S, Mankatittham W, Sirinak C, et al. (2009) HIV care and treatment factors associated with improved survival during TB treatment in Thailand: an observational study. BMC Infect Dis 9: 42.
  39. 39. Zhao H, Chen YMA, Lee C, Omar SFS, Phanuphak N, et al. (2014) Prognostic significance of the interval between the initiation of antiretroviral therapy and the initiation of anti-tuberculosis treatment in HIV/tuberculosis-coinfected patients: Results from the TREAT Asia HIV Observational Database. HIV Medicine 15 (2): 77–85.
  40. 40. Dos Santos APG, Pacheco AG, Staviack A, Golub JE, Chaisson RE, et al. (2013) Safety and effectiveness of HAART in tuberculosis-HIV co-infected patients in Brazil. International Journal of Tuberculosis and Lung Disease 17: 192–197+i.
  41. 41. Schmaltz CA, Sant'Anna FM, Neves SC, Velasque Lde S, Lourenco MC, et al. (2009) Influence of HIV infection on mortality in a cohort of patients treated for tuberculosis in the context of wide access to HAART, in Rio de Janeiro, Brazil. J Acquir Immune Defic Syndr 52: 623–628.
  42. 42. Dean GL, Edwards SG, Ives NJ, Matthews G, Fox EF, et al. (2002) Treatment of tuberculosis in HIV-infected persons in the era of highly active antiretroviral therapy. AIDS 16: 75–83.
  43. 43. Abdool Karim SS, Naidoo K, Grobler A, Padayatchi N, Baxter C, et al. (2010) Timing of initiation of antiretroviral drugs during tuberculosis therapy. N Engl J Med 362: 697–706.
  44. 44. Mfinanga SG, Kirenga BJ, Chanda DM, Mutayoba B, Mthiyane T, et al. (2014) Early versus delayed initiation of highly active antiretroviral therapy for HIV-positive adults with newly diagnosed pulmonary tuberculosis (TB-HAART): a prospective, international, randomised, placebo-controlled trial. Lancet Infect Dis 14: 563–571.
  45. 45. Sinha S, Shekhar RC, Singh G, Shah N, Ahmad H, et al. (2012) Early versus delayed initiation of antiretroviral therapy for Indian HIV-Infected individuals with tuberculosis on antituberculosis treatment. BMC Infect Dis 12: 168.
  46. 46. Walpola HC, Siskind V, Patel AM, Konstantinos A, Derhy P (2003) Tuberculosis-related deaths in Queensland, Australia, 1989-1998: characteristics and risk factors. Int J Tuberc Lung Dis 7: 742–750.
  47. 47. Dheda K, Lampe FC, Johnson MA, Lipman MC (2004) Outcome of HIV-associated tuberculosis in the era of highly active antiretroviral therapy. J Infect Dis 190: 1670–1676.
  48. 48. Lawn SD, Meintjes G, McIlleron H, Harries AD, Wood R (2013) Management of HIV-associated tuberculosis in resource-limited settings: A state-of-the-art review. BMC Medicine 11 (1).
  49. 49. Doherty M, Ford N, Vitoria M, Weiler G, Hirnschall G (2013) The 2013 WHO guidelines for antiretroviral therapy: Evidence-based recommendations to face new epidemic realities. Current Opinion in HIV and AIDS 8 (6): 528–534.
  50. 50. Abdool Karim SS, Naidoo K, Grobler A, Padayatchi N, Baxter C, et al. (2011) Integration of antiretroviral therapy with tuberculosis treatment. N Engl J Med 365: 1492–1501.
  51. 51. World Health Organization (2010) Treatment of tuberculosis: guidelines, 4th ed. 2010 (WHO/HTM/TB/2009.420) Available: Accessed August 14, 2014.
  52. 52. TB CARE I (2014) International Standards for Tuberculosis Care, Edition 3. TB CARE I, The Hague. Available: Accessed August 14, 2014.
  53. 53. Panel on Antiretroviral Guidelines for Adults and Adolescents (2013) Guidelines for the use of antiretroviral agents in HIV-1-infected adults and adolescents. Department of Health and Human Services. Available: Accessed August 14, 2014.
  54. 54. Thompson MA, Aberg JA, Hoy JF, Telenti A, Benson C, et al. (2012) Antiretroviral treatment of adult HIV infection: 2012 recommendations of the International Antiviral Society-USA panel. JAMA 308: 387–402.
  55. 55. World Health Organization (2013) Consolidated guidelines on the use of antiretroviral drugs for treating and preventing HIV infection, 2013. Available: Accessed August 14, 2014.
  56. 56. The Lancet Global Heath Blog (2013) Clinical trials and global health equity. Available: Accessed October 23, 2014.
  57. 57. Korenromp EL, Bierrenbach AL, Williams BG, Dye C (2009) The measurement and estimation of tuberculosis mortality. Int J Tuberc Lung Dis 13: 283–303.
  58. 58. The Stop TB Partnership (2014) WHA approves Post-2015 Global Strategy and Targets For Tuberculosis Prevention, Care and Control. Available: Accessed October 23, 2014.
  59. 59. Legido-Quigley H, Montgomery CM, Khan P, Atun R, Fakoya A, et al. (2013) Integrating tuberculosis and HIV services in low- and middle-income countries: a systematic review. Trop Med Int Health 18: 199–211.
  60. 60. Oxford Centre for Evidence-Based Medicine website. Available: Accessed October 23, 2014.