Tuberculosis (TB) remains an important public health problem with 8.6 million new cases and 1.3 million deaths in 2012.1 HIV has a profound impact on the TB epidemic, especially in Africa where an estimated 43% of people with TB are HIV infected.1
Despite the overwhelming evidence of reduced mortality in people with TB receiving antiretroviral therapy (ART), only 57% of HIV-infected TB patients initiated ART in 2012 and ART initiation is often delayed in patients with TB.1,2 Important obstacles to ART initiation include health-care provider concerns that concomitant treatment may result in increased risk of drug side effects and drug interactions, immune reconstitution inflammatory syndrome (IRIS), and poor adherence because of the high pill burden.3,4
Randomized clinical trials (RCTs) have convincingly demonstrated the benefit of early ART initiation in TB patients with CD4 count <50 cells per cubic millimeter. Questions remain about how to implement RCT findings into practice and the optimal ART timing for patients with CD4 counts >50 cells per cubic millimeter.5–7 Timely ART initiation in patients with TB requires integration of TB and HIV care, which has been challenging to achieve.8 TB treatment services are typically decentralized using a nurse-centered approach, whereas ART remains predominantly doctor centered and centralized.9 Although studies have demonstrated the benefit of task shifting for ART initiation,10,11 task shifting for integrated TB/HIV care has not yet been rigorously evaluated.
We performed a prospective study to evaluate nurse-centered, integrated TB/HIV treatment using a CD4-stratified ART timing strategy at primary care in the highly resource-scarce setting of Kinshasa, Democratic Republic of Congo (DRC).
We purposefully selected 5 clinics among the 65 primary care clinics providing TB treatment in Kinshasa. To comply with the legal requirement in the DRC that a physician initiates ART, we selected among the 24 clinics with a physician on-site at least once weekly.12 Among these, we selected clinics that routinely offered provider-initiated HIV testing and counseling for TB patients, had moderate to high patient volumes, and, to ensure local buy-in, clinics where the director was supportive of nurse-centered ART for TB patients. Primary care nurses with operational TB/HIV research experience were trained in general HIV care and initiation and monitoring of ART. Each study nurse was assigned to one clinic and worked together with the clinic TB nurse to provide integrated TB/HIV care.
Routine Care, Data Collection, and Study Procedures
Patients were diagnosed with TB and HIV by the routine clinic nurse using national DRC diagnostic algorithms. Persons presenting with symptoms suggestive of TB were assessed by the TB nurse using smear microscopy. Smear-negative or extrapulmonary TB was diagnosed by clinical assessment, repeat smear microscopy, and response to antibiotic treatment. All TB patients were offered provider-initiated HIV testing and counseling using 2 rapid tests. Culture for Mycobacterium tuberculosis, chest x-ray, and CD4 count were not routinely available.
Between August 2007 and November 2009, patients diagnosed with TB and HIV were offered participation in the Integrated Tuberculosis and Antiretroviral Treatment (ITART) study. After written consent, we collected sociodemographic and clinical data, performed a clinical examination and World Health Organization (WHO) clinical staging, and collected blood sample for CD4 count, full blood count, and alanine transferase. CD4 count was repeated before completion of TB treatment.
ART eligibility was based on clinical staging and CD4 count, in line with the 2006 WHO recommendations. ART was initiated at 1 month after the start of TB treatment if baseline CD4 count was below 100 cells per cubic millimeter and in patients with a clinical stage 4 illness other than extrapulmonary TB. ART was initiated 2 months after the start of TB treatment if CD4 count was between 100 and 350 cells per cubic millimeter. For patients with baseline CD4 count >350 cells per cubic millimeter, ART eligibility was reassessed at the end of TB treatment. When reagents for CD4 counts were not available at the national HIV reference laboratory (only laboratory with CD4 count capacity in Kinshasa during the study period), all TB patients were considered ART eligible.
The study nurse booked an on-site physician consultation at the time of the physician clinic visit scheduled closest to the proposed ART start date for every ART-eligible participant. The ART regimen consisted of stavudine, lamivudine, and efavirenz. Nevirapine was substituted by efavirenz for pregnant women and zidovudine by stavudine in cases of preexisting peripheral neuropathy. The presence of alanine transferase ≥3 times upper normal limit was considered a contraindication for ART initiation. At the ART initiation assessment visit, the study nurse and physician sequentially blinded from one another and completed a standardized ART eligibility and regimen selection form. Subsequently, the nurse and physician jointly made a final decision.
During the first 2 months of TB treatment, patients received their TB medication and morning ART dose under direct observation by the clinic TB nurse during weekdays and self-administered at home during weekends. All ART evening doses were self-administered at home. During the TB treatment continuation phase, patients collected TB and antiretroviral drugs from the TB clinic on a weekly and monthly basis, respectively. Patients who were more than 3 days late for a scheduled clinic visit were traced by phone or home visit.
At each study visit (weekly during intensive phase, monthly during continuation phase), the study nurse assessed the patient for new AIDS-defining illnesses, signs and symptoms suggestive of IRIS, adverse drug reactions, and adherence to medications by self-reported frequency of pill taking during the previous 4 days. Laboratory and radiographic examinations were only performed if clinically indicated. Severity of adverse events was classified using the Division of AIDS classification.13 Patients with grade 1 toxicities were managed by the study nurse, a physician consultation was booked for grade 2 and 3 toxicities, and the patient was referred for same-day physician assessment in case of grade 4 toxicity.
After completion of TB treatment, patients were referred to the nearest ART clinic for continuation of HIV care.
Data Analysis and Definitions
To assess the effectiveness of task shifting, integration of ART initiation into TB care at primary care clinics, and CD4 count–stratified ART initiation in increasing ART uptake and decreasing mortality, we compared results of the ITART cohort with a historical prospective cohort. This historical cohort consisted of patients starting TB treatment between January 2006 and May 2007 (pre-ITART) at 4 of the 5 clinics participating in the ITART study. Patients participating in the historical cohort received the same routine services as the ITART study cohort participants with regard to diagnosis and treatment of TB, provider-initiated HIV testing and counseling at the TB clinic, and initiation of co-trimoxazole prophylaxis in HIV-infected patients; but HIV-infected patients were referred to a centralized ART facility instead of receiving integrated ART.14
We defined the follow-up period for time-to-event analyses from study enrollment to TB treatment completion. Patients were censored at loss to follow-up, death, or 8 months after the start of TB treatment, the TB treatment duration for patients with a history of TB.
The complement of Kaplan–Meier survival curves was generated to estimate cumulative mortality during TB treatment stratified by baseline CD4 count category. Differences between groups were tested using the log-rank test.
Potential risk factors for mortality were assessed using Cox proportional hazard models and expressed as crude hazard ratio (HR) and adjusted HR (adj-HR) and their 95% confidence intervals (CIs). Adherence was categorized as high if the patient did not miss any clinic-based directly observed treatment visit and self-reported frequency of pill taking was “always” at all study visits; suboptimal if the patient self-reported frequency of pill taking was “never,” “sometimes,” or “most of the time” at any study visit; and default if the patient missed study visits during TB treatment. Patients with both extrapulmonary and pulmonary TB were categorized as having extrapulmonary TB. Anemia was defined as mild, moderate, and severe if hemoglobin (grams per deciliter) was 11.0–11.9, 8.0–10.9, and <8.0 for nonpregnant women; 10.0–10.9, 7.0–9.9, and <7.0 for pregnant women; and 1.10–12.9, 8.0–10.9, and <80 for men, respectively.15
Ethics and Protection of Human Subjects
The study was approved by the institutional review boards of the University of North Carolina at Chapel Hill and the University of Kinshasa. All patients provided written informed consent for study participation.
ITART Study Population
A total of 599 patients with TB and HIV were enrolled in the ITART study. Only 3 eligible patients refused participation. We sequentially excluded patients <18 years of age (n = 31), those enrolled >30 days after the start of TB treatment (n = 19), and those on ART at enrollment (n = 36).
Among the 513 patients included in the analysis, enrollment occurred after a median of 2 days [interquartile range (IQR), 0–5] of TB treatment. Median age was 38 years (IQR, 32–45), 59.1% were female, 45.8% were married, 70.2% completed secondary education, and 75.1% walked to the clinic (Table 1). Financial insecurity was high with 53.0% being unemployed and 92.2% not having enough money to meet their needs most of the time. One in 4 patients (25.7%) had a history of TB treatment. Median body mass index was 18.0 (IQR, 16.6–19.9), and 56.7% of patients were underweight (body mass index, <18.5). Overall, 85.0% of patients had anemia at baseline and 18.1% had severe anemia.
Participants were diagnosed with smear-positive pulmonary TB (39.7%), smear-negative TB (40.0%), and extrapulmonary TB (20.3%). Smear-negative pulmonary TB was more frequent as CD4 count category decreased (P < 0.0001), ranging from 24.5% among patients with CD4 count >350 to 53.9% among those with CD4 count 50–100 cells per cubic millimeter. Most common sites of extrapulmonary TB were the pleura (47.1%) and lymph nodes (34.6%); 42.3% (n = 44) also had pulmonary involvement. Six patients had documented multidrug-resistant TB.
Diagnosis of HIV infection was made at the time of TB diagnosis in most (90.5%) patients. Almost all (92.2%) had CD4 testing done within 2 weeks of study enrollment; baseline CD4 count was missing in 37 (7.2%) patients because of lack of reagents. Median baseline CD4 count was 165 cells per cubic millimeter (IQR, 82–307).
Historical Prospective Cohort Population
The median age (38.1 years), proportion female (62.1%), proportion smear negative (37.4%), and proportion extrapulmonary tuberculosis (23.4%) among HIV-infected patients in the historical cohort were similar to that in the ITART cohort.
ART Eligibility and ART Initiation
Of the 513 ITART patients, 419 (81.7%) were eligible for ART initiation during TB treatment. About one-third (27.7%, n = 142) were eligible for ART initiation at 1 month of TB treatment based on CD4 cell count <100 cells per cubic millimeter (n = 140) or WHO clinical stage 4 (n = 2, both Kaposi sarcoma). Another 37 were eligible for early ART because of lack of CD4 count within the first 21 days of enrollment. Almost half (46.8%, n = 240) were eligible for ART initiation at 2 months because of CD4 cell count between 100 and 350 cells per cubic millimeter.
Overall, 354 (69.0%) ITART participants initiated ART during TB treatment (Fig. 1). Among those ART eligible at month 1 of TB treatment, 136 (75.1%) participants started ART during TB treatment, after a median of 35 days (IQR, 29–49). The proportion starting ART was similar among those with CD4 cell count <50 and CD4 cell count 50–99 (78.2% vs 85.4%, P = 0.38). Median time to ART was also similar in both the groups (34 days). Among those ART eligible at month 2 of TB treatment, 206 (85.8%) participants started ART during TB treatment, after a median of 63 days (IQR, 57–74). Among the 94 patients with baseline CD4 count >350 cells per cubic millimeter, 65 (69.1%) were reassessed by CD4 count before the end of TB treatment. Of them, 22 (33.8%) participants were eligible to initiate ART based on CD4 count <350, of which 8 (36.4%) initiated ART after a median of 178 days (IQR, 107–200) after TB treatment initiation.
CD4 data count for patients in the historical cohort was not available. Among the 373 HIV-positive TB patients, only 16.6% initiated ART during TB treatment after referral to a centralized ART clinic.
Nurse and Physician ART Decision Making
Records of both nurse and physician ART decision making were available for 375 participants, including 349 (84.1%) patients ART eligible at baseline. Nurses and physicians were in agreement for almost all patients regarding the decision on ART initiation (98.1%) and the ART regimen to be initiated (95.3%).
Outcomes of Decentralized Integrated TB/HIV Care
Most participants had a successful TB treatment and ART outcome, but optimal adherence was more common for TB treatment than ART (84.6% vs 76.1%) (Table 2). Among those initiating ART, the median CD4 count increase was 101 cells per cubic millimeter (IQR, 46–167). Only 5 patients experienced a possible IRIS event, and 2 of these occurred in patients with CD4 count <50 cells per cubic millimeter. Although almost all (96.1%) reported at least one side effect of TB or ART medications, severe side effects were rare with grade 3 side effects in 23 (4.5%) patients and no grade 4 side effects.
Overall, 44 patients died during TB treatment after a median of 59 days (IQR, 39–102). Median baseline CD4 count among those who died was 108 cells per cubic millimeter (IQR, 43–239), and the majority of deaths (63.6%) occurred in patients not on ART. One patient died with drug-resistant TB. Cumulative mortality during TB treatment was strongly determined by CD4 count: 21.0% in CD4 stratum <50 cells per cubic millimeter, 16.0% in CD4 count 50–99, 4.7% in CD4 count 100–199, 6.7% in CD4 count 200–350, and 5.4% in CD4 stratum >350 cells per cubic millimeter (Fig. 2B). Cumulative mortality among those with baseline CD4 count <50 cells per cubic millimeter was higher than that in any other CD4 count stratum (P < 0.0001).
Among ITART participants, walking to clinic (adj-HR: 0.42, 95% CI: 0.21 to 0.86), higher CD4 count per 100 cells per cubic millimeter (adj-HR: 0.64, 95% CI: 0.50 to 0.82), and ART during TB treatment (adj-HR: 0.09, 95% CI: 0.04 to 0.19) were independently associated with lower mortality (Table 3). Female gender tended to be associated with lower mortality (HR: 0.56, 95% CI: 0.31 to 1.02) and extrapulmonary TB (HR: 1.74, 95% CI: 0.91 to 3.32) and being underweight (HR: 1.82, 95% CI: 0.95 to 3.47) with higher mortality in bivariate but not multivariate analysis.
Overall cumulative mortality during TB treatment among 513 HIV-infected ITART patients was lower than among 373 HIV-infected patients in the historical cohort (9.8% vs 20.1%, P < 0.0003) but higher than the mortality observed among 3577 HIV-negative patients in the historical cohort (9.8% vs 6.3%, P = 0.04) (Fig. 2A). During TB treatment, cumulative mortality among ITART patients with CD4 count >100 cells per cubic millimeter was similar to the mortality observed in HIV-negative TB patients (5.6% vs 6.3%, P = 0.65).
Our findings demonstrate that nurse-centered, integrated, CD4-stratified ART initiation at primary care level is effective in increasing ART uptake and reducing mortality but may not be adequate to prevent mortality among TB patients presenting with severe immunosuppression.
Uptake of ART among HIV-infected TB patients increased from 16.6% during the referral for ART model to 69% during the integrated model. High uptake rates have also been observed in other settings implementing TB/HIV integration.16,17 ART adherence was lower compared with TB treatment adherence (84.6% vs 76.1%), possibly because of the monthly (ART) vs weekly (TB) distribution of drugs during the continuation phase of TB treatment. Program harmonization, with ART and TB drugs distributed together, may further improve adherence in this population.
Primary care nurses correctly implemented the CD4-stratified timing of ART initiation for the majority of patients, with 76% of those eligible at month 1 of TB treatment starting at a median of 35 days and 86% of those eligible at month 2 starting ART at a median of 63 days. In contrast, ART initiation among those with CD4 count >350 cells per cubic millimeter at baseline was suboptimal, with only 69% being reassessed by CD4 count before the end of TB treatment and 36% of those eligible initiating ART.
Nurse decision making on ART initiation was validated by physicians in nearly all patients, supporting task shifting as a strategy to deliver integrated TB/HIV care at primary care. This is similar to findings in rural Uganda where nurses demonstrated strong agreement with physicians in deciding whether to initiate ART in patients without TB.18 Our results also expand the findings of RCTs and observational studies of task shifting in general HIV care clinics to TB patients at primary care level in extremely resource-limited settings.10,11,19–26
Mortality during TB treatment was relatively low in our study population and dropped from 20.1% during the referral for ART period to 9.8% during the fully integrated HIV/TB program. The overall 9.8% cumulative mortality during TB treatment was comparable with that reported at integrated clinics in Kenya (6.8%) and South Africa (8.0%–9%).16,17,27 Among ITART participants, walking to clinic (adj-HR: 0.42), higher CD4 count per 100 cells per cubic millimeter (adj-HR: 0.64), and ART during TB treatment (adj-HR: 0.09) were independently associated with lower mortality.
Mortality was disproportionately high (18.8%) in patients with CD4 count <100 cells per cubic millimeter, despite prioritizing this population for early ART initiation at 1 month after the start of TB treatment. Findings from RCTs suggest that starting ART in individuals with CD4 count <50 cells per cubic millimeter at 1 month of TB treatment may be too late.6,7 Our findings support the WHO recommendation that ART should be initiated as an emergency, within the first 2 weeks of TB treatment, in people with CD4 count <50.28 A recent study in South Africa found that most patients on ART in whom M. tuberculosis is detected on autopsy have additional comorbidities contributing to the cause of death, including bacterial infections, fungal infections, neoplasm, and noninfectious organ failure.29 Similarly, studies of cause of death among TB patients with low CD4 count in Thailand found that about 15% of deaths are caused by conditions other than TB, including bacterial infection, liver disease, and Stevens–Johnson syndrome.30,31 This suggests that the management of TB patients with severe immunosuppression may require resources and clinical expertise beyond those available at primary care. Future research is needed to determine whether ART initiation at primary care is effective for this population, whether immediate ART needs to be combined with additional interventions such as broad spectrum antibiotics other than or in addition to co-trimoxazole, or whether referral to higher level care settings is needed for all TB patients with severe immunosuppression presenting at primary care clinics in resource-limited settings.
Despite delayed initiation of ART in patients with CD4 count >100 cells per cubic millimeter, mortality was relatively low in this group and approached the mortality observed among HIV-negative patients with TB. This suggests that ART may not need to be initiated as soon as possible in this population. Initiation of ART at the start of the continuation phase of treatment for ambulatory patients with CD4 count >100 cells per cubic millimeter may be a simple, safe, and effective strategy. This pragmatic strategy could reduce the complexity of patient management at primary care,4 as it allows patients to stabilize clinically, provides time to prepare patients for lifelong ART, reduces the risk of IRIS which has been shown to be highest in patients started on ART immediately after the start of TB treatment,5,7,32 reduces the pill burden from 7 to 5 drugs, and reduces the risk of overlapping toxicity which is highest in the first weeks of treatment initiation.33
Delaying ART in patients with CD4 count >350 cells per cubic millimeter beyond the start of the continuation phase was counterproductive given the poor fidelity of health-care workers to reassess ART eligibility before the end of treatment and low rates of ART initiation in those who become eligible at TB treatment completion. This finding is consistent with evidence from the Starting Antiretroviral Therapy at Three Points in Tuberculosis RCT study, which found that deferring ART beyond the period of TB treatment significantly increased mortality.34
Our study had several limitations that must be considered in interpretation of the findings. First, our study was observational, using a historical cohort for comparison. Differences in ART uptake or mortality could therefore have been influenced by factors other than the intervention, including increase in resources and infrastructure for HIV care and treatment. Second, viral load data were not available to assess effectiveness of ART or validate self-reported adherence data. Third, overall mortality among patients with TB/HIV may have been underestimated as patients may have died before study enrollment, during the TB diagnostic process, or between TB and HIV diagnoses.
A crucial question in managing HIV-infected TB patients concerns when and how to initiate ART. Although findings of RCTs have provided key evidence for policy and guidelines, it is important to acknowledge that they were carried out under highly controlled conditions and provided results that were only conclusive for TB patients with CD4 count <50 cells per cubic millimeter.4 Our study expands the clinical trial evidence and translates findings to a real-world, highly resource-deprived setting. We show that CD4-stratified ART initiation by primary care nurses is feasible and associated with increased ART uptake and decreased mortality. In our setting, initiating ART in patients with CD4 counts >100 cells per cubic millimeter at the start of the continuation phase of TB treatment was safe and could help reduce the complexity of TB/HIV care that has been associated with initiation of ART soon after the start of TB treatment.4 Delaying ART beyond the continuation phase among patients with CD4 counts >350 was not a successful strategy as most of these patients were never reassessed for ART eligibility or were assessed but never initiated ART. Further research is needed to determine the optimal management at primary care level of TB patients with CD4 counts <100 cells per cubic millmeter, as mortality in this group remained disproportionately high despite prioritization for early ART.
The authors are grateful for the patient care, program administration, and coordination contributions of Drs. Roger, Kitenge, Bisuta, Bandale, Nzungani, Kapembi, Mbenza, and Seladi.
1. World Health Organization. Global Tuberculosis Report 2013. Geneva, Switzerland: World Health Organization; 2013. WHO/HTM/TB/2013.11.
2. Lawn SD, Campbell L, Kaplan R, et al.. Delays in starting antiretroviral therapy in patients with HIV-associated tuberculosis accessing non-integrated clinical services in a South African township. BMC Infect Dis. 2011;11:258.
3. Lawn SD, Harries AD, Williams BG, et al.. Antiretroviral therapy and the control of HIV-associated tuberculosis. Will ART do it? Int J Tuberc Lung Dis. 2011;15:571–581.
4. Naidoo K, Baxter C, Karim SSA. When to start antiretroviral therapy during tuberculosis treatment? Curr Opin Infect Dis. 2013;26:35–42.
5. Abdool Karim SS, Naidoo K, Grobler A, et al.. Integration of antiretroviral therapy with tuberculosis treatment. N Engl J Med. 2011;365:1492–1501.
6. Blanc FX, Sok T, Laureillard D, et al.. Earlier versus later start of antiretroviral therapy in HIV-infected adults with tuberculosis. N Engl J Med. 2011;365:1471–1481.
7. Havlir DV, Kendall MA, Ive P, et al.. Timing of antiretroviral therapy for HIV-1 infection and tuberculosis. N Engl J Med. 2011;365:1482–1491.
8. Legido-Quigley H, Montgomery CM, Khan P, et al.. Integrating tuberculosis and HIV services in low- and middle-income countries: a systematic review. Trop Med Int Health. 2013;18:199–211.
9. Friedland G, Harries A, Coetzee D. Implementation issues in tuberculosis/HIV program collaboration and integration: 3 case studies. J Infect Dis. 2007;196:S114–S123.
10. Sanne I, Orrell C, Fox MP, et al.. Nurse versus doctor management of HIV-infected patients receiving antiretroviral therapy (CIPRA-SA): a randomised non-inferiority trial. Lancet. 2010;376:33–40.
11. Fairall L, Bachmann MO, Lombard C, et al.. Task shifting of antiretroviral treatment from doctors to primary-care nurses in South Africa (STRETCH): a pragmatic, parallel, cluster-randomised trial. Lancet. 2012;380:889–898.
12. Martinot A, Van Rie A, Mulangu S, et al.. Baseline assessment of collaborative tuberculosis/HIV activities in Kinshasa, the Democratic Republic of Congo. Trop Doct. 2008;38:137–141.
13. National Institute of Allergy and Infectious Diseases, Division of AIDS. Table for Grading the Severity of Adult and Pediatric Adverse Events. Bethesda, Maryland: National Institute of Allergy and Infectious Diseases; 2004.
14. Henegar C, Behets F, Vanden Driessche K, et al.. Mortality among tuberculosis patients in the Democratic Republic of Congo. Int J Tuberc Lung Dis. 2012;16:1199–1204.
15. World Health Organization. Haemoglobin Concentrations for the Diagnosis of Anaemia and Assessment of Severity. Geneva, Switzerland: World Health Organization; 2011. WHO/NMH/NHD/MNM/11.1.
16. Kerschberger B, Hilderbrand K, Boulle AM, et al.. The effect of complete integration of HIV and TB services on time to initiation of antiretroviral therapy: a before-after study. PLoS One. 2012;7:e46988.
17. Huerga H, Spillane H, Guerrero W, et al.. Impact of introducing human immunodeficiency virus testing, treatment and care in a tuberculosis clinic in rural Kenya. Int J Tuberc Lung Dis. 2010;14:611–615.
18. Vasan A, Kenya-Mugisha N, Seung KJ, et al.. Agreement between physicians and non-physician clinicians in starting antiretroviral therapy in rural Uganda. Hum Resour Health. 2009;7:75.
19. Bedelu M, Ford N, Hilderbrand K, et al.. Implementing antiretroviral therapy in rural communities: the Lusikisiki model of decentralized HIV/AIDS care. J Infect Dis. 2007;196(suppl 3):S464–S468.
20. Bemelmans M, Van Den Akker T, Ford N, et al.. Providing universal access to antiretroviral therapy in Thyolo, Malawi through task shifting and decentralization of HIV/AIDS care. Trop Med Int Health. 2010;15:1413–1420.
21. Callaghan M, Ford N, Schneider H. A systematic review of task-shifting for HIV treatment and care in Africa. Hum Resour Health. 2010;8:8.
22. Cohen R, Lynch S, Bygrave H, et al.. Antiretroviral treatment outcomes from a nurse-driven, community-supported HIV/AIDS treatment programme in rural Lesotho: observational cohort assessment at two years. J Int AIDS Soc. 2009;12:23.
23. Colvin CJ, Fairall L, Lewin S, et al.. Expanding access to ART in South Africa: the role of nurse-initiated treatment. S Afr Med J. 2010;100:210–212.
24. Ivers LC, Jerome JG, Cullen KA, et al.. Task-shifting in HIV care: a case study of nurse-centered community-based care in Rural Haiti. PLoS One. 2011;6:e19276.
25. Shumbusho F, van Griensven J, Lowrance D, et al.. Task shifting for scale-up of HIV care: evaluation of nurse-centered antiretroviral treatment at rural health centers in Rwanda. PLoS Med. 2009;6:e1000163.
26. Zachariah R, Ford N, Philips M, et al.. Task shifting in HIV/AIDS: opportunities, challenges and proposed actions for sub-Saharan Africa. Trans R Soc Trop Med Hyg. 2009;103:549–558.
27. Gandhi NR, Moll AP, Lalloo U, et al.. Successful integration of tuberculosis and HIV treatment in rural South Africa: the Sizonq'oba study. J Acquir Immune Defic Syndr. 2009;50:37–43.
28. World Health Organization. WHO Policy on Collaborative TB/HIV Activities: Guidelines for National Programmes and Other Stakeholders. Geneva, Switzerland: World Health Organization; 2012. WHO/HTM/TB/2012.6.
29. Wong EB, Omar T, Setlhako GJ, et al.. Causes of death on antiretroviral therapy: a post-mortem study from South Africa. PloS One. 2012;7:e47542.
30. Cain KP, Anekthananon T, Burapat C, et al.. Causes of death in HIV-infected persons who have tuberculosis, Thailand. Emerg Infect Dis. 2009;15:258.
31. Kantipong P, Murakami K, Moolphate S, et al.. Causes of mortality among tuberculosis and HIV co-infected patients in Chiang Rai, Northern Thailand. HIV AIDS (Auckl). 2012;4:159.
32. Lawn SD, Myer L, Bekker LG, et al.. Tuberculosis-associated immune reconstitution disease: incidence, risk factors and impact in an antiretroviral treatment service in South Africa. AIDS. 2007;21:335–341.
33. McIlleron H, Meintjes G, Burman WJ, et al.. Complications of antiretroviral therapy in patients with tuberculosis: drug interactions, toxicity, and immune reconstitution inflammatory syndrome. J Infect Dis. 2007;196:S63–S75.
34. Abdool Karim SS, Naidoo K, Grobler A, et al.. Timing of initiation of antiretroviral drugs during tuberculosis therapy. N Engl J Med. 2010;362:697–706.