Integration of HIV and Tuberculosis in the Community

Abstract

Current strategies are insufficient to contain the growing tuberculosis (TB) epidemic in areas of high HIV prevalence such as sub-Saharan Africa. Due to the increased risk of morbidity and mortality among those coinfected, early detection is critical. However, strategies dependent on passive, facility-based case finding have failed due to severe limitations in the HIV-positive population. There is growing evidence from multiple clinical trials that early initiation of antiretroviral therapy (ART) in patients coinfected with HIV and TB reduces mortality. Integration of community-based distribution of ART and TB medicines should be considered for coinfected patients to help improve retention in care and to off-load busy health systems. Several models of integration of HIV and TB care in sub-Saharan Africa have been successful. This review article examines the concepts of HIV and TB integration of testing and treatment at the community level.

Introduction

Africa, particularly Southern Africa, has among the highest rates of HIV and tuberculosis (TB) coinfection in the world. Current strategies are insufficient to contain the growing TB epidemic in the setting of high HIV prevalence. The World Health Organization (WHO)’s Stop TB Partnership recently published a mathematical model detailing interventions that could reduce the rate of HIV-associated TB deaths by 50% to 80% by 2015, saving up to 1 million lives. ¹ The study showed that it is cost effective to treat both diseases in coinfected individuals, as the cost per life year gained through addition of antiretroviral therapy (ART) to TB treatment is only double the cost of TB treatment alone and reduces the case fatality rate of TB from 6% to 3%. ¹ Integration of TB-HIV services can help improve diagnosis and timely treatment of this deadly syndemic. However, in the setting of overburdened health systems, it is critical to bring TB-HIV integration to the community level in order to increase case detection and shorten time to initiation of treatment for patients coinfected with HIV and TB.

Diagnosis of HIV and TB in the Community

Currently, resources are becoming increasingly constrained and there is a renewed focus on finding efficiencies. ² Particular attention should be paid to populations where HIV is likely to be most prevalent, such as those enrolled in TB treatment programs. In high-burden TB countries like sub-Saharan Africa, HIV prevalence among incident patients with TB is 40% to 60% and may be as high as 88% in retreatment patients with TB. ^3,4 Testing for HIV within TB populations also represents a unique opportunity to access a patient subgroup with high levels of receptiveness to HIV testing. ³ In studies from South Africa, 77% to 100% of patients with TB accept HIV testing. ^5,6 Thus, increasing HIV testing among patients with TB represents a particularly important opportunity to diagnose HIV earlier.

In a large randomized control trial comparing standard clinic- or facility-based HIV testing and counseling with community-based testing, the community-based testing detected almost 4 times more HIV cases. This was despite the fact that the clinic-based testing populations had higher HIV prevalence rates. Interventions for the community-based testing included mobile testing, community mobilization activities, and support services following testing. ⁷ Patients may be more comfortable with community-based testing, as it removes the stigma of attending clinics that are known to provide HIV care and in addition, may be more accessible for patients in rural areas.

Knowledge of HIV status is also important in TB diagnosis. Smear-negative TB is more common among coinfected patients, the proportion of cases of smear-negative pulmonary TB in HIV-positive patients with TB is reported to range from 24% to 61%. ⁸ Thus, it is crucial to know the patient’s HIV status at the time of diagnosis of TB. With knowledge of the patient’s HIV status, clinicians may choose to recommend additional workup or have a lower threshold to initiate TB treatment based on symptomatology despite a negative smear.

Active versus Passive Case Finding for Diagnosis of TB among HIV-Positive Patients

In order to maximize the benefit of TB measures to obtain control at the population level, HIV-TB integration must not only be obtained at the facility but also at the community level for case finding. The WHO now recommends intensified case finding as a strategy for identifying TB in HIV-positive patients as part of their “Three Is for HIV/TB” strategy which also includes isoniazid preventative therapy and infection control. ⁹ Strategies dependent on passive, facility-based case finding have failed in areas of high HIV prevalence. For example, South Africa’s program has focused on passive presentation of patients with TB, which has resulted in progress to meet treatment targets, but South Africa simultaneously has seen a growth in the TB epidemic. ¹⁰ South Africa now has the third highest TB burden in the world.

Focusing on HIV-infected populations, passive case finding has been shown to be inadequate. A survey conducted in South Africa demonstrated that passive case finding identified TB in 67% of HIV-negative patients but in only 33% of HIV-positive patients. ¹¹ Globally, performance of HIV testing in patients with TB is poor, with only 34% of patients with TB tested for HIV. The African region performs slightly better, but still only reports that 59% of patients with TB are tested for HIV. ⁹ However, of those tested, nearly half were HIV positive. ⁹ In contrast to passive case finding, active case finding screens high-risk communities for TB using a symptom survey, sputum samples, and other methods.

Many studies have demonstrated success with community-level active case finding in areas of high HIV prevalence. A study in Malawi found the prevalence of TB utilizing passive case finding was only 0.19% as opposed to 1.74% with active case finding (P = .01). ¹² In addition, studies have compared various methods of active case finding. For example, a randomized study comparing 2 methods of active case finding at the community level, DETECTB, found a mobile approach superior to door-to-door inquiries. Overall, both methods lead to a decline in the prevalence of culture-positive TB from 6.5 per 1000 adults (95% confidence interval [CI] 5.1-8.3) to 3.7 per 1000 adults (absolute risk reduction [ARR] 0.59, 95% CI 0.40-0.9, P = .0112). ¹³

The Zambia-South Africa TB and AIDS Reduction (ZAMSTAR), a recent large-scale, community-randomized trial, highlighted the value of home-based testing for HIV and TB. In this study, passive enhanced case finding (ECF) did not show population-level benefit. ¹⁴ The ECF did not include HIV testing at the household level; and, thus, as ECF relied on sputum smear positivity, this may have missed cases of smear-negative disease in patients with HIV, thereby not getting people living with HIV into care. This study also highlighted the cost-effective strategy of using community health workers, which has been demonstrated in other studies on the efficacy of case finding. ^15,16 Other benefits of community-based case finding include improved infection control, whereby those who may carry active, undiagnosed TB are not required to attend clinics where they could transmit the disease to susceptible patients. ¹⁷

Community versus Facility Initiation of ART for Coinfected Patients

Early initiation of ART is of particular importance for coinfected patients. Growing evidence from low-income countries supports the initiation of ART earlier in the disease course of HIV, in order to reduce mortality and rates of infectious and noninfectious causes of morbidity. ¹⁸ Additional data confirm that ART can dramatically reduce HIV transmission, given further impetus to scale-up HIV treatment. ¹⁹ As TB accounts for a substantial portion of HIV-related mortality and morbidity, it is encouraging that use of ART not only reduces incident TB among those living with HIV by 54% to 92% but also substantially reduces case fatality rates for those who do contract TB by as much as 64% to 95%, in observational studies. ^20,21 Several recent randomized controlled trials, Cambodian Early versus Late Introduction of Antiretroviral Drugs, Immediate Versus Deferred Start of Anti-HIV Therapy in HIV-Infected Adults Being Treated for Tuberculosis, and Starting Antiretroviral Therapy at Three Points in Tuberculosis found significant reductions in mortality for those with HIV and TB coinfection who were initiated on early ART. ^19,22,23

There is significant attrition between testing positive for HIV and initiating on ART, including for patients who access HIV care via TB services. ²⁴ One study, from South Africa, found that almost 30% of patients with TB considered eligible for ART were initiated within 6 months of their HIV test; among those not eligible for ART, only 35% had a repeat CD4 count. ²⁴ In Kenya, only 34% of patients with HIV-TB coinfection started ART while being treated for TB between 2006 and 2009. ²⁵ A study in South Africa also demonstrated significant delays to initiation of ART existed among patients with TB who qualify for ART. ²⁶

It is critical to examine the integration of TB and HIV treatment and care services, not only within health services but also at the level of community-based interventions. ²⁷ A small study comparing the initiation of TB treatment and ART for coinfected patients before and after an integrated clinic for both services found that time to ART initiation decreased from 147 days (95% CI 85-188) to 75 days (95% CI 52-119) postintegration. There was also a 60% increased chance of coinfected patients started on ARTs, as compared with before the intervention. ²⁸ A recent study in Malawi demonstrated good outcomes with home-based ART initiation, with significantly more people starting ART as compared with facility-based ART. ²⁹ However, significant data on the benefits of community-based ART and TB treatment initiation are lacking.

Community-Based Linkage to Care and Treatment Success

Evidence has demonstrated that community-based distribution and monitoring of ART can lead to excellent patient outcomes in HIV-positive patients. For example, Médecins Sans Frontières cohorts in Mozambique, Malawi, and South Africa have all shown that for patients who are stable on ART, medication can be distributed at the community level by community health workers or via rotating patient groups, thus allowing increased spacing between clinic visits. ^30,31,32 Retention in care and virologic outcomes for these groups have been excellent. For example, a nurse-driven community treatment program in Lesotho found the proportion of HIV-positive adults arriving sick to clinic (CD4 count < 50 cells/mm³) decreased from 22.2% in 2006 to 11.9% in 2008. ³² A cluster randomized equivalence trial of home care versus facility-based care for HIV in Uganda found that mortality rates and virologic failure were similar between groups (relative risk [RR] 0.95 (0.71-1.28) and 1.04 (0.78-1.40), respectively). ³³ Home-based care involved a trained field officer visiting patients monthly to deliver ART and to monitor side effects with referral to the health system as needed.

The ZAMSTAR demonstrated the importance of strengthening the link between communities and clinics to effectively diagnose and treat TB and HIV coinfection. The study involved almost a million people in 24 communities across Zambia and in the Western Cape Province of South Africa. Household-based interventions that included multiple home visits, contact tracing and home HIV and TB counseling, and testing with linkage to care resulted in a 22% reduction in culture-positive patients with TB in the populations studied. ¹⁴ These household-based interventions reduced the prevalence of TB in a community with a high burden of TB and HIV at a cost of less than US$1 per person per year. ¹⁴

Several successful models of integration of HIV and TB care in the community have come from the KwaZulu-Natal region of South Africa, where more than 80% of active patients with TB are HIV positive. ³⁴ The Sizonq’oba study utilized the directly observed treatment, short-course model for both ART and TB treatment for 119 HIV/TB coinfected patients over 12 months and found that 88% of patients had undetectable viral loads and 84% completed treatment with 99% adherence to ART. ³⁵ Key components were the use of community health works and adherence education for patients and families. This study was replicated among 80 patients with multi-drug-resistant tuberculosis from 2008 to 2010, where 66% of patients were HIV coinfected. A total of 77% of patients were cured or on treatment, 82% had undetectable viral loads, and only 6% died. ³⁶

There is also evidence that decentralization of care from the hospital to the community for MDR-TB, in areas of high HIV prevalence, is beneficial at a community and individual level. Hospitals often have shortages of beds for the traditional extensive inpatient treatment for those with MDR-TB, thus treatment is often delayed and disease is therefore spread among the community. A study done in the Hlabisa health subdistrict of Kwa-Zulu Natal found that community-based treatment after brief hospitalization led to shorter conversion time to negative sputum smear as compared with the traditional model, 85 versus 119 days, respectively (P = .002). ³⁷ Comparable findings were noted in another prospective study, where more patients were put on ART within 2 weeks and culture conversion occurred more frequently at the decentralized site, at a rate of 54% versus 24% (P < .001). ³⁸ Conversion was hypothesized to be lower at the centralized site due to longer time to initiation, more retreatment cases, less patients on ART, and lack of community support.

Integration of community-based distribution of ART and TB medicines should be considered for certain coinfected patient groups to help improve retention in care and off-load busy health systems. This model of care for coinfected patients requires further study.

Community-Based Adherence Support for Coinfected Patients

Improving retention in HIV care for patients coinfected with TB is critical. Studies on DOT therapy for TB have shown that tracing participants who default on treatment can retain patients in care and reduce transmission to the community. A study of community tracer teams for TB treatment in South Africa found a significant decrease in patients defaulting from treatment in tracer verses nontracer subdistricts (P = .02). ³⁹ In addition, there is evidence that a similar approach is valuable in treatment of HIV-positive patients on ART. A randomized trial assessing the effects of community-based peer health workers on long-term treatment with ART found decreased virologic failure rates and a reduction in lost to follow-up (LTFU) rates. ⁴⁰

Patients coinfected with HIV and TB have a higher pill burden and are therefore more at risk of suboptimal adherence than monoinfection patients. Therefore, aiming to reduce LTFU among this patient population is critical to treatment success. A study from Kenya demonstrated utilization of contact tracing for those patients with HIV and TB. They found that HIV patients LTFU declined from 21.2% to 11.5% in a 3-year period, while TB patients’ LTFU rates declined from 19.3% to 12% in the same period. A total of 85% of patients were able to be contacted, meeting the WHO goal of <20% LTFU. ⁴¹ Integration of care in the community for patients coinfected with HIV and TB could be enhanced with dual tracking of defaulters through the same community health worker.

Conclusion

Improved integration of HIV-TB testing and treatment beyond the health facility represents an opportunity to further promote the timely diagnosis and treatment of coinfected patients, with the aim of controlling the HIV-TB syndemic. To this end, future programs should include active TB and HIV case finding not only in the clinic but also in the community. ⁴² Community health workers and other community systems can improve linkage to ART services and provide integrated follow-up for defaulters from both TB and HIV treatment programs. Community-based initiation and distribution of HIV and TB treatment offer the ability to scale-up the number of patients in treatment while not overwhelming health systems. However, data are lacking in examining the benefits of initiating ART and TB treatment in the community setting. and this should be a focus of future research. Improving the diagnosis, care, and outcomes for HIV-TB coinfected patients is an important step toward turning back the tide of HIV.