Isoniazid preventive therapy for tuberculosis occurrence in HIV-positive patients in Lesotho

Abstract

Tuberculosis (TB) remains a main public health problem, particularly in people living with HIV (PLHIV) in sub-Saharan Africa (SSA). This persistent problem may be an indication of underlying problems in national health policies, and their implementation in SSA. The Southern African country of Lesotho, with one of the highest TB incidences in the world, is facing a catastrophic syndemic of HIV and tuberculosis (TB). The effectiveness of isoniazid preventive therapy (IPT), which has the potential to reduce the incidence of TB in PLHIV, has not been adequately evaluated since its introduction in 2011. This study evaluated the uptake of IPT, its effectiveness and the associated factors in Lesotho, with the aim of establishing the necessary policy changes required to prevent the occurrence of TB in PLHIV. To determine the effectiveness of IPT and the factors underpinning the implementation of this intervention in Lesotho, a quantitative evaluation as well as a qualitative study of the implementation of IPT was used. The study was therefore based on a triangulation of quantitative and qualitative research methods in two phases. The qualitative phase of the study identified health system challenges affecting the implementation of IPT, based on a cross-sectional qualitative analysis of interview responses of healthcare workers and stakeholders of the TB/HIV programmes, which included the Ministry of Health officials and support partners, purposively selected for their roles in IPT implementation in Lesotho. The rationale of this study phase was based on the premise that the rate of initiation of IPT and its effectiveness is largely dependent on the quality of implementation of the IPT programme. The qualitative study phase revealed that seven factors in the health system were affecting the implementation of IPT in the country, namely poorly decentralised HIV services; inefficient monitoring and evaluation systems; ineffective service delivery; interrupted supply chains; an undertrained and inadequate health workforce; insufficient health system financing; and inefficient health information systems. The implementation of IPT was therefore a complex task which needed certain sectors of the health system to change. The most important lesson from this is that key health interventions need a ‘health system approach’ for success. The quantitative phase of the study was a quasi-experimental review of HIV-positive medical records randomly selected from eight health institutions in six districts of Lesotho. This study phase selected two patient groups, one enrolled into antiretroviral therapy (ART) before (2004-2010 cohort), and the other one after the launch of IPT (2011-2016 cohort), to establish the rate of initiation of IPT and its effectiveness in preventing the occurrence of TB in the country. IPT uptake and its effectiveness were evaluated using an analytical model based on Cox’s proportional hazards regression analysis, an approach often used to determine the relative risk of contracting a disease and the associated factors. The quantitative study phase included 2 955 randomly selected records that met the inclusion criteria set for the study. Overall, 68.8% of the 2 955 patients had received IPT over a course of six years (2011-2016), which translated to a sluggish IPT uptake rate of 20.6 per 100 person-years over the six-year period. Notably, only 135 (6.6%) patients defaulted IPT, which is a small proportion. Comparatively, the 2011-2016 cohort had a significantly (p=0.000) higher rate of IPT initiation (27.0 per 100 person-years) than the 2004-2010 cohort, (15.8 per 100 person-years), implying that patients newly enrolled into ART had a higher rate of IPT initiation. Findings indicated that the most significant predictors for initiation of IPT were age group, district category and duration of ART. Furthermore, based on odds ratios (OR), patients in the sparsely populated districts (OR=1.6) and males (OR=2.1) had significantly (p<0.05) higher odds of defaulting IPT, compared to those in the densely populated districts and females, respectively. Whereas higher defaults in the sparsely populated districts were associated with long distance from hospitals and the mountainous terrain associated with these districts, higher rates of defaulting by males were most likely due to migrant work in South Africa. The TB incidence rate reduced from 2.3 per 100 person-years in 7 985 person-years in the 2004-2010 cohort, to 1.6 per 100 person-years in 4 223 person-years in the 2011- 2016 cohort, implying that the IPT intervention had considerably reduced the occurrence of TB. However, the use of IPT was not without adverse effects. By proportion, the most common side effects to IPT were skin rash (37.2%), peripheral neuropathy (25.4%) and liver toxicity (9.4%). In addition, out of 246 patients who had developed TB and were discovered during a follow-up, 15.9% of the patients developed TB after exposure to IPT. Further findings indicated that prescribing IPT within one year of ART commencement, which reduced TB incidences to only 1.3 incidences per 100 person-years, was the most effective intervention for preventing the occurrence of TB, compared to other commencement timing of IPT intervention. Other TB incidences per 100 person-years by timing of IPT were as follows – IPT before ART (1.7), IPT after ART (1.8), no IPT (2.6), and IPT 3-5 years after ART initiation (2.3). Gender, baseline WHO clinical stage, district category and time to IPT relative to ART commencement emerged as significant predictors of TB occurrence. Notably, increasing commencement time for IPT by one six-month interval increased the risk of contracting TB by between 6% and 59%, depending on the cohort, with the 2011-2016 cohort having a higher risk compared to the 2004-2010 cohort. The findings of this study indicate that the implementation of IPT in Lesotho has notable challenges. Clearly, there is a need to improve the rate of IPT initiation in the patient groups with the most sluggish rate of IPT uptake, and to improve retention of some patient groups with poor adherence to IPT. The findings of this study also indicate that delayed IPT commencement after ART initiation significantly affects the effectiveness of IPT. Furthermore, the study reveals that IPT is a complex health intervention, and its implementation therefore needs a health sector-wide or ‘health systems’ approach.