Peer-led active tuberculosis case-finding among people living with HIV: lessons from Nepal

Abstract Problem People living with a human immunodeficiency virus (HIV) infection have a high risk of tuberculosis and should undergo regular screening. However, they can be difficult to reach because they are stigmatized and discriminated against. Approach In Nepal, the nongovernmental organization Naya Goreto implemented a peer-led tuberculosis screening project in which people living with HIV volunteered to contact others in this high-risk population. Volunteers took part in a short training course, after which they attempted to contact people living with HIV through existing networks and self-help groups. Tuberculosis screening and testing were carried out in accordance with national guidelines. Local setting In Nepal, the prevalence of HIV infection is 0.3% in the general population but is much higher, at 6%, in people in Kathmandu who inject drugs. To date, the health system has not been able to implement systematic tuberculosis screening in people living with HIV. Relevant changes Between May 2014 and mid-September 2015, 30 volunteers screened 6642 people in 10 districts, 5430 (82%) of whom were living with HIV. Of the 6642, 6046 (91%) were tested for tuberculosis and 287 (4.3%) were diagnosed with the disease, 240 of whom were HIV-positive. Of those with tuberculosis, 270 (94%) initiated treatment. Lessons learnt Using peers to contact people living with HIV for tuberculosis screening resulted in a high participation rate and the identification of a considerable number of HIV-positive tuberculosis patients. Follow-up during treatment was difficult in this highly mobile group and needs more attention in future interventions.


Introduction
The prevalence of tuberculosis is elevated in people living with a human immunodeficiency virus (HIV) infection and, in 2014, almost 400 000 HIV-infected people died from tuberculosis globally. 1 Consequently, the World Health Organization (WHO) recommends systematic tuberculosis screening for people living with HIV. 2 In Nepal, the National Tuberculosis Center struggles to screen these people even though it adapted its strategies on tuberculosis and HIV infection in 2009. A better way of reaching people living with HIV for tuberculosis screening and treatment would reduce the burden of disease and death. The Stop TB Partnership recommends that people from key populations affected by tuberculosis should be involved in tuberculosis care. 3 In Nepal, a nongovernmental organization established by and working with people living with HIV and people who use drugs -Naya Goreto -recruited volunteers among people living with HIV for tuberculosis screening of their peers. The project targeted 10 districts with large numbers of people living with HIV. The aims were to screen around 7050 people living with HIV and to ensure that those diagnosed with tuberculosis started treatment. In addition, it was hoped that the project would increase awareness of the importance of tuberculosis screening in people living with HIV among both those affected and health-care workers. In 2013, the 10 project districts accounted for 10 472 tuberculosis patients -31% of the country's total.

Local setting
In Nepal, HIV infection is a relatively small problem: the estimated prevalence in the adult population is 0.3%. 4 However, the prevalence is much higher in certain groups, such as drug users, migrant workers and sex workers. In Kathmandu, the prevalence in injection-drug users is 6%. 5 Among female and male sex workers, it is 2% and 9%, respectively. 5 Many people living with HIV or in these high-risk groups are marginalized, stigmatized and experience discrimination. [6][7][8] For 2013, WHO estimated that the number of incident, HIV-positive tuberculosis patients in Nepal was 1587. 9 However, of the 33 834 new patients actually reported in the country, only 11% knew their HIV-infection status and only 65 tuberculosis patients were known to be HIV positive, which indicates that tuberculosis is often missed in people living with HIV.

Approach
In 2014, staff at Naya Goreto asked leaders of self-help groups for people living with HIV and drug users to act as volunteers for the tuberculosis screening project because they knew how to reach people living with HIV. Naya Goreto staff themselves are people living with HIV and former drug-users and are members of these networks and self-help groups. In addition, the organization has worked with some of these leaders in the past. For the intervention, Naya Goreto provided training for volunteers on tuberculosis screening and on how to reach out to HIV-infected people. A one-day training course was given in each district on: (i) the appropriate circumstances Lessons from the field Peer-led tuberculosis case-finding in Nepal Dipu Joshi et al.
for disclosing HIV-infection status; (ii) behavioural change techniques for use with self-help groups and individuals; (iii) maintaining confidentiality; (iv) ensuring the availability of, and access to, diagnostic services without fear of discrimination; and (v) referral for tuberculosis treatment and follow-up.
In collaboration with the National Tuberculosis Center, Naya Goreto developed a screening tool for tuberculosis. The criteria were: (i) a cough for 2 weeks; (ii) fever; (iii) loss of appetite; (iv) weight loss; and (v) no tuberculosis test within the last six months. People living with HIV were contacted by volunteers through existing networks and self-help groups. The volunteers applied the screening tool and encouraged people who met one or more of the five criteria to request further tuberculosis screening at HIV treatment centres or care homes for HIV-infected people. If necessary, Naya Goreto provided money for transport. For the few people who did not want to go to health facilities because of previous poor experiences with formal health-care services, volunteers carried out screening and collected sputum samples using specially provided containers at a place chosen by the individual being screened.
Tuberculosis tests were carried out in accordance with national guidelines: two sputum samples were examined by smear microscopy and a third underwent the rapid molecular diagnostic test GeneXpert ® MTB/RIF (Cepheid Inc., Sunnyvale, United States of America). If necessary and available, a chest X-ray or fine-needle aspiration was carried out. Physicians diagnosed some tuberculosis patients clinically. Naya Goreto paid the diagnostic costs, which included a nominal fee for the hospital visit that all patients must pay, and fees for additional examinations. If a person was diagnosed with tuberculosis, a volunteer informed the person and arranged for him or her to attend the tuberculosis treatment facility of their choice. Thereafter, the volunteer tried to remain in contact with the patient and collect information on treatment outcomes. Naya Goreto collected data on screening, the diagnostic tests carried out, tuberculosis diagnoses, the treatment given and the risk group of each individual who underwent tuberculosis testing. Project coordinators in each district collected information from volunteers and reported to the project man-ager in Kathmandu, who maintained a database on the test results, treatment and outcomes for all patients with tuberculosis. Participation in screening was voluntary and Naya Goreto obtained informed consent from all participants for use of their data.

Relevant changes
The tuberculosis screening project was implemented between May 2014 and mid-September 2015. The 30 volunteers (21 male) screened 6642 people, of whom 5430 (82%) were living with HIV. In total, 6046 (91%) of the 6642 were tested for tuberculosis: 5402 were living with HIV, 331 were drug users, 170 were family members of tuberculosis patients, 138 were migrant workers and 5 were slum dwellers. Table 1 shows the tests performed and the results obtained. Overall, 287 tuberculosis patients (205 male and 240 HIV positive) were identified, 270 (94%) of whom started treatment. By the end of the project, the outcome of tuberculosis treatment was known for 178 patients: 39 (22%) successfully completed treatment, 15 (8%) died and 124 (70%) had transferred out of the intervention districts. The remaining 92 patients were still on treatment. The total cost of the project was 132 596 United States dollars, which comprised 61% for diagnostic and transport costs, 20% for Naya Goreto's expenditure and 12% for training volunteers.

Lessons learnt
Box 1 summarizes the main lessons learnt from this project. The peer-led, active, tuberculosis case-finding intervention had several successes. First, within a short period the volunteers contacted many people living with HIV, a marginalized group subject to discrimination in Nepal. Second, many more HIV positive tuberculosis patients were identified than notified at a national level in 2013. 9 Third, no major problem was encountered with using peer volunteers to reach this key population. Fourth, reports from volunteers indicated that some screening was still taking place after the project ended.
Many people living with HIV in Nepal are former or current drug users and know each other through drug users' networks. This familiarity contrib- Lessons from the field Peer-led tuberculosis case-finding in Nepal Dipu Joshi et al.
uted to the success of the intervention because the target group trusted the volunteers, whereas they do not always trust health-care workers because of stigmatization and discrimination. Although our project focused on people living with HIV, because volunteers contacted these individuals through their networks and self-help groups, they also reached other groups with a high risk of tuberculosis, such as drug users and migrant workers. We were unable to assess the effectiveness of peerled screening in these groups. Working with peers has also proved successful in other programmes involving individuals at a high risk of tuberculosis or HIV infection. In the Democratic Republic of the Congo, a peer-led strategy led to increased tuberculosis case-finding among internally displaced and migrant people. 10 Peer educators chosen from among presumptive tuberculosis patients in these communities identified other community members with presumptive tuberculosis for referral to screening and testing services in health centres.
One limitation of the intervention is that it was externally funded. Most of the project's expenditure was for diagnostic costs and transport support for patients, who may not have been able to cover these costs themselves. This may have contributed to the project's success. Although costs could be reduced by integrating tuberculosis screening into routine care for people living with HIV, they may still present a barrier to the successful diagnosis and treatment of tuberculosis. Additional support, such as cash transfers to patients, may help. 11 A worrying observation was that 70% of people whose tuberculosis treatment outcomes could be assessed at the end of the project had moved out of the intervention districts. Further follow-up of these patients was not possible within the project and they may not all have continued treatment at a health-care facility. Many people living with HIV in Nepal move frequently from place to place because they face discrimination or are migrant workers. 6 Efforts should be made in future interventions to ensure that these people finish tuberculosis treatment. Contacting peer volunteers at their destination to ensure follow-up may be one solution. ■ Funding: The project was funded by a grant from the TB REACH programme, which is supported by Global Affairs Canada and which operates as part of the Stop TB Partnership.