Information and educational materials.

Education materials were used to give information on various aspects of MDR-TB and its treatment to all patients and family members (S1 Appendix). The materials included a flipbook (an A4 spiral bound book with pictures on the patient-facing sheet and explanatory text for the health worker on the back), and two information leaflets, one for patients and one for family members. While both the leaflets had information on MDR-TB, the family member leaflet had additional information on their supportive roles. These materials were designed to be used by the health workers in their routine consultations with patients.

Screening process.

Screening for depression and anxiety was carried out using the 25 item Johns Hopkins Symptom Checklist (“HSCL”) [14], and screening for low social support using the Multidimensional Scale of Perceived Social Support (MSPSS) [15]. The screening was planned to take place during patients’ monthly follow-ups at the centre and to be conducted by the health workers. Patients who screened positive on at least one tool (anxiety ≥17 or depression ≥24 on HSCL [16]; or MSPSS≤3 [15]) were referred for counselling. Patients who screened negative across both tools were rescreened a month later. HSCL was selected as it has previously been translated and culturally adapted for use in Nepal for identification of depression and anxiety among internally displaced people, has been shown to have moderate validity [16], and was recommended by TPO Nepal, a Nepalese mental health NGO. There was no validated tool to capture the perceived need for social support in the needs assessment phase of the study [13]; the study team judged MSPSS was the most appropriate tool for the context. MSPSS was translated from English to Nepali by the research team and verified by back translation.

Brief counselling.

Counselling was provided to those found to be depressed using the Healthy Activity Program (HAP) approach. HAP is individual counselling based on behavioural activation (BA) psychological therapy, developed in India for treating depression within primary care using lay counsellors [17]. The core principle of BA is the use of activity scheduling (planned timetables) to address inactivity and problems that the person avoids. The recipient of BA identifies activities to undertake that they are likely to find enjoyable in line with their goals and values. Behavioural activation is equivalent in effectiveness to cognitive behavioural therapy for depression, but significantly cheaper [18]. HAP was selected as it is an intervention with promising evidence from neighbouring India [19] and was already being tested in a few districts of Nepal in primary health care settings. We worked with the local mental health NGO who were testing HAP [20], to adapt HAP for MDR-TB-related context and issues: the principles and process remained unchanged with the exception of removing detailed recording of the counselling sessions to reduce the already heavy burden of documentation within the study. The HAP counsellor used this BA approach, as well as the activation (or re-activation) of social networks and a positive problem-solving approach (all aspects of the original HAP intervention from India), to help the patient identify the links between what they did and how they feel, and then used this information to plan specific activities to begin to improve their mood. Monitoring the progress of treating the patient’s depression through HAP was done using PHQ-9 [21], which has been validated in primary care populations in Nepal [22]. It was administered before the first session and thereafter before every HAP session. If the first PHQ-9 screen score was below 10, patients did not receive HAP but were returned to monthly screening. Patients whose first PHQ-9 score was over 19 (indicating severe depression) or who expressed suicidal intent were referred to specialist psychiatric or medical services (as available at each study site). For patients who started HAP, sessions were continued until the PHQ-9 score reduced to less than 10, indicating restoration of the patient’s mental wellbeing or for up to 8 sessions (lasting for 30 to 60 minutes each). If the PHQ score was still 10 or above after 8 sessions then referral to psychiatric services was made.

HAP counsellors are typically lay people with at least high school education who receive specific training in delivering HAP from experienced counsellors. The initial plan was for the TB health workers to conduct the counselling, however, early discussions with participating centres made it clear that this would not be feasible. In light of this two additional staff from HERD were trained to deliver the counselling in the centres. Both counsellors had public health backgrounds: one had extensive experience of TB programme work, and the other had a physiotherapy bachelors degree, but neither had previous formal counselling experience. Although most patients in Kathmandu speak and understand Nepali, in Rupandehi a significant minority do not, so the counsellor there was chosen for a familiarity with several local languages. Different tools were used for monthly screening (HSCL and MSPSS) and HAP monitoring (PHQ-9) as we wanted to identify anxiety as well as depression in the monthly screening, which the HSCL tool does. The PHQ-9 was the recommended and evidenced tool to monitor HAP and so was used for this purpose.

Group support.

Interactions among patients in groups were used to increase patients’ understanding of MDR-TB and its treatment, reduce their negative emotions and improve their social support. Attendance at a group session was offered to all patients. A group of 8–15 people (mixed age and gender) with MDR-TB met regularly for up to an hour at the health facility, where they shared experiences and supported each other to face and overcome the challenges they were living with as a result of MDR-TB and its treatment. We planned a structured approach where the same group of patients would attend four consecutive sessions, covering important issues which would build trust and rapport between group members, increasing their ability to support one another. The group was facilitated by a counsellor, and often included a cured patient who shared experiences to provide motivation and instil hope in the group members. Group support has been shown to be beneficial in Peru where it has been a mainstay of MDR-TB care [23]. This aspect of the intervention package was developed in the second half of the study period after reviewing progress.

Training.

Six days training was provided to MDR-TB health workers and counsellors by the mental health NGO, TPO. This consisted of a recap of MDR-TB modular training provided by the NTP, orientation to the information materials we developed for the study, basic skills of counselling including rapport building, using screening tools and delivering HAP. A further two days training was provided to the counsellors by the study team covering recording and process documentation of the research study.

Quantitative data.

The quantitative data was collected from four sources between March 2015 and May 2016. The first was routine TB registers at the TB clinic. Data collected included: date diagnosed with MDR-TB, date started treatment, new or relapse case and sputum culture status. The second source was a data collection tool designed for this study, which collected self-reported socio-demographic data from the patient such as age, sex, education and occupation. The third source was the screening tools (HSCL, MSPSS and PHQ-9) which produced scores at each screening point for each patient. The fourth source was a researcher diary which collected process data from the counsellors and was used to identify such things as the percentage of patients who received different parts of the package and the time taken to deliver interventions. This measured fidelity to the planned intervention.

All quantitative data were manually input into a spreadsheet. Any outliers or gaps were checked with the counsellor and/or clinic registers. The whole dataset was then manually checked for consistency and completeness. Subsequently SPSS (version 20) was used to analyse the data and calculate descriptive statistics. As this was a feasibility study, no formal sample size estimation was carried out.

Qualitative data.

A semi-structured format of the researcher diary with space for free-format text was provided to enable the counsellors to reflect on the implementation of the pilot intervention (S2 Appendix). The diary was completed by the counsellors initially on a daily basis, and later reduced to weekly, bi-weekly and then monthly reporting. The counsellors also wrote reflective notes to summarise their observations of delivering the intervention. After each HAP counselling session, the counsellors also made some brief notes on the content of the session to assess fidelity to the model.

Interviews were conducted with the counsellors, some of the MDR-TB patients who received HAP and TB health workers. Interview guides were developed for each group based on the information within the daily diary, informal conversation with the counsellors, aspects of the intervention that we wanted to assess and areas where there were information gaps (S3 Appendix). Two rounds of interviews were conducted with the counsellors, one on month 9 of the intervention and the other at the end of the intervention, to more thoroughly understand the implementation and the context at the MDR-TB centres. The number of interviews was fixed with counsellors and TB health workers as there was a limited number of these staff (all were interviewed). Interviews with patients were continued until the interviewers considered data saturation had been reached.

Qualitative data from interviews were translated from Nepali into English by experienced translators and transcribed to English as literally as possible. These transcripts, daily diaries, reflective notes and HAP session comments were uploaded into NVivo (version 11). In accordance with the framework approach in qualitative analysis [25], two researchers thematically coded two interview transcripts and from this an agreed thematic framework was developed. This was then checked independently by a third researcher. The framework was then used to code all subsequent transcripts and diaries. Any deviant data were identified; further themes were considered and agreed between the two researchers, and the framework adjusted accordingly.

Ethics.

Ethics approval for the study was obtained from Nepal Health Research Council (Approval number: 130/2012) and University of Leeds Medical Research Ethics Committee (Approval number: HSLTM/11/026). Written informed consent was obtained from all research participants. Illiterate patients were informed verbally and confirmed consent using thumb prints. No incentives were provided to participants.

The screening process.

Within the first month of implementation the MDR-TB health workers stated that they were unable to perform the role of initial screening of all MDR-TB patients due to the heavy patient flow and time constraints. This was in spite of the fact that the intervention was developed with their full engagement and they had agreed this role. With limited opening hours between 10am and 2pm and MDR-TB patients from the sub-centres presenting on just two days per month, health workers could not manage the screening of the patients. Therefore, the counsellors themselves undertook the screening process which they had enough time to do. For the last month of the study the MDR-TB health workers were asked to use a locally developed ultra-brief screening tool[22] using two questions similar to PHQ-2, but using Nepali idioms of psychological distress. They did this and reported having enough time to undertake this for every MDR-TB patient, indicating the use of two questions as an initial screen is feasible without additional resources in the current TB service.

“If there is a separate person [for screening], then it is good to go through HOPKINS and MSPSS, but if he [MDR-TB health worker] has to implement in the regular system … then the two questions will be better. It is short and doesn’t take too much time.” (Counsellor 2)

One of the benefits of this revised delivery method using the counsellors was there was minimal waiting time for the patient to be screened. The counsellors managed to perform over 70% of the planned monthly screening sessions, providing plenty of opportunity to identify patients with depression, indicating the feasibility of this approach. Counsellors were also positive about using HSCL and MSPSS (compared to just PHQ-9) for screening because it provided richer information for the HAP counselling session as they knew more about the patient’s state of mind and social support. They reported it took them 10–15 minutes to complete the screening process for each patient.

Counsellors reported that, due to cultural factors, one question in the HSCL tool (relating to sexual interest) was uncomfortable to ask and for patients to answer. Other questions in the tools were acceptable. Counsellors also reported that patients referred for HAP complained about the repetition of the questions in the two screening tools (HSCL and PHQ-9).

Information and educational materials.

The convenience of giving the leaflet to patients led to the MDR-TB health workers providing one to a majority of patients. However, although the MDR-TB workers initially used the flipbook in their contact with new patients, this practice soon stopped. Although it only took 5–10 minutes to go through, this seemed to become a barrier to them using the flipbook. They stated that they gave the same messages but more briefly.

“It [the flipbook] is very difficult for us [to use for every patient]. It is practical in places that have a small number of patients. On the other hand, this facility has a large volume of patients here.” MDR-TB Health worker 2

“The things I used to give are also from the flip book. I finish up saying the things in two lines.”MDR-TB Health worker 2

All the information materials were in Nepali and for a majority of patents these could be understood; however, for a minority of patients at Rupandehi who did not speak Nepali, this presented a barrier. This was overcome by the counsellor translating the content into a local language.

Most of the patients found the information materials (particularly the pictures) helpful to understand MDR-TB and its management better, which reduced their mental stress (see quote from Counsellor 2 below). Some also stated that the intervention should be available in all MDR-TB centres.

“In the beginning, I was interested to know what it [the leaflet] was. I felt like reading it and I read it too. When I found out that it was about my own [condition], my ‘interest’ grew even more.” Rupandehi Pt 3, Female

“If we screen some patients and if we counsel them they said that their results came good. They also said that, ‘Sir, we didn’t know about TB and we didn’t know with whom to share. Because of this, we were depressed but after you came, we could share with you everything, we knew everything regarding its transmission. And after sharing these things I felt relaxed…’” Counsellor 2

The HAP counselling.

It took 30–45 minutes to deliver each HAP session. Due to the availability of the counsellor as an extra member of staff, the time patients spent waiting to receive HAP counselling was minimal.

Twelve of the 135 (9%) patients received the HAP counselling. The median number of HAP sessions delivered per patient was 2.5 (range 1–4). Only one patient brought a family member to the HAP counselling (for two sessions). With the counsellors also undertaking the screening, there was generally no waiting time between screening and delivery of the HAP sessions as they occurred consecutively. There was a single instance where two patients were identified to receive HAP on the same day resulting in a longer waiting time for one of them. However, a key limitation of delivering the HAP sessions was the lack of an appropriate, private space in which to conduct the counselling at the centre (as was the case with screening), which often created distractions for the patients and counsellors during counselling. The counsellors had to improvise, and either used rooms that were shared with other services or an outside space that was as private as possible.

A review of the counsellors’ reflections indicated the focus of the HAP sessions was on patient activities in line with the behavioural activation approach of HAP. Typical examples of activation included patients doing breathing exercises, intentionally talking with neighbours, starting to undertake manageable household chores, ensuring a good sleep routine and starting to work part-time in the family business. As the counsellors took a purely oral approach to activity planning without the use of paperwork, the counsellors reported that the patients with low literacy were able to engage in the counselling.

There was high attendance for HAP sessions. All patients appreciated the availability of someone with sufficient time in the centres to be able to share their problems and sorrows. This was particularly important for those who, due to the stigma of TB, were unable to talk openly about their condition with family and friends. In addition, all the patients interviewed also stated that they were willing to see our counsellors and follow the agreed activities of the HAP session.

“Before this, other people have come asking, but I never showed them my worries. And when counsellor asked us very nicely, giving us enough time … I shared all my problems with him … after talking with him I have improved a lot. I have felt very good after I met the counsellor … and I used to follow every little thing said by him.” Kathmandu Patient 2, Male

“What happens is we don’t share with other friends from outside about our disease. But he asks us about how our family is and what is going on. At least there is a person in the hospital who asks us about our family … and that’s a big thing for us” Kathmandu Patient 3, Male

Most of the patients identified that the behavioural activation approach using changes to daily activities was particularly beneficial.

“Before meeting the Counsellor, I used to think a lot … I used to spend time gazing a lot and continue walking unknowingly. Then the counsellor told me not to overthink and not to take “tension” … and used to teach me to do these things and told me if you will do these things then this [improvement] will happen. From then I didn’t take any kind of tension … after talking with him I have improved a lot” Kathmandu Pt 2, Male

“After the counselling, I tried to change myself. While I was in depression, I used to walk around thinking, ‘I do not care whatever happens’. Because of the counselling, I got the chance to change myself.” Rupandehi Pt 3, Female

“In my experience, one thing is that, they [patients] came out searching for him [counsellor] and no more psychological problems came to me … the problem has been reduced compared to last year. Not only the suicidal thoughts but also suicidal attempts had been done here in mine [my facility]” MDR-TB HW 1

Group support.

It was not possible to retain the same members for subsequent group sessions as planned because patients attended the clinics on different days of the month and times of day. Therefore topic-based group sessions proved challenging. Instead, the counsellors facilitated participants to identify the challenges they were willing to discuss in a group setting on a session-by-session basis. Groups were limited to 30 minutes, based on the willingness of patients to stay that long. As the group could only start once all the invited patients had seen the MDR-TB health workers, the session start was often delayed, resulting in patients requesting shorter group sessions. Overall, patients were positive about group counselling; however, some thought it may not suit everyone as some people would not like to express differences in opinions and some might worry more after hearing others’ problems. In addition, many patients expressed how beneficial it was to hear the experiences of a cured patient in the group.

“Seeing the people who have completed their course of medication will motivate the people who are continuing their medicines. On seeing such people, they will realize that they too can survive if they take medicines. They get motivated and they go on completing their course” Rupandehi Patient 3, Female

“In the beginning, I used to go there, bring medicines and take them. There was no one like that [with MDR-TB] in the village, so I used to stress out thinking that I was the only one with that disease. When I saw all the other people here [the treatment centre], I realized that there were others like me who were suffering from it. I got courage thinking that others have also gone through it and they have survived.” Rupandehi Pt 5, Male

Other challenges and benefits.

According to the intervention protocol, patients whose scores indicated anxiety or severe depression were supposed to be referred to a psychiatrist. No psychiatrists were available within the study health facilities so cases were notified to the MDR-TB health worker, who then followed up with a physician or psychiatrist at another facility.

The MDR-TB health workers expressed their views that the intervention was useful and a valuable part of the service that they were offering. While they had limited time to deliver the intervention they did provide support to facilitate the intervention process by arranging space for counselling and encouraging patients to engage in the components of the intervention.

“The things that we have been implementing with our patients [the intervention package] have been really good for the patients and we have had no defaulters at all. They now come to inform us immediately whenever they have a problem. They used to hesitate to come and talk to us before. We feel that these things [the intervention] have made the situation far better than before.” MDR-TB Health worker 1

Preliminary HAP outcomes.

Our sample size was not powered to report effectiveness of HAP. That said, the 9 patients who completed HAP had a mean reduction in scores on both depression tools (7.9 points on HSCL-Depression and 2.7 points on PHQ-9). This is an encouraging signal of the possible impact of HAP on depressive symptoms in this setting, but no firm conclusions on effectiveness can be made. Minimal change occurred for the HSCL-Anxiety scores after HAP, compared with before HAP.