Acceptability and Impact of Group Interpersonal Therapy (IPT-G) on Kenyan Adolescent Mothers Living with HIV: A Qualitative Analysis

Background Task shifting is a well-tested implementation strategy, within low- and middle-income countries (LMICs), that addresses the shortage of trained mental health personnel. Task sharing can increase access to care for patients with mental illnesses. In Kenya, community health workers [CHWs are a combination of community health assistants (CHAs) and community health volunteers (CHVs)], have played a crucial role in this front. In our study, we seek to assess the acceptability and feasibility of IPT-G delivered by CHWs among depressed postpartum adolescents (PPAs) living with HIV. Method Twenty-four PPAs were administered IPT-G by trained CHWs from two health centers. A two-arm study design (IPT-G intervention and treatment as usual) with an intent to treat was used to assess acceptability and feasibility of IPT-G. Participants who scored >10 on Edinburgh postnatal depression scale (EPDS) and who were 6–12 weeks postpartum were eligible for the study using purposeful sampling. Participants were equally distributed into two groups: one group for intervention and another for wait-list group. This was achieved by randomly allocating numerical numbers and separating those with odd numbers (intervention group) and even numbers (wait-list group). Focus group discussions (FGDs) and in-depth interviews ascertained the experience and perceptions of the postpartum adolescents and the CHWs. In addition to weekly face-to-face continuous supportive supervision for CHWs, phone calls, short messages services, and WhatsApp instant messaging services were also utilized. Results The CHWs found the intervention useful for their own knowledge and skill-set. On participation, 21out of the 24 adolescents attended all sessions. Most of the adolescents reported improvement in their interpersonal relationships with reduced distress and lessening of HIV-related stigma. Primary health care workers embraced the intervention by availing space for sessions. Conclusion Our study demonstrates possible benets of task shifting in addressing mental health problems within low-resource settings in Kenya and group IPT is demonstrated to be both acceptable and feasible by health workers and adolescents receiving care. Acceptability and feasibility of IPT-G when delivered by a non-medical specialist in routine clinical settings was assessed. We conducted our intervention between August 2018 and July 2019. Twenty-four postpartum adolescents (PPAs) aged 15-24 years and living with HIV were recruited from two sites: Kangemi health center and Kariobangi health center. They were also expected to be 6-12 weeks postpartum. The two study sites were chosen since it is situated on the two opposite poles of Nairobi city, and both are home to a population of low economic status. CHWs included CHAs, (n=2), and CHVs, (n=6) delivered IPT-G to PPAs. Both study sites had same numbers of CHWs.


Background
Shortage of mental health personnel and associated challenges Chronic shortage of well-trained health personnel in Sub-Saharan Africa (SSA) necessitated task shifting as a key implementation strategy. Task shifting in SSA health systems typically mean CHWs are trained to help increase number of health services being provided to reduce cost and improve delivery of care (WHO, 2007). The concept of task shifting involves rational distribution from well-trained to less specialized health workers or those who have been trained on a limited time duration on speci c skills in a given area of need (WHO, 2006). Several barriers to task shifting revolve around the need to strengthen health systems by improving systemic and physical structures (Dawson et al., 2014). CHWs undergo challenging experiences while task-shifting: undertaking tasks that may drain them emotionally and physically, persistent problems of inadequate training, unstructured supervision, and poor remuneration or complete lack of reward or entitlements to any form of bene t in some instances (Mundeva et al., 2018).
Task shifting is meant to reduce workloads for overburdened specialist health workers and improve patients' linkage to services (Mwai et al., 2013). In Kenya, CHWs are not well compensated and sometimes expected to work on voluntary basis with poorly structured responsibilities causing them to take roles requiring more skills than their abilities (Angwenyi et al., 2013). Embracing task shifting of key preventive and promotive activities in the HIV program using CHWs promises to be a good step towards achieving the 90-90-90 goals. UNAIDS, too, has identi ed the engagement of community workers as essential in HIV prevention and advocacy (UNAIDS and Stop AIDS Alliance, 2015).

Prevalence of HIV in peripartum adolescents in Kenya
Our study is targeting one of the most vulnerable youth populations: adolescent mothers living with HIV. In SSA, adolescent girls aged 15-24 years represent 10% of the population, and using 2017 estimates, this group accounts for 25% of the new HIV infections (UNAIDS, 2018). The prevalence of HIV in Kenya among female adolescents aged 15-24 is 4% (NACC, 2016), and these young girls are two times more likely to contract HIV than their male counterparts (NASCOP -Kenya, 2016).
The overall prevalence of adolescent pregnancy in Africa is 18.8% and 19.3% for the Sub-Saharan region (Kassa, Arowojolu, Odukogbe, & Yalew, 2018). By the age of 18, 42% of adolescents from SSA living in urban areas will have become pregnant, and more than 50% of their rural counterparts would be so (UNAIDS, 2019). This early pregnancy also increases their chances of HIV infections (Christo des et al., 2014). Most adolescents are seen to be infected with HIV by older men aged late 20s and early 30s who may not even be aware of their status and thus, unlikely to be on anti-retroviral therapy (ART) (de Oliveira et al., 2017).
A study in Malawi found that adolescents living with HIV had a depression prevalence of 18.9% (Kim et al., 2014). A similar study in Kenya focusing on mental health outcomes among adolescents living with HIV documented almost same depression prevalence of 17.8% (Kamau et al., 2012). Worldwide, postpartum depression (PPD) prevalence in adolescents is higher than that for adults, and it is estimated

Role of psychosocial interventions for adolescents
Psychological interventions have been recommended for persons living with HIV to mitigate common mental health illnesses, including depressive illnesses (Sherr et al., 2011), with no side effects against ART (Cruess et al., 2003). Psychological interventions for adolescents should aim at addressing issues related to psychosocial development, training on social skills including life skills, and shaping their behaviors towards a future productive life through livelihood and vocational training (Martinez et al., 2014). It has been found that adolescents have signi cant risks of frequent unprotected sex with an equally high risk of contracting HIV and having an unplanned pregnancy or both at some time (Schunter et al., 2014). Adolescents engage in risk-taking behaviors and girls may be less assertive to negotiate for condoms; hence, higher chances of unprotected sex are serious considerations (Januraga et al., 2014) in trying to address well-being of adolescent girls.
Prevention of mother-to-child transmission clinic (PMTCT) has provided a conducive environment for addressing life challenges associated with HIV infection among perinatal women. In the year 2017, global ART coverage among men aged 15 years and above was 53% compared with 65% of women for the same age (UNAIDS, 2018). In a recent study from six sub-Saharan countries, HIV-related stigma has been associated with delays in treatment and di culties in adhering to ART, as shown in systematic review studies (Ammon et al., 2018;Croome et al., 2017). An emerging nding of note is that depression treatment improves adherence to ART, which is key to improved quality of life (Sin & DiMatteo, 2014;Wagner et al., 2020).
A Nigerian study on psychological intervention for adolescents living with HIV utilized support groups using Facebook groups for a 5-weekly session. The ability for adolescents to interact, learn more about HIV, share experiences, and their fears on social media was seen to help them in coping with their status (Dulli et al., 2018).
Group Interpersonal Psychotherapy (IPT-G) and its relevance for this vulnerable population A study assessed interpersonal relationships between youth and their families and found out that poor relationships led to depression (Okawa et al., 2018). Lower caregiver supervision was also associated with higher depression (Bhana et al., 2016). IPT-G is well-poised to help address interpersonal di culties that adolescents are predisposed to. It is conceptualized around four problem areas: grief and loss, interpersonal role disputes, role transitions, and interpersonal de cits/social isolation. During adaptation of IPT-G for depressed adolescents, Mufson et al (2004) reported that the intervention targets an individual's interpersonal relations with other persons in a given family/society where therapeutic bene ts are achieved (Mufson et al., 2004). For example, when postpartum adolescents living with HIV are put together in a therapy session, they appreciate that their unpleasant experiences in life also affect other persons in similar situations. This motivates the adolescent to try new interpersonal interactions that enhance better social functioning in society. IPT-G provides adolescents with peers who have similar di culties and utilizes synergies in groups to understand their interpersonal problems and develop new ways of coping.
Our study is in line with a call by WHO to embrace strengthening of Universal Health coverage. Primary health care depends on health system structures acknowledging various levels of health care services where specialized health workers are deployed at the referral levels and CHWs are based at the community level and both are interdependent (Ministry of Health, 2020; WHO-Unicef, 1978; WHO, 2017).
We aim to assess qualitatively the acceptability and feasibility of IPT-G for postpartum adolescents living with HIV being delivered by CHWs within routine clinical settings in Nairobi, Kenya.
Screening for depressive symptoms was carried out using EPDS (Cox et al., 1987) to monitor change in scores following IPT-G intervention. Interpersonal inventory was used to assess the interpersonal relations (emotional attachments and social interaction) among the PPAs. IPT Knowledge test with 15 questions with a maximum score of 30 was used to assess CHWs skills and performance. If 70% of the items (21 points) are answered correctly, it a rms competence of an IPT practitioner.
Acceptability and feasibility of IPT-G delivered by CHWs was assessed through FGDs among PPA (n=19) and CHWs (n=7) for them to share their experiences and perception about IPT-G. In-depth interviews were conducted with the medical staff within the two sites: nursing o cer-in-charge (n=2), mentor-mother (n=2), and CHAs (n=2). The audio was recorded for all sessions to enable auditing of each weekly session as they progress to improve therapy quality. Field notes were kept during every session to document observations, experiences, and perceptions of CHWs, PPA, and other health staff within the facility (see Table 1).

Data analytic approach
Transcription of audio-recorded qualitative data and eld notes were coded to identify common emerging themes highlighting barriers and facilitators of IPT-G. We conducted FGDs for CHWs and PPAs who had participated in the study and in-depth interviews for CHAs and health care management staff within the two health centers to understand their experiences and perceptions (McCain, 1988). Finally, we determined feasibility by evaluating the recruitment process, weekly observations during sessions, and completion rate of the intervention session whereas acceptability was determined by inquiring how individual participants felt about the intervention, participant's knowledge, perceptions, barriers, and gains achieved from the intervention (Sekhon et al., 2017). (See Table 2).

Results
Among the 24 participants, the majority (21,87.5%) were aged between 21-24 years and were mostly residing with their partners (17,70.8%) with parity of less than 2 children (20,83.3%). About half had attained secondary and above education level (despite unemployment ranking high at 79.2%), and with most of them earning less than 100 USD per month (22,91.7%) (see Table 1).

Qualitative acceptability and feasibility data
CHWs found the intervention useful in terms of how the intervention built their knowledge and skills and successfully delivered IPT-G. During the training of lay health workers, we found that they demonstrated a good understanding of depression and reported competency in delivering IPT-G. The adolescent mothers bene ted from the IPT-G as they narrated how they could now function better, communicate better with their families and partners, interact socially, manage their anger, and even resume work for an income or education (see Table 3).
Our adolescent participants acknowledged that IPT-G helped alleviate social isolation, anger, hopelessness, and low mood, which are typical depressive symptoms. After IPT-G intervention, our adolescent participants can now acknowledge they lived with horrible thoughts and a feeling of hopelessness around their past and felt liberated and at great ease with their new situation. They narrated how they were able to socialize with others and perform their family responsibilities effectively, including looking after the baby without negative self-perceived stigma despite them living with HIV (see Table 4).
The study had a retention and follow-up rate of 21 (87.5%) out of the 24. Among the 8 CHWs recruited to deliver IPT-G, one of the CHV unfortunately suddenly passed away though remaining 7 CHVs who participated successfully completed the delivery of intervention. One of the CHA from Kangemi health center stated that the IPT-G delivery process was workable considering it was held weekly; hence that sort of time allocation was possible (see Table 5).
Capacity building through a collaborative care approach was used, which involved engaging Director of Mental Health Services and cascading the partnership along the health care management team downwards to the level of CHWs. The entire health management team was very supportive of our intent by issuing us with clearance and linking us to all the relevant clinical staff (See Figure 1).

Pre and post IPT-G changes observed
In both study sites, there was a shortage of trained mental health personnel. Kangemi health center has only one nurse in training under sponsorship of a non-governmental organization. In Kariobangi health center, there are only two health care workers with post-diploma training in psychiatry. There was no mental health designated space, and counselors under HIV testing services (HTS) were operating in mobile tents. After our study, most health care providers within the two health centers had appreciated the impact of IPT-G on adolescents attending prevention of mother-to-child transmission clinic (PMTCT), citing improved social functioning, better communication, and appealing personal hygiene/grooming by the adolescent mothers.
Training of CHWs on IPT-G Screening questionnaires were administered to CHWs before training to assess their knowledge around mental health concepts, such as stress and depression. They all demonstrated a fair understanding of the difference between these. The CHWs appreciated continuous supportive supervision, and they felt that all their concerns arising from weekly sessions were being addressed promptly and adequately. Furthermore, it was very encouraging to hear from the CHWs that our participants after intervention became role models in the community by imparting skills of managing their day to day issues of life (see table 5).
The loss of one of the CHVs emotionally affected most of our participants, and it could be because he was youthful and could easily identify with him during subsequent sessions. The lead researcher, clinical supervisor, and research assistant organized a loss and grief therapy session for the group (adolescents and CHWs) using one of the IPT problem areas. We also visited the family and arranged for tree planting within the facility, whereby family members and clinical staff where invited for the function, as a means of bringing closure for all who knew the deceased CHV. We noti ed the ethics o ce of the incident, which was documented in our protocol.

Discussion
CHWs found the intervention useful in addition to their existing knowledge and skills and were able to deliver IPT-G successfully. Our research team found that through task shifting, IPT-G can be disseminated to other adolescents in similar settings so long as recommended by WHO, training and continuous supervision of CHWs is upheld (WHO, 2007). We can now embrace ndings from other studies where nonspecialist (CHWs) can be trained on speci c skills to deliver an intervention with similar effectiveness just like when being delivered by specialists in mental health (Kredo et al., 2014;Murray et al., 2017). In cognizance of Universal Health Care, we too lend our voice towards the need to consider CHWs to be involved in mental health care delivery to cover for the shortage of trained mental health specialist (Ministry of Health, 2020). The CHWs were very satis ed with their achievements, and some shared their experiences on the milestones towards IPT-G delivery: "At the beginning of the sessions, there were challenges because all the people were still new, so it is later that people came to know each other and developed trust in the group and everyone could say all her issues". (SK, age 50, Kariobangi) The intervention empowered some of the adolescent mothers to help others, which was one of our intended purpose towards disseminating bene ts to the community level.: "Even the sessions were very good; because some used to come and tell you, I have this and that problem and I passed through this, but I went and did what you told me, and I have succeeded, you get it? One member in our group, I recall, and even at the moment, she normally rings me and tells me, 'that thing helped me, and I am able to help others'." (NG, age 43, Kangemi) The IPT-G sessions formed part of the entry point to enlarging the scope of their social support and even after the intervention they continued to share their issues: "So, it was a very big problem at the beginning, but I am glad that as the sessions were proceeding, we became good friends; actually, we developed a rapport, and even until now, some still call 'when are we meeting again?' [Laughter] --." (EO, age 24, Kariobangi) Others were of the view that follow-up is of importance because most of the participants open-up to their issues at the middle of the sessions: "What I can say; when I joined-let me talk about group two, we came well but after attending like three sessions is when you could see them now pouring out [Laughter] they tell you all their troubles in life--." (FC, age 30, Kangemi) The CHWs were now able to connect easily with the adolescent mothers after bonding with them during the sessions: "So, we are also still doing follow-ups. Some even call by themselves, sometimes I call them, some even when they come for clinics, they just come looking for us which is good so, so they trust us. So, there is that trust, there is that friendship, there are so many bene ts that came out of this, so I was just gladly sharing." (EO, age 24, Kariobangi) Our adolescent participants bene ted from the IPT-G as they narrated how they could now communicate, interact with others, manage their anger, and even resume work for an income. It is worth noting that in a study that assessed interpersonal relationships between youth and their families or caregivers it was found that poor relationships were strongly related to depression symptoms (Miller et al., 2016;Spence et al., 2016).
"I am called JI, and it helped me now I can talk to people, I was hopeless but now…. yes, I am of importance. When I saw that I am this way, I felt that there is no life and that I am not important." (JI, age21, Kangemi) One of the adolescent mothers narrated how HIV-related stigma used to torment her before the before our intervention: " I know myself since I know how I used to feel; when I sit down, I ask myself, 'what kind of life is this?' I used to feel guilty. I don't want people's stories. I just feel that a story may arise and reach the point of the infection, so what will I say? You know at that point you will just be forced to remain silent; you will not talk-you will never have what to say because they are negative, and you are positive, and they want to talk about that infection --." (JA, age22, Kariobangi) It was clear that irritability was negatively affecting communication with their partners as one of the participants narrated how she used to live while having depressive symptoms: "In the past, whenever I could get angry, I could not cook…. I cannot eat, if it is talking to people, I cannot talk to him; everybody sits apart when it comes to washing, everybody washes their clothes." (SK, age23, Kangemi) Besides, adolescent participants could appreciate that IPT-G helped them alleviate depressive symptoms (social isolation, anger, hopelessness, and low mood). Most of them reported better sleep patterns, good appetite, increased social interactions, and decreased HIV-related stigma. Our study population was noted to have several challenges which seemed to support previous studies where they found out that adolescents living with HIV are more vulnerable to mental health disorders due to medical and psychosocial stressors associated with HIV/AIDS (Mellins & Malee, 2013;Nanni et al., 2015). There is no doubt that our intervention addressing depression will improve adherence to ART among our study participants, which is key to good quality of life (Sin & DiMatteo, 2014). "It has helped me when I am with people I have accepted myself the way I am and then anger issues nowadays are not there, at least I can make friends; in the past, I could not make friends, but now at least I can sit with somebody and share something with her that is helpful." (SA, age23, Kariobangi) Isolation was one of the maladaptive behaviors associated with depressive symptoms as indicated by on of the adolescent mother: "I used to lock myself in the house, but currently, I can get out and make stories with neighbors, when I am called for a job I can go." (RA, age 18, Kangemi) The male partner's role towards how relationships evolve through pregnancy and motherhood among adolescents is worth looking into to help adolescent girls in abusive marriages. One qualitative study suggested that postpartum adolescents living with their partners or caregivers will bene t from social support as postpartum adolescents transit early motherhood challenges (UNESCO, 2017). See vignette from our participants here highlighting this point further: "I am FW; it has helped me I am not the same as I used to be; I used to be angry; I could get out of the house at night due to anger. Now we talk well in the house… yes [Laughter]." (FW, age21, Kariobangi) We noted the devastating effects of depression as highlighted by one of the participants on how isolation, marital con ict, and persistent distress used to affect her daily living: "I think I was depressed; I used to sit in the house and had no friends. But when I started coming to this group, I found friends here, outside I have also made friends. I used to be stressed as to why my husband doesn't go to work and frequent disagreements. But since I started coming here for advice, I realized that I am not the only one who has problems, so for me, it has gotten out; I just feel that it has helped me a lot, stress is usual, but for now, I feel stress free." (FW, age21, Kariobangi) Our study acknowledges that adolescents living with HIV are vulnerable to negative community perceptions, which agrees with a similar study in Uganda (Ashaba et al., 2019). These could manifest with suicidal ideations. "I can remember there is one who said after the sessions, 'nowadays I can get out and talk with women, I can go out of the gate because when I was alone, I felt that I am not okay because I have HIV and I have given birth at a younger age. But now I have gotten that courage after these sessions to go outside I can talk, I have that courage,' so that made her feel that they are many and she is not alone." (LN, age 34, CHA, Kangemi) One of the CHV encountered sudden death outside the study area, which caused a lot of grief among the participants. Besides, we realized most of our CHWs had experienced the loss of a loved one and were also struggling with healing, which our sessions also supported them as they narrated to us. The CHWs were affected by the loss of a colleague (CA), and in addition to other previous bereavement worsened their state (Shear, 2012), and IPT-G helped them cope: "CA also left me, he was my friend, and it drew me so much down, but now I am ne, isn't it? --During CA's time of demise, it was like everybody in Kangemi felt like was carrying something [Silence]." (LN, age 34, CHA, Kangemi) One of the CHW expressed how IPT-G helped her process the thoughts and emotions of loss and grief associated with past incidents: "It has helped me a lot, the second thing let me say I was given a husband by Kangemi people, when we were doing training, and I lost the husband (refers to CA, deceased CHV), so IPT helped me to go through the loss and grief, anyway it was not a real husband [Laughter]. But he was a very good friend of mine.
Anyway, when I was undergoing through the sessions, I lost most of very important people, but IPT helped me. You know when we were talking to these girls, and they are also expressing, 'I lost my kid, I lost my daddy,' and I was also like it was like me losing the people whom I care about, so we were going through these together. It was a process for all of us, so it healed me, and it healed them; that was very good." (EO, age 24, CHV, Kariobangi) The community health assistant was mourning the loss of her child during the sessions and had to say this in appreciation of the intervention: "--and when I came here, I was so much stressed, and it could have resulted in depression because there is something, I lost my child on delivery, then I lost both parents at the same time. So, when we started talking, I felt that I had put the load down, you get it? so when we continued and reached the middle, and I realized that I am good." (NG, age 43, CHV, Kangemi) Continuous supportive supervision was appreciated by the CHWs and they felt that all their concerns arising from the weekly session was addressed promptly and adequately. During our group sessions, supportive supervision emphasized joint problem solving, mentoring, and two-way communication "It was just like Kangemi; at rst for the people with whom you are not familiar enough, they could not open up; you ask her a question, and she feels like where do you want to lock and take her? But now the second time they will be free, and the third time she will feel like 'I can remember what happened to me and how it is on somebody else,' because we had supervision like from CHA and MK and could correct us when wrong and it becomes normal." (SN, age 65, Kariobangi) Furthermore, it was very encouraging to hear from the CHWs that our participants after intervention became role models in the community by imparting skills of managing their day-to-day life issues. For CHWs and adolescents to be able to understand the concept of IPT-G and even use it to help others, the learning cycle was achieved by both actual experiences and empowering them to practice through supportive supervision (Huber, 1991;Newell, 2005) "I think even these girls have become role models in the community…we told them 'if you nd somebody who is in a situation maybe you can assist and feel that it closer to yours or it is similar to yours, you are now empowered, and you can help this person at your level and if you feel that it is di cult is when you can refer.' Still, now they are doing on their own, which is good." (EO, age 24, CHV, Kariobangi) One of the CHA a rmed that time for IPT-G sessions was easily embedded into their routine activities since it consumes 90 minutes only once a week.
"Not a burden because it was usually once a week, and if you are meeting those clients later, it is the one in which you can negotiate the time that you will meet, so I don't think, maybe if others think otherwise." (LN, age 34, CHA, Kangemi) Nevertheless, several CHWs acknowledged the lack of designated space for mental health services as one of the limitations towards adopting IPT-G delivery to routine PMTCT services: "I think for us (in) Kariobangi we had a big challenge when it came to the venue of the meeting because sometimes you nd we are here sometimes we are displaced at the tent, maybe the other tent is very dirty, sometimes we are in this other tent we come here displaced, so it was a very big challenge [Cross talk] --." (EO, age 24, Kariobangi) The space to administer mental health services was missing within the health centers as shown by how we had to improvise a room for the sessions: "We were okay; we had been given the maternity, a place somewhere…. You see, if we close the middle of the room too and pull the curtains, it was so good; it didn't have an issue." (LN, age 34, Kangemi) Another CHW seemed helpless of the situation and consented to the state of the available space despite all the challenges associated with it: "But let's just say it was okay; even if it is bad, it is our place, so we cannot say that it is bad [Laughter], but it was ne." (JA, age 24, Kangemi) The lack of speci ed venue for mental health services posed uncertainty on effective service delivery for the CHWs: "The space was okay; it was somewhere where we don't have issues with noise, but we don't know next." (SN, age 65, Karionbangi) We were able to form a strong collaborative team by creating a WhatsApp group and took issues of the CHWs and PPA with a lot of importance by demonstrating practical empathy, thus improving commitment for all of us in the study process. The use of technology proved equally useful, just like a Nigerian study on psychological intervention for adolescents living with HIV where support groups were engaged using Facebook groups for a 5-weekly session. The ability for them to interact, learn more about HIV, share experiences, and their fears a rmed that social media could help them cope with their status (Dulli et al., 2018).
"When I lost my child on delivery, I also want to thank YO, I don't know what to tell him, but there is a time I looked and realized that he is a very different person. We hadn't known each other-we had known each other for only three days, but when I had a problem he came with MK to my house and consoled me, and I felt that it wasn't even about friendship, they are just part of like my family, and he helped me." (LN,age 34,CHA,Kangemi) One of the CHW underscored the potential impact IPT-G could have in the community as indicated in her explanation: "--it has changed us, as much as we were the teachers, but we feel that it helped us. So I just pray that more of that kind comes and the way I said that when you just change one person he/she will change ve others and those ve others will change others and at the end of the day you may nd that you have changed the whole country and even the world, so thank you". (JA,age 24,Kangemi) Generally, our ndings of challenges affecting PPA did not come as a surprise considering previous studies had also identi ed that perinatal adolescents, who are also parenting, face several di culties, such as social stigma, lack of emotional support, poor healthcare access, and stresses around new life adjustments (Kumar et al., 2018). Our IPT-G intervention helped adolescents by improving their interpersonal relationships, communication processes, and overall mental health.

Strengths
We were able to qualitatively establish acceptability and feasibility of IPT-G delivered to postpartum adolescents living with HIV by CHWs in primary health care settings. The two study sites make the ndings more reliable for a representative outcome for low-resource urban settings.

Limitations
Despite all the exciting, positive impact on the participants' lives from this intervention, as seen from the qualitative ndings, the sample size was not powered. Thus, other quantitative results may not offer strong enough argument when discussing the intervention's impact on this vulnerable population.

Conclusion
The shortage of trained mental health workers has led to inaccessibility of mental health services both in urban and rural settings. To enhance mental health services to the broader population, our study sought to assess acceptability and feasibility of IPT-G being delivered to postpartum adolescents living with HIV by CHWs in Nairobi area at primary health care within routine clinical settings. Our ndings a rm that this intervention has a promise of being delivered by CHWs for this speci c population and the need for supportive supervision to enhance delity. We recommend a follow-up study with a larger representative sample size from diverse community within urban and rural settings with a view to scaling-up IPT-G at the primary health care level in all counties.    "I had an issue with my parents, but now we are okay, and we communicate well". (RJ, age23, Kangemi) "About my estranged partner, I can say that at the moment, even though we have not yet met, we are ne because we are communicating." (MM, age24, Kangemi) "This group has helped me in terms of communication." (JF, age24, Kangemi) "It helped me to interact with people." (RA, age21, Kangemi) Improved anger management "I am called S, but that one is informal, but in reality, I am called NC; by the way, even me, it has helped me so much, I have seen that several things have changed. The anger I used to have is no longer there; I just feel that I am okay." (NC, age24, Kariobangi) "For me, I got assistance because even now I see that getting angered is not so much there, I just see that life is okay; I don't want-I mean I know how to control anger, and it cannot rise the way it used to happen to me." (JA, age22, Kariobangi) IPT-G delivery process "They were nice on my side, caring, they used to concentrate on us, they were social; let me say they were just nice." (JF, age24, Kangemi) "They used to understand us." (JI, age21, Kangemi) "They were nice people; in case you got stuck on something they could help you [Inaudible]." (SK, age23, Kangemi) "They used simple terms….and for those that were di cult they used to elaborate." (MR, age24, Kangemi) "They were just using simple terms." (RA, age18, Kangemi)

Relevant Quotations
Overall Perceptions towards IPT by postpartum adolescents "I can tell him(lead researcher) thanks because he has helped us so much, we had stress, we were lonely but the way he organized this group it has helped us and we have found means of helping our colleagues out there and that he should just continue that way without giving up and God grant him strength and life." (FW, age21, Kariobangi) "I can tell him (lead researcher) that he assisted us so much because if some of us could still be where we used to be then, we wouldn't exist till now; but he did an important thing and assisted us so he should continue that way and may God bless him." (NC, age24, Kariobangi) "I can tell him (lead researcher) thanks for creating this organization it has helped all of us on how we can also educate other people to be happy, and he should just continue with that spirit." (LA, age19, Kariobangi) "I can tell him thanks so much, he helped us a great deal because we came here with stress, but we have been assisted, and we say thank you so much and it should just continue this way [Crosstalk]." (SA, age23, Kariobangi) "I want to tell him (lead researcher) thanks for he has made me able to believe in myself, and may God give him strength to proceed with this program for others who are behind and are like us to get assistance." (EA, age24, Kariobangi) "For me, it had helped me, I have lived positively and I am not bothered by what people say [Some silence]." (LA, age19, Kariobangi) "I am EAO and for me this content made me more aggressive, I mean I am not the way I used to be; I have many differences such that when I walk in other places I am not afraid I know who I am now, I have decided to live a positive life and I am used to it, and if I see someone else I teach her the way I have been taught." (EO, age24, Kariobangi) "It was worse until hey! I used to keep quiet so much because I used to feel whenever somebody talks to me just a little then I get angry and I just wish that we ght [Laughter] I mean I used to get extremely angry --." (JA, age22, Kariobangi) "I was depressed but I had several regrets as I was just questioning myself but nowadays it is over. The regrets were 'why have I become pregnant early?" (LA, age19, Kariobangi) "I am another one; whenever someone could anger me just a little I could moody, sitting in the bed with tears, and when I cook the food cannot even be eaten because of too much salt [Laughter] or raw ugali; but for now I have changed and I am good, even if you talk I act like I am not hearing or I go and sit outside and come back when he is quiet." (JN, age21, Kariobangi) "Before I joined this group, I was so depressed; I used to be an angry person, my anger was so much near, even if someone wrongs me where we live then I carry it into the house on the kids and even on the husband; but since I came here, I have changed." (LA, age19, Kariobangi) "--at the moment I am stress-free and I just know how I can handle it; I sleep and wake up when I have forgotten those things but I normally thank God because I have never had stress." (RJ, age23, Kangemi) "It has been a nice thing, every time I see you. I normally get impressed and just smile because at least there is progress in our lives, we are not the same as the way we used to be okay?" (EO, age24, Kariobangi) "I think that will help a lot because you see for the young girls you might nd that when they do HIV test the husband is negative and the lady is positive so you see the girl gets stressed because she is harassed by the husband "where did you get the virus?" And whatever, it will help a lot [Silence]." (JA,age 24,Kangemi) Building capacity and acknowledging the team "Follow-up of that client, you have done IPT for this client you have done termination session; you have referred that client from facility to the community, who will do that follow-up? The community health workers. They are the linkage between the facility and the community and community to the facility." (LN,age 34,Kangemi) Third IPT session is when disclosure happens "--yes, they exist; when we started, we had a challenge because one we had not been trained, we had a challenge but we reached where we reached. I can say that even those girls when we start with them there are usually problems but the good thing is that mid-way they pick and can listen to you and sessions go well." (JA,age 24,Kangemi) IPT Impact in the community "In fact we always talk about that group wherever I go and maybe I meet a partner I always talk about that group, that's why I am saying I am still looking for a partner to support them, I always talk about it because we felt that we changed them and we wouldn't like them to get lost on the way, we would love them to come back and help the others." (LN,age 34,Kangemi) "I think there is a time when we began IPT, those girls most of them were locking themselves in the house.…. but now after the sessions in Kariobangi, we realized that they went back to work. Most of them have gone back to work, some other have started businesses, MA got a job and also has a business, EL has a business. So, people have gone back to work, there is SA has gone back to work; people are going back to work, people have started businesses but for one or two we are still following." (EO, age24, Kariobangi) Group strengthens cohesion and empowers "She (PPA) realizes that she is not the only one and that they are several, don't you even see that the rst thing they were accepting so much whenever she hears that this one is also like me, even this one and the other one so she feels that 'we are many." (NG,age 43,Kangemi) "It used to help them to unwind because maybe somebody has an issue and feels like 'maybe it is only me who has this issue' so when she nds somebody in the group who says almost like similar to hers then she also opens her heart and speaks." (EO, age24, Kariobangi) Figure 1 Collaborative structure for IPT implementation