Barriers and facilitators to chemotherapy initiation and adherence for patients with HIV-associated Kaposi’s sarcoma in Kenya: a qualitative study

Background Kaposi sarcoma is one of the most prevalent HIV-associated malignancies in sub-Saharan Africa and is often diagnosed at advanced stage of disease. Only 50% of KS patients who qualify for chemotherapy receive it and adherence is sub-optimal. Methods 57 patients > 18 years with newly diagnosed KS within the AMPATH clinic network in Western Kenya were purposively selected to participate in semi-structured interviews stratified by whether they had completed, partially completed, or not completed chemotherapy for advanced stage KS. We based the interview guide and coding framework on the situated Information, Motivation, Behavioral Skills (sIMB) framework, in which the core patient centered IMB constructs are situated into the socioecological context of receiving care. Results Of the 57 participants, the median age was 37 (IQR 32–41) and the majority were male (68%). Notable barriers to chemotherapy initiation and adherence included lack of financial means, difficulty with convenience of appointments such as distance to facility, appointment times, long lines, limited appointments, intrapersonal barriers such as fear or hopelessness, and lack of proper or sufficient information about chemotherapy. Factors that facilitated chemotherapy initiation and adherence included health literacy, motivation to treat symptoms, improvement on chemotherapy, prioritization of self-care, resilience while experiencing side effects, ability to carry out behavioral skills, obtaining national health insurance, and free chemotherapy. Conclusion Our findings about the barriers and facilitators to chemotherapy initiation and adherence for KS in Western Kenya support further work that promotes public health campaigns with reliable cancer and chemotherapy information, improves education about the chemotherapy process and side effects, increases oncology service ability, supports enrollment in national health insurance, and increases incorporation of chronic disease care into existing HIV treatment networks.


Methods
This qualitative study is nested within a larger longitudinal epidemiological study of newly diagnosed adults with HIV-associated KS in western Kenya, where participants were enrolled using rapid case ascertainment (RCA) methodology promptly after diagnosis [11]. Of this group of patients, we purposively sampled patients for in-depth semi-structured interviews.

Study setting
This study was conducted at the Academic Model Providing Access to Healthcare (AMPATH), a network of HIV primary care clinics in western Kenya from March 2019 through September 2019. This network includes 60 different clinical sites; participants were recruited for interviews from the following seven AMPATH locations: Moi Teaching and Referral Hospital (MTRH), Chulaimbo, Busia, Kitale, Webuye, Mukhobola, and Usain Gisu District Hospital (UGDH). MTRH is the tertiary referral center for AMPATH, which provides oncology care.

Participant selection
Eligible participants were HIV-positive, > 18 years, and had been enrolled in our parent study to rapidly identify cases of Kaposi's sarcoma [11]. Participants were recruited by purposive sampling to include patients who i) had not initiated chemotherapy after a period of > 30 days from when a provider recommended chemotherapy initiation, ii) who had initiated but not completed a standard chemotherapy course, and iii) completed chemotherapy. Potential participants were approached by the research team in person or by telephone. Those willing to participate were screened for inclusion criteria.

Data collection
The interview guide and initial coding framework were based off of the situated Information Motivation Behavior model (sIMB), developed by Amico et al. as a model for care initiation and maintenance for chronic medical conditions [12]. In this model, each of the core IMB model constructs of information, motivation, and behavioral skills are situated into the socioecological context of receiving care, which we adapted to our care setting and study question (Fig. 1).
Data was collected from March 2019 through September 2019. In a private setting, a Swahili and Kalenjin speaking researcher (LC) obtained verbal consent and conducted one-on-one, hour-long semi-structured interviews. The interview guides included knowledge and attitudes towards KS, cancer, and HIV, barriers and facilitators to chemotherapy initiation and completion, and possible interventions for treatment. Participants were reimbursed for travel. The Massachusetts General Hospital Institutional Review Board and the Kenyan Institutional Review Committee approved the protocol.

Data analysis
Audio recordings were professionally transcribed and translated into English. Data were analyzed with NVivo using a framework approach. After each interview, authors LC and EF created a list of major themes generated during the interviews. These themes were used to develop an initial codebook. This codebook was then applied to 15 interviews by authors DM and MG. This initial coding structure was iteratively compared during Keywords: Kaposi sarcoma, HIV/AIDS, Chemotherapy, Treatment, Adherence, Africa South of the Sahara, Kenya, Qualitative weekly meetings, and discrepancies were resolved by senior author EF. Codes were compared until each had a Kappa score of greater than 0.6 to ensure high interrater reliability. Coding discrepancies were resolved through a group consensus. The remaining interviews were split in half and coded by DM and MG individually.
The approach to analysis combined deductive and inductive analysis in an iterative fashion. A deductive framework analysis approach was used during the initial phase of coding in order to assign interview themes to the IMB and socioecological model. Inductive analysis was also implemented to apply our research question within the Western Kenyan setting.

Results
57 participants underwent semi-structured interviews. Participants were median age 37 (IQR 32-41), majority male (68%), and were interviewed at a variety of health centers and one referral hospital throughout Western Kenya (Table 1). Most themes within the sIMB model applied to both chemotherapy initiation and adherence, as described below and outlined in Table 2 and depicted in Fig. 2.      Knowledge of chemotherapy side effects "That these drugs, if you have already started them, they will turn your body black and will shave the whole of your hair with pain, the stomach will swell, you know what I hear from people, that the stomach will swell, your hair will fall of like someone who has shaved with a razorblade and walking will be a problem.

Situated-economic/political/societal
There were multiple overarching situated barriers and facilitators which influenced chemotherapy initiation and adherence. One of the most common barriers was not having the financial resources to afford transportation to chemotherapy, chemotherapy medications, or food, with one patient reporting "My parents said let us sell this cow for you to at least get one injection…I told them not to because I was wondering what will happen to my younger siblings… I felt like I will leave them in poverty. " (#86, M, 36, MTRH, treatment non-starter) Transportation in particular was very costly for many patients, who often had to travel long distances to reach one of few centers providing chemotherapy. A prominent facilitator to chemotherapy initiation and adherence was having the Kenyan government sponsored National Health Insurance Fund (NHIF) health insurance, which paid for the majority of patients' treatment. A few patients also had work-sponsored sick leave or disability insurance to cover loss of income while receiving treatment. Additionally, some patients were part of microfinance groups, often either through HIV support groups or faith-based networks. Lastly, external funding sources such as through AMPATH or private donors who sponsored the cost of KS treatment was another notable facilitator. As one patient reported, "My NHIF card helped me…When I went to Webuye, they first asked me if I had an NHIF card. I told them I had one. So whenever I came here, I would only incur cost of transport, but treatment and everything else was covered by NHIF. " (#52, M, 34, Webuye, treatment completer) In particular, at the AMPATH center, Chulaimbo, all chemotherapy was provided free of charge for a period of time during this study through a donor sponsored program. Lack of functional skills "And now let us say when you are coming here for your treatment, how will you be planning for your transport?" "Planning?" "Yes, because you've said that you are thinking of starting, how will you be planning for your transport" "God knows, I can't say anything on that. I leave everything to God." (#84, F, 30, Chulaimbo, treatment non-starter) Both Low Total: 2 Completer: 1 Non-Completer: 0 Non-Starter: 1 Type of theme indicates whether the theme was for chemotherapy initiation, adherence, or both. Incidence density indicates whether the theme was high density (> 100 responses), medium density (50 to 100 responses) or low density (< 50 responses)

Situated-community
At the community level some patients reported limited community social support and in fact reported a large amount of stigma after being diagnosed. Some patients believed this stigma was related to having visible skin lesions, while others believed the stigma was more related to the diseases processes itself including HIV and cancer. As one patient stated, "In the village… someone could come to visit you at home, but his intentions are to spy on you, when they go away, you get to hear other things. You hear them say that so and so has bewitched you. They are not of any help; they come to visit you but are not of any help. " (#43, M, 40, Chulaimbo, treatment completer) A few patients additionally reported that they were the first person in their community to be diagnosed with KS, which they felt limited community support and understanding of how to proceed with treatment. This was often a barrier specific to chemotherapy initiation rather than adherence.
In contrast, there were many themes relating to facilitators for starting and adhering to chemotherapy. Multiple patients found social support in their larger community, including meeting other patients in the community with KS or having community, often faith-based fund raisers for treatment. For example, one patient stated, "The church too supported me…The also came to visit me and pray for me, if I told them I did not have transport to the hospital, they could contribute and give me, and they also encouraged me that I should go to hospital. " (#63, F, 35, Chulaimbo, treatment non-completer) Some patients also found community health workers who would monitor their HIV care an important source of advice and support.

Situated-health centers/system
Within the health care system, many larger barriers and facilitators influenced a patient's access to chemotherapy initiation and adherence. Prominent barriers within the health system were relatively limited sites from which patients could obtain chemotherapy, which led to issues with transportation. Furthermore, many sites only offered chemotherapy every week or every other week on a specific day, leading to frequent scheduling conflicts, as one patient described "I wasn't happy at all… I am a widow, I have children and I don't have anyone supporting me, and my appointments are very close, after every two weeks. Looking for that transport is very hard. So, it can discourage you and you can stop the medication. " (#71, F, 33, Chulaimbo, treatment non-completer).
At the larger tertiary hospital in particular, some patients found it difficult to navigate among different specialist services, such as from HIV care to oncology care. Fig. 2 Major themes for initiation of and adherence to chemotherapy, structured within the Situated Information, Motivation, Behavior (sIMB) framework as developed by Amico et al. [12] When patients did manage to navigate to the oncology center, they regularly had to wait for many hours to see a provider and oftentimes there were additional delays for scheduling the chemotherapy start date.
Facilitators to chemotherapy initiation and adherence included an easy chemotherapy referral process and free chemotherapy. Furthermore, at the large teaching and referral hospital there were social workers available to help with costs of chemotherapy and help patients enroll in NHIF, allowing their chemotherapy costs to be covered by insurance.

Information
In addition to the larger structural barriers and facilitators identified, within the ecologic framework there were also multiple individual level themes related to chemotherapy adherence and initiation. The first related to patient's knowledge of KS and its treatment modalities. Prominent barriers were limited knowledge or false beliefs, including that chemotherapy is deadly, that cancer cannot be cured, and not understanding the purpose of chemotherapy. As one patient stated, "I just know that when you receive chemo, it is death and there is no healing. " (#75, F, 38, MTRH, treatment non-starter).
An additional barrier was the knowledge of chemotherapy side effects-some of which were accurate, and others were likely over-estimations of the negative consequences of chemotherapy. An additional barrier was the use of traditional medicine instead of chemotherapy.
Facilitators were patients understanding the positive impacts of chemotherapy to improve or cure KS, understanding the chemotherapy regimen and timing of doses, and having knowledge about health centers where chemotherapy was available. An additional facilitator was when patients were aware of which chemotherapy side effects to expect, and then were able to tolerate these side effects to continue to return for further rounds of chemotherapy.

Motivation
Patients additionally described multiple interpersonal and intrapersonal motivations related to chemotherapy initiation and adherence. Barriers included having multiple different healthcare providers with inconsistent information, lack of appropriate provider follow up, lack of education from providers about chemotherapy schedule and side effects, general mistrust of the healthcare system, and negative interactions with healthcare workers such as being reprimanded by providers. For example, one patient described "That made me anxious about going to the hospital…because of your status [HIV status], they will not speak to you well… that would make you want to give up. " (#78, M, 35, Kitale, treatment non-starter) Negative motivators from family and friends were stigma surrounding chemotherapy and a cancer diagnosis, as well as lack of support for pursuing KS treatment.
Interpersonal motivation often occurred between the patients and their providers as well as between the patients and their support network. Motivators for chemotherapy initiation and adherence included positive experiences with providers, consistent and accurate provider information, positive experiences with research staff, and both emotional as well as material support from family and friends.
Patients additionally described intrapersonal motivation barriers and facilitators. Intrapersonal barriers to chemotherapy were having poor health, losing hope, being fearful of chemotherapy, experiencing side effects of chemotherapy, and having co-morbid disease processes take priority (ex. Managing HIV care or pregnancy care). Many participants who reported being fearful or the chemotherapy or associated side effects noted this to be a large barrier to chemotherapy initiation.
Facilitator themes were motivation to seek or continue treatment from their severity of KS, faith in the healthcare system, and motivation to support their family. For example, a patient explained "What helped me was the disease that I suffered from…It was smelling…It made me come for treatment. " (#54, M, 40, Busia, treatment non-completer).

Behavioral skills
Lastly, patients either acquired or lacked various behavioral skills which impacted their initiation and continuation of chemotherapy.
The activation skills which facilitated chemotherapy were coping skills, resilience, garnering social support, prioritizing oncology care, acceptance of disease, and accepting some of the hassles of the healthcare system such as long lines and wait times from treatment.
Functional skills which facilitated chemotherapy were setting and remembering chemotherapy appointment times, arranging leave for work for chemotherapy, arranging money for transportation and chemotherapy costs, and communicating concerns and wishes to their healthcare provider. As one patient with many functional skills described "As soon as I clear one [appointment], I start looking for resources for the next [appointment]…So first of all, I make sure the NHIF card is active… and then I look for money for transport. " (#34, M, 45, Chulaimbo/ Busia, treatment completer) Some patients lacked these functional skills, which impeded their ability to initiate and adhere to chemotherapy. Themes included challenges planning for chemotherapy, arranging time off, arranging money, managing medications, and communicating with healthcare providers.
Lastly, system navigation skills that facilitated chemotherapy were a patient's ability to navigate the logistics of chemotherapy and the complex referral system.

Discussion
In this qualitative analysis, patients with HIV-associated KS in Kenya faced many barriers to initiation of and adherence to chemotherapy. The most significant barriers were lack of financial means, difficulty with convenience of appointments such as distance to facility, appointment times, long lines, limited appointments, intrapersonal barriers such as fear or hopelessness, and lack of proper or sufficient information about chemotherapy.
One of the common overall barriers to initiation and adherence was lack of financial means to afford not only chemotherapy but also transportation to appointments, and necessities such as food. Despite the fact that there are systems in place such as Kenya's NHIF health insurance, which can often cover full chemotherapy treatment costs, and some external funding sources through AMPATH, patients with HIV-associated KS continue to face challenges with obtaining necessary funds. Prior studies in sub-Saharan Africa report that the path to obtaining health insurance (NHIF) can often be long and complicated with further limited access due to unaffordable monthly premiums [13,14]. Financial insufficiency and poverty will likely continue to play large roles in the lives of people living with HIV-associated KS until these larger systems are able to implement initiatives that are both easy to access as well as maintain by all members of the population.
Despite local efforts to develop facilities and resources for providing care for KS, many patients still have to travel many hours to reach a care site, only to wait in line for extended periods of time. The infrequency and lack of consistency of appointments proved to be a major deterrent for patients to adhere to their chemotherapy regimen. In the future, resources should be directed towards expanding the number of trained medical personnel and equipment, increasing chemotherapy availability in terms of both location as well as appointment times, and creating a patient-friendly navigation system to avoid confusion and long lines [15,16].
Lack of motivation and intrapersonal barriers such as fear and hopelessness played a large role in the delay or lack of chemotherapy initiation. The misconception that chemotherapy is associated with death and the fatalistic mindset that cancer is a death sentence were common during the interviews. Studies assessing the role of fear of disease and treatment in people living with HIV in SSA report the need for increased supportive and management counselling for patients and their support networks [17,18]. The role of HIV, cancer, KS, and chemotherapy education also ties in closely with this barrier. Prior studies on HIV-associated malignancies in SSA note the need for education about oncological treatment and treatment benefits to improve treatment access and adherence [19,20]. Educational interventions such as easy to understand pamphlets and public health campaigns should be expanded to alleviate feelings of fatalism, fear, and hopelessness.
Behavioral skills such as activation and functional skills played an important role in the initiation and adherence of chemotherapy regimens. Prior studies note that interventions to assist patients in acquiring, self-cueing, and self administering medications as well as an intervention to incorporate the treatment regimen into everyday life are important behvaioral skills to promote adherence [21,22].
The findings in this study are essential to creating tailored interventions for this patient population. The multiple levels of the sIMB framework include national policy, community, health center, and individual targets. At the level of policy, our findings related to the difficulty accessing and maintaining NHIF may be relevant to ministry of health officials. At the level of community, community-wide education regarding KS and its treatment would enhance social support and potentially reduce KS-related stigma. At the level of health centers, the need for a navigation system and the importance of provider training are most applicable to improving the function of care programs. At the individual level, patient and caregiver education can enhance understanding, motivation, and support to initiate and complete chemotherapy. The original IMB model created by Fisher and Fisher, was designed specifically for HIV prevention behaviors in the United States, and presented an individual focused rather than structural focused framework, which might be less applicable in most low resource settings [21]. Since this time, the original framework has been adapted to include frameworks such as sIMB and motivational interviewing based IMB (CLIMB), which address some of the gaps in the original framework [23][24][25]. These adapted IMB frameworks, specifically the sIMB model, better represented our patient interviews and provided a suitable framework for our study question.
A major limitation to this study was the population from which the patients for this study were recruited. This study population is a subset of a larger study population which was comprised of patients within AMPATH, who likely face fewer barriers to chemotherapy initiation and adherence than the majority of the population with HIV-associated KS.