Setting

We carried out the situation analysis in Neno district, Malawi. Neno district, with a population of 130,612, is a rural, hard-to-reach area bordered by Mozambique to the West and six Malawi districts along its remaining boundaries (Figure 1) [33]. The district’s main health and economic indicators are representative of national averages, with male and female life expectancy slightly higher in Neno at 48.6 and 53.1 years, respectively, than Malawi overall, and 56% of the population living below the World Bank poverty line of 116 USD of income per person per year [33], [34]. The Malawi Ministry of Health, through the Neno District Health Office (DHO), provides essential health services free of charge at 13 public health facilities, including Neno District Hospital situated in the administrative capital of the district–Neno boma [33].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 1. Map of Neno District, Republic of Malawi.

The inset picture depicts Neno district, located in the southwestern portion of the country. The district shares an international border with Mozambique to the west and borders with Ntcheu district to the north; Balaka and Zomba districts to the northeast; Blantyre district to the east; Chikwawa to the south; and Mwanza district to the southwest. The inset legend illustrates the locations of major health facilities, including the district’s two hospitals–Neno District Hospital and Lisungwi Community Hospital–as well as the remaining 11 health centers.

https://doi.org/10.1371/journal.pone.0110457.g001

Data Collection Overview

Employing REM, we collected data from three principal sources: 1) chart review of paper-based and electronic medical records for adult palliative care patients enrolled in the NPCP during its first 9 months; 2) structured interviews with patients and their caregivers; and 3) semi-structured interviews with key stakeholders. Collecting data from three sources enabled verification of information through triangulation [25], [35]. We defined caregivers to be any accompanying family member, friend, or life partner identified by the patient as their primary custodian or “guardian” at the time of NPCP enrollment. We defined key stakeholders as frontline health workers and leaders of organizations involved in any aspect of palliative care delivery in Neno district or those involved in palliative care policy at the national level.

Chart Review

We reviewed paper-based medical records and registers for all adult patients enrolled in the NPCP during its first 9 months of existence (mid-January through early October 2012). We assessed the first 9 months to better understand our patient population and referral practices and to evaluate the quality of service delivery at the start of the program. We used the APZU OpenMRS-based electronic medical record to augment data abstracted from paper-based records [36]. We collected de-identified data on the following patient dimensions: demographic information, primary diagnosis, HIV status, pain, other symptoms, psychosocial distress, encounter location (i.e. in the outpatient clinic, hospital ward, and/or home visit in the patient’s home), retention-in-care, and referral for psychosocial support services. Palliative care providers documented the severity and quality of patient pain, other symptoms, and psychosocial distress at the time of each clinical encounter using a visual hand-scale representation of the African Palliative Care Association (APCA) African Palliative Outcome Scale (African POS) [37]–[40]. The African POS is a multidimensional palliative care scale designed for and validated in African palliative care populations [38], [39]. Responses to African POS questions are scored using a 6-point Likert scale ranging from 0 to 5 [39]. When used to assess pain in adults, an African POS score of 3 or greater suggests moderate-to-severe pain, with a score of 5 representing the most severe or worst pain.

Structured Interviews: Patients and Caregivers

We developed two interview instruments–one for patients and a second for caregivers–to qualitatively assess the following participant domains: 1) awareness of diagnosis and preferences for palliative care; 2) pain and other physical symptoms; and 3) psychological, social, and spiritual needs and sources of support. Both instruments incorporated structured items and open-ended questions with interviewee prompts. The patient and caregiver interview instruments included all items intended for patients and caregivers, respectively, of the African POS [37], [39]. The African POS has been used to measure patient outcomes on the three domains enumerated above; to evaluate caregiver knowledge, confidence, and anxiety; and to assess the quality of palliative care services [37]. The structured interview instrument was translated from English into the local language of patient and caregiver participants, Chichewa. The translation and preparation of the final Chichewa version of the instrument were performed at the APZU field office by trained local staff fluent in both English and Chichewa.

We planned to interview a sample of 80 individuals, comprising 5 patient participants at each of 4 antiretroviral therapy clinics (20 total), 10 hospitalized patients, and 10 patients receiving chemotherapy for Kaposi sarcoma, along with a matching number of accompanying caregivers (i.e. 40 caregiver participants total). We devised this sampling strategy in an attempt to capture a diverse range of patient and caregiver perspectives regarding palliative care services and experiences with serious chronic illness in Neno, Malawi. With this strategy in place, we recruited and interviewed a consecutive sample of patients and their caregivers 18 years of age or older who presented for routine HIV care, in-patient hospital care, or chemotherapy for Kaposi sarcoma, and who were newly referred for enrollment in the NPCP at 4 public health facilities in Neno District on selected days from January 14 to February 27, 2013. The facilities included one district hospital, one community hospital, and two health centers, chosen to represent the breadth of facility-based healthcare delivery in the district. We approached potential participants face-to-face at health facilities to maximize the efficiency of the recruitment process given the accessibility of the target population at health facilities and the cost and logistical limitations involved with alternative recruitment approaches (e.g. telephone, paper flier, etc.). While study investigators helped recruit interview participants, they had not met any participants prior to the time of the interview. Recruitment was stopped at 47 participants (36 patient and 11 caregiver participants) after an interim analysis revealed that saturation of major themes had been achieved. No patients or caregivers declined study participation.

Study investigators with formal medicine and public health training (S.N.E., E.B.W., and A.M.) conducted patient and caregiver interviews with the assistance of an independent, trained study interpreter fluent in both English and Chichewa. We obtained written informed consent from all participants prior to commencing interviews. As part of the informed consent process and prior to the structured interviews, study investigators (S.N.E., E.B.W., and A.M.) explained the rationale for the study but did not disclose any personal information regarding their interests in the research topic or their assumptions about possible study findings. Patient and caregivers were interviewed separately in the absence of the other person, whenever possible. However, at the discretion of patient participants, caregivers were allowed to be present during interview sessions and vice versa. All interviews were conducted in private areas at one of the four health facilities described above in a single session lasting approximately 30 to 45 minutes. An investigator led each interview session and read questions aloud sequentially, allowing enough time to fully capture participant responses and adhering to the format of the interview instrument as it progressed through the following sections: 1) Patient demographics, knowledge about their diagnosis and treatment, and medical care preferences; 2) African POS items; and 3) Psychosocial support and relevant information–needs, preferences, and ongoing sources of support (Appendix S1). For each participant, the study interpreter translated all questions aloud into the local language, Chichewa, and subsequently translated participant responses into English. The investigator then transcribed English-translated responses verbatim. No audiovisual recording equipment was used during interviews because of local restrictions on the use of such technology in patient care areas. No repeat interviews were carried out at any time during the study period.

Semi-structured Interviews: Key Stakeholders

We interviewed a purposive sample of key local and national stakeholders, including at least one individual from all organizations in Neno district working in the palliative care arena. The specific sites were as follows: 1) an integrated cancer and palliative care clinic at Neno District Hospital; 2) a MoH ART clinic; 3) the Neno District Health Office home-based care program; 4) the APZU/PIH Program on Social and Economic Rights (POSER); 5) the APZU/PIH Village Health Worker Program; 6) Chiyembekezo Community-based Organization (CBO); and 7) the Palliative Care Association of Malawi (PACAM).

Stakeholders were interviewed at each site by one of the study investigators (S.N.E.) and asked to discuss the palliative care infrastructure and service delivery practices at their institution or organization. Transcripts were reviewed with key stakeholder participants for comment and correction. Palliative care policy documents, clinical encounter forms, data recording tools, and training materials were reviewed by the investigator prior to stakeholder interviews to provide context, among them the NPCP Initial Assessment Form (Appendix S2), Malawi National Palliative Care Guidelines (Appendix S3), the Introduction to Palliative Care: Health Care Workers’ Service Providers Manual (Appendix S4), and the Malawi Guidelines for the Clinical Management of HIV in Children and Adults (Appendix S5) [13], [41]–[43]. To minimize the influence of palliative care documents on stakeholder responses, these materials were not provided to stakeholders before or during study interviews. Preliminary study findings were shared at a dissemination meeting to refine our understanding and representation of perspectives expressed by key stakeholders during interviews.

Data Analysis and Reporting

For qualitative data we employed content analysis methodology to develop thematic codes for patient, caregiver, and stakeholder interview responses using Dedoose software (SocioCultural Research Consultants, LLC, Manhattan Beach, CA, USA) [44], [45]. Two study investigators (S.N.E., E.B.W.) coded participant responses together (beginning with “participant 0” as designated automatically by Dedoose software) and reconciled differences by consensus. We calculated frequencies and simple proportions for all thematic codes generated, including root and child codes. We then used this information to develop a taxonomy of themes, providing a structure to interpret patient and caregivers’ experiences with illness and to describe palliative care knowledge and preferences among all interview participants. The results of structured interviews are reported in accordance with accepted criteria for qualitative research [46]. For quantitative survey and chart review data we calculated frequencies, proportions, means or medians, and measures of dispersion. For continuous variables, we compared means between two independent groups using the two-sample t-test and means of repeated measurements in the same group using the paired t-test. For variables with more than two categories, we compared means with the one-way analysis of variance. We considered a two-sided p-value<0.05 to be statistically significant. All quantitative analyses were performed in SPSS (Version 20, 2011, IBM, Armonk, NY, USA) or Stata (Version 12.1, 2011, StataCorp, College Station, TX, USA).

Ethics Statement

All components of this study were conducted in accordance with the principles expressed in the Declaration of Helsinki. All patient and caregiver interview participants provided written informed consent prior to participation in the study. All key stakeholder interview participants, who were asked generic questions about palliative care programs and policies only and not about themselves or individual patients, provided verbal informed consent, which was documented on de-identified semi-structured interview instruments. All individuals depicted in this manuscript are NPCP service providers, not study participants, and have given written informed consent (as outlined in the PLOS consent form) to publish their likeness in photograph form. Human subjects research approval for all study procedures was granted by the National Health Sciences Research Committee of Malawi and by the Institutional Review Boards of the University of North Carolina, Chapel Hill, USA, Partners Healthcare System, Boston, USA and Harvard Medical School, Boston, USA.

Overview

We first describe the Neno Palliative Care Program (NPCP) model and early results from the NPCP. We then report major characteristics of patient and caregiver interview participants. Lastly, we present the results of patient, caregiver, and key stakeholder interviews, organized by our taxonomy of major and minor interview themes (Figure 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 2. Major thematic codes generated from patient and caregiver interviews.

The coding tree depicts major themes identified from structured patient and caregiver interviews. Codes relate to patient and family caregivers’ experiences with illness, palliative care services provided, and palliative care services desired. Subordinate child codes elaborate descriptions for each root code. All root and child codes were generated using content analysis methodology and Dedoose software (SocioCultural Research Consultants, LLC., Manhattan Beach, CA, USA).

https://doi.org/10.1371/journal.pone.0110457.g002

The Neno Palliative Care Program Model

The joint MoH–APZU/PIH Neno palliative care team, first formed in late 2011 to implement the NPCP, consists of: 1) 1 physician; 2) 2 clinical officers (mid-level health workers common in Malawi); 3) 1 palliative care coordinator; 4) 1 nurse; 5) approximately 150 MoH home-based care volunteers; and 6) 691 community health workers, known locally as village health workers (VHWs), paid and trained by APZU/PIH to provide health-related services throughout the district. An experienced international palliative care physician-educator provides longitudinal training and clinical mentorship to NPCP clinical providers [47], [48]. All clinical providers received a week-long palliative care training in early January 2012 prior to NPCP implementation, followed by a week-long advanced training in March 2012.

NPCP clinical providers train VHWs to recognize pain, other symptoms, treatment side effects, major disease complications, and psychosocial distress, and to request an immediate home visit for patients with one of these issues (Figure 3). The palliative care clinical team manages a weekly outpatient clinic at Neno District Hospital, conducts daily inpatient hospital consultations, and oversees a weekly district-wide mobile outreach clinic. Clinical officers working closely with the supervising palliative care physician conduct clinical evaluations, administer disease-modifying chemotherapy for Kaposi sarcoma, and prescribe symptom-relieving medications (including morphine), following MoH guidelines and the three-step WHO pain relief ladder [13], [40], [41], [43], [49]. The Malawi MoH, with support from APZU/PIH, procures essential palliative care medicines and supplies for the NPCP and all district health facilities–including morphine liquid, slow-release morphine tablets, codeine phosphate tablets, ibuprofen, amitriptyline, diazepam, anti-emetics, antibiotics, incontinence pads, sterile gauze, and bandages, among other commodities–in accordance with local regulations, including the Malawi Controlled Drug Act [13]. MoH and APZU/PIH health workers and programmatic staff refer patients to the NPCP from patient support programs and health facilities located throughout Neno district, including the wards and clinical departments of the district’s two hospitals–Neno District Hospital and Lisungwi Community Hospital (Figure 1). At the discretion of the referring health worker, patients may start analgesia before or at the time of NPCP referral, or during the first NPCP encounter. For patients living with HIV, clinical officers provide integrated HIV treatment and primary care services, including assessment for ART initiation and evaluation for possible ART side effects in accordance with national guidelines [43].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 3. Neno Palliative Care Program team members conduct a home visit in a rural village.

Five members of the Neno Palliative Care Program (NPCP) team travel by foot to reach the home of a palliative care patient living in rural Neno district, Malawi. An Abwenzi Pa Za Umoyo/Partners In Health Medical Officer with specialized training in palliative care (foreground) leads the home visit. He is accompanied by (from left to right) a palliative care nurse, a Program on Social and Economic Rights social worker, a Village Health Worker and a NPCP clerk. NPCP team members work together to assess and treat patient pain and other physical symptoms, provide disease-specific care for serious chronic illnesses, and offer psychosocial support to patients and their families.

https://doi.org/10.1371/journal.pone.0110457.g003

The NPCP responds to patient and family psychosocial needs through the APZU/PIH Program on Social and Economic Rights (POSER). POSER provides counseling and emotional support, as well as socioeconomic assistance for selected patients, in partnership with local community-based and faith-based organizations and government Ministries, including the Ministry of Gender, Child Development and Community Development and the Neno Office of Social Welfare. POSER staff evaluate the vulnerability of referred patients using a standardized assessment tool, and provide qualifying patients with one or more of the following types of socioeconomic assistance: food packages, home repairs, transportation reimbursement, and school support for children.

Early Results from the Neno Palliative Care Program

Over a 9-month period from mid-January through early October 2012, the NPCP enrolled 63 patients with serious chronic illnesses (Table 1). Most patients had cancer as a primary diagnosis (n = 50, 79%), and 32 of these (64%) had Kaposi sarcoma (Table 1). The majority of patients were HIV-infected (37 of 61, 61%). Of those with HIV infection, 31 (84%) lived in Neno district and received ART at 1 of the district’s 13 public clinics.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Characteristics of patients at enrollment in the Neno Palliative Care Program (NPCP, January through October 2012).

https://doi.org/10.1371/journal.pone.0110457.t001

Neno palliative care providers conducted 112 patient encounters during the 9-month evaluation period. Most encounters occurred in the outpatient palliative care clinic based at Neno District Hospital (n = 57, 51%) (Table 2). NPCP providers conducted 22 home visits, comprising 20% of all patient encounters. Complementing clinical service delivery, APZU/PIH VHWs provided longitudinal psychosocial support to 35 (56%) patients in the form of counseling during regularly scheduled home visits. Similarly, 34 (54%) patients were referred to POSER for psychosocial support, which included emotional support and socioeconomic assistance (Table 2). In total, providers referred 48 (76%) patients for psychosocial services in the form of VHW home visits, POSER support, or both.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Summary of palliative care services provided by the Neno Palliative Care Program (NPCP, January through October 2012).

https://doi.org/10.1371/journal.pone.0110457.t002

Overall, 56 (89%) patients had a documented baseline pain assessment from their initial NPCP encounter, which included a record of the presence and severity of pain (Table 3). Among these 56 patients, 13 (23%) patients were found to be pain free (i.e. African POS score = 0) and 43 (77%) reported the presence of any pain (i.e. African POS score>0) (Table 3). Among the 43 patients reporting any pain at baseline, the mean African POS pain score was 3.2 (standard error, SE: 0.2), and 31 (72%) patients reported moderate-to-severe pain (i.e. African POS score≥3). Of the 43 patients with documented pain at baseline, 35 (81%) had a cancer diagnosis. Patients with documented baseline pain and cancer had a significantly higher mean African POS pain score than those without cancer (3.3 with cancer vs. 2.4 without cancer, p = 0.04) (Table 3). Among patients with baseline pain, pain severity did not differ significantly by age, gender, HIV status, or presence of Kaposi sarcoma (Table 3).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Patient pain status, severity, and management at the initial Neno Palliative Care Program (NPCP) encounter (January through October 2012).

https://doi.org/10.1371/journal.pone.0110457.t003

Among the 43 patients with documented pain at NPCP intake, 29 (67%) reported ongoing use of pain medications initiated by a referring health worker (Table 3). Of these 29 patients: 13 (45%) described using non-opioid analgesics exclusively, such as ibuprofen, paracetamol, aspirin, amitriptyline or some combination thereof; 2 (7%) reported taking codeine–a weak opioid–with or without an accompanying non-opioid agent; and 14 (48%) reported using oral morphine–a strong opioid–with or without a co-administered non-opioid medication (Table 3).

For the 43 patients with documented pain at baseline, NPCP providers newly started 10 (23%) patients on analgesia, added a drug or changed the current analgesic regimen for 15 (35%) patients, and continued the previously prescribed pain medications in 9 (21%) patients; 9 (21%) patients had no documentation of receiving a prescription for an analgesic agent during their initial NPCP encounter (Table 3). Morphine–available free of charge from the Malawi MoH–was prescribed for 23 of 43 (53%) patients with documented baseline pain. Having documented moderate-to-severe baseline pain was significantly associated with receiving a prescription for morphine at the initial NPCP encounter (p = 0.001). The mean African POS pain score was significantly higher for patients prescribed morphine (3.7, SE: 0.2) than for patients prescribed codeine or a non-opioid analgesic (2.1, SE: 0.3) (p<0.001) (Table 3).

Most patients (n = 35, 56%) had at least one recorded follow-up encounter (Table 4). Of these, 26 (74%) patients had a documented pain assessment, with 5 (19%) noted to be pain free and 21 (81%) reporting any pain (Table 4). Among the 21 patients reporting any pain at follow-up, the mean African POS pain score was 2.9 (SE: 0.3) and 11 (52%) patients reported moderate-to-severe pain. Pain severity did not differ significantly by age, gender, HIV status, or the presence of cancer, including KS, for patients reporting pain at follow-up (Table 4). For patients with pain at baseline and documented pain assessment at enrollment and follow-up (n = 18), we observed a small, non-significant decrease in mean African POS pain score (3.0 vs. 2.7, p = 0.5) over the follow-up interval. We recorded 4 deaths (6%) that occurred before the patient’s first scheduled follow-up encounter, each transpiring within 1 month of program enrollment (Table 4).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Patient pain status, severity, and management at the first follow-up Neno Palliative Care Program (NPCP) encounter (May through October 2012).

https://doi.org/10.1371/journal.pone.0110457.t004

Characteristics of Patient and Caregiver Interview Participants

We interviewed 36 patients at the time of NPCP enrollment who were not in the early NPCP cohort, including 9 patients interviewed in the presence of their caregiver. Patients interviewed were, on average, 38 years old and 19 (53%) were female (Table 5). Most patients interviewed had cancer (n = 19, 53%), with Kaposi sarcoma being the most prevalent diagnosis in the sample (n = 13, 36%). HIV/AIDS was the primary diagnosis for 10 (28%) patients, with every known HIV-infected patient interviewed already enrolled in MoH HIV care. Those surveyed reported being ill for about 1 year, on average (median: 13 months; interquartile range, IQR: 0–156 months), at the time of the interview. Several patients (n = 7, 19%) were unable to work due to illness.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 5. Characteristics of patient and caregiver interview participants.

https://doi.org/10.1371/journal.pone.0110457.t005

We interviewed 11 of the 12 (92%) caregivers who had accompanied a patient interview participant and were available for study recruitment (Table 5). All caregivers surveyed were family relatives of a patient enrolled in the study and were predominantly female (n = 8, 73%) (Table 5). Most caregivers reported working, typically having jobs as subsistence farmers or market vendors. Caregivers reported having served in this role for a median of 13 months (IQR: 0–54 months).

Prevalence of Pain and Other Symptoms among Patient Interview Participants

A majority of interviewed patients reported pain (n = 31, 86%), with 23 (64%) experiencing moderate-to-severe pain (African POS score≥3) (Table 6). Of those with moderate-to-severe pain, 13 (57%) had a cancer diagnosis. The mean African POS pain score of the 31 interview participants reporting pain was 3.3 (SE 0.2). Most interviewed patients reported additional physical symptoms, with numbness, cough, difficulty sleeping, and shortness of breath cited most frequently.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 6. Patient symptom presence and severity over the three days prior to Neno Palliative Care Program enrollment (N = 36).

https://doi.org/10.1371/journal.pone.0110457.t006

Psychosocial Needs and Support Reported by Patient and Caregiver Interview Participants

In response to questions about desired psychosocial services, patients frequently mentioned socioeconomic needs (Table 7). Most patients cited a need for generating income (n = 24, 67%), receiving food (n = 22, 61%), or having their homes repaired (n = 22, 61%). Many patients mentioned poverty and lack of food as important barriers to their well-being:

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 7. Patient-reported psychosocial needs and sources of psychosocial support (N = 36).

https://doi.org/10.1371/journal.pone.0110457.t007

‘I was told to eat a well-balanced diet, which I can't do because of financial constraints.

First I was given food rations, but I'm not getting them now and I'm losing weight.’ (Participant 20)

‘[My house is] very debilitated. It was built by the church. [I am] happy to be in the hospital with rain coming because I won’t get soaked. I sleep in mud at home.’ (Participant 26)

‘What I need most in my life is a job; I have always wanted to have a job. I would like help with that.’ (Participant 0)

Most (n = 21, 58%) patients interviewed reported a desire for emotional support. Few patients (n = 2, 6%) mentioned HIV-related stigma when discussing their need for emotional support.

‘There are people who still don’t understand. Neighbors. And I was told I would not be hired for a landscaping job because of my status.’ (Participant 19)

Similar to patients, most caregivers (n = 7, 64%) stated that they needed emotional support. A high percentage of caregivers reported needs in helping meet the demands of caring for patients, including financial support (n = 10, 91%), food (n = 8, 73%), and physical assistance in providing care (n = 5 45%). A smaller number of caregivers reported the following additional needs: affordable medical care for the patient (n = 2, 18%) and help finding employment or help with transportation (n = 1, 9%).

All patients interviewed reported receiving some form of psychosocial support (Table 7). A majority of patients (n = 27, 75%) described receiving psychosocial support from family members. This most frequently consisted of emotional support, but often also included food, housing, caregiving, or financial support. Caregivers noted that close friends (n = 5, 45%), or a parent or neighbor (n = 2, 18%) were important sources of assistance. Several caregivers (n = 4, 36%) reported that they had received material support in the form of food, sharing of caregiving duties, and financial support. When provided, this support came from immediate or extended family members.

Community churches provided many patients (n = 18, 50%) and caregivers (n = 6, 55%) with psychosocial support consisting of home visits by clergy or congregation members or, occasionally, housekeeping and cleaning services. APZU/PIH was also a frequently cited source of psychosocial support, with 12 (33%) patients reporting longitudinal assistance from a VHW and an equal number (n = 12, 33%) mentioning that they had received POSER services. POSER provided patients with emotional support and socioeconomic assistance in the form of food packages, income generating activities, school fee funding, and housing renovations. Patients reported that village health workers in their communities provided them with frequent counseling during home visits.

The Role of Caregivers

All caregivers reported that they took on the role because of familial obligation. However, a minority of caregivers interviewed also noted that volunteering, friendship, and being the only caregiver available (n = 1 for each response) were also valued reasons. All caregivers described a large time commitment involved in the caregiver role. Caregivers reported attending to patients from 6 to 24 hours per day (mean = 17 hours). Caregivers assisted with a variety of activities of daily living, including feeding, bathing and dressing. Caregivers also frequently gave medications (n = 7, 64%), offered emotional support (n = 7, 64%), and provided symptomatic care (n = 5, 45%). Transportation to the clinic and care for dependents were provided by most caregivers (n = 6, 55%), and several (n = 3, 27%) provided housing.

Caregivers reported considerable confidence in their abilities to provide care but also considerable worry (Table 8). Caregivers noted the following as their greatest fears: losing a loved one (n = 5, 45%); not having enough money or resources to care for the patient (n = 4, 36%); an undesirable future for the patient’s children (n = 3, 27%); an unwanted future for themselves (n = 2, 18%); and discrimination or not receiving any assistance in caring for the patient (n = 1, 9%).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 8. Caregiver-reported confidence, worry, and information over the past three days (N = 11).

https://doi.org/10.1371/journal.pone.0110457.t008

‘I [as a caregiver] worry about the illness because I don’t have enough money at home because I can’t work and take care of my brother, and I’m also taking care of two orphaned children.’ (Participant 7)

‘We cannot be open about this type of illness, so I am her only guardian…. Our neighbors think that they will get the disease if they come near, since they’re not sure what type of illness she has.’ (Participant 4)

Palliative Care Knowledge

With respect to knowledge about their illness, most patients interviewed were able to describe the nature of their illness and the names of, or indications for, any treatments that they were receiving. The most frequent gaps in knowledge existed for hospitalized patients. These patients often had a new or still unconfirmed diagnosis. One patient–who knew that he was HIV-positive but was not aware that he had just been diagnosed with Kaposi sarcoma–illustrated this point as follows:

‘I have some sores on the chest, edema in the legs, and some shortness of breath,

and some bad pains. I haven't been told anything else.’ (Participant 33)

Despite attending to the many physical and psychosocial needs of patients, few caregivers (n = 2, 18%) had received formal training or counseling in palliative care. Most caregivers (n = 7, 64%) identified training in patient care as a major need.

We interviewed 7 key stakeholders. All key stakeholders reported being familiar with the concept of palliative care, even those working for organizations providing psychosocial support without an explicit palliative care mission. All were aware that national palliative care guidelines existed and that they had been updated in 2011. However, most stakeholders not directly involved with the NPCP had not seen the updated guidelines, and noted that they had not yet been fully disseminated at the district level. Stakeholders noted that additional training was needed for all levels of palliative care providers.

Preferences for Palliative Care Setting and Models of Service Delivery

Patient and caregiver interview participants were asked to select one or more settings where they would most like to receive palliative care from among five possibilities and to explain the rationale for their selections (Appendix S1). Patient responses most frequently included home-based care (46%), community-based care at a nearby health post or health center (30%), and hospital-based care (24%); only 1 response (2% of the 50 total responses) favored intensive palliative care provided in a designated hospital-based ward (a notion approximating the idea of a hospice). Caregivers reported split preferences for where they would most like patients to receive palliative care, with 46% of responses mentioning home-based care, 46% citing community-based care, and 36% referencing hospital-based care. Most caregivers (n = 9, 82%) reported wanting the support of a village health worker, home-based care volunteer, or other form of home-based care assistance.

‘If I can get a doctor to visit at home that would be good because I came a very long way here and I have many other duties that I am missing at home.’ (Participant 13)

‘Assuming I will be discharged, will I be able to get care at home? I cannot walk, or use my arms and legs, so it was very difficult to get here, and I need care at home.’ (Participant 3)

While most patients and caregivers expressed a preference for home-based care, hospital care was still important to many, particularly patients who felt that their illness was serious or required higher level care, and patients who expressed a distrust of the level of expertise that would be provided in a health center or in their home. For others, the convenience of receiving care in an easily accessible place was most important. Several patients noted that travel to the hospital could be difficult, time-consuming, and disruptive to work and home responsibilities. One caregiver summarized the conflicts between accessing healthcare and the other areas of his life this way:

‘It would be good to have health care close to where we live, because then I would be able to take my brother for care and then go out and do my farming and work. Then I would be able to take care of our family.’ (Participant 6)

All stakeholders who identified as health workers expressed that palliative care should be widely accessible and integrated with inpatient and outpatient clinical services.

‘Palliative care should be part of the routine services, not a separate project. It is part of essential healthcare delivery.’ (Stakeholder 7)

Health workers at public facilities noted that the integration of palliative care with other clinical services proved difficult because of limited health worker time and training. Most stakeholders favored a model in which a mobile palliative care team combined home-based care–where symptom assessment and psychosocial support would be provided by home-based care volunteers or village health workers in the community–with a dedicated hospital-based palliative care clinic or inpatient service for very sick or complex patients. They stated that a major success of palliative care in Neno District was the linking of palliative care in the home, the community, and health facilities. All stakeholders noted that trained village health workers and home-based care volunteers had been crucial in making palliative care, and healthcare in general, accessible to vulnerable patients in Neno district.

Future Palliative Care Priorities

Opinions about future palliative care priorities varied widely among key stakeholders. Policy-makers felt that the integration of palliative care into national policy would assure its eventual implementation in all districts. Health workers prioritized development and scale up of new models of palliative care service delivery, particularly mobile teams. Psychosocial service providers from APZU/PIH and Chiyembekezo CBO cited the need to further integrate clinical and non-clinical palliative care services and reinforce existing referral systems between these services. Most stakeholders noted that improving coordination–both among palliative care-relevant governmental and non-governmental organizations and across tiers of the public health system–should be a priority for future palliative care activities in Neno district. All stakeholders cited a lack of consistent funding for palliative care activities, amidst competing health priorities, as a threat to the sustainability and scale-up of palliative care services.

‘When donors provide money for certain activities, priorities shift and old plans can be forgotten. We need to have money [for palliative care] that is stable from year to year.’ (Stakeholder 5)