The needs, models of care, interventions and outcomes of palliative care in the Caribbean: a systematic review of the evidence

Background Palliative care provision is expanding in low and middle income countries. Services are developing in the Caribbean in response to the region’s ageing population, the significant burden of cancer, non-communicable diseases and HIV/AIDS. Appraisal of the existing evidence on palliative care needs, models of care, interventions and outcomes in the Caribbean is essential to inform emerging practice and future research. Methods Systematic review and narrative synthesis. Following implementation of a search strategy, titles, abstracts and full texts were screened. Data from nine studies were synthesized. The Qualsyst tools were used to assess the quality of quantitative and qualitative studies. Data were extracted into a common table, and themes were generated from the available peer review evidence using narrative synthesis. Results Nine papers were retained for appraisal. Eight papers described palliative care needs in the Caribbean. The needs for analgesia, support for patients, education and training of staff in palliative care and palliative care services were identified in the literature. Models of care for palliative care in the Caribbean were not described in great depth (n = 2 papers) and no intervention studies were found. Outcomes of palliative care such as quality of life, quality of care, and patient’s preferred place of care and death were identified from six papers. Quantitative methodology was used in seven of the nine papers in this review. One paper used a mixed methodology design, and one a qualitative approach. Conclusions Research from the Caribbean highlights the need for health care policy, training of staff, education, and access to analgesia and palliative care support services in this region. This sparse evidence must be taken into consideration with cultural beliefs and preferences of the Caribbean population in order to achieve improved outcomes for patients, their caregivers and health care professionals. This underscores the importance for more research in the field of palliative care in the Caribbean.


Background
Of the 58 million people who die yearly, 45 million die in low and middle income countries (LMIC) [1]. In the last decade, 99 % of the population growth took place in the developing world [2]. It is expected that by 2050, the population in developing countries will expand by 55 % [2]. The Caribbean is defined as the Caribbean islands that are located in the Caribbean sea. These islands are The WHO projects greater life expectancy for the population of Latin America and Caribbean countries between 2004 and 2030 [12]. This gain in life expectancy brings with it complex demands on the health care system [12]. The health care needs of the elderly may be chronic, time consuming and costly [12,13].
With the morbidity and mortality trends seen in Latin America and the Caribbean, it is estimated that approximately a half million people require palliative care services [14]. The latest global categorisation of countries in the Caribbean in terms of their palliative care integration into health systems is as follows: Group 1 (no known activity) n = 10; Group 2 (capacity building activity only) n = 4; Group 3a (isolated provision) n = 8; Group 3b (generalised provision) n = 0; Group 4a (preliminary integration into mainstream) n = 1; Group 4b (integrated provision) n = 0 [15].
There appears to be a discrepancy between the need for, and existence of, palliative care services in the Caribbean, since the majority of Caribbean islands have no or only isolated palliative care provision. A bibliometric analysis looking at research activity in palliative care in Latin America and the Caribbean (LAC) found that there was a modest contribution of research from LAC to international palliative care literature [16].
LMIC have conducted and reported systematic reviews to appraise the current state of evidence, and synergise research activities to improve the availability and quality of palliative care [17][18][19]. This study aimed to identify and appraise the evidence needs, models of care, interventions and outcomes of palliative care in the Caribbean.

Research question
A systematic review of the literature was conducted to answer the question: "What are the needs, models of care, interventions and outcomes of palliative care in the Caribbean?"

Design statement
A systematic literature review was conducted in line with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [20] statement and the guidelines set out from the National Health Service (NHS) Centre for Reviews and Dissemination (CRD) [21]. The synthesis of the retained data was carried out using the narrative synthesis framework by Popay et al [22]. There were no ethics required as this was a systematic review of published literature.

Definitions
The definitions of terms used in this review are presented in Table 1.

Search strategy
The biomedical databases of MEDLINE, EMBASE, CINAHL, Web of Science and Global Health were searched from 1987 (the recognition of palliative care as a specialty in medicine) to October 2014.
Search terms used were the union of the two groups of search terms as represented in Table 2. The first search strategy was the separate run of group 1 and group 2's search terms. The second search strategy was the union of group 1 and group 2 results with 'and'.
Apart from database searching, key experts in the field of palliative care in the Caribbean were contacted to enquire about relevant published research. These persons were sourced by the regional palliative care societies, a search of contacts from the website page of the International Association for Hospice and Palliative care, and recommendations of health professionals from the Caribbean.
The  The term "palliative" when used to describe patients, covers a varied heterogeneous population [45]. Patients with certain primary diseases have a palliative period, the period when the disease has become progressive and emphasis is not placed on cure but quality of life that follows [45].
Health Care needs Health care need can be defined as the "capacity to benefit from health care" [46]. The WHO further defines it as perceived health care needs as experienced by the patient, professionally defined health care needs as those services defined by health professionals and scientifically confirmed needs [47]. Palliative care outcomes A health outcome is defined as "Change in a patient' s current or future health status that can be attributed to antecedent health care" [50] End of Life The General Medical Council (GMC) defines patients as "approaching the end of life" when they are likely to die within the next twelve months [51]. This is broadly accepted cross-nationally [52].

Models of Care
Advanced disease "Connected with the active and progressive disease and has a limited prognosis. Prognostication in advanced disease relates to different factors such as symptoms, performance status and disease trajectory. As disease trajectories vary depending on whether the patient is suffering from malignant or non-malignant disease [53], advanced stages have to be defined independently for every disease" [54]  Once the literature search was completed, all the articles were exported to the bibliographic software tool Endnote Version X7.1 for the removal of duplicate articles and then scrutiny of the remaining articles.

Inclusion and exclusion criteria
Both qualitative and quantitative research papers were included for analysis. The criteria used for selection of the studies are demonstrated in Table 3.
A first researcher (SM) screened all full abstracts to assess for eligibility against the inclusion/exclusion criteria. The full text of eligible articles was retrieved if the paper met inclusion criteria, or if the abstract did not contribute adequate detail to warrant rejection. If there was any uncertainty after the full text of the article was examined, a second researcher (RH) assisted in making the final decision on agreement for inclusion of the paper.

Data extraction
Data extraction is required to accurately extract the relevant features and the results of the selected studies [23]. Data were extracted by one researcher (SM) and reviewed by a second researcher (RH) if necessary.
Publication details, study aim, sample characteristics, and main findings were extracted from each retained paper into a common table.

Quality assessment
Quality assessment allows both the internal validity and external validity of a study to be considered [23]. The quality assessment tools called the QualSyst tools [24], constructed by researchers from the Alberta Heritage Foundation were used for assessment of the quality of both qualitative and quantitative papers using any study design.
A numerical final score derived from a scoring system provides a systematic, reproducible and quantitative means of assessing the quality of research amongst different study designs [25].
The Qualsyst tool for assessment of the quality of quantitative studies is a checklist of 14 questions. A score from 0 to 2 can be awarded to each item on the tool. A summary score is calculated for each paper when the tool is applied. The Qualsyst tool for assessment of the quality of quantitative studies is a checklist of 14 questions. A summary score is calculated for each paper when the tool is applied by calculating a total sum then dividing it by the total possible sum [24]. The creators of the tool suggested that a cut off of 0.75 as a score to be the threshold for allowing a paper to be included in a review. The definition of the quality of a paper was defined as follows by Lee et al [25] in the use of this tool in their systematic review as; strong (summary score of >0.80),good (summary score of 0.71-0.79),adequate (summary score of 0.50-0.70) and limited (summary score of <0.50). The method of assessing the quality suggested by Lee et al [25] was appropriate for our goal of assessment of the quality of the data rather than using a threshold for including a paper into the review.
The Qualsyst tool for assessment of qualitative studies is a validated checklist that is made up of 10 criteria to be assessed. The scores for each item range from 0 to 2 with a maximum total score being 20. A summary score can be attained by summating the total score across the ten items and dividing it by 20. A threshold of 0.55 was used by the creators for inclusion of qualitative studies into their systematic review [24]. For the purposes of this review, the researcher defined a score of ≥0.55 as an 'adequate quality' paper. A score of ≤0.54 was deemed as a 'low quality' paper.

Data synthesis
The method of data synthesis used for this systematic review was narrative synthesis. This method was chosen because of the decision to include both qualitative and quantitative data in the final analysis. The guidance from Popay et al [22] allows synthesis of a wide array of research designs. The general framework for conducting narrative synthesis as suggested by the authors, Popay et al, consists of four main elements [22]: 1. Element 1: developing a theory of how the intervention works, why and for whom. A theory was not developed for this systematic review as all of the studies were non-intervention studies and descriptive in nature that were synthesized.  Table 4. In addition to tabulation, thematic analysis was also performed. The school of Psychology at the University of Auckland proposes six steps for conduct of thematic analysis [26] which were applied in this review. These steps include familiarization with the data, coding, searching for themes, reviewing the themes, defining and naming the themes and then producing the report. 3. Element 3: exploring relationships in the data. Idea webbing and concept mapping were used to analyse the findings from this review. 4. Element 4: assessing the robustness of the synthesis.
Critical reflection of this review process was highlighted in the discussion section of this paper.
This framework allowed the researcher to carefully synthesize the data.

Results
The search yielded a total of 10,026 references after duplicate references were removed. After application of the inclusion and exclusion criteria, the data of nine studies [27][28][29][30][31][32][33][34][35] were retained and synthesized in this review. The PRISMA flow chart representing this selection process is shown in Fig. 1. The characteristics of these nine studies are represented in Table 4.

Countries reported
Four studies reported data from Jamaica [32][33][34][35], three each from Barbados [27,34,35] and Cuba [29][30][31], two studies each from Trinidad and Tobago [34,35], St Lucia [34,35] and Grenada [28,35] and one study reported from each of the following islands: Antigua and Barbuda, Montserrat, St Kitts and Nevis, St Vincent and the Grenadines [35], Anguilla, Dominica and the Dominican Republic [34] . Please refer to Table 4 to view the studies in detail. There were no studies from Aruba, Bonaire, British Virgin Islands, Curaçao, Guadeloupe, Martinique, Saba, Saint Barthelemy, Saint Martin, Saint Eustatius, US Virgin Islands, The Bahamas, Turks and Caicos, Haiti and Cayman Islands. Regrettably, a few studies from Puerto Rico were written in Spanish and therefore did not meet our inclusion criteria.

Study methods
Seven studies [27, 29-32, 34, 35] used quantitative methodological approaches, and cross sectional survey designs were utilized within these studies. One study used a mixed methods design with qualitative and quantitative components [33]. One study used a qualitative approach using interviews and focus groups [28].

Study sample sizes
Sample sizes from seven studies, [27][28][29][30][31][32][33] were able to be determined. There were 312 patients across four studies [27,[31][32][33], 75 caregivers from two studies [28,33] and 104 health practitioners interviewed in four studies [28][29][30]33]. One study from Jamaica [33] interviewed seven local community members and held four focus groups ranging from 8-10 participants for each group. A further study [34] interviewed physicians from Caribbean countries, however the actual numbers of physicians were difficult to identify since the results from the Caribbean countries were combined with physicians reports from Latin America. One study [35] identified ten participants from different Caribbean islands; however it was difficult to determine the professions of those interviewed.

Diagnostic groups and health professionals
Diagnoses of the patients included in the review were patients with advanced cancer in two studies [31,33], end stage renal disease [32] in one study, and moribund intensive care [27] patients from one another. Two studies included eight physicians and fourteen nurses [28,33]. The specialties of these health professionals were not mentioned in the study. The health practitioners from two studies [29,30] comprised the same study subjects (deciphered from methodology, data presented and authors of the papers) and presented percentages suggesting that respondents specialties were medical (44 %), surgical oncology (25 %), anesthesiology (13 %), radiotherapy (8 %), paediatric oncology (5 %), nursing (4 %) and other (1 %). Meta-analyses of quantitative data were not performed in this review since the studies were heterogeneous in nature.

Thematic analysis
Eight themes arose from this review based on its broad scope . A significant number of themes were generated since the papers provided rich viewpoints from a diverse population. The themes generated in this review were, The majority of patients, even after the futility of treatment was established, withdrawal of any form of therapy was not attempted. Factors influencing the decisions on futility of care were: age of the patient, legal considerations, family influences and surgeon's refusal to accept futility in a post operative patient.

Age:
Withdrawals of no forms of therapy were made in seven brain dead patients under the age of 30 years.
In eight patients who were unresponsive after laparotomy, the four patients above 80 years, some form of therapeutic intervention was withheld. In the four patients who were in their sixties no support was withdrawn.

Legal considerations
Medico legal cases received more therapeutic support than non medicolegal cases.

Family influences
In 24 cases where family had hopes of their relatives recovery and requests to surgeons contributed to the maintenance of therapy. In two cases, there was fear of litigation on withdrawal of therapy since the patient's relatives live abroad. Surgeons refusal to accept futility of care in the post operative patient. Support was continued on 12 patients based on the surgeons inability to accept futility of care

Caregivers
More than half of the deaths of the patients occurred at home as reported by the caregivers. The themes uncovered were: Place of death The participants revealed that many of the deceased preferred to remain at home.

Pain and Suffering
Participants responses to pain and suffering of their deceased relative included: "pain and suffering is a must", "there was a time when I thought she (my wife) was going to die because she had so much pain", "bore her pain well, but whenever she was in great pain you would hear it", "he was afraid of the pain" and Qualitative score:0.70 "there is no real policy in place for that kind of thing", "we do not have an official, like support services, but we do have a hospital chaplain" and "facilities need improvement, more social workers, improved treatment to extend lives, more readily available care. We give lots of pain medications. Sometimes we run out, but rarely." Nurses thought they are not hostile but remarked: "patients don't tend to follow rules. I make sure they follow the doctors prescription and maintain the patients' visiting hours, and I don't think they like that.. we are short staffed", "with a dying patient we try and make the patient feel as comfortable as possible." Others thought compassionate care is hindered by low salaries.

Physicians opinions
Thought home care is often better than hospital care due to lack of hospital resources. One doctor pointed out that although dying persons abroad receive more pain relief, Caribbean people tend to "believe that illness is a god given destination, so they don't mind suffering. They believe that maybe it was something Availability of end of life services and access to cancer care in practice settings were identified as predictors of the quality of advanced cancer care in practice settings. Strong positive associations between the report of institutional availability of key opioid analgesics and higher quality of care ratings.The specialty of the provider was related with the quality of care ratings.Barriers to optimal cancer pain management predicted the quality of care ratings. Respondents who cited "medical staff reluctance to prescribe opiates" as a top barrier, provided quality of care assessments that were much lower than those who did not identify this as a top barrier. The use of the WHO three step analgesic ladder for cancer pain relief was positively associated with higher ratings of quality of care   Table 4 Data Extraction Table (Continued) viewed as an immediate "death sentence". Beliefs such as "only evil people get cancer" and that cancer is a form of "god's punishment" or cancer is a curse are held. These beliefs cause patients to believe that "obeah" (refers to "witchcraft, evil magic or sorcery by which supernatural power is invoked to achieve personal protection or the destruction of enemies") was the way to treat these disorders.Some patients believe there is a stigma attached to certain cancers, for instance cervical cancer may be linked to sexual promiscuity or prostate cancer heralds threats to masculinity and sexuality.

Theme 1 patients need for appropriate pain control and access to analgesics
Two studies [28,33] reported patients and/or their caregivers view that there was a need for analgesics and also access to these medications. In one study [28], caregivers mentioned "she thought she was going to die because she had so much pain" and "he was really scared about that pain" when asked if the deceased suffered much pain. In another paper [33], seventy five percent of the caregivers identified pain management as the most critical need of the person for whom they were caring.
Three papers [28,33,34] highlighted the issue of difficulty in accessing opioids in the Caribbean for cancer pain. One paper [34] drew attention to the opioid formulary deficiencies in some Caribbean islands that reduce access to some of these drugs. The paper illustrated the lack of particular opioids in some Caribbean islands, and restrictions enforced for prescribing and dispensing opioids. In addition, a study based in Cuba [29], health practitioners suggested that international restriction on the import and export of medications and the fear of diversion of opioids to the illegal market as two of the top barriers to optimal cancer care in that country.

Theme 2 patients need for support
The need for emotional, spiritual, financial and social support to patients and their caregivers were described in three papers [28,32,33]. Caregivers in Grenada [28], divulged the financial hardships secondary to medical and/or funeral costs for the deceased. They were unaware of any bereavement services such as counselling, financial assistance and home care assistance. One participant suggesting: "our system is not geared toward helping people; there is no social thing here to help you." In Cuba [29] health practitioners noted that the affordability-of-care predicted the quality of advancedcancer care. Spirituality was mentioned by the majority of the participants in the survey in Grenada [28] as being important to them. A participant relates "I think the best solution with no money or anything is the lord in prayer." They also identified that emotional support came from friends and family.
One study [31] used the Palliative Outcome Scale to discover that patients with advanced cancer in Cuba report their most burdensome problems as patient and family anxiety.
In Jamaica [32], results from a study on quality of life in end stage renal disease, attention was drawn to low patient satisfaction, general health, physical functioning, physical role and emotional role scores when compared to a reference population.

Theme 3 lack of knowledge of palliative care by health care professionals
Some of the physicians that participated in the needs assessment survey in Jamaica [33] poorly understood the meaning of palliative care and its constituents. In Cuba [29], health professionals identified that palliative care is not a priority in health care education as one of the top barriers to optimal cancer care.

Theme 4 lack of trained staff and need for more staff in palliative care
A cross sectional study [35] carried out amongst eight of the Caribbean islands included, found that there were five specialists in palliative care in total amongst three of these islands. In Jamaica, health care practitioners [33] believed there was not enough training in palliative care. Health care providers in Cuba [30] cited that the specialties of the provider were predictors of the quality of advanced cancer care. They also noted medical staff reluctance to prescribe opiates as a barrier to the quality of advanced cancer care. A study in an Intensive Care Unit in Barbados [27] reported that anesthesiologists and surgeons found it difficult to withdraw treatment even though there was an agreement of futility of treatment.
The survey participants in Grenada [28] voiced that there was not enough kindness from the nurses towards the deceased. The nurses in this same study allude to there being isolated cases of mistreatment of patients. Limited support services and the need for additional staff, for instance, the chaplain, counsellors and social workers were emphasised by the nurses interviewed in this study. Physicians who were interviewed conveyed concerns about the nurses' low salaries, workload, long working hours and also difficulty in accessing patients' homes to care for dying patients at home.
The participants in the Jamaican study [33] felt that there were communication issues between patients and health practitioners. They suggested that patients often do not understand what is being said to them by doctors. They felt disrespected and had lack of trust for the physicians.

Theme 5 need for specialized palliative care services
Of the eight Caribbean islands surveyed [35], two of the islands do not offer any palliative care facilities. Health care providers in Cuba [30] propose availability of endof life services as the main predictor of quality of care.

Theme 6 deficiency in health care policy
In two of the papers [29,33], health practitioners reported that palliative care appears not to be a priority in the formulation of health care policy. The participants in the Jamaican study went on to comment that "the capacity to deliver effective care to the terminally ill was curtailed by gaps in government support at both regional and national levels" [33]. This reinforced the absence of protocols and policies in the formation of palliative care services.
The paper which identified the greatest issues with health care policy is a report from the Global Opioid Policy Initiative (GOPI) [34]. This study highlights not only the absence of certain opioids in the formulary in some Caribbean islands, it also lists regulatory barriers for accessing opioids in these countries. A few barriers in some Caribbean islands were that physicians require special licence or permission to prescribe opioids, the limits in the number of days opioids can be dispensed, limitations on which pharmacies can dispense opioids and negative connotations of opioids in some countries, describing them as "drugs of addiction" or dangerous drugs". Lack of health care resources such as pain relief and personnel were noted by physicians from Grenada [28].
Finally, legal considerations must also be considered since ICU practitioners in Barbados [27] voiced the fear of litigation if they were to withdraw futile therapy from a patient.

Theme 7 cultural beliefs of patients
Some participants in the study in Jamaica [33] considered that the belief in folk wisdom and the patients' lack of knowledge had a part to play in their late presentation or unwillingness to present to health care facilities for suspected cancer. Some patients believe a diagnosis of cancer equals an "instant death sentence" and "only evil people get cancer". By the same token, reports from the physicians in the Grenadian study [28] state Caribbean people "believe that illness is a God given destination, so they don't mind suffering. They believe that maybe it was something they did, and it is a question of almost purification before they go beyond. So they accept it." Since cancer was deemed to have spiritual roots, patients felt that witch doctors or "obeah" (witchcraft or evil magic) persons or "madda" were better able to diagnose and treat spiritual problems rather than a medical doctor [33]. Physicians in the Grenadian study [28] suggested with regards to herbal medication, the terminal and chronically ill put faith into that approach so that they would have explored all avenues of treatment.
Another point generated by the Jamaican study [33] was the stigmatization of certain cancers. For instance, a diagnosis of prostate cancer was seen as a threat to masculinity. The participants of the focus groups mentioned that education for patients with regards to clarifying misconceptions about cancer and addressing the myths was required to assist with stigmatization.

Theme 8 patients preferred place of care and death
The majority of caregivers in the Jamaican survey [33] suggested that they preferred to care for their relative at home. The patients in this survey cited the preference for hospital care followed by home care. They expressed the preference for hospital care since there were inadequate home care services. In Grenada [28], many of the deceased preferred to stay/die at home for different reasons as reported by their caregivers.

Exploring relationships in the data
Idea webbing and concept mapping were used for exploring relationships in the data. This was chosen because of heterogeneity of the identified papers. The quantitative and qualitative data sets can be represented conceptually and link various findings from the individual papers [22]. Please see Fig. 2 for the model demonstrating the relationships from the studies included in this review.
Using the themes generated from the included studies, a model was also inductively formed from them and represented in Fig. 3. The use of idea webbing generated this second model and took a step further from Fig. 2 by attempting to make a connection between the themes induced from the different papers. The themes could be arranged into broad categories of needs, models of care, health care policy, outcome measures and cultural beliefs. The combination of health care needs and services are necessary for provision of palliative care. However, overarching health care policies and cultural beliefs must also be taken into consideration for an intervention such as palliative care.

Discussion
This is the first systematic review of the peer reviewed literature on palliative care from the Caribbean. A literature review [36] of models of palliative care delivery globally was undertaken as part of the Barbados Palliative care Needs assessment project. This article, together with the survey of cancer patients, caregivers and health professionals carried out in Jamaica [33] included in this review, represent two types of needs assessment methods; comparative and corporate approaches.
The results from our review described patients' needs and caregivers' requests for pain relief, access to analgesics, holistic support and assistance with their wishes for preferred place of care at the end of life. Community members interviewed, echoed the need for emotional support and counselling for patients and their caregivers. They added that education on diagnosis and treatment options was necessary. Health care practitioners identified their needs for more education and training in palliative care. Moreover, a lack of healthcare policy regarding laws, formulary deficiencies and legal considerations were significant findings in this review. These findings highlight many issues in the provision of palliative care in the Caribbean. The needs underscore the importance of developing sustainable models of care in palliative care in the Caribbean since the demand for it is apparent.
The World Health Organization (WHO) Public Health Strategy describes four components or pillars which must be addressed for integration of palliative care into society [37]. Researchers with an interest in palliative care in Africa propose in addition to the four pillars of policy, education, drug availability and implementation that a fifth pillar of research activity is also required to support this strategy in Africa [38]. The recognition by regional and world health bodies of the importance of palliative care integration into the health systems of the developing countries is mammoth. Therefore, for palliative care to be integrated into the Caribbean health care system a lot of work still needs to be done at the level of the core pillars. It is known that the understanding of a population's need for palliative care is crucial in planning services [39].
In order to plan palliative care services, health policy must be in place. Authors postulate that to contribute to effective policy, there must be a bridge in the gap in between theory, policy and practice [40]. The WHO Public Health Strategy provides the theory for integration of palliative care in health systems. The practice of palliative care in the Caribbean was only superficially described in the published literature. However, there is a notion of more established services in some of the Caribbean islands that is not documented in peer reviewed literature. In the process of reviewing literature for this research piece, apart from finding peer reviewed publications, documents such as opinion pieces, articles, and pamphlets were also discovered. Opinion pieces on services available in Trinidad and Tobago [41], surgical palliative care in Haiti [42], bereavement practices and their outcomes amongst religious and ethnic groups in the English speaking Caribbean [43] did not meet the inclusion criteria for this systematic review.
Trinidad and Tobago has a Palliative care Unit opened in 2014, which has not been reported in peer review literature as yet. Antigua has a hospice which has not been documented by the International Observatory on End of life care at Lancaster University in the mapping of global provision of palliative care. These are just a couple of examples which are known to the researcher that provide palliative care services and not acknowledged in research articles. Fig. 2 Demonstrating concept mapping of the themes extracted from the multiple studies included in this systematic review and their relationship with outcomes in palliative care generated from the studies Although, outcomes (quality of life and quality of death) were alluded by the papers in this systematic review, it was reassuring to find a validated tool (the Palliative Outcome Scale :POS) was used in one of the studies included.
Health care practitioners from this review voiced the need for health care policy on palliative care. Although, all the themes found in this systematic review are extremely important, addressing policy appears to be a consistent principle amongst most themes. The seminal paper produced by the Global Opioid Policy Initiative on formulary availability and regulatory barriers to accessibility of opioids in the Caribbean and Latin America tackles many crucial issues that are correlated with palliative care needs in the Caribbean [34]. One author from Grenada suggests the need for stakeholder consultation to produce 'locally effective and sustainable' services in palliative care [44]. She recognizes that there is a need for palliative care in the Caribbean; however its provision must be in a feasible framework of cost, cultural beliefs and the national health systems available in the Caribbean [44].
The idea of successful integration of palliative care into the health care system of the Caribbean stimulated the production of the model portrayed in Fig. 3. However, as more research is conducted in the Caribbean this model can be developed further.

Strengths and limitations
The studies included for this systematic review were retained following appraisal against the inclusion and exclusion criteria by one reviewer, a potential bias. In Fig. 3 Idea webbing representing the possible relationships across studies and their integration of these ideas into a proposed conceptual model using abstract and inductive reasoning mitigation, a second senior researcher was consulted if any doubts were encountered in retaining or rejecting a study for the review.
Overall, the studies were quite similar in design, with the majority utilizing cross sectional surveys. Cross sectional studies are well suited to the research question since they aim to be descriptive, determine prevalence and reveal associations with regards to outcomes. There was one qualitative and one mixed methodology cross sectional survey design. Sample sizes across individual studies were noted to be small. The numbers could not be pooled due to study/sample heterogeneity.
The QualSyst tool was an appropriate tool for assessing study quality; however it did lack the ability to identify specific biases in the papers. This may have resulted in papers having inflated quality grades. On reflection, other tools could have been considered especially for assessment of the qualitative studies. For instance, the Qualsyst tools were designed for the purpose of producing a threshold for inclusion of a study into a systematic review based on quality. However, a question posed by this review was to appraise the quality of available evidence. A decision was made for assessing the scores from the qualitative papers by dividing them into two main groups based on the cut-off point provided by the authors of the tool: adequate quality and low quality. There was limited application of the tool in the evidence base for assessing qualitative work which therefore restricted its validity and reliability for use in this type of research piece. The Qualsyst tool when used for assessing the methodological quality of quantitative studies was found to be applied in a few studies in the literature base. This allowed for an established way of using the tool by grading the literature.
In general, biases and methodological flaws were observed in many of the studies, which included: selection bias with the use of non-probability sampling, reporting bias for those studies that collected narratives from participants, and observer bias where data collectors may have reported favourably or unfavourably on certain outcomes subliminally or interpreted the narrative incorrectly. In addition, the characteristics of the participants who did not respond to the survey were sometimes not presented, introducing potential non-participant bias. Measurement bias occurred in a few studies since non validated tools were used.
In addition, some papers were excluded since they were in Spanish and others could not be retrieved, instituting publication bias and also the inability to include possibly important information.
Non probability sampling methods and where the sample was drawn from may have created unrepresentative samples for ethnicity, education level, socio economic status and abode (whether urban or rural areas) in most of the studies included in this review. In addition, physicians who would have given information on their countries palliative care services, policies and opioid records may not have reported fully accurate or recent data. The studies included in this review ranged in publication from 2003-2014. The paper published in 2014 [35] describes palliative care services in the Caribbean; however, many of the earlier published papers helped shape the palliative care needs in the Caribbean. These needs may have changed over time.
Furthermore, five out of the nine papers included in this review focused on a diagnosis of cancer, with one on end stage renal disease. Therefore, the needs of patients with other disease processes requiring palliative care at some point during their illness trajectory were not described.

Conclusions
This systematic review has revealed that the peer reviewed literature offers little evidence on palliative care needs of the Caribbean population. The available evidence was broadly divided into health care practitioner's needs, patients' needs and health care institutional needs. They included patients' needs for access to analgesia, preferred place of care /death and multidimensional aspects of support needed for patients and their caregivers. Health care practitioners spoke about their need for health policy and education in palliative care. In addition, needs of more trained staff and the services in palliative care were cited as institutional needs. The models of care existent in the Caribbean were not well described in the peer reviewed literature. There were no intervention studies. It is timely to now develop and evaluate evidence-based palliative care services in the Caribbean.