Exploring Barriers to Primary Health Care Services on First Nations Reserves in Manitoba By Geoffrey Johnson BSc., University of Victoria, 2009 THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE IN COMMUNITY HEALTH THE UNIVERSITY OF NORTHERN BRITISH COLUMBIA May 2012 © Geoffrey Johnson, 2012 1+1 Library and Archives Canada Bibliotheque et Archives Canada Published Heritage Branch Direction du Patrimoine de I'edition 395 Wellington Street Ottawa ON K1A0N4 Canada 395, rue Wellington Ottawa ON K1A 0N4 Canada Your file Votre reference ISBN: 978-0-494-87535-3 Our file Notre reference ISBN: 978-0-494-87535-3 NOTICE: AVIS: The author has granted a non­ exclusive license allowing Library and Archives Canada to reproduce, publish, archive, preserve, conserve, communicate to the public by telecommunication or on the Internet, loan, distrbute and sell theses worldwide, for commercial or non­ commercial purposes, in microform, paper, electronic and/or any other formats. L'auteur a accorde une licence non exclusive permettant a la Bibliotheque et Archives Canada de reproduire, publier, archiver, sauvegarder, conserver, transmettre au public par telecommunication ou par I'lnternet, preter, distribuer et vendre des theses partout dans le monde, a des fins commerciales ou autres, sur support microforme, papier, electronique et/ou autres formats. The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. L'auteur conserve la propriete du droit d'auteur et des droits moraux qui protege cette these. Ni la these ni des extraits substantiels de celle-ci ne doivent etre imprimes ou autrement reproduits sans son autorisation. In compliance with the Canadian Privacy Act some supporting forms may have been removed from this thesis. Conformement a la loi canadienne sur la protection de la vie privee, quelques formulaires secondaires ont ete enleves de cette these. While these forms may be included in the document page count, their removal does not represent any loss of content from the thesis. Bien que ces formulaires aient inclus dans la pagination, il n'y aura aucun contenu manquant. Canada ii Abstract This study examines some of the barriers facing the delivery of effective, culturally appropriate primary health care services to First Nations people living on reserves in Manitoba. The data for this study comes from a data set made up of the responses to the Manitoba portion of the First Nations Regional Longitudinal Health Survey (RHS). The relationships between reported barriers to accessing health care and self-rated health were used to inform this work. Analysis relied on the use of Chi-square, Cramer's V, and adjusted residuals to determine statistically significant relationships and their respective strengths. The results indicate that barriers to access to primary health care on-reserve could be classified as: underfunding, limited human resources, lack of culturally appropriate care, and inadequate levels of local control over service delivery. iii Contents Abstract ii List of Tables vi Acknowledgement vii Chapter One: Introduction 1 Conceptual Framework 1 Research Objectives 2 Chapter Two: Literature Review 5 Background 5 The Role of Primary Health Care 6 Human Resources in Primary Health Care 10 Non-Insured Health Benefits 12 Health Transfer Policy 14 Health Transfer Outcomes 17 First Nations Regional Longitudinal Health Survey 21 Summary 23 iv Chapter Three: Data and Methods 24 Study Population 24 Data Collection 26 Variables of Interest 26 Ethics Approval 28 Objectives and Hypotheses 28 Statistical Analysis 32 Summary 33 Chapter Four: Results 35 Findings 35 Summary 43 Chapter Five: Discussion of the Findings 44 Funding 44 Human Resource and Facility Availability 45 Culturally Appropriate Care 47 Community Control 50 V Limitations 55 Summary 59 Chapter Six: Conclusion Implications 60 61 References 64 Appendix 1: Adjusted Residuals for Barriers and Community Characteristics 69 Appendix 2: Adjusted Residuals for Health Transfer Agreement 71 vi List of Tables Table 1 Facility designation criteria 9 Table 2 Community types and populations 25 Table 3 Statistical relationship between self-rated health and barriers to access to care 36 Table 4 Statistical relationship between self-rated health and community characteristics 41 vii Acknowledgement I would like to thank my supervisor Dr. Josee Lavoie for all of the support and opportunities that she gave to me over the course of my graduate studies. This thesis would not exist without her guidance. My thesis committee also provided invaluable guidance in the design of the study and the writing of the thesis. I would also like to thank the Assembly of Manitoba Chiefs for their support of this research. In particular I would like to thank Leona Star, Dr. Kathi Kinew, Chief Norman Bone, Ardell Cohcrane, and Leanne Gillis for the time that they have put into helping me with data, arranging logistics, and discussing the RHS. Finally I would like to thank my family and friends for putting up with all of the times I missed events or left early with this thesis as my excuse. This document is the end of that excuse. Chapter One: Introduction Health care in Canada is a source of national pride and universal access to medically necessary care has been available for decades, leading to some of the best health outcomes in the world. The system is ranked twelfth, for life expectancy at birth (80.7 years), out of thirty-one OECD countries (OECD, 2011). In spite of this world class health care system there still exists within Canada marginalized populations who remain unable to access necessary care at the same level as mainstream Canadians. Studies from throughout Canada indicate that First Nations people lag behind the rest of society in many indicators of health (MacMillan et al., 2003; Martens, Sanderson, Jebamani, 2005; Lavoie et al., 2010). In response to this, First Nations people have developed and implemented a national survey through which health needs can be measured and appropriate steps can be taken to address those needs. Conceptual Framework The poor health status experienced by many First Nations people relative to the rest of the Canadian population raises the issue of equity of access to care. Inequity can occur either within a country (e.g. between Caucasians and First Nations people) or between different countries (e.g. between Canada and Somalia). In this study, the concept of equity will be approached as it pertains to the Canadian context of First Nations people and mainstream society. Whitehead (1992) argues that inequity refers to differences in health that are not only avoidable and unnecessary but also considered to be unjust. Whitehead acknowledges that variation in health due to biological variation, freely-chosen health damaging behaviour, and transient advantages due to a certain group adopting a healthy behaviour would not be 2 considered inequities. However, differences due to restricted lifestyle choice, unhealthy/stressful living and working conditions, insufficient access to health care, and the tendency for sick people to move down the social scale as a result of their sickness are considered to be avoidable and therefore inequitable. Culyer (2001) recommends that if policy is to reflect the principles of equity, there must be equal and universal access to assessment, but unequal access (i.e. access only when deemed necessary by assessment) to the treatments. The unequal access to the approved treatments reflects the reality that health care in Canada must operate within the constraints of scarcity of resources. Culyer suggests that approved treatments should be the ones that make the greatest strides to reducing gaps in health. Thus this study will consider areas in which policy reform has the potential to improve the health of First Nations people living on reserve. Research Objectives A starting point for reducing the gap in health status comes in the form of primary health care (PHC)1, which has been shown repeatedly to provide an efficient way of improving the 1 Primary health care is defined by Starfield (1996) as encompassing primary, secondary, and tertiary prevention as well as primary care. Starfield (1998) defines primary care as "that level of a health service system that provides entry into the system for all new needs and problems, provides person-focused (not disease-oriented) care over time, provides care for all but very uncommon or unusual conditions, and co-ordinates or integrates care provided elsewhere by others." 3 health of a target population (Starfield, Shi, & Macinko, 2005). Improvement of PHC services on-reserve is one step towards improving the health status of First Nations people in Manitoba. The objective of this study is to explore barriers to accessing primary health care services on First Nations reserves in Manitoba. The barriers will be identified by the analysis of data coming from the 2008-20010 First Nations Regional Longitudinal Health Survey (RHS), as well as exploring what is already within the literature. The use of the RHS data brings an end user perspective into the literature, since all of the participants are reporting on their experiences with health care in their communities. The study will endeavour to answer the following questions: 1. What is the current state of primary health care services in First Nations communities? 2. What steps forward are indicated by the literature and by First Nations communities? The current state of on-reserve PHC services will be assessed via a review of the current literature, specifically looking at health outcomes, service level designation, human resources, and non-insured health benefits. The literature review will also bring some context to the current state of PHC services by looking at the Health Transfer Policy and its relationship to community control and how these shape the landscape of PHC services. The literature review will finally look at how First Nations are bringing their knowledge and opinions to the PHC discussion via the RHS. The analytical portion of this study delves into data from the RHS regarding community characteristics and barriers that are preventing people from accessing PHC services. This analysis brings the collective experience and voice of many First Nations people in Manitoba into PHC discussion. This perspective along with data from the literature are then brought together in the discussion, where they provide an critical examination of barriers to primary health care services on First Nations Reserves in Manitoba. 5 Chapter Two: Literature Review In this chapter, the literature pertaining to the status of First Nations health in Canada will be explored to provide background and rationale for the current study. Following this will be a discussion of primary health care, service level designation, human resources, and non-insured health benefits and the role they all play in on-reserve health outcomes. This leads into a discussion of the Health Transfer Policy which is the current structure under which on-reserve health care is provided in Manitoba and in most First Nations across Canada. The chapter then concludes with an overview of the First Nations Regional Longitudinal Health Survey. Background First Nations people in Canada are overrepresented for a multitude of health problems when compared to the mainstream population. A recent study of Manitoba First Nations health services use found that, relative to all other Manitobans, First Nations had: (1) higher rates of ambulatory physician visits (6.13 v. 4.85 visits/person), (2) higher rates of hospital separation (0.348 v. 0.156 separations/person), and (3) lower rates of specialist visits (0.895 v. 1.284 visits/person) (Martens, Sanderson & Jebmambi, 2005). The higher rates of hospitalization for ambulatory care sensitive conditions2 are indicative of inadequate access to PHC services, which may be due to a lack of culturally appropriate PHC models or even a lack of PHC services 2 Ambulatory care sensitive conditions (ACSC) are conditions "for which timely and effective outpatient care can help to reduce the risks of hospitalization by either preventing the onset of an illness or condition, controlling an acute episodic illness or condition, or managing a chronic disease or condition" (Billings et al., 1993, p. 163) 6 altogether (Brown et al., 2001; Jiwa, Kelly, St Pierre-Hanson, 2008; Lavoie et al., 2010). These findings support the results of an earlier study in Ontario which found that residents of First Nations communities were 2.54 times more likely to be admitted to hospital for ambulatory care sensitive conditions (ACSC) than the general population and 0.64 times as likely to utilize specialist services (Shah et al., 2003). Variables such as geography and socioeconomic status were only able to partially explain the risk (Shah et al., 2003; Martens, Sanderson & Jebmambi, 2005). In 2006-2007 asthma was responsible for 15% of all ACSC hospitalizations in Canada and chronic obstructive pulmonary disease was responsible for another 29%; COPD was the only ACSC related hospitalization to increase between 2001-2002 and 2006-2007 (Sanchez et al., 2008). This is relevant since First Nations people in Alberta were found to be 2.1 and 1.6 times more likely to visit emergency departments and physicians' offices for asthma and COPD (Sin, Wells, Svenson, & Man, 2002). Diabetes, which accounts for a further 13% of the total ACSC hospitalizations, is prevalent in some First Nations communities at three to four time the rate observed in the general population (Dyck, Osgood, Lin, Gao, & Stang, 2010; Green, Blanchard, Young, & Griffith, 2003; Sanchez et al., 2008). The Role of Primary Health Care PHC has been reported to have positive effects on the well-being of the population served as well as on the efficiency of the health care systems (Starfield et al., 2005). The basis of this claim rests in three bodies of literature: studies involving the distribution and supply of primary care physicians, studies involving people who rely mainly on primary care physicians, and studies which assess the relationship between health status and the utilization of primary care 7 services (Starfield et al., 2005). Each of these groups of studies showed positive influences on the health of those served by PHC systems. The effects of PHC were also found to transcend time, race, geographical regions (national and international), and economies (Starfield et al., 2005). The one exception noted was that a large supply of primary care physicians in urban settings did not always ensure access to all sub-populations, indicating that barriers to access differ between rural and urban settings (Starfield et al., 2005). Starfield (1996) differentiates PHC from secondary and tertiary care, stating that PHC encompasses primary, secondary, and tertiary prevention as well as primary care, none of which require hospitalization. Currently PHC services are provided on reserves through health offices, health stations, health centers, or nursing stations, each of which provides different levels of services. The level of services is determined by the number of registered First Nations people living on-reserve, proximity of provincial services, quality of roads, and distance to referral centers (see Table 1). The different facilities, in the order listed above, provide increasing levels of health services where each step up the ladder provides the same services as the one below plus additional services; health offices provide screening and prevention services on a part time basis, health centers improve with provision of emergency services and increased hours (limited or no offhours coverage), and nursing stations provide additional treatment services and nursing care on a 24/7 basis (Lavoie et al., 2010). This translates into differing levels of available PHC services between reserves; where some communities have access to nursing stations which can provide a large range of PHC services, more than half of reserves do not qualify for nursing stations and instead have health centers or health offices which can provide only limited services (Lavoie et al., 2010). Since the health care system operates within the confines of limited resources it is not 8 reasonable to expect that the same level of care offered in Winnipeg can be offered in every community in Manitoba. Still the higher rates of hospitalization reported by Martens, Sanderson and Jebamani (2005) and the higher rates of hospitalization for ACSCs reported by Lavoie and colleagues (2010) indicate that investments in PHC might yield some benefits. Furthermore, elevated avoidable hospitalization rates ultimately increase costs in secondary and tertiary health care sectors leading to further pressures on these systems (McDermott & Segal, 2006). Thus this study will consider how primary health care services can be improved in First Nations communities. 9 Table 1 Facility designation criteria Type of Facility Community characteristics (the community should meet a majority of the following criteria). On-reserve health services funded: Health office Criteria: Population: 0 to 750 total on-reserve population, non-isolated and semi-isolated community (isolated under favourable conditions). Proximate health services: other health services available in nearby communities/cities, hospital accessible by road in less than 2 h. Transportation: All weather road/air access. Community infrastructure: Adequate community services. Part-time, often non-resident screening and prevention services only, Health station Criteria: Population over 100 on-reserve, remote isolated to semiisolated community, over 150 km from a service centre but within 50 km of a nursing station or other FNIHB facility. Proximate health services: hospital accessible by road less than 2 h, occasional unavailability of local ambulance and first response services. Transportation: Accessible by air or road from FNIHB facility, poor road conditions. Community infrastructure: Limited community services. Part-time screening and prevention services only, Health center Criteria: Population over 100 on-reserve. Non-isolated and semiisolated community, less than 350 km from a service centre. Proximate health services: hospital accessible by road less than 2 h, occasional unavailability of local ambulance and first response services. Transportation: All weather road/air access, poor road conditions. Community infrastructure: Limited community services. Emergency, screening and prevention available 5 days per week, with limited or no after hour care locally. Nursing station Criteria: Population over 500 on-reserve. Remote or isolated community, over 350 km/3 h travel to a service centre. Proximate health services: nearest hospital more than 2 h away, limited availability of local ambulance and first response services. Transportation: No year round road access to other health care facilities. Community infrastructure: Limited community services. Treatment and prevention, accessible 24/7. Note. Adapted from Lavoie et al., 2010. 10 Human Resources in Primary Health Care One of the major difficulties facing the provision of adequate PHC services on reserves is the recruitment and retention of health care professionals. This means that communities are often faced with shortages and/or high rates of turnover of health care professionals. Many reasons are cited for this high turnover, including lack of opportunities for spouses, lack of family ties and excessive workloads (Aylward, Gaudine, & Bennett, 2011; Manahan & Lavoie, 2008). The process of "Othering" by health care professionals, defined as differentiation based on perceived or presumed differences from the dominant culture, can also contribute to the marginalization of patients (Tarlier, Brown, & Johnson, 2007). One way of dealing with these issues is through the use of paraprofessional health workers. First Nations communities in Canada have been using community health representatives (CHR) since 1962 as a way to enhance the provision of PHC services. The originally intended role of CHRs was to provide some level of health promotion and disease prevention, with little direct care to patients; however, in some communities CHRs took on expanded roles out of necessity (Minore, Jacklin, Boone, & Cromarty, 2009). CHRs are often members of the community being served and are therefore much more likely to commit to a longer term within the community providing a source of continuity within the community health care system. In spite of the services they provide to First Nations communities, CHRs have started to face opposition to their provision of direct care to patients, due mainly to concerns of professionals and administrators (Minore et al., 2009). These concerns are based mainly around issues of liability and accountability, since there is no formal accreditation, certification and regulation of CHRs (Minore et al., 2009). 11 Another area of health human resources to consider is traditional healers. Traditional healers are not a well-defined group with a specific scope of practice; rather, there are many types of traditional healers working from various points of knowledge ranging from the spiritual to extensive knowledge of plant uses (Waldram, Herring, & Young, 2007). This diversity makes it difficult to assess the specific levels of traditional care available on reserve. Other First Nations health care professionals also play an important role in providing health care on-reserves. It was estimated that as of 2005, Aboriginal (Inuit and Metis included) physicians made up approximately 0.25% of the physician population while Aboriginal people made up about 4% of the entire Canadian population (Anderson & Lavallee, 2007). Macaulay (2009) points out that it is not only physicians but all health professions that are underrepresented in terms of the number of Aboriginal people working within the various professions. The low numbers of First Nations health care professionals means that people from outside of the community must come in to deliver health care, which reduces the probability that people will receive care that is culturally appropriate. This is not to say that non-First Nations people cannot provide culturally appropriate care, rather that in many cases the high turnover and unfamiliarity with the community can lead to a lack of understanding of the local culture (Tarlier, Brown, & Johnson, 2007). In response to these under-representations, some nursing and medical schools have implemented initiatives to try and increase the numbers of Aboriginal students enrolling in health care professions (Anonson, Desjarlais, Nixon, Whiteman & Bird, 2008). Even with these initiatives in place enrolment has not yet reached optimal levels. A large part of this is likely due to early educational experiences of First Nations students. In their Call to Action on Education, 12 the Assembly of First Nations states that only half of First Nations youth complete high school. This factor alone already cuts in half the population available for further training in health care professions. On top of this First Nations schools receive, on average, over $2000 less per child than provincially funded school. This underfunding often leads to inadequate learning materials as well as increasing difficulty in recruiting and retaining qualified educators (Call to Action on Education, n.d.). A better early educational experience has the potential to significantly increase the number of First Nations high school graduates and in turn increase the number of students eligible for university training. Non-Insured Health Benefits Non-insured health benefits (NIHB)3 refers to a benefit plan for registered First Nations people, funded by the First Nations and Inuit Health Branch of Health Canada. The program provides prescription drugs, over-the-counter medication, medical supplies and equipment, shortterm crisis counselling, dental care, vision care, and medical transportation to qualified First Nations and Inuit people. Although these benefits do not fall within the realm of PHC, they are designed by Health Canada to help First Nations in attaining health outcomes similar to those of other Canadians (RHS, 2005; Health Canada, 201 la). 3 NIHB - The term non-insured is a reference to services and supplies that fall outside of the coverage of provincial health insurance plans (Waldrum, Herring, Young, 2007). 13 Prior approval is required on all vision care, some dental care, all medical transportation, some drugs, and most medical supplies/equipment (e.g. catheters, respiratory supplies). Health Canada (201 lb) states that denial of approval can be appealed by the person, their parent, a legal guardian, or some other representative. The information required by NIHB is: 1. the condition for which the benefit is being requested; 2. the diagnosis and prognosis, including what other alternatives have been tried; 3. relevant diagnostic test results; and 4. justification for the proposed treatment and any additional supporting information. Once the appeal is received it is reviewed by a relevant health care professional and a decision is made based on the needs of the person, the medical justification, the availability of alternatives, and NIHB policy. If the person does not agree with the outcome of the appeal process they can request to have the appeal reviewed at the second level, and even a third and final level if they also disagree with the outcome of the second appeal process. The 2002/2003 RHS national survey indicated that First Nations people often find that these approval and appeal processes act as barriers to receiving the benefits that are designed to help them (RHS, 2005). In 2009/10 NIHB expenditures were the highest in Manitoba with 19.8 percent of national expenditures occurring here for 16.2 percent of the eligible population (Health Canada, 201 lb). This disproportional expenditure has been attributed to the high proportion (29.5%) of medical transportation used by Manitoba First Nations, resulting from high levels of clients living in remote and/or fly-in only communities (Health Canada, 201 lb). As discussed before, effective PHC services can reduce the rates of hospitalization for ACSCs. This means that improved PHC services will not only result in better health outcomes, but will also reduce the level of expenditure on avoidable medical transportation, freeing up funding for other necessary 14 services. Manitoba is also the second largest user of pharmacy benefits, which are responsible for the largest portion of total expenditures (Health Canada, 201 lb). Furthermore, these expenditures are growing every year and it has been projected that if current trends continue and service levels are not changed, there will be a 57 percent increase in expenditures for First Nations people living on-reserve (Lavoie & Forget, 2006). Thus having been found to be a barrier across the nation, having a close relationship to PHC services, and comprising a large portion of the health care expenditures, NIHB will be included in this study. Health Transfer Policy The Health Transfer Policy (HTP) was introduced in 1989 and it was viewed as the answer to dealing with the existing health inequalities (Lavoie, 2004). The three objectives of the HTP were: to enable First Nations to design establish and allocate funds to health services that are relevant to the communities' priorities, to improve accountability of local leadership to community members, and to ensure that mandatory programs are delivered in order to maintain public health (Lavoie et al., 2005). The HTP was also an answer to First Nations' desires to participate in the development and delivery of health programs that were relevant to the needs of individual communities (Lavoie, Forget, & O'Neil, 2007). Part of the driving forces that contributed to the formation of HTP was the success of the Esketemec First Nation in British Columbia. Plagued with alcoholism and high unemployment rates, the Esketemc people designed and implemented a plan in which the sale of alcohol was prohibited, and recovery programs were made available; as a result the unemployment rate was greatly ameliorated (Warry, 1998). In a ten year span (1971-1981) the Esketemec First Nation was able to reduce the prevalence of alcoholism from 95 percent to 5 percent (Guillory, Willie, Duran, 1988). This 15 success illustrated the ability of community members to not only act in leadership roles, but also to show that they could identify, prioritize and respond to issues in an effective manner. Prior to the HTP, there were two programs that were run by First Nations, Community Health Representatives and the National Native Alcohol and Drugs Addictions Program. The HTP, combined with the programs already in place, allowed for an expanded scope of services which could be provided under the direction of First Nations; however, the policy did not allow for an expansion in the types of services provided on-reserve (Lavoie et al., 2007; Warry, 1998). Effectively, while the government did allow for more control at the local level, it did not allow First Nations to offer more PHC services than were historically offered. Due to the different demographics and capacities of First Nations organizations, there were different options as to the level of transfer in which an organization could participate. These options were expanded in 1994 when the First Nations and Inuit Heath Branch of Health Canada (FNIHB) introduced the Integrated Community option (Lavoie, 2009). The integrated option allowed smaller communities, which were previously not eligible for transfer, to participate (Lavoie et al., 2007). By 2003, 78 percent of communities eligible for transfer were involved in some level of control of their local health services and by 2008, 83 percent were involved (Lavoie et al., 2007; Transfer Status as of March 2008, n.d.). Despite the high level of participation by First Nations in the HTP, the policy was not without critics. The government's policy focus of transferring control of existing services to First Nations was seen by some as a way in which the federal government could move away from their responsibility for First Nations health care (Warry, 1998). Critics argued that the level of self-determination was small given the inflexible guidelines and externally dictated rules; 16 questions were also raised about whether the inflexibility would defeat the purposes of local sensitivity to health priorities (Warry, 1998). It may have appeared to the critics that the federal government was simply paying lip-service to the idea of self-determination. Despite the validity for some of these criticisms, there was acceptance of the policy by many First Nations, and some of the limitations in service expansion were overcome via the documentation of need for those services (Warry, 1998). In some cases the imposed limitations were even ignored during local planning of what would be required for relevant quality health care, exemplifying the resolve of First Nations to realize self-determination in spite of restrictive policies. In one health transfer process, a group of volunteers who represented First Nations communities came together to form a Health Planning Circle which was used to identify the particular needs of the various communities. One of the issues that arose during the identification of community needs, was the fear that once transfer was completed funding schemes would become entrenched (Warry, 1998). Politics between various communities and the tribal council, as well as lack of clarity about some members' roles in decision making, also proved to be barriers that slowed the process. This lack of clarity regarding roles has also been documented in another study that assessed post-transfer issues (Lavoie et al., 2005). An additional issue that arose and ended up having serious implications was the tension between local and regional needs. This tension was exploited during the negotiation process such that rather than coming to regional agreements, the various communities had their own separate transfer agreements (Warry, 1998). Despite all of these issues and fears, 97.1 percent of Health Directors who responded to the national evaluation of the HTP indicated support for continuing a mechanism which encourages community control (Lavoie et al., 2005). 17 Health Transfer Outcomes A study was conducted in the early days of the HTP within the First Nations community of Montreal Lake in northern Saskatchewan. Band members had to travel one hundred kilometers to access health services. The Montreal Lake Band conducted a needs assessment and based on that assessment determined that a primary health care facility, controlled by the First Nation, would be able the best approach to address many of the findings from that assessment. The Band went through a three year planning and negotiation process, so that in the fall of 1988, the William Charles Health Center was opened and the band became the first to take control over local health care via a transfer agreement (Waldram, Herring, & Young, 2007). Moore, Forbes and Henderson (1990) found that within a year there were positive changes in health awareness, health maintenance behavior and health status. One observed change in health awareness was Band members' perception of being safer. In the first year that the health center was open many medical emergencies (e.g. child birth, coronary events, injuries) that previously would have required transportation to larger urban centers were handled by the Health Center. Additionally, given the new found sense of security, community members became more willing to attempt to manage minor illnesses at home, knowing that if things did get worse there was help close at hand. This sense of security was credited with reducing the number of early hospitalizations. Furthermore, the increased health awareness, in concert with the freedom granted under a transfer agreement, allowed a Band council member (also a health coordinator) to recognize an emerging need within the community and respond to it in a timely and effective manner. Health maintenance was also improved with the sudden ability to access a culturally relevant health care system. Community elders reported that treatment staff members were familiar with their home 18 and family circumstances, making them feel safer and more welcome than when they accessed health services in an urban center where only English was spoken. Another particularly important change noted was the earlier visits from prenatal patients. The increase was attributed to ease of access, lack of embarrassment and feeling accepted (Moore, Forbes, Henderson, 1990). Additionally, community nurses were able, through trial and error, to find effective ways of increasing the rates of immunization. There were multiple areas in which health status was improved; and these improvements were attributed to people seeking care sooner and thereby avoiding more serious illness as well as the ability of nurses to tailor treatment due to their knowledge of living conditions and family history. With improved health came fewer hospitalizations and fewer emergency outpatient visits (Moore, Forbes, Henderson, 1990). The experiences of Montreal Lake illustrate the beneficial results that can occur with increased levels of community control over health service delivery. More recent studies have also analyzed the extent to which local control over health services affects health status within transferred communities. The findings from these studies are positive in that communities that have been transferred the longest are showing improved health outcomes (Lavoie et al., 2010). However, in 2005 there was a national evaluation of the HTP and some concerns were raised. Using data from the evaluation, Lavoie and colleagues (2007) tested three hypotheses framed around the larger question of whether there were differences in funding between transferred communities. The hypotheses were: 1) inequalities exist in per capita funding between communities; 2) the inequalities are not explained by the level of transfer; and 3) communities that transferred in the early 1990s have less funding than those communities that transferred more recently. Each of these hypotheses was found to be supported. The authors 19 suggest that inequality in funding between communities might be reflecting the differential characteristics of each community (e.g. some communities may have higher levels of need due to geographical location or prevalence of health conditions). The differences in funding between years may have resulted from federal funding cuts in the mid-1990s as well as the practice of basing transfer negotiation on expenditures from the previous year. The idea behind basing funding on the previous year's expenditures was that the level of need was reflected by what was being spent (Lavoie et al., 2007). However, this logic is flawed as it assumes that historical funding was already at an adequate level and that health care costs would remain static. Recall that studies of hospitalization rates for ACSC indicate that there are still needs which are unmet in First Nations communities (Martens et al., 2005; Shah et al., 2003). This concern was echoed in the Evaluation of the First Nations and Inuit Health Transfer Policy, where more than 85 percent of respondents reported that funding does not match the needs of the community, provide for population growth, or take into account the use of services by off-reserve and non-status users (Lavoie et al., 2005). There are also differing levels of control that are allowed under the HTP: selfgovernment, transfer, integrated, and other. One of the major differences between these agreement types is the level of control that the First Nation has over the delivery of health care services. The most common agreements are transferred or integrated, with self-government agreements being less common (twenty-three communities Canada-wide, none of which are in Manitoba) (Transfer Status as of March 2008, n.d.) In a transfer model, a 3 to 5 year Contribution Agreement allows a community (or communities in the case of multi-community transfer agreements) to administer a number of programs in a flexible way that allows for the 20 prioritization of local needs (Lavoie et al., 2005). In addition, this funding can be carried over from one year to the next to be used to address local health priorities. An integrated model is similar with two important exceptions: first, communities can allocate funding according to their work plan and if they want to make changes must obtain permission from FNIHB; second, there is no allowance for the carryover of resources. These two differences lead to a decrease in the flexibility of the community to respond in a timely manner to changing priorities in the delivery of health services. Concerns have been raised regarding renegotiation; specifically, that renewal is not guaranteed and that funding becomes entrenched without room for renegotiation (Jacklin & Warry, 2004). An example comes from the experiences of the Wikwemikong Unceded Indian Reserve where when it came time to renegotiate, the only offer was a 3% increase aimed at wage increases with no option to expand services regardless of reports by Health Canada indicating increases in reserve population and rates of disease (Jacklin & Warry, 2004). Perhaps the most alarming factor in the renegotiation was the lack of reference to self-government and the lack of opportunity to renegotiate the terms of the contract, effectively removing the Wikwemikong people from the decision making process and forcing them to accept the externally dictated policy (Jacklin & Warry, 2004). Residents of the reserve felt that government did not understand local needs, failed to recognize the uniqueness of the community, and failed to provide funding in response to demonstrated need. This has led to feelings of disempowerment from a policy that was supposed to set the stage for self-determination (Jacklin & Warry, 2004). The transfer agreements include non-binding dispute resolution provisions that many feel are futile and those who have pursued dispute resolution have been left unsatisfied with the unilateral decisions 21 (Lavoie et al., 2005). Despite of all this, self-government agreements are proceeding. In Manitoba there are no communities that have finalized a self-government agreement; however, the Sioux Valley Dakota First Nation signed a comprehensive agreement in principle in 2001 (Canada & Sioux Valley Dakota Nation, 2001). First Nations Regional Longitudinal Health Survey In 1994 Statistics Canada began three national longitudinal surveys which explicitly excluded First Nations peoples living on-reserve as well as Inuit Communities (RHS National Team, 2007). In 1996, the Assembly of First Nations Chiefs Committee on Health responded to this exclusion by mandating a First Nations health survey to be implemented across Canada; the First Nations and Inuit Regional Longitudinal Health Survey (FNIRLHS) was the result of this mandate. In 1997, the first survey took place and 14,008 surveys were collected in 186 First Nations and Inuit communities (FNIGC, n.d.). Eventually, the FNIRLHS became the First Nations Regional Longitudinal Health Survey when the Inuit opted for Inuit specific research and the survey is now commonly known as the RHS (RHS National Team, 2007). The RHS is the only national health survey that is designed, delivered and governed by First Nations. It is guided by a code of ethics central to which are the First Nations principles of Ownership, Control, Access, and Possession (OCAP). The RHS code of ethics (2007) defines these principles as follows: Ownership: The notion of ownership refers to the relationship of a First Nations community to its cultural knowledge/data/information. The principle states that a community or group owns information collectively in the same way that an individual 22 owns his or her personal information. It is distinct from stewardship or possession (see below). Control: The aspirations and rights of First Nations to maintain and regain control of all aspects of their lives and institutions include research and information. The principle of "control" asserts that First Nations, their communities and representative bodies are within their rights in seeking to control research and information management process which impact them. This includes all stages of research projects, and more broadly, research policy, resources, review processes, formulation of conceptual frameworks data management, and so on. Access: First Nations people must have access to information and data about themselves and their communities, regardless of where these are currently held. The principle also refers to the right of First Nations communities and organizations to manage and make decisions regarding access to their collective information. Possession: While "ownership" identifies the relationship between a people and their data in principle, the idea of "possession" or "stewardship" is more literal. Although not a condition of ownership, possession (of data) is a mechanism by which ownership can be asserted and protected. When data owned by one party are in the possession of another, there is a risk of breach or misuse. This is particularly important when trust is lacking between the owner and possessor. Phase one of the RHS occurred in 2002/2003 and made many improvements on the 1997 pilot survey. The phase one iteration underwent an independent review by the Harvard Project on 23 American Indian Economic Development and was found to have high technical quality in sampling design, data collection, data analysis, and dissemination (Harvard, 2006). The 20082010 iteration of the RHS was the second phase of the study and it underwent extensive revisions to deal with comparability, non-response, and redundant questions. New themes were added, and the adult survey included new questions regarding migration, food security, violence, care giving, depression, gambling, and new health indicators (FNIGC, 2011). Summary Inequalities in health status experienced by Manitoba First Nations are unnecessary and believed to stem partially from inadequate access to and utilization of PHC services. Investing in PHC services on-reserve has been suggested as a method to improve health outcomes and the HTP has provided a step in the right direction for this objective. Local community control over health services is also an area that is showing a lot of promise in making health services relevant and effective. Underfunding acts as a large barrier to both primary health care and education on reserves, this is particularly important because the two are intimately entwined with health human resources. This study will add to the literature by bringing in the opinions and views of First Nations people from Manitoba. 24 Chapter Three: Data and Methods The current literature illustrates the poor health outcomes experienced by First Nations people in Manitoba as well as identifies ways (i.e. investing in PHC services and increasing control of health services at the local level) to improve health outcomes. However, there is very little information within the literature indicating what First Nations perceive to be wrong with the system. This work begins to answer that question by assessing the barriers to health care services that are reported by First Nations people in Manitoba. The study was conducted using a retrospective, cross sectional design, utilizing previously gathered data collected at a single point in time. This method has multiple advantages in that several variables can be analyzed at the same time, a large sample size can be used and perhaps most important the pre collected data can be used saving the researcher time and money. Low costs and minimal time spent in data collection make the retrospective cross-sectional approach particularly useful given the time and resource constraints inherent in a Master's thesis. A cross sectional design has one major drawback in that due to exclusion of the time dimension, it is impossible to establish causality. However, even in the face of this limitation, the data presented here are useful in that they can provide evidence that either supports or detracts from theories developed with previous research, as well as provide direction for future research. Study Population The data that was analyzed in this study came from the First Nations Regional Longitudinal Health Survey (RHS). More specifically, the data analyzed was the 2008-2010 Manitoba First Nations (MFN) subset of this nation-wide survey. The MFN RHS targeted 4527 25 people from 33 First Nations in Manitoba (there are 63 First Nations in Manitoba) and had a response rate of over 70%. The MFN RHS employed four different survey tools: adult, youth and child questionnaires, and a community survey. This study made use of the responses to the adult questionnaire and the community survey. The use of these two surveys gave a sample size of 1739 individuals. The individuals making up this sample represent the on-reserve population not only because they are drawn from over half of the First Nation in Manitoba, but also because they come from a diversity of both language backgrounds (e.g. Cree, Ojibway, Dene) and geography with First Nations from all throughout Manitoba. The First Nations that participated also represented diversity in their levels of remoteness and population size. There were 17 non­ isolated First Nations that participated and 16 isolated (made up of remote-isolated, semiisolated, and isolated) communities. Isolated and non-isolated First Nations had similar average populations of 1649 and 1694 respectively (INAC, 2011). Isolated communities showed a greater diversity in community size ranging from 113 to 5379 compared to non-isolated which ranged from 319 to 3681 (INAC, 2011). Table 2 Community types and populations Isolation Number of First Nations Average Population Minimum Population Maximum Population Isolated (includes Remote- and SemiIsolated) 16 1649 113 5379 Non-Isolated 17 1693 319 3681 Source: Indian and Northern Affairs Canada (2011). 26 Data Collection Data used in this study was collected between July 2008 and March 2010 by local interviewers from each community. The interviewers were identified by First Nations Health Directors and the MFN RHS team and then trained to conduct survey interviews using computerassisted-personal-interviewing technology. Following collection the data was "cleaned" by validating against the consent database to ensure that participants had given informed consent. The data was weighted by RHS technicians to ensure that age group, sex, and community were in fact representative of the larger population. This was achieved by first, weighting individuals to the known population for their age/gender group within their community. Then community weighting was done with via a ratio of the number of communities within their stratum to the number of communities sampled in that stratum. Finally, they were weighted by the ratio of the stratum population for their age gender group to the stratum population estimated by applying the first two weights. The final overall weight was a product of those three stages. Variables of Interest The variables that this study analyzed were drawn from the Manitoba portion of the RHS and look specifically at the RHS Community Survey 08/10 and the RHS Adult Questionnaire 08/10. Self-rated health was analyzed to determine its relationship with multiple barriers to access, community health care agreement type, and access to health care providers. Specific details of the variables are provided below. 27 Adult Questionnaire responses to the following questions were assessed: In general would you say that your health is: Excellent Very Good Good Fair Poor During the past 12 months, have you experienced any of the following barriers to receiving health care? Doctor or nurse not available in my area Health facility not available in my area (e.g. nursing station or hospital) Waiting list is too long Unable to arrange transportation Difficulty in getting traditional care (e.g. healer, medicine person, or elder) Not covered by Non-Insured Health Benefits (NIHB) Prior approval of Non-Insured Health Benefits was denied Could not afford direct cost of care/services Could not afford transportation costs Could not afford childcare costs Felt health care provided was inadequate Felt service was not culturally appropriate Chose not to see heath care professional Service was not available in my area Responses to the following questions from the Community Survey were also assessed in relation to self-rated health: Which of the following agreements is in place for healthcare in your First Nation? Self-Government agreement Single-community Transfer Other Multi-community Transfer Integrated Agreement 28 Does the First Nation have any of the following? Physicians stationed in the community every day Nurses stationed in the community every day Traditional healers stationed in the community Physicians visiting the community at least weekly Nurses visiting the community at least weekly Traditional healers visiting the community at least twice/year Dietician/nutritionist services Ethics Approval Due to the geographical location of this study ethics approval was sought from four separate ethics bodies. The University of Northern British Columbia Research Ethics Board, the Health Information Research Governance Committee (Assembly of Manitoba Chiefs), the Health Information Privacy Committee (Province of Manitoba) and the University of Manitoba Bannatyne Campus Research Ethics Board all gave ethical approval for this study. Objectives and Hypotheses The following questions about self-rated health were analyzed to inform the objectives of this study. 1. Is there a difference in self-reported health between those who do and those who do not report transportation costs as a barrier to accessing health care? HI: People who reported transportation costs as a barrier to accessing health care will report lower health status. 2. Is there a difference in self-reported health between those who do and those who do not report being denied prior approval for non-insured health benefits? 29 H2: People who were denied prior approval will report lower health status. 3. Is there a difference in self-reported health between those who do and those who do not report that there is not a doctor or nurse available in their area? H3: People who report that a doctor or nurse is not available in their area will report lower health status. 4. Is there a difference in self-reported health between those who do and those who do not report that a health facility is not available in their area? H4: People who report that a health facility is not available in their area will report lower health status. 5. Is there a difference in self-reported health between those who do and those who do not report that waiting lists are too long? H5: People who report that waiting lists are too long will report lower health status. 6. Is there a difference in self-reported health between those who do and those who do not report that they were unable to arrange transportation? H6: People who report that they were unable to arrange transportation (for health care) will report lower health status. 7. Is there a difference in self-reported health between those who do and those who do not report difficulty getting traditional care (e.g. healer, medicine person, or elder)? H7: People who report difficulty getting traditional care will report lower health status. 8. Is there a difference in self-reported health between those who do and those who do not report that service/supplies are not covered by Non-insured Health Benefits (NIHB)? H8: People who report that service/supplies are not covered by NIHB will report lower health status. 30 9. People who report that they could not afford direct cost of services will report lower health status. H9: People who report that they could not afford direct cost of services will report lower health status. 10. Is there a difference in self-reported health between those who do and those who do not report that they could not afford child care costs? H10: People who report that they could not afford child care costs will report lower health status. 11. Is there a difference in self-reported health between those who do and those who do not report that they felt the health care provided was inadequate? HI 1: People who report that they felt the health care provided was inadequate will report lower health status. 12. Is there a difference in self-reported health between those who do and those who do not report that they felt the service was not culturally appropriate? HI2: People who report that they felt the service was not culturally appropriate will report lower health status. 13. Is there a difference in self-reported health between those who do and those who do not report that they chose not to see a health professional? HI 3: People who report that they chose not to see a health professional will report lower health status. 14. Is there a difference in self-reported health between those who do and those who do not report that the service they needed was not available in their area? 31 HI4: People who report that the service they needed was not available in their area will report lower health status. 15. Does the availability of a doctor, nurse, dietician or traditional healer relate to self-rated health? HI 5: People with in communities with physicians stationed there every day are more likely to report a better health status. HI 6: People with in communities with physicians stationed there at least weekly are more likely to report a better health status. HI 7: People with access to nurses every day are more likely to report a better health status. HI8: People with access to a nurse at least weekly are more likely to report a better health status. HI9: People with access to traditional healers at least twice a year are more likely to report a better health status. H20: People from communities in which a traditional healer is stationed are more likely to report a better health status. H21: People from communities in which dietician/nutritionist services are available are more likely to report a better health status. 16. Are there associations between different levels of local control over health care and selfrated health? H22: Different types of health care control will result in different levels of reportedhealth. 32 Statistical Analysis The hypotheses were tested using contingency tables which compared each of the barriers and community characteristics with self-rated health. A chi-square test of independence was conducted to determine which variables had a statistically significant (alpha of < .05) relationship (Table 3, Table 4). The chi-square analysis was chosen because it is a useful tool for analysing cross-tab data. Chi-square output was also desirable because further analysis indicates strength of association and which variables are contributing the most to statistical significance. The chi-square test has two important assumptions: first that the sample is randomly selected and second that all expected cell sizes are greater than 5. Potential participants for the survey were chosen based whether their name was on the membership list of the selected First Nation. Once potential participants were identified they were randomly selected from within their age/gender groups. The second assumption was met when SPSS outputs indicated that there were no expected cell sizes less than 5. Cramer's V, a measure of association ranging from 0 (no association) to 1 (perfect association), was also calculated to determine the level of association between variables. A Cramer's V score of less than 0.100 is considered weak, and further analysis of association was conducted only for variables with a Cramer's V of 0.100 or higher. Since Chi-square tests are omnibus, that is they test for a number of null hypotheses at the same time, further analysis of individual cells is needed to determine which variables are responsible for the association. Simply put Chi-square analysis does not tell us how much of a role different categories of self-rated health play in the association with the barrier. In order to better understand the relationship between the barriers and the specific ratings of health, adjusted residuals were calculated for each set of variables that had a Cramer's V greater than 0.100 33 (Appendix 1). Debate exists in the literature as to whether a Bonferroni correction needs to be applied when determining the significance of the adjusted residuals. MacDonald and Gardner (2000) make the case for utilizing a Bonferroni correction as a means of reducing the potential for a Type 1 error4 and acknowledge that this leads to a conservative experimentwise error rate. Opposed to this view is Davis (2001) who points out that as the Bonferroni correction reduces the probability of a Type I error, it also increases the probability of a Type II error. If the probability of a Type II error5 becomes too high real relationships will be missed. This increase in probability of Type II errors is just as unacceptable and has led some researchers to consider the use of a Bonferroni correction useless for hypothesis testing (Perneger, 1998). In light of this, a Bonferroni correction was not used in the calculation of adjusted residuals; rather, as per convention, any adjusted residual with an absolute value greater than 2 will be considered to indicate a significant relationship with the acknowledgement that there is the potential for a Type I error. Summary Data drawn from the 2008-2010 Manitoba RHS was used to answer questions around the relationship of self-rated health to reported barriers to access and select community characteristics, such as human resources and local control of health care services. These 4 A Type I error occurs when the null hypothesis is rejected even though it was true (false rejection of null). 5 A Type II error occurs when the null hypothesis is not rejected when it should be rejected (false acceptance of null). 34 relationships were assessed via Chi-square, Cramer's V, and adjusted residuals. Chi-square indicated whether there was a statistically significant relationship, Cramer's V informed the strength of the relationship, and the adjusted residuals indicated the specific variables that were contributing the most to the relationship. 35 Chapter Four: Results The relationship between self-rated health and barriers to access to care (HI-HI4) were assessed to determine whether there was any statistically significant relationship. The statistically significant relationships were then assessed for strength and specific relationships that were contributing the most of that strength. Following this, the relationship between self-rated health and community characteristics (H15-H22) were analyzed in the same manner. Findings As reported in Table 3, with the exception of choosing not to see a health professional X2(4,) = 3.883, p = .422, and affordability of child care costs x2(4,) = 1.415, p = .842, each of the barriers showed a statistically significant relationship with self-rated health. Feeling that care provided was inadequate (Cramer's V = 0.098, p = .005), being unable to arrange transportation for care (Cramer's V = 0.081, p = .029) and feeling that the wait list was too long (Cramer's V = 0.094, p = .006) were all statistically significant in their relationship to self-reported health but had Cramer's V's of less than 0.100. This means that the level of association was very weak and as a result, the outcomes were not pursued for further analysis. 36 Table 3 Statistical relationship between self-rated health and barriers to access to care Hypothesis Barrier Chi-square value pvalue Cramer's V HI transportation costs 30.580 .000 .139 H2 denied prior approval for non-insured health benefits 23.167 .000 .124 H3 doctor or nurse not available in their area 23.519 .000 .120 H4 health facility is not available in their area 16.293 .003 .100 H5 waiting lists are too long 14.365 .006 .094 H6 unable to arrange transportation 10.753 .029 .081 H7 difficulty getting traditional care 21.442 .000 .118 H8 service/supplies are not covered by non-insured health benefits 24.449 .000 .127 H9 could not afford direct cost of care/services 39.818 .000 .160 H10 could not afford child care costs 1.415 .842 .030 HI 1 felt the health care provided was inadequate 15.096 .005 .098 HI2 felt the service was not culturally appropriate 36.201 .000 .156 HI3 chose not to see a health professional 3.883 .422 .051 H14 service they needed was not available in their area 15.552 .004 .102 37 The remaining barriers: transportation costs (Cramer's V = 0.139, p < .000), being denied prior approval for non-insured health benefits (Cramer's V = 0.124, p < .000), not having a doctor or nurse available (Cramer's V = 0.120, p < .000), not having a health facility available in their area (Cramer's V = 0.100, p = .003), difficulty getting traditional care (Cramer's V = 0.118, p < .000), services not being covered by non-insured health benefits (Cramer's V = 0.127, p < .000), not being able to afford direct costs of care (Cramer's V = 0.160, p < .000), feeling that services were not culturally appropriate (Cramer's V = 0.156, p < .000), and not having a needed service available in their area (Cramer's V = 0.102, p = .004) each had a small association with self-reported health (Table 3). In light of the low levels of associations found (Cramer's V ranging from 0.100-0.160) adjusted residuals (AR) were calculated to compare the observed and expected frequencies in each cell, in order to determine which specific cells were responsible for the association (Appendix 1). In the analysis of transportation costs association with self-reported health we found that the biggest contributor to the Chi-square value was those with poor health and very good health. People reporting transportation costs as a barrier were underrepresented in the very good health category (AR = -3.207)6. The negative value indicates that there are fewer people than expected in this cell, while the absolute value being greater than 2 indicates that this is significantly 6 Please note that all AR values can be found in Appendices 1 and 2. 38 different than expected. People who reported not being able to afford transportation were found to be overrepresented in the poor health category (AR = 4.525) (Appendix 1). Conversely, those who did not report transportation costs as a barrier were underrepresented in the poor health category (AR = -4.524) and overrepresented in the very good health category (AR = 3.207) (Appendix 1). Overall, people who reported that they could not afford transportation to access health care were significantly overrepresented in poor health and significantly underrepresented in very good health. The significant difference between those reporting that prior approval for NIHB services was a barrier to accessing health care was in the numbers reporting fair and poor health. Those who reported the barrier were overrepresented in the fair (AR = 2.891) and poor health categories (AR = 3.362) and those who did not report the barrier had fewer cases of fair (AR = 2.892) and poor health (AR = -3.368). Those reporting that a doctor or nurse was not available in their area were significantly less than expected in excellent (AR=-2.797) and good health (AR = -2.480) and higher than expected in fair (AR = 2.027) and poor health (AR = 2.990). Conversely, those who did not report the barrier reported significantly higher than expected excellent (AR = 2.795) and good health (AR = 2.480) and significantly lower than expected fair (AR= -2.067) and poor health (AR = -2.990) (Appendix 1). People from communities where health facilities were not available were significantly higher than expected in poor health (AR = 3.017), while those who did not report the barrier were significantly fewer than expected in poor health (AR = -3.020) (Appendix 1). 39 The group that reported difficulty accessing traditional care (e.g. healer, medicine person, elder) was significantly higher than expected in fair (AR = 2.673) and poor health (AR =3.319) and significantly lower than expected in good health (AR = - 2.449) (Appendix 1). Those who reported not being covered by NIHB showed a similar trend, being significantly higher than expected in fair (AR = 2.563) and poor health (AR = 3.795) and significantly lower in very good health (AR = -2.026) (Appendix 1). Those who could not afford direct costs of care or services had significantly higher than expected poor health (AR = 5.674) and significantly lower than expected very good health (AR = -2.389), while those who did not report the barrier were significantly higher than expected in very good health (AR = 2.389) and significantly lower than expected in poor health (AR = 5.665) (Appendix 1). Those who felt that service that was not culturally appropriate was a barrier to accessing health care were also significantly higher than expected in fair (AR = 2.889) and poor health (AR = 4.736) and significantly fewer than expected in excellent (AR = -2.106) and good health (AR = -2.538) (Appendix 1). Those who did not report this barrier were significantly higher than expected in good health (AR = 2.538) and significantly lower in fair (AR = -2.887) and poor health (AR = -4.727) (Appendix 1). People who reported that service was not available in their area were significantly higher than expected in poor health (AR = 3.667) while those who did not face this barrier were significantly less than expected in poor health (AR = -3.666) (Appendix 1). Community characteristics were also analyzed in conjunction with self-reported health to determine whether any significant associations were present (Table 4). There was no significant 40 relationship between self-rated health and any of the following community characteristics: physicians being stationed in the community every day (p=.369), nurses visiting the community at least weekly (p=.594), and having traditional healers stationed in a community every day (p=.8195) (Table 4). However, physicians visiting the community at least weekly (p=.002), nurses stationed in the community every day (p<.001), traditional healers visiting the community at least twice yearly (p=.005), having dietician/nutritionist services (p=.041), and the type of health agreement in place (p< 001), were all significantly related to self-rated health (Table 4). 41 Table 4 Statistical relationship between self-rated health and community characteristics Hypothesis Community Characteristic Chi-square value p-value Cramer's V H15 Physicians are stationed in community every day 4.28 .369 .0519 H16 Physicians visiting the community at least weekly 16.613 .002 .098 H17 Nurses stationed in the community every day 29.635 .000 .131 H18 Nurses visiting the community at least weekly 2.79 .594 .0419 H19 Traditional healers visiting the community at least twice/year 14.88 .005 .1026 H20 Traditional healers stationed in the community 1.54 .8195 .0314 H21 Dietician/nutritionist services 9.958 .041 .077 H22 Type of agreement in place for First Nation (all types) 30.71 .000 .107 Again taking into account Cramer's V even though the above variables were significantly related to self-rated health, the following had Cramer's V values less than 0.100: physicians visiting the community at least weekly (Cramer's V = 0.098, p = .002), traditional healers stationed in the community (Cramer's V = 0.031, p = .005) and dietician/nutritionist services (Cramer's V = 0.077, p = .041) (Table 4). These community characteristics were not considered 42 for further analysis. The remaining community characteristics, nurses stationed in the community every day (Cramer's V = 0.131, p < .000), traditional healers visiting the community at least twice yearly (Cramer's V = 0.103, p = .005) and the type of heath agreement (Cramer's V = 0.107, p < .000) each had a Cramer's V greater than 0.100 and adjusted residuals were calculated to determine which specific variables were contributing to the significant difference (Table 4). Analysis of the adjusted residuals shows that there were significantly higher than expected numbers of people reporting fair (AR = 3.852) and poor health (AR = 2.764) in communities that did not have a nurse stationed there every day. Furthermore people from these communities reported significantly lower than expected levels of excellent (AR = -2.874) and very good health (AR = -2.011) (Appendix 2). In communities that had nurses stationed there every day the opposite was true with significantly higher than expected numbers of people reporting very good (AR = 2.012) and excellent health (AR = 2.873) and lower than expected numbers of people reporting fair (AR = -3.850) and poor health (AR = -2.766) (Appendix 2). In communities with traditional healers visiting at least twice annually, significantly more than expected people rated their health as poor (AR = 3.118) and communities that did not have this characteristic reported significantly fewer than expected cases of poor health (Appendix 2). The final relationship to be analyzed was between level of community control and selfrated health. People from communities that had a transfer agreement in place reported excellent health at significantly higher than expected numbers (AR = 2.966) (Appendix 2). Communities with an integrated agreement had significantly higher than expected levels of fair (AR = 2.042) and poor health (AR = 2.152) and significantly lower than expected levels of excellent health (AR = -2.712) (Appendix 2). People residing in communities that had other agreements reported 43 significantly higher than expected good health (AR = 3.240) and significantly lower than expected poor health (AR = -3.052) (Appendix 2). Summary With the exceptions of child care costs and choosing not to see a health care professional, all of the barriers to accessing health care were significantly related to self-rated health. Transportation costs, being denied prior approval for NIHB, not having a doctor or nurse available in their area, not having a health facility available in their area, difficulty getting traditional care, service/supplies are not being covered by NIHB, being unable to afford direct cost of care/services, feeling health care provided was inadequate, and not having the service needed in their area were all found to have a strong enough relationship to self-rated health to warrant further analysis. Physicians visiting the community at least weekly, having nurses stationed in the community every day, having traditional healers visiting the community at least twice/year, availability of dietician/nutritionist services, and the type of health agreement in place for the First Nation were all significantly related to self-rated health. Of these relationships nurses stationed in the community every day, traditional healers visiting the community at least twice yearly, and the type of heath agreement each had a strong enough relationship to warrant analysis of the specific factors contributing to the relationship. 44 Chapter Five: Discussion of the Findings The main purpose of this study was to identify the current state of PHC in First Nations communities in Manitoba as well as determine what the next steps forward should be in trying to gain better health outcomes. The literature review provided suggestions for more investment in PHC services on-reserve as well as shifting more control to the local community level. The results of this study can be organized into four main topics: funding, resource availability, culturally appropriate care, and community control. Due to the low levels of association found in the analysis, the relationships are compared to the wider literature to determine whether they are consistent with previous findings. Funding Transportation costs, direct costs of care/services, being denied for prior approval for NIHB, and "not covered by NIHB," were each significantly related to higher than expected levels of fair and/or poor health. All except prior approval were also significantly related to lower than expected levels of very good health. Taken together these barriers indicate that funding health care is a very real challenge for many First Nations. Underfunding is not something novel to First Nations health care in Manitoba. In fact many areas of social services ranging from education, to housing, to health are chronically underfunded. Lavoie, Forget and O'Neil (2007) reported that the earlier a First Nation entered into a transfer agreement, the less funding they had in relation to their responsibilities. This is attributed to the levels of funding being based on historical expenditures that became entrenched once an agreement is signed. The funding levels also do not take into account the rapidly 45 expanding populations in First Nations communities or the ever-changing landscape of health care needs. As we will see below, the Assembly of Manitoba Chiefs also has reported that education on reserves is severely underfunded relative to provincial counterparts. The implications of this underfunding are discussed below in relation to the development of human capital and a workforce that can meet the health care needs of First Nations. Human Resource and Facility Availability The results of this study pertaining to resources can be broken down into two themes, staffing and facilities. People who reported that there was no health care facility available in their area also reported higher than expected levels of poor health. Similarly, people who reported that a health care service (non-specific) was not available in their area also reported higher than expected level of poor health, when compared with people who did not cite this particular barrier. The fact that people were reporting that a health facility was not available in their area may be a bit misleading, since each community that participated in the Manitoba portion of the survey has at least some sort of facility. What this question may have been capturing is the fact that these facilities were not sufficient to provide the required level of care. With regards to staffing, those who reported that a nurse or doctor was not available also reported higher than expected levels of fair and poor health in conjunction with lower than expected excellent and good health. Higher than expected levels of poor health were also found when people reported that they had difficulty accessing traditional care as well as when they reported low visitation rates of traditional healers to the community. Not surprisingly, when nurses are stationed in the community more people are reporting excellent and very good health with the corresponding low reporting of fair and poor health. 46 The difficulties associated with recruitment and retention of health care professionals to work in rural and remote locations has received much attention in Canada (Aylward, Gaudine, Bennett, 2011; Pitbaldo & Pong, 1999). Commonly cited reasons for not wanting to work in rural areas are: lack of family ties in the area, limited employment opportunities for spouses, lack of services, diminishing populations, decreased opportunity for professional development, attitudes towards rurality, and excessive workloads that often accompany the shortage of health care professionals. Many First Nations in Manitoba are located in what can be considered to be rural and remote locations and the issue of recruitment and retention is also a barrier to accessing health care in First Nations communities. The solution then is to determine what factors contribute to health care providers wanting to live and work in these communities. Another issue that faces many smaller communities is economies of scale. Within the context of providing care for a small community (e.g. population = 131) it becomes impossible to employ a health care professional full time. There is no easy solution to this; however, one approach being taken is the formation of partnerships between First Nations that are located near to each other. Such a partnership can allow for full-time positions in which the health care worker serves multiple communities and travels between them. Unfortunately, this solution does not work for all communities as some are so remote and small that it becomes impossible to have full time health care provided. Manahan and Lavoie (2008) report that job satisfaction, opportunities for children and spouses to access education, employment, and recreational activities all play major roles in whether a nurse will stay in a rural practice. The authors recommend that to increase retention of nurses the work environment should include opportunities for autonomy, task variety, peer- 47 support and stress management mechanisms. Cameron, Este and Worthington (2010) reported that there were a number of factors that made physicians want to practice in rural Alberta, among them were connection and active support. Connection was defined as having a sense of belonging and integration into the community and active support was referring to community mobilization to assist the medical community via fundraising, volunteering, political advocacy, and befriending physicians and their families. Miedema (2009) also reported that in New Brunswick one of the benefits cited by rural physicians was getting to know the patients and their families well. It was also reported that physicians who grew up in rural areas were more likely to report that the benefits of practicing in a rural community offset the disadvantages. For Manitoba First Nations this means that the expansion of the First Nations health services workforce could be particularly useful in addressing recruitment and retention issues. The development of this workforce will also help in dealing with the next issue of culture. Culturally Appropriate Care Having a traditional healer in the community every day was not significantly related to self-reported health. Having a traditional healer visit the community at least twice per year was significantly associated with elevated rates of poor health. At first one might assume that accessing care through a traditional healer may be detrimental to one's health. However, people who reported difficulty accessing traditional care also reported significantly higher than expected levels of fair and poor health; in addition, they reported significantly lower than expected levels of good health. Taken together these findings give a picture of what might be happening. If a community has access to a traditional healer as little as two times per year this could easily fit the description of "difficulty accessing traditional care." Thus it seems that both questions are asking 48 a very similar question and arriving at essentially the same answer. Not having access to traditional care was associated with poor health so one would expect to find an association between regular access to traditional healers and good health. In fact no such association was found. One reason for not finding this link might be due to the way the questions were presented within the survey itself. In the RHS Community Survey a traditional healer was not defined and there may be some confusion as to what constitutes a traditional healer. The Adult Questionnaire addressed traditional care on an individual basis by asking whether people had experienced difficulty getting access to traditional care in the last 12 months. The AQ gives the examples of healer, medicine person, or elder as routes to accessing traditional care. Presumably people answering this question had to decide for themselves what constitutes traditional care and a wide variety of potential answers may have come to mind (Waldrum, Herring, & Young, 2007). For example, two people from the same community may have different interpretations of what constitutes traditional care and therefore, answered the same question differently. Essentially, this negates the ability of the question to act as a descriptor of a community characteristic. Opposed to this, when respondents are asked about nurses being stationed in a community there is no ambiguity about what a nurse is and whether or not they are stationed within the community. The lack of a significant relationship between traditional healers stationed in the community and self-rated health may then be due to the use of multiple poorly-defined questions all targeting the same variable. The discussion of what constitutes traditional care and a traditional healer are well beyond the scope of this study and no attempt will be made to define either. 49 The variable that had one of the strongest relationships to self-rated health was whether or not the person felt their health care service was culturally appropriate. Those who reported that their care was not culturally appropriate reported significantly higher than expected levels of fair and poor health and significantly lower than expected levels of excellent and good health. Conversely, those who did not report this barrier reported significantly higher than expected levels of good health and significantly lower than expected levels of fair and poor health. Culturally appropriate care is important if the medical community is going to gain the trust of First Nations people. One study found that First Nations women's encounters with mainstream health care services could be classified as either invalidating or affirming encounters (Browne & Fiske, 2001). The main themes within invalidating encounters were: being dismissed by health care providers, having to transform one's self to gain credibility, facing negative stereotypes, being marginalized by the mainstream (e.g. sense of being outside of the health care system), having sense of vulnerability, and facing a lack of understanding about personal circumstances (Browne & Fiske, 2001). Alternatively, First Nations women also had some affirming encounters such as: being an active participant in their health care decisions, encountering health care providers who gave exceptional care, having their personal and cultural identity affirmed, and being able to form positive, lasting relationships with health care providers (Browne & Fiske, 2001). Browne and Fiske point out that these positive encounters are what most people in mainstream society would expect to encounter, but that in the context of a First Nations woman, these encounters mean much more: "they represent unexpected exceptions to the ubiquitous form of racism and discrimination that shape women's everyday social experiences" (pi43). 50 Unfortunately, as indicated by the findings of the current study, care is not often being provided in a way that is culturally appropriate. There are however, ways in which the issue can be addressed. One way that is recommended by Browne and Fiske (2001) is using cultural safety as an analytical lens through which health care providers can question their assumptions and better place First Nations peoples' health concerns within a proper context that takes into account the various factors that make the person who they are. Another way in which culturally appropriate care can be approached is through the development of the First Nations health care workforce. Due to the low rates of high school graduation this will not be an easy task, but the effects beyond the immediate benefits of cultural safety provide all the more reason to invest in education. Recruitment and retention of qualified health professionals is difficult for many First Nations and some of the factors (e.g. lack of family ties) that lead health professionals to not want to practice in this setting do not have the same effect for people who have grown up in and are a part of the community. Additionally, having an expanded First Nations workforce has positive implications for the local economy. Community Control The analysis of community control over health care services and delivery showed some intuitive results. The RHS survey included self-government as one of the health agreement options and many people selected it as being representative of the agreement in their community. This group of people was included with those indicating a transfer agreement for reasons already discussed, and it was found that First Nations people coming from a community with a transfer/self-government agreement had higher than expected levels of excellent health. An opposite pattern of health is seen in First Nations that have an integrated agreement, where lower 51 than expected levels of excellent health and higher than expected levels of fair and poor health were reported. This local control allows for better long term planning and increased ability for timely responsiveness to changing health priorities (Lavoie et al., 2005). People accessing services in an integrated agreement community showed significantly lower than expected levels of excellent health and significantly higher than expected levels of fair and poor health. One of the main reasons for this may be the limited flexibility to respond to changing health care priorities. Furthermore, the inability to roll surplus funds over into the next year for locally identified priority health services decreases the ability of the community to respond adequately to health care needs. Another potential contributor to the differences in health between communities with more and less control is the added employment opportunities that come with having a greater amount of health care delivery to administer. Communities with greater local control have been found to be more likely to indicate that signing a transfer agreement led to an increase in employment opportunities and in workforce stability. This is not to say that integrated communities did not experience this as well, they did, just not to the same extent (Lavoie et al. 2005). As a whole, the findings on the relationship point towards better than expected health when communities have greater levels of control. Of course it cannot be claimed from this study that more control causes better health, given that this data is cross-sectional. The opposite could also be argued, that better health contributes to the ability for a community to gain more control over their health care resources. Each of these possibilities is plausible; however, at present there is not enough evidence to convincingly support either theory. 52 In their 2003 Working Paper Series, Cornell and Kalt wrote about the link between selfgovernment and service delivery. The research was not health specific, but rather looked at economic development, self-government and service delivery together. Despite not being a health-centric analysis, the findings are still relevant to the current discussion. The conclusion reached was that there are three different factors that were linked to development success: practical self-rule, capable governing institutions, and cultural match. Practical self-rule (First Nations control over First Nations affairs) was concluded to be necessary, yet not sufficient for sustained economic development. The authors state that they have yet to find a case of a Canadian First Nation or an American Indian nation that demonstrated "sustained, positive economic performance in which somebody other than the Indian nation itself is making the major decisions about governing institutions, governmental policy, development strategy, resource allocation and use, internal affairs and related matters" (Cornell & Kalt, 2003, p.13). The authors explain that when First Nations are able to move outsiders from the role of decision makers to the role of resource providers, and then take over the role of decision makers, several things happen that are conducive to successful development. In particular, self-rule promotes citizen engagement, locates the development agenda in local hands, and finally creates a link between decisions and the resultant consequences. Capable governing institutions were cited as a second key to development success. In order to have a capable governing institution there needs to be dispute-resolution mechanisms that are non-politicized, the eradication of corrupt practices, opposition to opportunistic behavior on the part of politicians, development of efficient decision-making mechanisms, and the adoption of policies that are developed by the First Nation (Cornell & Kalt, 2003). The 53 competence of these governing institutions is important mainly for the people served by them, but is also important for the development of nation to nation relationships as well as with potential investors. Within the wider Canadian population support for First Nations selfgovernment was found to be lowest in the Alberta, Saskatchewan and Manitoba (Martin & Adams, 2000). Many capable governing institutions already exist and the development of more, in addition to the maintenance of those currently in place, will serve to change public opinion for the better and put First Nations in a place where they can better negotiate with other nations and attract investors to develop sustainable economies. Cultural match is defined by Cornell and Kalt (2003) as the fit between the governance institution and the local conception of how authority should be organized. This means that the governance institution needs to be in synch with the First Nations historical and contemporary conception of governance. In recent history and still today Canadian government has designed and imposed euro-centric policies that are not in line with the culture of many First Nations. Most of these policies were not developed with the concept of a nation to nation relationship and as such are often doomed to fail (Cornell & Kalt, 2003). Other studies have looked at the relationship between self-government and health. Specifically, suicide, a major contributor to low life-expectancy, and ambulatory care sensitive conditions, indicators of primary health care, were analyzed in their relationship to selfgovernment and local control. In 1998, Chandler and Lalonde analyzed the relationship between a variety of factors meant to approximate cultural continuity and the rates of youth suicide within communities that had and did not have these indicators. The indicators included: land claims, defined as the 54 community having taken steps towards securing traditional territory; self-government, defined as having been successful in gaining economic and political independence within their territory; education services, dependent on whether a majority of students attended a band school; police and fire services, defined by whether or not the band had substantial control over the services; health services, defined broadly by either having some measure of local control or not; and cultural facilities, which was broadly defined as a facility designated for cultural use. Of these indicators self-government was the best predictor of low suicide rates within the community. Another more recent study that provides even greater support for the link between selfgovernment and good health, found that the longer a First Nation had control (at the transfer level) over their health services, the lower the rate of hospitalizations for ambulatory care sensitive conditions (Lavoie, 2010). The study assessed what happened in the years following the signing of an agreement (transfer and integrated as a single cluster) and found that the rate of hospitalization for ACSCs decreased each year. The researchers then took the study one step further and compared transferred and integrated communities; they found that the annual decrease in ACSC hospitalizations was attributable to the transferred communities and that integrated communities did not show a significant decrease in ACSC hospitalization rates. Additionally, this study found that having increased local access to primary health care was also associated with lower rates of hospitalization for ACSCs. At the time of writing there are multiple First Nations in the process of negotiating selfgovernment agreements as well as some communities that have already entered into selfgovernment agreements. Using the ACSC method used by Lavoie and colleagues (2011) one could take a retrospective look at hospitalizations for ACSCs among people accessing their 55 primary health care in these communities. Alternatively, baselines could be measured in the communities that are currently in the negotiation phase and then a prospective analysis of reported health and reported barriers to accessing health care could be undertaken to determine whether any changes take place following the negotiation of a self-government agreement. In fact, this second analysis method is already underway via the RHS which has been monitoring a number of the First Nations throughout Canada that are in the process of negotiating selfgovernment agreements. Now would be the ideal time for identifying any specific questions that could be included in the RHS specific to these communities so that data collection can better serve the needs of First Nations representatives and policy makers. In terms of the literature, it would be useful to continue to closely monitor the relationship between self-rated health and community control of health services. This further investigation would be very useful in distinguishing whether healthy communities are simply better able to negotiate a selfgovernment agreement or whether having control over health services is in fact leading to improved health outcomes. Limitations Post-hoc analysis of the Chi-square tests of independence consisted of calculating the adjusted residuals and regarding any absolute value greater than 2 as significant. Caution should be applied when interpreting these numbers as no Bonferroni correction was used. The rationale for not using the Bonferroni correction was that the test becomes so conservative in its attempt to reduce Type I errors that it loses power and would increase the likelihood of committing a Type II error. This choice was made in light of: the literature, which provided arguments for and against the use of a Bonferroni correction and the severity of making a Type I error relative to 56 making a Type II error. The literature was unable to provide a convincing argument for the use of Bonferroni and several peer-reviewed journal articles did not show any indication of using a Bonferroni correction during residual analysis (Busca, Moras, Pena & Rodriguez-Jimenez, 2011; Fassoulaki, Paraskeva, Papilas & Karabinis, 2000). Making a Type I error during the residual analysis would result in identifying an association when it does not exist in reality, whereas making a Type II error would mean missing an association that truly does exist. Ultimately, the adjusted residuals are only being used to give a picture of directionality and overall, each of the barriers is pointing towards the same thing, a lower health status. Even if a few of these relationships are spurious it would not detract from the overall direction that the results are pointing. If the relationships in the individual cells are deemed important on their own, future studies of the relationship will be able to provide evidence that either supports or detracts from the results of this study. The implications of making at type II error means that variables that are in fact related to the health-status of First Nations people could be missed and as a result future studies will be more likely to overlook these variables. Making a type I error, however, does not have the same issue since replication of the study and further analysis into these variables will eventually root out any spurious relationships. For the purposes of this study, the individual relationships are not important, but rather the whole picture that they paint. Having taken the risk of making type II, errors the findings presented here should be considered in light of the possibility that some of the relationships may in fact be spurious. Given the high likelihood that at least some of these findings may be spurious, results should be considered in light of the wider literature and more credence given to studies that employed more robust statistical analysis. 57 Another methodological limitation, that has implications for the interpretation of these results, is the use Chi-square as the main form of analysis. Chi-square does not enable the user to control for outside factors that may be confounding the relationships. For example, the finding that not being able to afford transportation for medical care is related to elevated rates of poor health, may be confounded by the fact that the analysis did not control for socioeconomic status or degree of remoteness. This lack of control does not invalidate the results, just the level of specificity we can apply in our interpretation. That is we would be wrong to claim, based on this study alone, that transportation costs are cause for poor health; however, I think that it is safe to claim that financial barriers are playing a role in poor health outcomes. If future studies want to look at the extent to which specific factors are contributing to health status, the employment of a multiple logistic regression would be a useful method. Another of the limitations of using data that has been previously gathered and that is only available in a secure facility out of province is that the data could only be accessed once and for a limited period of time. Analytical methods had to be predetermined before the trip and modified once the data were actually available. Additional questions that arose out of the results which might have been further pursued were left unexplored. Limitations also arose due to some of the survey questions that had some level of ambiguity, making the interpretation of results difficult. The definitions of traditional care and traditional healers were unclear and broad. To improve in the next iteration of the survey it would be useful to have greater specificity as to what is meant by traditional care or traditional healer. The improved specificity would be useful to local health care administrators when they 58 are deciding which programs to implement. It is beyond the scope of this study to make any recommendations as to how to define or categorize traditional care or traditional healers. The finding on nurses being stationed in the community every day, although useful, does not provide a totally clear picture. The question fails to take into account the scope of practice of the nurse. For example, if community A has a nursing station and community B has a health office, both will have nurses available, but the nurse at the nursing station will have a greater scope of practice than the nurse at the health office. It is possible that the effects of the nurses with expanded scopes of practice are being diluted in the present method of collection. The use of the term self-government when asking what type of health agreement is very useful in the national survey; however, regionally there are no self-government agreements in place within Manitoba. Since there were some people who indicated that their community was under a self-government agreement, their responses were assessed in the same pool as transferred communities. There is currently one First Nation in Manitoba that is in self-government negotiations and it has a transfer agreement in place. It is not unreasonable to believe that many of the people who reported self-government were from this and other transferred communities. The data used in this study was cross sectional so no form of temporality can be claimed. Also, this study was based on data collected from the adult on-reserve population of First Nations in Manitoba. As such, the statistics may not have generalizability to other populations such as children and youth, or those living not living on reserves. 59 Summary The findings of this study in combination with the literature highlight several issues to consider during development of stronger primary health care systems in First Nations communities. There is underfunding of both primary health care and education, which is resulting in lower health and educational attainment. Improving the funding system so that it reflects need rather than historical expenditure, will help to meet the immediate health care needs of First Nations communities as well as aid in the development of a larger First Nations health care workforce. This First Nations workforce will be well equipped to provide care that is culturally appropriate and consistent and may help to decrease the need for the often inconsistent health care workforce that is not in tune with First Nations cultures and needs. These findings also highlight the need to continue to explore the role of local control over health care services. 60 Chapter Six: Conclusion The purpose of this study was to assess the current state of PHC on First Nations reserves in Manitoba and to identify some of the next steps towards improving health outcomes for people living on-reserve. The literature indicated that, in general, First Nations people have poorer access to primary health care than other Canadians. The findings indicate that this is a result of many things including underfunding, remoteness of some communities, difficulty with recruiting and retaining health care workers, and a lack of culturally appropriate care. This study analysed data from the RHS in order to bring the experiences and perspectives of First Nations people to the process of determining the best approach to improving health outcomes of First Nations people in Manitoba. The analysis led to the findings that there are multiple barriers to accessing primary health care services. These reported barriers were classified into four themes: funding, resource availability, culturally appropriate care, and community control. This study demonstrated that people with poorer self-rated health are experiencing financial barriers to accessing health services. Specifically, people with low selfrated health were more likely to also report being unable to afford transportation and direct costs associated with health care. This group of people also reported obtaining the approval process for NIHBs was a barrier to receiving care. Lack of human resources (i.e. nurses/physicians), insufficient health facilities, and lack of health service availability at a local level were also found to be significantly related to poorer health outcomes. Lack of access to culturally appropriate care, traditional healers, and traditional care were all significantly associated with poorer health outcomes. Finally, lower levels of local control over health service delivery was significantly related to poorer health outcomes. 61 Implications In the discussion recommendations from other studies state that in the short term these barriers can be addressed via the implementation of funding models that take into account population growth, recruitment and retention incentives, the changing health profiles of First Nations, and the need for expansion of services. In addition to this, an expansion of the scope of practice that nurses are allowed to provide in communities has potential to show immediate benefits. In conjunction with an expanded scope of practice, education of all health care workers regarding cultural safety and sensitivity will serve to reduce some of the reported barriers arising from cultural insensitivity. It would also be useful for better understanding the relationships between health care barriers and health status if RHS data could be linked with outside health data-bases. Specifically, an assessment of the relationship of ACSCs and access barriers, could be facilitated via the linkage of RHS Manitoba data with ACSC data collected by the Manitoba Center for Health Policy. This research would provide a clearer picture of which barriers are leading to poor health outcomes, and appropriate policy responses could be made. In the long term, improvement of on-reserve primary health care services is going to require a significant change from the status quo. The expansion of the First Nations health services workforce has the potential to address some of the recruitment and retention issues unique to First Nations communities, cultural sensitivity issues, and to some extent funding issues, via the improvement of on-reserve economies. Many university level support systems are in place for First Nations students to help with the transition to what is often a foreign world. There is however, one very large obstacle in the way of developing this workforce: only about half of First Nations youth are graduating from high school and thus the potential group of 62 university students is drastically reduced. Underfunding of on-reserve education needs to be addressed immediately so that students can be successful and have a positive educational experience. Improved access to education also has implications beyond health care. The development of sufficient capacity, needed to run and deliver many types of programs, will be greatly supported by an increase in the number of people attaining post-secondary education. As capacity increases delivery of services will improve in all sectors. Although this study was only able to find a modest association between local control and good health outcomes a potentially effective long-term strategy for improving primary health care on-reserve is increasing the level of local control over the delivery. Increasing the level of local control over local affairs and decreasing the level of control by outside entities (i.e. the transformation of the Federal government from decision maker into a resource role), will enable First Nations to develop and implement culturally relevant solutions to local priorities. These locally developed strategies and programs could work much more effectively than the current method of developing policies and programs from a top-down outside source. Additionally, a self-government approach to education will also allow for the development of strategies that address locally determined priorities and as a result improve the educational experience of First Nations students. Increased education and the resultant development of capacity will contribute to increasing the ability of a First Nation to self-govern and provide high quality relevant services. Ultimately, a greater level of control at the local level warrants further exploration as a potentially effective method for improving the ability of administrators to react to the dynamic ever-changing needs of the people accessing health care within their community. Over the long 63 term this approach has the potential to be both effective and efficient in that self-government may lead to a better health status of First Nations in Manitoba and this better health status can translate into a reduction of the need for expensive hospitalizations. The solution to better primary health care services will not come from a government office or the ivory tower of academia, but rather it will develop on the front lines of health care patients and practitioners. Our job then as researchers and informers of policy is to support a system that fosters the ability of First Nations to determine what is needed in their community and the manner in which they will respond. 64 References Anonson, J. ML, Desjarles, J., Nixon, J., Whiteman, L., Bird, A. (2008). Strategies to support recruitment and retention of First Nations youth in baccalaureate nursing programs in Saskatchewan, Canada. Journal of Transcultural Nursing, 19(3), 274-283. Aylward, M., Gaudine, A., & Bennet, L. (2011). Nurse recruitment and retention in rural Newfoundland and Labrador communities: the experiences of healthcare managers. Online Journal of Rural Nursing and Health Care, 11(1), 54-69. Billings, J., Zeitel, L., Lukomnik, J., Carey, T. S., Blank, A. E., & Newman, L. (1993). Impact of socioeconomic status on hospital use in New York City. Health Affairs, 72(1), 162-173. Brown, A. D., Goldcare, M. J., Hicks, N., Rourke, J. T., McMurtry, R. Y., Brown, J. D., & Anderson, G. M. (2001). Hospitalization for ambulatory care-sensitive conditions: a method for comparative access and quality studies using routinely collected statistics. Canadian Journal of Public Health, 92(2), 155-159. Browne, A. J., & Fiske, J. A. (2001). First Nations women's encounters with mainstream health care services. Western Journal of Nursing Research. 23(2), 126-147. Busca, B., Moras, G., Pena J., & Rodriguez-Jimenez, S. (2011). The influence of serve characteristics on performance in men's and women's high standard beach volleyball. Journal of Sports Sciences, 30(3), 269-276. Call to Action on Education, (n.d.) Retrieved from http://www.afn.ca/index.php/en/newsmedia/current-issues/call-to-action-on-education). Canada & Sioux Valley Dakota Nation. (2001). Comprehensive agreement-in-principle between the Sioux Valley Dakota Nation and Her Majesty in Right of Canada. Ottawa: Indian Affairs and Northern Development. Chandler, M.J. & Lalonde, C.E. (1998). Cultural Continuity as a hedge against suicide in Canada's First Nations. Transcultural Psychiatry, 35(2), 193-211. Cornell, S., & Kalt, J. (2003). Alaska native self-government and service delivery: What works?. Tucson, AZ: Udall Center for Studies in Public Policy, University of Arizona. Culyer, A.J. (2001). Equity - some theory and its policy implications. Journal of Medical Ethics, 27, 275-283. Dyck, R., Osgood, N., Lin, T. H., Gao, A., & Stang, M. R. (2010). Epidemiology of diabetes mellitus among First Nations and non-First Nations adults. Canadian Medical Association Journal, 182(3), 249-256. 65 Fassoulaki, A., Paraskeva, A., Papilas, K., & Karabinis, G. (2000). Self-citations in six anaesthesia journals and their significance in determining the impact factor. British Journal of Anaesthesia, 84(2), 266-269. FNIGC. (2011). RHSphase 2 (2008/10) preliminary results. Ottawa, ON: The First Nations Information Governance Centre. FNIGC. (n.d.). RHS 1997: The Pilot Survey. Retrieved from FNIGC website: http://www.rhsers.ca/node/38 Green, C., Blanchard, J. F., Young, T. K., Griffith, J. (2003). The epidemiology of diabetes in the Manitoba-registered First Nation population. Current patterns and comparative trends. Diabetes Care, 26(7), 1993-1998. Guillory, B.M., Willie, E., & Duran, E.F. (1988). Analysis of a community organizing case study: Alkali Lake. Journal of Rural Community Psychology, 9(1), 27-36. Harvard. (2006) Review of the First Nations Regional Longitudinal Health Survey (RHS) 2002/2003. Cambridge, MA: Harvard Project on American Indian Economic Development. Health Canada. (2011a). Your health benefits - a guide for First Nations to access non-insured health benefits. Retrieved from Health Canada website: http://www.hc-sc.gc.ca/fniahspnia/pubs/nihb-ssna/201Orpt/index-eng.php Health Canada. (201 lb). Non-Insured Health Benefits Program Annual Report 2009/2010. Retrieved from Health Canada website: http://www.hc-sc.gc.ca/fniahspnia/alt_formats/pdf/pubs/nihb-ssna/2010_rpt/2010_rpt-eng.pdf Indian and Northern Affairs Canada (2011). Registered Indian population by sex and residence 2010. Minister of Department of Indian Affairs and Northern Development, Ottawa. Jacklin, K. M., & Warry, W. (2004). The Indian health transfer policy in Canada. In A. Castro & M. Singer (Eds.), Unhealthy health policy (pp. 215-234). New York: AltaMira Press. Jiwa, A., Kelly, L., & St Pierre-Hansen, N. (2008). Healing the community to heal the individual. Literature review of aboriginal community-based alcohol and substance abuse programs. Canadian Family Physician, 54(1000.e), 1-7. Lavoie, J. G. (2004). Governed by contracts: the development of Indigenous primary health services in Canada, Australia and New Zealand. Journal of Aboriginal Health, 7(1), 624. Lavoie, J.G., 0"Neil, J., Sanderson, L., Elias, B., Mignone, J., Bartlett, J., Forget, E., Burton, R., Schmeichel, C., & McNeil, D. (2005). The Evaluation of the First Nations and Inuit 66 Health Transfer Policy Final Report: Volume 1, Executive Summary. Winnipeg: Manitoba First Nations Centre for Aboriginal Health Research. Lavoie, J.G., & Forget, E. (2006). A Financial Analysis of the Current and Prospective Health Care Expenditures for First Nations in Manitoba. Manitoba: Centre for Aboriginal Health Research. Lavoie, J. G., Forget, E., & O'Neil, J. D. (2007). Why equity in financing First Nations onreserve health services matters: findings from the 2005 national evaluation of the health transfer policy. Health Care Policy, 2 (4), 79-95. Lavoie, J. G. (2009). The values and challenges of separate services, First Nation in Canada. In J. Healy & M. McKee (Eds.), Accessing healthcare, responding to diversity, (pp. 325-349). Oxford Scholarship Online: September 2009. Lavoie, J., Forget, E., Prakash, T., Dahl, M., Martens, P., O'Neil, J. (2010). Have investments in on-reserve health services and initiatives promoting community control improved First Nations' health in Manitoba? Social Science and Medicine, 71, 717-724. doi: 10.1016/j.socscimed.2010.04.037 MacMillan, H. L., Walsh, C. A., Jamieson, E., Wong, M. Y., Faries, E. J., McCue, H., et al. (2003). The health of Ontario First Nations people: results from the Ontario First nations Regional Health Survey. Canadian Journal of Public Health, 94(3), 168-171. Manahan, C., & Lavoie, J. G. (2008). Who stays in rural practice?: An international review of the literature on factors influencing rural nurse retention. Online Journal of Rural Nursing and Health Care, 5(2), 42-53. Martens, P., Bond, L., Jebamani, L., Burchil, C., Roos, N., Derksen, S., et al. (2002). The health and health care use of registered First Nations people living in Manitoba: a populationbased study. Winnipeg: Manitoba Center for Health Policy, Department of Community Health Sciences, Faculty of Medicine, University of Manitoba. Martens, P. J., Sanderson, D., & Jebamani, L. S. (2005). Health services use of Manitoba First Nations people: is it related to underlying need? Canadian Journal of Public Health, 96, S39-S44. Martin, D., & Adams, C. (2000). Canadian public opinion regarding Aboriginal self-government: diverging viewpoints as found in national survey results. American Review of Canadian Studies, 50(1), 79-88. McDermott, R., & Segal, L. (2006) Cost Impact of Improved Primary Level Diabetes Care in Remote Australian Indigenous Communities. Australian Journal of Primary Health, 12, 124-130. 67 MacDonald, P. L., & Gardner, R. C. (2000). Type I error rate comparisons of post hoc procedures for Ij chi-square tables. Educational and Psychological Measurement, 60, 735-754. Miedema, B., Hamilton, R., Fortin, P., Easley, J., & Tatemichi, S. (2009). The challenges and rewards of rural family practice in New Brunswick, Canada: lessons for retention. Rural and Remote Heath, 9, 1141-1150. Minore, B., Jacklin, K., Boone, M., & Cromarty, H. (2009). Realistic Expectations: The changing role of paraprofessional health Workers in the First Nation communities in Canada. Education for Health, 22(2), 1-9. Retrieved June 16, 2010 from http://www.educationforhealth.net/publishedarticles/article_print_298.pdf Moore, M., Forbes, H., & Henderson, L. (1990). The provision of primary health services under band control: the Montreal Lake case. Native Studies Review, <5(1), 153-164. Organization for Economic Co-Operation and Development. (20011). Health at a glance 2011: OECD Indicators. Retrieved February 9, 2012 from http: //dx.doi.org/10.1787/health_glance-2011-en Perneger, T. V. (1998). What's wrong with Bonferroni adjustments. British Medical Journal, 375(7139), 1236-1238. Pitbaldo, J. P., & Pong, R. W. (1999). Geographic distribution of physicians in Canada. Sudbury, ON; Center for Rural and Northern Health Research Laurentian University. RHS National Team. (2007). First Nations regional longitudinal health survey, Code of research ethics. Retrieved December 5, 2010 from http://www.rhs-ers.ca/english/pdf/rhs-code-ofresearch-ethics-2007.pdf Sanchez, M., Vellanky, S., Herring, J., Liang, J., & Jia, Hui. (2008). Variations in Canadian rates of hospitalization for ambulatory care sensitive conditions. Healthcare Quarterly, 11(4), 20-22. Shah, B. R. S., Gunraj, N., & Hux, J. E. (2003). Markers of access to and quality of primary care for Aboriginal people in Ontario, Canada. American Journal of Public Health, 93 (5), 789-802. Sin, D. D., Wells, H., Svenson, L. W., Man, S. F. P. (2002). Asthma and COPD among Aboriginals in Alberta, Canada. Chest, 121, 1841-1846. Starfleld, B. (1996). Public health and primary care: a framework for proposed linkages. American Journal of Public Health, 86, 1365-1369. Starfleld, B. (1998). Primary Care: balancing health needs, services and technology. 2nd ed. New York, NY: Oxford University Press; pp 8-9. 68 Starfield, B., Shi, L., & Macinko, J. (2005). Contributions of Primary Care to Health Systems and Health. The Milbank Quarterly, 83(3), 457-502. Tarlier, D. S., Browne. A. J., & Johnson, J. (2007). The influence of geographical and social distance on nursing practice and continuity of care in a remote First Nations community. Canadian Journal of Nursing Research 39(3), 126-148. Transfer Status as of March 2008. (n.d.). Retrieved from http://www.hc-sc.gc.ca/fniahspnia/finance/agree-accord/trans_rpt_stats-eng.php#signed Waldrum, J. B., Herring, D. A., & Young, T. K. (2007). Aboriginal health in Canada. Historical, cultural, and epidemiological perspectives (2nd ed.). Toronto: University of Toronto Press Incorporated. Warry, W. (1998). Unfinished Dreams: Community healing and the reality of Aboriginal selfgovernment. Toronto: University of Toronto Press Incorporated. Whitehead, M. (1992). The concepts and principles of equity and health. International Journal of Health Services, 22, 429-445. Young, T. K., Reading, J., Elias, B., & O'Neil, J. (2000). Type 2 diabetes mellitus in Canada's First Nations: status of an epidemic in progress. Canadian Medical Association Journal, 163(5), 561-566. 69 Appendix 1: Adjusted Residuals for Barriers and Community Characteristics Adjusted residual analysis of barriers/community characteristics in relation to self-rated health. Did you experience the following? Answer Excellent Very Good Good Fair Poor Could not afford transportation yes -1.69484 -3.20692 0.705553 1.277284 4.525121 no 1.696351 3.206655 -0.70557 -1.27676 -4.52378 Prior approval denied yes 1.434 -1.832 1.496 2.891 3.362 no 1.433 1.832 1.495 -2.892 -3.368 Doctor or nurse not available yes -2.797 1.135 -2.480 2.027 2.990 no 2.795 -1.135 2.480 -2.027 -2.990 yes -1.553 0.111 -2.157 1.777 3.017 no 1.554 -0.111 2.156 -1.778 -3.020 Difficulty accessing traditional care yes -1.059 -0.856 -2.449 2.673 3.319 no 1.059 0.856 2.450 -2.675 -3.313 Services not covered by NIHB yes -1.255 -2.026 -1.438 2.564 3.795 no 1.257 2.026 1.439 -2.564 -3.797 Facility not in area 70 Could not afford direct costs yes -1.399 -2.389 -1.423 1.912 5.674 no 1.398 2.389 1.423 -1.911 -5.665 Felt services was not culturally appropriate yes -2.106 -0.932 -2.538 2.889 4.736 no 1.877 0.932 2.538 -2.887 -4.727 Services not available in area yes -1.627 0.028 -1.042 0.241 3.667 no 1.628 -0.028 1.042 -0.241 -3.666 Nurse stationed in community everyday yes 2.873 2.012 0.834 -3.850 -2.766 no "2-874 "2-011 "°-834 3-852 2-764 Traditional healers visiting at least twice annually yes -0.375 -1.555 -1.475 1.732 3.118 no 0449 1 559 1478 -1-806 "3119 71 Appendix 2: Adjusted Residuals for Health Transfer Agreement Adjusted residual analysis of health transfer agreement types in relation to self-rated health. Agreement Type Excellent Very Good Good Fair Poor Self-government agreement/Transfer Agreement 2.966 -0.331 -1.976 -0.237 0.692 Integrated agreement -2.712 -0.009 -0.884 2.042 2.152 Other Agreement -0.658 0.399 3.240 -1.847 -3.052