"Women's health involves emotional, social, cultural, economic, spiritual and physical well-being and is determined by the social, cultural, political and economic context of women's lives as well as by biology." S. Phillips, 1995

Table of Contents
Introduction
Methods
Summary Analyses
Conclusions
Recommendations
Exemplary Practices for Applying Gender Analysis to the Health Sector

Introduction
The mandate of the Prairie Women's Health Centre of Excellence (PWHCE) includes generating new knowledge through the identification and analysis of research on women's health issues; and providing policy advice, analysis and information to governments, health organizations and non-governmental organizations. The purposes of this project were:

• to generate new knowledge about the impact of the regionalization of health planning and service delivery by examining the degree to which gender sensitivity and women's health issues were reflected in the planning processes of regional health bodies in Manitoba and Saskatchewan; and
• to provide information which the PWHCE could use in advising governments, regional health bodies and others on how to make regional needs assessments and health plans more sensitive to the needs of women.

The research team developed evaluation frameworks based on the relevant literature and discussions with PWHCE staff and members of the PWHCE Theme Advisory Group on the Effects of Health Reform on Women. The frameworks were used to analyze needs assessment and health plan documents. This process was followed by interviews with key stakeholders within the regional health bodies that had provided written documents.

The framework for evaluating health plans addressed the following issues:

• evidence of women's health as a priority;
• recognition of context and determinants of women's health;
• approaches to women's health issues (primarily illness-focussed, or inclusive of health promotion and gender analysis of social conditions);
• sensitivity and proactive approach in addressing diversity;
• accessibility (to services, types of providers, community settings);
• types of collaborative relationships (with women, agencies);
• recognition of informal caregiver issues;
• recognition of effects of health care reform on employees (mostly female);
• evidence-based decision making and evaluation.

The needs assessment evaluation framework addressed the following issues:

• inclusivity of consultations;
• minimizing barriers to participation;
• inclusion of data related to health determinants;
• disaggregation of data by sex;
• discussion of findings for specific groups of women;
• verification of findings with communities.

Guiding questions for representatives of regional health bodies (key informants) addressed:

• how decisions are made about health priorities;
• if/how gender-related issues are included in the health planning process;
• perceptions of the most important influences on women's health;
• ways of including women and organizations that work with women in health planning processes;
• use of evidence-based decision making in planning;
• differential influences on women and men of determinants of health and health care reform (e.g., institution to community shift);
• collaborative initiatives with other organizations serving women;
• ways of addressing diversity in the health planning process, and related challenges;
• ways of including women and organizations that work with women in evaluation of health reform efforts;
• ways to be responsive to women's needs as health reform proceeds.

A. METHODS

1. NEEDS ASSESSMENT AND HEALTH PLAN DOCUMENTS

Each region/district received a letter from the Director of the Prairie Women's Health Centre of Excellence introducing the project and the project team members. They then received a follow-up letter from a member of the research team which requested that they provide:

• the most recent health plan for their region/district;
• health needs assessments done for their region/district; and
• any other work which they may already have done on the health needs of women in their region/district.

This documentation was analyzed using two frameworks -- one for the needs assessment documentation and one for the health plan documentation.

In all, eight of 11 Manitoba Regional Health Authorities (RHAs) responded. Seven of these provided health plan documents for review and analysis, and eight provided needs assessment documents. The two Winnipeg RHAs were excluded from the survey because they began operation in April 1998 and have not yet published their first needs assessment documents. In Saskatchewan, 17 of 32 Health Districts responded to the request. Sixteen of those provided health plan documentation, and 12 provided needs assessment documentation. Documents submitted by regions/districts, in response to the request for information about any work undertaken on women's health, have also been included from both provinces.

In the information and analysis of the needs assessment and health plan documents which follow, the data from the two provinces has been combined. There were no substantial differences between responses from Manitoba and those from Saskatchewan.

2. INTERVIEWS WITH REPRESENTATIVES OF REGIONAL HEALTH BODIES

Regional health bodies were selected for interviews using the following criteria: the sample should be representative of both provinces (3 Manitoba, 5 Saskatchewan); and the sample should be geographically representative (north, south, rural, urban). In several cases, representatives of regional health bodies expressed interest in being interviewed for this project. All of these representatives were interviewed. In order to arrange the interviews, the CEO's office was contacted and the purpose of the interviews was explained. The interviewee was selected by the CEO. Their positions in the regional health bodies varied from Health Educator to Vice President to Medical Officer of Health. All interviews were conducted by telephone, by the same member of the research team. She took verbatim notes during the interview.

These transcribed interviews were imported into the QSR NUD*IST software program for qualitative data analysis. All transcripts were read by all of the research team members. One member conducted an analysis using the constant comparison method. The analyst moved back and forth between transcripts and analysis, uncovering similarities and differences, within and between interviews. Codes were applied to a section of a transcript, and then all sections of transcripts from interviews with the same codes were reviewed. The analyst went back and forth between sections of the transcript and the whole transcript to check context of quotes and verify interpretations. As sections were coded, categories and subcategories became apparent. As a result of this process, the interviews were reviewed several times. The analyst then wrote a narrative description of the data, following the coding. The process of writing and interpreting lead to returning to the transcripts, re-checking context, and searching the interviews for other codes and categories. The preliminary analysis was then sent to the other two researchers who checked the credibility of interpretation. Differences of opinion were few, but were discussed until consensus was reached. The variety in backgrounds of the three researchers enhances the transferability of the findings; in other words, there is less chance that the interpretation is narrowed by discipline or experience. In qualitative methodology, the term credibility is equivalent to validity and transferability to generalizability.

B. SUMMARY ANALYSES

1. NEEDS ASSESSMENTS

a. The needs assessment documents reviewed indicate that gender was rarely considered as a variable in assessing local health needs and that consequently, the health needs of women rarely were considered separately from those of men.

b. Regional health bodies published little sex- disaggregated data. While gender analysis is much more than simply looking at health data for men and women both separately and together, the lack of availability of sex-disaggregated data makes gender analysis impossible. Regional health bodies are also limited, since they did not have additional funds to order sex-disaggregated data from other sources (such as Statistics Canada) for their areas, nor did either province undertake to provide this to them.

c. Although Manitoba Health has set women's health as one of its priorities, RHAs were given no background information about women's health, nor any guidance about bow to specifically assess the health of women in their communities. This lack of information is reflected in their responses. In both provinces, only 25% of those participating included any data about gender in their needs assessments.

2. HEALTH PLANS

a. Based on the review of these documents, it is evident that regional health bodies have not given a high priority to women's health. While four of the seven responding Manitoba RHAs listed women's health as a priority, and referenced Manitoba Health in doing so, there was little evidence of such prioritization in their health plans. Only one regional health body--in Saskatchewan--expressed a written commitment to gender equity.

b. Where women's health issues were considered, the most frequent references were to gender-specific health needs (reproductive health, breast and cervical cancer screening) and to women's role as mothers.

c. While both provinces officially promote a determinants of health approach, there is little evidence of it in the health plans reviewed for this project. Manitoba health plans contained, on average, reference to 2.4 health determinants, while Saskatchewan plans included, on average, only 1.5 of the 11 health determinants used in this project. Health plans tend to emphasize financial reporting and funding requests.

d. The documents reviewed do not demonstrate an appreciation for the differing health needs of diverse groups of women, including Aboriginal women, women from ethnic and visible minorities, lesbian women and women with disabilities.

e. Consistent with all of the above, none of the regional health bodies surveyed reported any training on gender issues for either staff, management or Board members.

f. There was no evidence that women's organizations, and organizations providing services to women are included in the health planning process.

g. Rather than recognizing the additional burden on women of providing informal care to family members and friends, regional health bodies have promoted this by emphasizing women's presumed role as gatekeepers of family health.

3. INTERVIEWS WITH REPRESENTATIVES OF REGIONAL HEALTH BODIES

Themes from the qualitative analysis of interview transcripts indicated that women's health was discussed in the context of three categories--reproduction, family members, and use of health services--rather than as a valued outcome in and of itself. Within these categories there was very little gender analysis with a few exceptions. Similarly, there was widespread understanding of the social determinants of health, but gender was seldom mentioned and the other determinants lacked a gender analysis. Major women's health issues were identified, and this problem focus formed one of the main themes in talking about women's health, along with recognition of different populations of women, and of women's roles. Again, in general, there was, at best, the beginning of a gender analysis in these conversations. Many informants seemed reluctant to address gender at all. In some instances, "backlash" was noted: that is, people who believed that "all this attention to women's health" represents a loss for men and a threat to their health. As would be expected, the discussions of health planning and participation of women were not rich in examples of equity strategies used or gender differences addressed.

C. CONCLUSIONS

Conclusion 1. There was no significant difference in findings for Manitoba and Saskatchewan. This is noteworthy given their different political environments at the time the study was conducted; and the fact that Manitoba identifies women as one of its four priority populations, and Saskatchewan identifies women's special health needs as a priority budget area.

Conclusion 2. There is little evidence of gender analysis or gender-sensitive strategies among the regional health bodies participating in this study, as indicated by review of needs assessment and health planning documents, and interviews with representatives of the participating health bodies. For example, only 25% of the participating regional health bodies included any data about gender in their needs assessments.

Conclusion 3. While the reasons for this lack of evidence are multi-faceted, the primary reason is a lack of value placed on women's health in general, and therefore, on gender analysis in particular, as legitimate areas of concern. This is corroborated by our finding that there was no evidence of training on issues related to gender inequality that effect women's health. Where women's health issues were considered, the most frequent references were to biological sex-specific health needs (reproductive health, breast and cervical cancer screening) and to women's role as mothers.

Conclusion 4. Some reasons for this lack of value placed on gender analysis are as follows:

• 4.1 Many of the participants believed that women's primary health role is as gatekeepers and informal care givers, responsible for the health of their families and communities. Regional health bodies have not recognized the additional burden on women of providing informal care to family members and friends. Rather, they have potentially added to this burden by emphasizing women's presumed role as the gatekeepers of family health. Women's health did not appear to be valued in its own right.
• 4.2 Gender analysis did not appear to be valued by the provincial governments which fund the regional health bodies. For example, although Manitoba Health has set women's health as one of its four priority areas, RHAs appear to have been given no background information about women's health, nor any guidance about how to specifically assess the health of the women in their communities.
• 4.3 The overwhelming financial pressures faced by regional health bodies dealing with provincially-imposed funding restraints encourage a crisis-management focus (e.g., emergency staffing issues). Gender analysis is not seen as high priority in this environment.
• 4.4 There did not seem to be widespread anti-feminist sentiment or hostility toward women's health. Rather, women's health issues (beyond those related to reproduction) and gender analysis did not appear to be priorities to the health bodies participating in this project. This could be changed by involving women's organizations and organizations providing services to women in the health planning process. However, there was no evidence of such a collaborative approach in the documents reviewed.

These conclusions are, unfortunately, consistent with much current work in the field of population health. As Patricia Kaufert has noted in her analysis of four of the key texts in population health:

" ... [the authors'] decision to ignore women cannot be explained as a matter of chance or academic absent-mindedness. At some level, conscious or unconscious, the decision was made to ignore these differences, to treat them as taken for granted, 'no longer questioned, examined or viewed as problematic'." Kaufert, P., "The vanishing woman: gender and population health" in Sex, Gender and Health, Cambridge University Press, 1999, p. 123.

Conclusion 5. Neither province requires that health data be disaggregated by sex, although Manitoba does require that the sex of survey respondents be recorded. (Manitoba RHAs can therefore report the percentage of male and female respondents, but they have not reported if and how the responses of men and women differed.) While gender analysis is much more than simply looking at health data for men and women both separately and together, the lack of availability of sex-disaggregated data makes gender analysis impossible. Regional health bodies are also limited, since they did not have additional funds to order sex-disaggregated data from other sources (such as Statistics Canada) for their areas, nor did either province undertake to provide this to them.

Patricia Kaufert has described this tendency, found in the work of most population health experts, as follows:

"For epidemiologists and statisticians, the aggregation of data, or their adjustment for age or sex, are simply routine procedures. This approach is so commonplace I did not question it myself until deliberately hunting for the women and finding they were missing or hidden within an aggregated data set." Kaufert, op. cit. p. 125.

In the health needs assessment surveys which were examined for this project, all of the regional health bodies which reported the sex of their respondents reported that more respondents were women. Their published results may therefore not adequately reflect the health needs of men in their local communities.

Conclusion 6. The documents reviewed do not demonstrate an appreciation for the differing health needs of diverse groups of women, including Aboriginal women, women from ethnic and visible minorities, lesbian women and women with disabilities.

Conclusion 7. The decision of the Manitoba government, and of those health districts in Saskatchewan which collected survey data, to use household rather than individual data also created problems of data interpretation. One does not know who is represented by the responses. Is the respondent speaking for her/himself or others in the household when answering a question about a particular health need, behaviour or interest? This type of proxy data is particularly questionable, for example, when obtaining information about reproductive health or mental health issues. It makes the disaggregation and analysis of data by sex more problematic.

Conclusion 8. While both provinces officially promote a determinants of health approach, there is little evidence of this in the health plans reviewed for this project. Manitoba health plans contained, on average, reference to 2.4 health determinants, while Saskatchewan plans included an average of only 1.5 of the determinants used in our framework. Health plans tend to emphasize financial reporting and funding requests.

Conclusion 9. Regional health bodies vary considerably in their level of technical expertise in assessment planning, data collection and analysis. Rural regions are at a serious disadvantage with regard to both research literature and access to technical assistance. Internet access is not sufficient to address their information needs.

D. RECOMMENDATIONS

Recommendation 1. Consistent with Canada's international commitments and in order to accurately assess community health needs, and to develop policies, programs and strategies to promote good health and meet health service needs, we recommend that the provincial ministries of health:

• require that regional health bodies collect and report gender disaggregated data in their needs assessments and health plans; and include gender analyses in their health plans; and
• provide regional health bodies with the necessary training, expertise and funds to accomplish these tasks.

Recommendation 2. We recommend that provincial governments ensure that regional health bodies, especially those in rural areas, have affordable access to information sources such as relevant research-based journals and ongoing information about gender analysis and women's health.

Recommendation 3. We recommend that in order to provide the necessary leadership, each provincial government should establish an appropriately-staffed office with expertise in gender analysis and women's health. The expertise of this office should be made available to the regional health bodies and to other government departments the policies of which directly effect women's health, such as finance, social/family services, housing and seniors' services.

Recommendation 4. We recommend that both the provincial governments and regional health bodies broaden their perspective on women's health beyond reproductive and family caregiving to encompass a broad determinants of health approachùincluding gender as a separate determinantùin practice as well as in their public relations materials. In addition, we recommend that eligibility for community-based health services not be based on the assumption that women are willing to provide unpaid caregiving services to family members.

Recommendation 5. The need to develop skills in gender analysis exists across Canada is not limited to the two provinces examined in this project. Following from Canada's signature to the Beijing Platform for Action, we recommend that the Federal government establish a Federal/Provincial/Territorial working group to synthesize and adapt existing policies and gender analysis frameworks and tools for use by regional health bodies across the country. In order to make the best use of existing knowledge, this group needs to work with the Centres of Excellence for Women's Health and other experts in the field.

Recommendation 6. There is a need to incorporate gender analysis throughout the whole planning process, especially at the policy-making and senior planning level, so that there is a systematic approach to addressing women's health needs and gender sensitivity. Though this research focused on needs assessment and the development of health plans, we recommend applying the methods and tools of gender analysis to program implementation, evaluation and resource allocation as well. Some of the gender analysis tools and model approaches presented in this report can provide guidance.

Recommendation 7. We recommend that regional health bodies institute processes for ongoing input and feedback from diverse groups of women regarding their policies, programs and strategies and how well they meet the needs of women in their regions. Regional health bodies can draw from the expertise of community organizations that work with women, as well as researchers with expertise in gender issues and participatory research approaches.

Recommendation 8. Some regional health bodies developed a keen interest in gender analysis during this project. We recommend that the PWHCE pursue opportunities to facilitate and promote gender-sensitive approaches by continuing to work with those regional health bodies which expressed an interest in gender analysis during the course of this project.

Recommendation 9. In addition to training for regional health bodies and provincial governments, it is important that women and organizations that work with women in the community have access to educational materials and events (e.g., workshops) on gender-based analysis and gender-sensitive health planning. Community organizations and concerned individuals often link with decision makers in their various community roles, and would benefit from gaining the expertise to analyze policies and programs and as citizens hold decision-makers accountable for their actions. PWHCE could work with community stakeholders on this issue, as well as with regional health bodies.

Recommendation 10. In order to monitor change and progress, we recommend that regional health bodies be studied again in five years regarding their use of gender analysis and gender-sensitive planning.

E. EXEMPLARY PRACTICES FOR APPLYING GENDER ANALYSIS TO THE HEALTH SECTOR

The most detailed tools to date applying gender analysis to the health sector have been produced by Schalkwyk, Woroniuk, and Thomas (1997) for the Swedish International Development Corporation Agency, the Gender and Health Group (1999), and the Pan American Health Organization. These are reviewed in depth in the full report. Some key issues they address include:

• increased representation of women in decision-making and opportunities for advancement in the health sector (e.g., as employees);
• recognition of social context influences on health (e.g., social and economic disadvantages);
• broadening the focus of women's health beyond reproduction, women's role as mothers, and conditions specific to or more prevalent among women (e.g., cervical and breast cancer);
• inclusion of men as well as women in redressing inequities and promoting women's health and equality (e.g., safer sexual practices);
• gender-sensitivity in all programs, not just those specifically for women;
• equality of outcomes, rather than sameness of activities or treatment (e.g., an equity focus), and inclusivity in developing indicators of success;
• disaggregation of data by sex as well as other demographics;
• training in women's health and gender issues (in both practice and research) for both decision-makers and staff;
• use of inclusive public consultation processes that take barriers to participation into account (e.g., child care, transportation);
• links to women's organizations and other sources of expertise in gender analysis, as well as to organizations that address broader health determinants (e.g., food security);
• equitable distribution of resources, access and quality of services by gender, as well as attention to the impacts of health reform on unpaid caregiving and out-of-pocket costs (e.g., user fees);
• sensitivity to diversity (e.g., cultural); and
• inclusion of women in research--both as participants and in the planning process of research.