4.3.1 beth jackson
•Download as PPS, PDF•
0 likes•249 views
4.3.1 AWHN Conference 6 2010 Chancellor 2: Evidence for Equity: Population Statistics, Narrative Accounts and Gender-sensitive Indicators
Recommended
Recommended
More Related Content
Similar to 4.3.1 beth jackson
Similar to 4.3.1 beth jackson (20)
More from Australian Women's Health Network 6th Conference 2010
More from Australian Women's Health Network 6th Conference 2010 (20)
Recently uploaded
Recently uploaded (20)
4.3.1 beth jackson
- 1. Evidence for Equity:Evidence for Equity: Population Statistics,Population Statistics, Narrative Accounts andNarrative Accounts and Gender-sensitive IndicatorsGender-sensitive Indicators M Haworth-BrockmanM Haworth-Brockman A PedersonA Pederson B JacksonB Jackson May 21, 2010May 21, 2010
- 2. BackgroundBackground For more than 10 years we have beenFor more than 10 years we have been working towards policies and standardsworking towards policies and standards that will:that will: 1.1. record and monitor women’s healthrecord and monitor women’s health adequately and appropriatelyadequately and appropriately 2.2. integrate gender considerations in healthintegrate gender considerations in health surveillance, planning & policysurveillance, planning & policy This presentation will describe our contributionsThis presentation will describe our contributions to the international discourse.to the international discourse.
- 3. Three Parts Our experience with gender-sensitive (or not) indicators A case study of wait times Understanding the process to understanding indicators
- 4. Health authorities at allHealth authorities at all levels have been grapplinglevels have been grappling with accountabilitywith accountability frameworks in general,frameworks in general, evidence-based decisions asevidence-based decisions as a concept, and indicators ofa concept, and indicators of outcomes for both.outcomes for both.
- 5. Part of an on-going process …Part of an on-going process … …… to getto get gendergender andand women’s healthwomen’s health onon federal, provincial &federal, provincial & regional agendasregional agendas Our role has included:Our role has included: back and forth of policyback and forth of policy conversations,conversations, new research, andnew research, and trainingtraining
- 6. For example …For example … developed with support, interestdeveloped with support, interest and funding from both theand funding from both the provincial & federal governmentsprovincial & federal governments serves asserves as baseline reporting, alsobaseline reporting, also an example of how to integratean example of how to integrate different kinds of evidence in adifferent kinds of evidence in a “report card”, and of course,“report card”, and of course, how to do both of these with SGBAhow to do both of these with SGBA
- 7. At the same time…At the same time… We conducted one of three tests of WHOWe conducted one of three tests of WHO gender-sensitive core set of leadinggender-sensitive core set of leading indicatorsindicators Was to be a technical feasibility testWas to be a technical feasibility test To really test the gender-sensitivity, had to do anTo really test the gender-sensitivity, had to do an SGBA of each of the 37 indicatorsSGBA of each of the 37 indicators in each case, need to interpret beyond the healthin each case, need to interpret beyond the health surveillance numbers to understand not onlysurveillance numbers to understand not only whatwhat was going on butwas going on but whywhy As Lin and others found, regionalAs Lin and others found, regional and other differences aboundand other differences abound (Lin et al 2005, Abdool et al 2002)(Lin et al 2005, Abdool et al 2002)
- 9. After refining our methods in theseAfter refining our methods in these analyses of gender indicators …analyses of gender indicators … Written guidelines on ourWritten guidelines on our methods in SGBA of healthmethods in SGBA of health information & for health profilesinformation & for health profiles
- 10. Additional Case StudiesAdditional Case Studies BelizeBelize HIV/AIDSHIV/AIDS Transport AccidentsTransport Accidents DiabetesDiabetes BrazilBrazil homicidehomicide GuatemalaGuatemala Body weightBody weight ManitobaManitoba CancersCancers Focus on ethnicity andFocus on ethnicity and regional differencesregional differences waswas eithereither explicitexplicit because requestedbecause requested oror because this isbecause this is thethe way we do SGBAway we do SGBA anywayanyway
- 11. Through repeated case studies it is clearThrough repeated case studies it is clear that getting at the gendered aspects ofthat getting at the gendered aspects of health indicators is a complicated process,health indicators is a complicated process, requiring different kinds of evidence.requiring different kinds of evidence.
- 12. Analysis & discuss literature Apply a gender lens to those determinants of health that are relevant to diabetes. Consider the biological, social (family, community), and systemic pathways (health care, policies) that are protective or detrimental for F&M. Show comparable or conflicting findings. Introduce the purpose The purpose is to determine if sexes differ for illness, death, & disability due to diabetes Name data sources & limitations Diabetes hospitalizations & death data are available. In what ways are data limited in responding to the stated purpose? Analyze & describe data What do sex-disaggregated data on diabetes tell us? What other questions are raised? Conclude on lessons Diabetes differs for F&M in these ways… The gendered pathways that contribute to the differences are… Gaps and needs are… Create a framework Community and experts (F&M) may call for a GBA of diabetes, within a framework of other topics and indicators. Profile’s other indicators add broader understanding Literature suggests data sources & critiques quality Literature & networks provide policy & program ideas Analyze and recommend policy Which policies are in effect? How can new or revised policies better address gender significant findings? Literature, community, personal knowledge inform gender and diabetes analysis Isfeld 2008
- 13. Analytical ProcessAnalytical Process DefinitionDefinition of issues & measuresof issues & measures Gathering InformationGathering Information —review of—review of data & add gender contexts, meaning,data & add gender contexts, meaning, experienceexperience Analytical InquiryAnalytical Inquiry —asking—asking challenging questionschallenging questions Implications & lessonsImplications & lessons to buildto build gender sensitive strategiesgender sensitive strategies
- 14. Gender and Wait Times for TJAGender and Wait Times for TJA Wait time is a health systemWait time is a health system performance indicator – usuallyperformance indicator – usually measured in days and then reportedmeasured in days and then reported by facility and/or physicianby facility and/or physician Tells us nothing about who isTells us nothing about who is waiting only how long nor what thewaiting only how long nor what the meaning of waiting is for those whomeaning of waiting is for those who waitwait Insensitive to social processes thatInsensitive to social processes that affect who is able to access care oraffect who is able to access care or who decides how care is allocatedwho decides how care is allocated
- 15. ““New Questions, New Knowledge”New Questions, New Knowledge”
- 16. ““Waiting to Wait”Waiting to Wait”
- 17. Systematic ReviewSystematic Review http://ebp.lib.uic.edu/nursing/?q=node/12 Library of the Health Sciences-Chicago, University of Illinois at Chicago
- 18. Realist review?Realist review? “The results of the review combine theoretical understanding and empirical evidence, and focus on explaining the relationship between the context in which the intervention is applied, the mechanisms by which it works and the outcomes which are produced.” Pawson, R., Greenhalgh, T., Harvey, G., & K. Walshe. Realist review – a new method of systematic review designed for complex policy interventions. J Health Serv Res Policy 2005;10:21-34
- 19. So this is what we know:So this is what we know: There is a demonstrated need for establishingThere is a demonstrated need for establishing both womenboth women’’s and gender-sensitive healths and gender-sensitive health indicators to monitor and improve womenindicators to monitor and improve women’’ss health and equity.health and equity. In health policy circles there have beenIn health policy circles there have been repeated calls to complement descriptiverepeated calls to complement descriptive quantitative surveillance data with qualitativequantitative surveillance data with qualitative research and analysis. But…research and analysis. But… Very rarely does this lead to theVery rarely does this lead to the systematicsystematic applicationapplication of qualitative research inof qualitative research in evidence-based decision-making. In otherevidence-based decision-making. In other words, there is ‘buy-in’ but little ‘uptake’.words, there is ‘buy-in’ but little ‘uptake’.
- 20. From ‘art’ to evidence…From ‘art’ to evidence… Shift in medicine & public health from caseShift in medicine & public health from case study & interpretation (the ‘art’ of medicinestudy & interpretation (the ‘art’ of medicine grounded in tacit & practical knowledge) togrounded in tacit & practical knowledge) to clinical epidemiology & evidence-basedclinical epidemiology & evidence-based decision makingdecision making Demonstrates a hierarchy of evidentiaryDemonstrates a hierarchy of evidentiary authority (interpretive practices & evidenceauthority (interpretive practices & evidence are trumped by scientific fact)are trumped by scientific fact)
- 21. Prioritization of positivist, evidence-basedPrioritization of positivist, evidence-based practice and policy making has led topractice and policy making has led to the disavowal of the individual case,the disavowal of the individual case, particular experience, and testimonyparticular experience, and testimony as credible sources of evidence.as credible sources of evidence. Moreover, standardization andMoreover, standardization and population statistics erase the “Other”,population statistics erase the “Other”, the remainder, the excess.the remainder, the excess. And yet, particular experience, on itsAnd yet, particular experience, on its own, is not a sufficient foundation forown, is not a sufficient foundation for knowledge -- when tacit knowledge isknowledge -- when tacit knowledge is isolated, cut off from analysis of socialisolated, cut off from analysis of social contexts and relations of power, it toocontexts and relations of power, it too is vulnerable to error.is vulnerable to error.
- 22. What to Do?What to Do? NeitherNeither devaluedevalue nornor valorize:valorize: a)a) the universalizing evidence of biomedicalthe universalizing evidence of biomedical experimentation and population statisticsexperimentation and population statistics oror b)b) individualized, experiential data.individualized, experiential data. Take seriously, on its own terms, each type ofTake seriously, on its own terms, each type of evidence and the methods & analysis thatevidence and the methods & analysis that produce it.produce it. Reinstate the authority of qualitative, case-Reinstate the authority of qualitative, case- based evidence (in the face of a persistentbased evidence (in the face of a persistent hierarchy of evidence).hierarchy of evidence). Foster sophisticated, ‘ecological’ analyses thatFoster sophisticated, ‘ecological’ analyses that engage diverse sources & types of evidence.engage diverse sources & types of evidence.
- 23. A. Recover, revive, reinstate theA. Recover, revive, reinstate the authority of the particular case andauthority of the particular case and testimonial evidencetestimonial evidence The ‘urge to generalize’ has a firm gripThe ‘urge to generalize’ has a firm grip on researchers and policy makers –on researchers and policy makers – but qualitative evidence can sensitizebut qualitative evidence can sensitize us to inequities otherwise overlooked.us to inequities otherwise overlooked.
- 24. Locate individual experience andLocate individual experience and testimony in relations of powertestimony in relations of power Social locations (e.g. gender, race, class) areSocial locations (e.g. gender, race, class) are notnot simply attributes of individuals, they are the productsimply attributes of individuals, they are the product of social relations.of social relations. When we situate local knowledge clearly withinWhen we situate local knowledge clearly within social structures (e.g. diagnostic practices,social structures (e.g. diagnostic practices, gendered relations of care), our analysis ofgendered relations of care), our analysis of testimony has power beyond the individual case.testimony has power beyond the individual case.
- 25. B. Develop sophisticated, ‘ecological’B. Develop sophisticated, ‘ecological’ analyses that move between differentanalyses that move between different levels of analysis and diverse sources &levels of analysis and diverse sources & types of evidencetypes of evidence
- 26. Case Case Case A Case B Particular Experience (Individuals) Micro-level Interactions Generalized Evidence (Population) Macro-level Structures
- 27. Arthritis & TJAArthritis & TJAThis mode of thinking allows us to link the incidence of disease, to the pain women feel, to how it is diagnosed with radiography, to how women’s pain and disease are perceived by physicians, and to why women’s responsibilities may prevent them from taking advantage of their position on a wait list for surgery. Who’s going to do all this if I go for surgery? ...and who will look after me?
- 28. A Mode of ThinkingA Mode of Thinking supports gender-based and intersectionalsupports gender-based and intersectional analysisanalysis demonstrates an ability to read differentdemonstrates an ability to read different sources of evidence against one anothersources of evidence against one another learns from analogies and disanalogieslearns from analogies and disanalogies across sites, across levels of analysis andacross sites, across levels of analysis and across practices of research and policyacross practices of research and policy makingmaking
- 29. Thank youThank you M. Haworth-BrockmanM. Haworth-Brockman Prairie Women’s Health Centre of ExcellencePrairie Women’s Health Centre of Excellence Ann PedersonAnn Pederson British Columbia Centre of ExcellenceBritish Columbia Centre of Excellence for Women’s Healthfor Women’s Health Beth JacksonBeth Jackson Women & Health Care ReformWomen & Health Care Reform . With thanks to Harpa Isfeld
Editor's Notes
- Attend to gender in each step of the analytical process.
- In Canada, we have had a national initiative to reduce wait times for several surgical and diagnostic procedures since the federal, provincial and territorial First Ministers agreed on priority areas in 2005. One of the identified priorities was wait times for hip and knee replacement surgery – total joint arthoplasty. Wait times are a type of health system performance indicator. They are usually measured in days, often by institution, occasionally by physician, and now they are commonly , reported in aggregate form on public websites. In the past, wait times were unsystematic, largely invisible, more rumour than fact, privileged information tracked by individual physicians. To assuage public fears about long waits for care, government established more publicly visible processes for measuring and reporting wait times. They also appointed a Federal Wait Times advisor, Dr. Brian Postl, to consult widely on remedies to reduce wait times for these priority areas. Our research group – Women and Health Care Reform – found ourselves in the interesting position of being asked to comment on what a gender analysis would contribute to the discussion of wait times. So we chose to look specifically at wait times for TJA as a case example and see what conducting an SGBA would tell us. When we initially looked at the provincial reports and websites on wait times, we could not discern anything about gender at all because the data were reported in aggregate.
- Given the available time and resources, we approached a sex- and gender-based analysis of wait times for TJA by conducting a narrative review of the published literature and key items in the grey literature, complemented by a small number of key informant interviews. The report, which we entitled “New Questions, New Knowledge”, eventually found its way into an appendix of the Federal Wait Times Advisor’s report. Briefly, our report argued that there was published literature suggesting that given that wait times were measured as the length of time from when a surgeon and a patient agreed that TJA was necessary and a referral for surgery was made, that this indicator was not very good at illuminating the processes involved in being referred. Published literature raised questions about women’s and men’s relative need for TJA, given differences in the prevalence of osteoarthritis (the condition underlying most referrals for TJA) and subsequent rates of referral. Studies using standardized patients, for example, raised questions about whether physicians demonstrated gender bias and were more likely to refer male than female patients for surgery. The literature suggested that this may reflect physicians’ different interpretation of reports of pain by women and men, as well as possible differences in their interaction with male and female patients. For instance, some studies showed that women were less likely to ask about TJA and physicians were less likely to raise the topic with women, even when women and men showed the same level of joint damage and reported the same level of disability and pain.
- We concluded that the operational definition of wait times as ….obscured the patient experience prior to the referral to surgery. We came to understand that there were numerous points along the patient journey at which various aspects of gender-related bias could contribute to a person either being facilitated through or impeded along the pathway to surgery. In short, we came to understand that more women than men were “ waiting to wait,” that is, that while both men and women underuse TJA relative to need, women underuse TJA 3 times more than men and this may possibly be accounted for by gender-related biases in patient decision making, physician referral practices, clinical and diagnostic assessment tools, and assumptions about the relative importance of male versus female activities – paid work versus caregiving, for example.
- This past year, we decided to try to approach the problem of gender and wait times for TJA through a different methodology. Rather than using a narrative review approach, we decided to see what we could learn about gender and wait times through the use of systematic review. We chose to do this because systematic reviews are a major tool for knowledge synthesis and we wanted to know how we might incorporate our interest in equity concerns, such as gender, within this approach to knowledge synthesis. So we put together a team, identified a research question, worked with a university-based librarian to conduct an extensive literature search, developed inclusion and exclusion criteria, and finally identified a set of literature to review. What we found was not very satisfying. Despite the care with which we conducted the literature search, we ended up with a collection of articles that had relatively little to say about wait times at all. What they focused on was the impact of waiting upon clinical outcomes – pain, disability, dysfunction – and, to a lesser extent…. When we tried to examine sex and gender, we were challenged by the way literature is organized by keywords, by the way that analyses often control for sex or report sex only as a demographic characteristic of the study sample, and by an overall lack of interest in examining wait times by variables other than clinical ones. It was even more difficult to examine the gender in this literature. Though the word “gender” was sometimes used, it was often in cases in which an author was presenting data and it was most often used as an alternative to word to ‘sex’. That is, gender was reported as being either male or female. The complex experiences of gendering that underpin social experience, and the intersection of gender with other important social processes such as racialization, classism, and heteronormativity, were impossible to explore. This way of understanding gender links it to history, culture and social practices, phenomena that are difficult to reduce to independent and dependent variables. We concluded that the methods of systematic review were not able to duplicate the SGBA we had been able to produce approaching the problem in a more narrative method.
- Our purpose in this part of the presentation was to illustrate the challenge of generating health evidence from statistical data. Our point is not to give an exhaustive account of the factors that influence wait times for TJA. But by trying to examine wait times for TJA using the methods of systematic review, we came to understand that the method was limited in its ability to address context and, by definition, gender is about context. Systematic reviews are designed to strip context away – to standardize an intervention so that we can isolate the factor that makes it “work.” Yet this approach may not be able to help us understand very much about the implications of sex and gender in a particular health system performance indicator. For that, we may need to consider alternative methods. Currently we are exploring the possibility of ‘realist review’ as an alternative approach to conventional systematic review. This approach explicitly acknowledges that interventions “work” in particular contexts and asks the analyst to generate theories by which the mechanisms in a complex intervention operate. This approach, which I’ve greatly oversimplified, may permit us to examine the particularities of a given context and to ask different questions – including questions about how a mechanism or intervention is gendered. In contrast to conventional systematic reviews, realist reviews do not purport to generate generalizable findings; rather, those who have written about realist reviews suggest a more modest goal, namely, that of generating insights and potential theories about why something works in a particular circumstance. To do this, realist reviews suggest incorporating qualitative data into the analysis and stress the value of gathering local expert opinion and experience. Realist reviews thus generate knowledge from the particular and it is the analysts’ role to identify the ways that a given case might have lessons to offer to another, including lessons about how gender shapes experience, practice, and even policy.
- Over the past several decades, we have witnessed a significant shift in the knowledge base of clinical medicine & public health, from case study & interpretation (the ‘art’ of medicine embedded in tacit & tacit knowledge) to clinical epidemiology & evidence-based decision making. This shift demonstrates a hierarchy of evidentiary authority (interpretive practices & evidence are trumped by scientific fact)
- One of the effects of this shift to evidence-based practice and policy making has been the disavowal of the individual case, particular experience, and testimony as credible sources of evidence. The exclusionary effects of this shift are profoundly damaging – the “Other”, the remainder, the excess, are erased by standardization and population statistics. And yet, particular experience, on its own, is not a sufficient foundation for knowledge -- when tacit knowledge is isolated, cut off from analysis of social contexts and relations of power, it too vulnerable to error.
- While we wish to guard against the urge to generalize, we also want to avoid the error of ‘experientialism,’ where experience is seen to be the most authoritative source of knowledge – so we must locate individual experience and testimony in an analysis of relations of power. Our analysis of individual/local experience is strengthened when we: understand that social locations (like gender, race, class) are not simply attributes of individuals, they are the product of social relations. situate local knowledge/experience clearly within social structures (e.g. diagnostic practices, gendered relations of care)
- Our analyses must move both horizontally (from situation to situation) AND vertically (from particular to general, micro to macro-structural) – tacking between empirical sites, between levels of aggregation, and between levels of analysis, looking for and learning from analogies and disanalogies among cases and conditions. Strong, multi-method, multilevel analyses will avoid the pitfalls of conceptualizing and analysing data at only one level, ignoring other important levels. (this speaks to the problem of using aggregated, individual-level data such as wait lists as evidence for the lack of gender bias in access to total joint arthroplasty) This is a mode of thinking, not simply a method; it requires attention to detail and an ability to take the long/broad view that situates local detail in socio-political-historical structures of power This mode of thinking acknowledges and respects commonalities and observable trends, but does not do so at the expense of particular experience/knowledge – it remains open to both generalized knowledge (e.g. prevalence of osteoarthritis in women and men) and particular experience (e.g. individual’s reports of pain and disability) It allows us to ‘read’ multiple kinds of evidence (qualitative, quantitative) at different levels of analysis (individual, insitutional, structural) through different lenses of analysis (e.g. sex- and gender-based analysis) – without privileging one approach to knowledge production *explanatory note re: levels of aggregation and levels of analysis: Individual-level variables (like income, smoking behaviour, attitudes, disease diagnosis) are described by individual-level data; these data may be aggregated and expressed as measures of central tendency (e.g. mean, median) or as rates/ratios (e.g. suicide rate, prevalence of infectious/chronic disease, level of income inequality); of course, these aggregated measures do not represent any particular individual’s experience some ‘structural’ or ‘relational’ variables are described by group-level data which reflect the pattern of relationships and interactions between individuals belonging to the group (e.g. social networks) other structural, ‘contextual’ or ‘global’ variables measure attributes of groups, organisations or places, and are not reducible to individual persons/units. They are fixed for all, or nearly all, individual group members (e.g. social (dis)organisation; social capital; legislation or regulation) So, individual-level variable data may be aggregated, but this does not mean the variable addresses structural-level phenomena/relationships NOTE: Aggregate variables and contextual variables should not be conflated: we must “disentangle the idea of aggregate variables from contextual variables which cannot be reduced to a lower level” (Subramanian et al., 2009, Internatl Journ Epidemiology 38 (2): 349) - This “underscores the need for multilevel thinking; i.e. we need to simultaneously examine the circumstances of individuals at one level, in the context of the different levels shaping their circumstances.” (my emphasis, Subramanian et al., 2009, Internatl Journ Epidemiology 38 (2): 349) - “Multilevel thinking, grounded in historical and spatiotemporal context, is thus a necessity, not an option. It is not that ecological effects are unconditionally important. However, continuing to do individual-level analyses stripped out of its context will never inform us about how context may or may not shape individual and ecological outcomes.” (Subramanian et al., 2009, Internatl Journ Epidemiology 38 (2): 355)
- We can make an important contribution to health services research and policy by fostering and modeling not simply a set of methods, but a mode of thinking – one that supports gender-based and intersectional analysis, and demonstrates an ability to read different sources of evidence against one another, learning from analogies and disanalogies across sites, across levels of analysis and across practices of research and policy making. Requires bridging & borderlands work, linking across epistemologies, methodologies, disciplines and methods. Requires a kind of multi-lingualism and a tolerance of the productive tensions between/among approaches to knowledge-production.