1 Perceived Discrimination in Health Care Among American Indian WomenBead work by Stella Washines Kelly Gonzales, MPH, PhD APHA-conference Denver, Colorado November 2010 Good morning. introduction
2 Scope of Presentation Introduction on perceived discrimination in health care Summarize study and findings on perceived experiences of discrimination in health care among American Indian women What are the potential mechanisms of such discrimination? Does it matter which measure is used to assess such discrimination? Does such discrimination negatively influence health care processes?
3 Discrimination and HealthExtensive evidence that discrimination negatively impacts health Physiological stress responses Cortisol Blood pressure Risk heart disease Behavioral responses Researcher suggest discrimination is a fundamental cause of racial/ethnic health disparities Limited information on discrimination and its impact among American Indian populations a population with some of the worst health disparities reported & experiences of explicit discrimination
4 Discrimination--definition“Differential and negative treatment of individuals because of their membership in a particular demographic group (eg, race, sex, class).” Source: Hausman et al., 2001; Smedley et al., 2003 Discrimination has been defined as the ‘‘process by which a member, or members, of a socially defined group is, or are, treated differently because of his/her/their membership of that group.’’
5 Medical DiscriminationDiscrimination is relevant in medical settings Levels of medical discrimination Medical systems levels-level Patient-level Care-process level Interpersonal experiences between patients and medical providers In recent years, more research examines the contribution of discrimination within health care systems—motivated by trying to understand whether or not discrimination is relevant in racial/ethnic health care experiences and if and how it contributes to persistent health disparities. A report by the Institute of Medicine indicates discrimination occurs at three levels within health care systems---including during the provider-patient interaction—and may harm the health care process, including patients health care decisions, and providers medical decisions and recommendations. According to the research, discrimination in healthcare settings may cultivate patient disengagement from the healthcare system, thereby negatively affecting future healthcare encounters and patient health.
6 Model of Perceived Discrimination in American Indian Health Care; Source Gonzales, K. 2010Patient Attitudes Shaped by Health Care System Factors & Social Determinants Historical/contemporary health care policies Trauma & Grief related to health care provisions Medical Mistrust Medical Access/Resources/Diversity of provider work-force Discrimination Health Care System Factors Provider bias, stereotypes (implicit/explicit) Provider training/mode of care/resources to support providers Patient-Provider Interaction Patient Health Care Decisions/Health Seeking Behaviors Disengagement Delay, forgo care Poor adherence Quality of Medical Care Treatment Recommendations Diagnoses Health Outcomes Unmet health care need Poor health Late stage diagnoses Coping—Cultural Buffers Cultural Identify Spiritual Practice Traditional Practices Source: Walters, K & Simoni, J , AJPHA Whitbeck: 2001 The mechanisms of discrimination in health care are not well known, and the majority of this research describes health care for African Americans. Distinct differences in culture, health policy and history regarding health care provisions may limit the generalizability of these studies for American Indians. For example, unlike other U.S. populations, the provisions of health care to American Indians is administered by the Indian health service, a federal agency that has a legacy of harming, oppressing and violating American Indians—including violations that occurred during the 1970s which involved the forced and unconsented sterilization of American Indian by the Indian Health Service. This legacy has resulted in many AIs having deeply rooted mistrust of health care systems, particularly the IHS---I propose that this history which involves unresolved grief and mistrust shapes contemporary patient-provider interactions--and may be manifested as perceived discrimination in American Indian health care experiences.
7 Perceived Discrimination in Health Care Among American IndiansTo date there are 5-studies available 1-specifically examined perceived provider discrimination Wide-range prevalence estimates: % Mixed associations on use of prevention services Limitations of the current research: Not specific to the Indian health care system Majority used a single-item measures None investigate the relationships between patient’s health seeking behaviors, health outcomes There are five studies available that examine AIs perceived experience of discrimination in their health care experiences. Despite the importance of these studies, there are methodological limitations that need consideration. First, these studies used measures of perceived discrimination that varied in wording, and used a single-item measure that may lead to underestimating the true experience of such discrimination, as well as its influence on outcomes of interests. In addition, none of these published studies have examin how such discrimination negatively influences AI/AN health seeking behaviors regarding adherence; nor did they select samples of American indains who seek medical care mainly from the Indian health system.
8 Specific Aims
9 Research Questions To what extent do American Indian women perceive experience of discrimination during encounters with medical care providers? What is the performance of measures used to assess such discrimination? How does such discrimination influence Individuals health seeking behaviors Use of standard diabetes care services Diabetes control Based on these considerations, I administered a survey and collected medical records data to assess among American Indian women, the following questions:. 1. What is the extent and reasons for perceived discrimination in AI womens health care encounters within the Indian health system? 2. Does it matter if using a single and multi-item to pick up such discrimination in this sample? 3. To what extent does such discrimination negatively affect health seeking behaviors among AI women? 4. What is the association between such discrimination on AI women use of standard diabetes care services and measures for diabetes management?
10 Methods Trail of Tears,
11 Sampling--Tribes 4 Northwest Tribes Purposively SelectedIndian Health Care Facilities IHS and Tribal facilities 4 Northwest Tribes Oregon Washington Purposively Selected Similar characteristics Surveillance Quality Assurance Staff Services In the Spring of 2008, I recruited American Indian women who receive their primary medical care from the Indian health care system from four reservation-based Indian health care facilities from Oregon and Washington. These facilities were selected because of similar criteria in surveillance, quality assurance of diabetes health data, medical staffing and services. For protection of the tribes, I cannot identify which tribes participated—but each tribe reviewed this project and provided approval to participate.
12 Sampling-ParticipantsSelection Criteria Female American Indian Age 18 years and older Resides within the service delivery area of the particular Indian health care facility Has had medical care within past five years at the particular Indian health care facility Documented with at least two codes for diabetes Diagnosed with diabetes one year before study I used the facilities electronic records systems, RPMS to generate a cohort of American Indian women who meet the selection criteria listed here. The selection criteria listed here to generate a sample of AI women, 18 years and older, and who are considered an active patient at one of the participating Indian health care facility. These lists were generated by using each participating tribes electronic medical record system (e.g., RPMS).
13 Data Collection Self-administered survey Medical record dataAbstracted from RPMS Diabetes Standards of Care Diabetes Audit Linked sources of data for analyses Next, after recruiting the women using both passive and targeted approaches, they completed a self-adminstered survey and I abstracted electronic medical records data for the cohort. The survey was pilot tested in one NW tribal community among Native women who met this studies eligibility criteria. In respect of time, I will not detail much of the methods—but, I’d be happy to give more information after this presentation.
14 Respondent CharacteristicsAge distribution 25.9% ages 18-49 42.4% ages 50-64 31.7% ages ≥65 Perceived health status 16.1% very good/better 42.9% good 41.0 fair/poor Source of health insurance 22.9% IHS or tribe only 77.1% IHS and other Duration of diabetes 44.4%, 1-5 years 29.8%, 6-10 years 25.9%, ≥11 years 79.2% (215/270) Response Rate 4.6% (10/215) Surveys Excluded 200 in analyses Here are few of the demographic summaries of the full sample of respondents-- The overall response rate was high at 79.2% 10 surveys excluded due missing survey data. The final sample in the study and used for analysis was next page do not read other percentages!
15 Approvals Institutional Review BoardsOregon State University Portland Area IHS National Area IHS Tribal Resolution of the 43 member tribes of the Northwest Portland Area Indian Health Board Participating tribes I’d like to mention that approvals for this research were obtained from the appropriate entities listed here.
16 Prevalence & Reasons for Perceived Discrimination in Medical EncountersBreath
17 Survey Measures—General PMDWhen you get health care, ANYWHERE, have you EVER felt the following from your medical provider(s), nurse(s) or doctor(s) for any reason? (Answer EACH Item) YES NO a. Discriminated against by your medical provider(s), nurse(s) or doctor(s) 1 2 d. Pre-judged or negatively labeled by your medical provider(s) , nurse(s) or doctor(s) c. Your illness or pain wasn’t believed by your medical provider(s) or doctor(s) A total of 15 measures to assess PMD. These measures were pulled from previously published research, and a broadly categorized as either general or race-based PMD. As shown here, the first PMD instrument asks about individuals specific perceptions about the behaviors of their medical care provider during clinical encounters. Read main stem—The main stem was..”when you get health care anywhere, have you ever felt…discriminated against; treated with disrespect; …
18 Thinking about your EXPERIENCE getting health care, ANYWHERE in general, how often do EACH of the following happen to you because of your RACE? (Answer EACH Item) Never Rarely Sometimes Most of the Time Always a. You are NOT treated courteously 1 2 3 4 5 b. You are NOT treated with respect c. You receive poor medical service d. A doctor or nurse acts as if he or she thinks you are not smart e. A doctor or nurse acts as if he or she is afraid of you f. A doctor or nurse acts as if he or she is better than you g.You feel like a doctor or nurse is not listening to you We used two main measures to assess perceived experiences of discrimination. The first is a multi-item measure developed in African American populations. This scale had high-internal reliability (alpha 0.94) and was determined a one-factor via one-way factor analyses. Good internal-reliability (alpha:0.94) *(Bird et al., 2001, 2003, 2004; Thorburn et al., 2005)
19 Results Breath
20 Item Characteristics of the Seven-item Scale of Perceived Racial Discrimination in Clinical Encounters Scale (n=200) Scale Item Mean =Never (SD) n (%) 2=Rarely n (%) 3=Sometimes 4=Most of the Time 5=Always You are not treated courteously (1.1) (49) 42 (21) 13 (7) 6 (3) You are not treated with respect (2.0) (48) 48 (24) 34 (17) 17 (9) You receive poor medical service (1.1) (48) 45 (23) 43 (22) 14 (7) 3 (2) A doctor or nurse acts as if he or she thinks you are not smart (1.2) (50) 36 (18) 37 (19) 21 (11) 7 (4) A doctor or nurse acts as if he or she is afraid of you (0.83) (64) 20 (10) 1 (.05) A doctor or nurse acts as if he or she is better than you (1.2) (51) 38 (19) 15 (8) 9 (5) You feel like a doctor or nurse is not listening to you (1.3) (40) 46 (23) 24 (12) 12 (6)
21 Published Prevalence among AIsPrevalence of Perceived Discrimination in Health Care Among American Indian Women (n=200) Percentages % Published Prevalence among AIs As you can see each instrament had high internal reliability. And two-thirds of American Indian women in thesample perceived medical discrimination by their medical provider because of their race. As shown, the prevalence of perceived medical discrimination documented in this particular study greatly exceeds the estimates previously published for AIs by as much as 6 times higher.
22 Measurement ConcordanceSingle-Item/Multi-Item Response Categories Case Concordance n (%) X2 Yes/Yes 81 (41) P<.001 No/Yes 52 (26) Yes/No 5 (3) No/No 62 (31)
23 Reasons for Perceived Discrimination in Health Care Among American Indian Women, Beyond Race (n=111)Note: Items not mutually exclusive. Percentages % In addition to race, 44.2% of the women indicated income and weight as the most common reasons for the discrimination. I was surprised to find that personal or a family member(s) prior or current use of drugs and alcohol was given as main reasons for the race-PMD. I should mention, these two items originate from our pilot testing feedback and to our knowledge have not been administered in the published research. This finding is compelling and implies that social stereotypes about Native populations may be present in AI health care experiences and may contribute to perceptions of medical discrimination and AI womens health care decisions.
24 Negative Health Seeking & Perceived Discrimination in Health Care Among American Indian Women. (n=111) Percentages % Note: Items not mutually exclusive. 82.9% of the women who reported PPD reported their health seeking behaviors were negatively affected. About two-thirds of women reported that perceived experiences of provider discrimination resulted in them postponing needed medical care. And, over half with affirmative PPD reported such discrimination resulted in them being more hesitant about their medical care.
25 I generated linear regression models, to examine the number of negative health seeking behavior items according to mean levels of PPD—which were low, moderate and high. Results show, a monotonic trend where negative health seeking behavior significantly increased as experience of PPD increased. For example, women with the highest amount of PPD report on average report 4.60 negative health seeking behaviors. Adjusted for: demographics, duration of diabetes, health insurance coverage, perceived health status.
26 Logistic Regression Model for Perceived Provider Discrimination and Use of Health Care ServicesNot Current OR (95% CI) Model 1: Clinical Breast Exam Model 2: Pap Testing Model 3: Mammography Reference group: current status Adjusted: age, perceived health status, duration of diabetes, source of health insurance Pap test & Mammography, limited to those eligible for screening based on national screening criteria.
27 Logistic Regression Model for Perceived Provider Discrimination and Diabetes ControlOptimum Care* OR (95% CI) ≥7 services: (ref) ≤6 services: (1.05, .12) *Based on mean estimate of 12 standard services for diabetes care Adjusted: age, perceived health status, duration of diabetes, source of health insurance Source of Data: electronic medical records
28 Logistic Regression Models for Perceived Provider Discrimination and Use of Standard Services for Diabetes Care Diabetes Control* OR (95% CI) Hemoglobin A1C ≤7.0% (ref) Hemoglobin A1C >7.0% (0.99, 1.82), p=.059 *Based on mean estimate of last 2 blood glucose measures one year prior to survey-data collection Adjusted: age, perceived health status, duration of diabetes, source of health insurance Source of Data: electronic medical records
29 Discussion & ImplicationsCoast Salish King Killer Whale Print by Floyd Joseph Breath
30 Conclusions American Indians perceive experiences of discrimination in their health care: During encounters with their medical care provider’s Such discrimination appears to influence process of care, including patients health care decisions Such discrimination may be related to unmet health care need among American Indian populations However, despite these limitations, this research provides new information that suggests race-PMD may be an important barrier to optimum diabetes care and result in unmet health care need among AI women—that may help to further explain reasons for the persistent mortality and morbidity disparities from diabetes in AI populaitons.
31 Next Steps in Research Interventions—will they work & howThe mechanisms by which perceived discrimination impacts health care processes & health for American Indians are not known Providers role—stereotypes, clinical decisions Patients role—meaning of discrimination, what contributes to meaning, responses Health care characteristics—resources/lack of resources & contributions Interventions—will they work & how Cultural competency Patient-centered care Other approaches?
32 Kelly Gonzales ([email protected])Mt. Shasta, Northern California ACKNOWLEDGEMENTS PhD Thesis Committee Anna Harding, PhD Oregon State University (OSU), Public Health, Major Professor Mathew Carlson, PhD Portland State University, Sociology Joe Finkbonner, RPh, MHA Executive Director, Northwest Portland Area Indian Health Board Brian Flay, PhD OSU, Public Health Bill Lambert, PhD Oregon Health Sciences University, Public Health Jeff McCubbin, PhD OSU, Nutrition and Exercise Science, Graduate Representative Husband, Children, Family, Friends, AI Women Participants, NW tribes, POA-IRB Kelly Gonzales I appreciate your attendance today, and I hope you have enjoyed this presentation. Thank you for your time. I would be happy to answer any of your questions.