Use of Race in Clinical Decision Making

CLINICAL SCHOLARSHIP Nurses’ Use of Race in Clinical Decision Making Sherrill L. Sellers, PhD1 , Melissa E. Moss, BA2 , Kathleen Calzone, PhD, RN, APNG, FAAN3 , Khadijah E. Abdallah, MPH4 , Jean F. Jenkins, PhD, RN, FAAN5 , & Vence L. Bonham, JD6 1 ∗ Associate Dean for Undergraduate Education and Professor, Department of Family Studies and Social Work, Miami University, Oxford, OH, USA 2 ∗ Postbaccalaureate Intramural Research Training Award (IRTA) Fellow, Health Disparities Unit, Social and Behavioral Research Branch, Division of Intramural Research, National Human Genome Research Institute, National Institutes of Health, Bethesda, MD, USA 3 Senior Nurse Specialist, Research, National Cancer Institute, Center for Cancer Research, Genetics Branch, National Institutes of Health, Bethesda, MD, USA 4 Research Analyst, Health Disparities Unit, Social and Behavioral Research Branch, Division of Intramural Research, National Human Genome Research Institute, National Institutes of Health, Bethesda, MD, USA 5 Clinical Advisor, Office of the Director, National Human Genome Research Institute, National Institutes of Health, Bethesda, MD, USA 6 Associate Investigator, Health Disparities Unit, Social and Behavioral Research Branch, Division of Intramural Research, National Human Genome Research Institute, National Institutes of Health, Bethesda, MD, USA Key words RACE (Racial Attributes in Clinical Evaluation) Measure, clinical use of race, clinical decision making, nurses Correspondence Dr. Sherrill L. Sellers, Miami University, Department of Family Studies & Social Work, 207 McGuffey Hall, Oxford, OH. E-mail: sellersl@miamioh.edu. The opinions expressed are the authors’ own and do not reflect the policies or positions of the National Institutes of Health or the U.S. Department of Health and Human Services. ∗ asterisk indicates co-first authorship. Accepted August 2, 2016 doi: 10.1111/jnu.12251 Journal of Nursing Scholarship, 2016; 48:6, 577–586.  C 2016 Sigma Theta Tau International Abstract Purpose: To examine nurses’ self-reported use of race in clinical evaluation. Design: This cross-sectional study analyzed data collected from three separate studies using the Genetics and Genomics in Nursing Practice Survey, which includes items about use of race and genomic information in nursing practice. The Racial Attributes in Clinical Evaluation (RACE) scale was used to measure explicit clinical use of race among nurses from across the United States. Methods: Multivariate regression analysis was used to examine associations between RACE score and individual-level characteristics and beliefs in 5,733 registered nurses. Findings: Analysis revealed significant relationships between RACE score and nurses’ race and ethnicity, educational level, and views on the clinical importance of patient demographic characteristics. Asian nurses reported RACE scores 1.41 points higher than White nurses (p < .001), and Black nurses reported RACE scores 0.55 points higher than White nurses (p < .05). Compared to diploma-level nurses, the baccalaureate-level nurses reported 0.69 points higher RACE scores (p < .05), master’s-level nurses reported 1.63 points higher RACE scores (p < .001), and doctorate-level nurses reported 1.77 points higher RACE scores (p < .01). In terms of clinical importance of patient characteristics, patient race and ethnicity corresponded to a 0.54-point increase in RACE score (p < .001), patient genes to a 0.21-point increase in RACE score (p < .001), patient family history to a 0.15-point increase in RACE score (p < .01), and patient age to a 0.19-point increase in RACE score (p < .001). Conclusions: Higher reported use of race among minority nurses may be due, in part, to differential levels of racial self-awareness. A relatively linear positive relationship between level of nursing degree nursing education and use of race suggests that a stronger foundation of knowledge about genetic ancestry, population genetics and the concept “race” and genetic ancestry may increase in clinical decision making could allow nurses to more appropriately use of race in clinical care. Integrating patient demographic characteristics into clinical decisions is an important component of nursing practice. Clinical Relevance: Registered nurses provide care for diverse racial and ethnic patient populations and stand on the front line of clinical care, making them essential for reducing racial and ethnic disparities in healthcare delivery. Exploring registered nurses’ individual-level characteristics and clinical use of 577 Nurses’ Clinical Use of Race Sellers et al. race may provide a more comprehensive understanding of specific training needs and inform nursing education and practice. More than 4 million healthcare professionals in the United States are nurses, making them the front line of health provision (National Council of State Boards of Nursing, 2016). Driven by the Patient Protection and Affordable Care Act of 2010, the discipline of nursing has entered a new era of clinical practice. The profession has been called upon to increase the number of baccalaureate-prepared and doctorate-prepared nurses, appropriately inform and educate the next generation of nurses, and aid in closing the health disparities gap (Fairman, Rowe, Hassmiller, & Shalala, 2011; Hassmiller, 2010;

Institute of Medicine, 2010; Levin & Bateman, 2012). The growing utilization of nurses as primary care providers highlights their important role in bringing a precision medicine approach to health care (Calzone, Jenkins, Nicol, et al., 2013; Cheek, Bashore, & Brazeau, 2015).

Although precision medicine may allow a more accurate approach to patient care that moves beyond race, the complicated relationship between race and genetic ancestry continues to stir an ongoing debate around the clinical utility of race (Dankwa-Mullan, Bull, & Sy, 2015). Race is a fluid concept used to group people according to various factors, including ancestral background and social identity (Smedley & Smedley, 2012). Because race is a crude proxy for certain underlying genetic risk, it remains a commonly used indicator in disease prevention, screening, and treatment strategies. Differential Health Treatment and Outcomes by Race and Ethnicity Persistent health disparities are apparent in the variation of disease incidence and mortality across racial and ethnic populations (Badve et al., 2011; Kaiser Family Foundation, 2015; U.S. Cancer Statistics Working Group, 2016). For example, complex diseases such as cardiovascular disease, type 2 diabetes, and prostate and colorectal cancer have disproportionally affected certain racial and ethnic minority populations (Centers for Disease Control and Prevention-National Center for Health Statistics, 2015). A debate has risen around whether race-based screening guidelines are needed to address these disparities and to what extent race has clinical utility, particularly as a proxy for genetic ancestry. One facet of the debate involves common practice and healthcare guidelines that specifically include the use of race as a proxy for ancestry, genetic risk, and response 578 in diagnostic and treatment decisions.

For example, a recently approved test for the Lp-PLA2 biological marker to predict risk for coronary heart disease is reported by the U.S. Food and Drug Administration (2014) to predict risk better in Black women. Additionally, there is a long history of using race in clinical decisions about the most effective type of drugs to administer (Ramamoorthy, Pacanowski, Bull, & Zhang, 2015). Given the current treatment guidelines and ongoing contention around the role of race in clinical practice, scientific discoveries that are beginning to illuminate the contributions of genomic variation and environmental factors to health outcomes for persons with complex chronic diseases hold the promise of guiding development of effective health interventions. Nurses’ Clinical Use of Race As research continues to clarify the contribution of social and genetic factors to racial and ethnic differences in health, disease, intervention choices, and outcomes, it is necessary to understand how healthcare providers use (i.e., collect, perceive, and interpret) race in public health practice and clinical care. Research indicates that individual background characteristics, personal beliefs, and biases influence the clinical encounter, often to the disadvantage of minority patients (Lawrence, Rasinski, Yoon, & Curlin, 2014; McKinlay, Piccolo, & Marceau, 2013; Sabin, Nosek, Greenwald, & Rivara, 2009). Research also indicates that patients receive differential treatment by race and may respond differently to treatment based on genomic profile differences (Keenan et al., 2015; Wandner et al., 2014). There remains a dearth of literature focusing on nurses in the context of race, interpretation of genetic risk and response, and healthcare. A few studies have explored nurses’ understanding and use of genomics (Calzone, Jenkins, Culp, Bonham, & Badzek, 2013; Coleman et al., 2014), while others have focused on race and revealed low levels of cultural competency with little awareness of personal biases in nurses (Haider et al., 2015; Hirsh, Jensen, & Robinson, 2010).

However, huge knowledge gaps remain about nurses’ behaviors and beliefs concerning race in clinical care. Given the importance of nursing practice and increased responsibilities of nurses in healthcare provision, our primary aim is to better understand nurses’ clinical use of race, which can affect clinical outcomes. Specifically, although patient race Journal of Nursing Scholarship, 2016; 48:6, 577–586.  C 2016 Sigma Theta Tau International Nurses’ Clinical Use of Race Sellers et al. along with other information can help guide diagnostic and treatment decisions in certain circumstances, it can also be used inappropriately, which could adversely influence patient care (Nelson, 2002). Methods Eligibility Study eligibility consisted of registered nurses (RNs) at all levels of academic preparation, role, or specialty employed at participating institutions. Licensed practical nurses and licensed vocational nurses were excluded. This was the only exclusion criterion for study participation. Study Sample The study sample includes 5,733 RNs across three separately recruited groups. The largest group is comR -designated posed of nurses employed by Magnet hospitals participating in the Method to Integrate a New Competency (MINC) study (Calzone, Jenkins, Culp, Caskey, & Badzek, 2014).

MINC included nurses from 23 American Nurses Credentialing Center–designated R hospitals located in Magnet Recognition Program 17 U.S. states. The second group was recruited by member associations of the National Coalition of Ethnic Minority Nurse Associations (NCEMNA), a national collaboration of five ethnic minority nurse associations. Four NCEMNA associations participated in this study: Asian American/Pacific Islander Nurses Association, National Association of Hispanic Nurses, National Black Nurses Association, and Philippine Nurses Association of America (Coleman et al., 2014). The third group is a sample of nurses recruited by the American Nurses Association (ANA; Calzone, Jenkins, Culp, et al., 2013). Instrument Data were collected in all three groups using online survey methods and convenience sampling. The survey instrument, called the Genetics and Genomics in Nursing Practice Survey (GGNPS), was developed from a tool for family practice physicians, scaled for nursing and pilot tested with a sample of RNs (Calzone et al., 2012). Instrument validation was performed using structural equation modeling (SEM), which found that the instrument items aligned well with the Rogers Diffusion of Innovations (DOI) domains (Jenkins, Steven, Kahn, & McBride, 2010; Rogers, 2003).

The items used in the present study were background demographic characteristics of the nurses and the Bonham & Sellers Racial Attributes in Clinical Evaluation (RACE) scale. Journal of Nursing Scholarship, 2016; 48:6, 577–586.  C 2016 Sigma Theta Tau International The RACE scale was adapted for nurses from an eight-item RACE scale for physicians designed to measure self-reported use of patient race in clinical decision making (Bonham, Sellers, & Woolford, 2014). Bonham et al. (2014) developed the RACE scale as a tool to measure healthcare providers’ use of race in clinical care, without applying any value-laden assessment of the benefit or harm of its use. Higher RACE scores indicate greater self-reported use of race in clinical care. Using a portion of the total nurses’ responses (n = 577) for confirmatory factor analysis (CFA) of the 8 items, a final five-item version of the scale was developed for evaluation of nurses’ clinical use of race (Figure 1). The five-item RACE scale had a Cronbach’s alpha of .86, the loadings were all positive and statistically significant, and other measures (GFI = .99, AGFI = .95, CFI = .99, RMSEA = .08) all indicated adequate fit. Data Collection The GGNPS was voluntary, took 15 to 20 min, and did not collect any participant personal identifiers. Recruitment strategies varied across the three participating groups. MINC institutions sent email announcements about the survey to their nursing staff. Some institutions employed recruitment methods such as advertising, supervisor encouragement, and intranet postings. Data were collected between July and October 2012, and the survey was open for completion 28 days at each institution (Calzone et al., 2014).

Although there was no standard protocol for providing compensation, incentives were offered by some institutions when approved by their individual institutional review boards (IRBs). Recruitment of NCEMNA study participants was done by each participating NCEMNA member association through email announcements and newsletter notifications (Coleman et al., 2014). The survey was open for completion from fall 2010 to early 2011, and a link to the survey was posted on the main NCEMNA website and each participating NCEMNA member association website. No compensation was provided for participation. The ANA recruited study participants by posting announcements on the main ANA website (“nursingworld”) and in The American Nurse, the official publication of the ANA (Calzone, Jenkins, Culp, et al., 2013). Additionally, study announcements were sent out to ANA SmartBrief and eNewsletter subscribers. The survey was open for completion from fall 2009 to early 2010. No compensation was provided for participation. Survey development was approved by the National Human Genome Research Institute IRB (05-HG-N196). The IRB at West Virginia University approved the MINC 579 Nurses’ Clinical Use of Race Sellers et al. Likert Scale: 0 1 2 3 4 1. I consider information from patients about their racial background. 2. I consider my patients’ race to better understand their genetic predispositions. 3. In my practice, I consider my patients’ race when administering medications. 4. I consider my patients’ race in determining genetic risk for common, complex diseases (e.g., kidney disease or diabetes). 5. In my practice, I consider my patients’ race when checking medication dosages as prescribed. Figure 1. Racial Attributes in Clinical Evaluation (RACE) Scale for Nurses, with score calculated as sum of responses (0–20), with higher scores signifying greater use of race in clinical practice. α = .86. Adapted from Bonham et al.’s (2014) RACE scale for physicians. (H-23491) survey, and the Cedars Sinai IRB approved the NCEMNA survey (PRO00018344). The National Institutes of Health Office of Human Subjects Research Protections considered the MINC (OHSRP#11366), ANA (OHSRP#4891), and NCEMA (OHSRP#4570) surveys to be exempt pursuant to 45 CFR 46 because respondents were anonymous and there were no risks to participants. Study Variables The outcome variable, use of race in clinical decision making, is measured using a five-item RACE scale (see Figure 1).Use of Race in Clinical Decision Making

Items in this 0- to 20-point scale gauge level of nurses’ clinical use of race using a 0–4 rating scale, with 0 as “none of the time” and 4 as “all of the time.” This measure was adapted from the RACE scale for physicians in order to measure clinical use of race by nurses (Bonham et al., 2014). As with the original scale, higher RACE scores indicate more self-reported use of race in clinical decision making; it measures the extent of clinical use of race without placing a positive or negative label on the phenomenon itself. Predictor variables included age, portion of work time spent with patients, nursing education level, primary nursing role, race and ethnicity, and views on clinical importance of certain patient demographics. Age was measured by subtracting participant birth year from the year in which the survey was completed by the participant. Portion of work time spent with patients was presented as a percentage-based write-in question. Nursing education level, determined by highest nursing degree attained, was measured with the following categories organized low to high: diploma nurse, associate’s, baccalaureate, master’s, and doctorate degree. Primary nursing role was broken down into five categories: patient care, administration, education, research, and other (including students). Selfidentified race and ethnicity categories included American Indian or Alaska Native, Asian, Black or African 580 American, Native Hawaiian or Pacific Islander, Hispanic or Latino, and White. However, because there were only 26 American Indian or Alaska Native participants and 39 Native Hawaiian or Pacific Islander participants, these two racial and ethnic groups were combined into one category called “Indigenous Populations.” Additionally, the racial and ethnic categories were simplified and combined into one variable with mutually exclusive categories, such that participants who identified as Hispanic or Latino were placed in that category and the other five categories are non-Hispanic in this study. The variables for opinions on clinical importance of certain patient characteristics are gender, race and ethnicity, genes, family history, and age, measured on a 1–7 Likert scale, with 1 meaning “not at all important” and 7 meaning “essential.” Survey group was included as a covariate, with MINC, NCEMNA, and ANA respondents as three categorical groups. Use of Race in Clinical Decision Making

 This variable was created specifically to show and account for between-group differences of the three nursing study subgroups. Data Analysis Respondents who were not RNs and those who did not answer all five RACE scale items (and therefore could not receive a valid RACE score) were excluded from data analysis (35% of respondents dropped from analysis). Before dropping these data from the final analysis, assessment showed no pattern of the missing RACE scale items, so 35% was deemed an acceptable amount of exclusion. Once cases that met the exclusion criteria were removed from the dataset, analysis was performed using R R SPSS Statistics 19.0 (SPSS, Inc., Armonk, NY, IBM USA). After basic evaluation of descriptive frequencies and means, the relationships between variables were assessed with bivariate analyses, namely correlation and one-way analysis of variance (ANOVA). The variables Journal of Nursing Scholarship, 2016; 48:6, 577–586.  C 2016 Sigma Theta Tau International Nurses’ Clinical Use of Race Sellers et al. were then placed in an ordinary least squares (OLS) regression model for multivariate analysis. Results Study Sample Description Table 1 presents the characteristics of the 5,733 nurses in our sample. The average age in our study sample was 44 years, which is slightly younger than the national average of 50 years (Budden, Moulton, & Cimiotti, 2013). Nurses reported spending an average of 72% of their work ti …Use of Race in Clinical Decision Making