Community-Level Interventions to Collect Race/Ethnicity and Language Data to Reduce Disparities
Published Online: September 22, 2012
Romana Hasnain-Wynia, PhD; Deidre M. Weber, BA; Julie C. Yonek, MPH; Javiera Pumarino, BA; and Jessica N. Mittler, PhD
The systematic collection and use of race/ethnicity and language (REL) data by healthcare organizations has long been recognized as a critical step in identifying and reducing healthcare disparities locally and nationally.1-3 The American Recovery and Reinvestment Act of 2009 urges the “use of electronic systems to ensure the comprehensive collection of patient demographic data, including, at a minimum, REL and gender information.”4 The 2010 Affordable Care Act has explicit expectations regarding the standardized collection and reporting of self-reported REL data. Private accreditation organizations such as the Joint Commission and the National Committee for Quality Assurance have developed standards regarding the collection of REL data by hospitals and health plans.5 Despite widespread consensus that these data are necessary for identifying, monitoring, and reducing disparities in care,6 few healthcare organizations systematically collect this information.
Kilbourne et al recommended a framework for addressing disparities in the healthcare system, which included 3 broad phases: “The first requires detection; the second, understanding; and the third involves the development, implementation, and evaluation of interventions that reduce or eliminate disparities.”7 Collecting REL data is largely about detection. In this regard, evidence of pervasive healthcare disparities is readily available through national surveys, reports, and peer-reviewed literature.1,3,8 However, among healthcare providers and leaders, including even those familiar with the national data pointing to disparities in care, there is a tendency to believe that disparities exist “somewhere out there” but not within their own organizations.9,10 One reason such views can persist is the frequent lack of data to demonstrate or disprove the existence of disparities at the local, organizational level. Organizations need to collect data to detect potential disparities within their communities and be accountable for addressing this pervasive problem.
A substantial body of work has examined the barriers and facilitators to collecting REL data in individual healthcare organizations such as hospitals, physician practices, and health plans.2,3,11-13 This paper examines a set of community-level interventions for collecting REL data. The Aligning Forces for Quality (AF4Q) initiative, a national program of the Robert Wood Johnson Foundation (RWJF), is designed to help targeted communities improve the overall quality of healthcare including reductions in racial and ethnic health disparities.14 It provides an unprecedented opportunity to examine a range of community-level strategies to institute standardized REL data collection to detect disparities in local health system performance. The multi-stakeholder AF4Q alliances (the generic term used for the multi-stakeholder partnership in each community) are focused on propelling communitywide efforts to build the local infrastructure to collect patient demographic data for disparities detection and reduction activities. As federal and state mandates and accreditation standards begin to require the collection of REL data, the experiences of these AF4Q alliances might provide important lessons.
Even under ideal circumstances, meeting the multiple AF4Q programmatic goals would be daunting. Each of the AF4Q alliances comprises volunteer partner organizations and individuals, bringing distinct perspectives and sometimes competing interests to the table,11 which can present opportunities and challenges in accomplishing goals. Given resource limitations and competing demands within these voluntary alliances, the challenges and opportunities they have faced in collecting REL data might presage experiences soon to arise in other communities nationwide. This article describes why collecting these data has remained difficult despite the fact that many alliance leaders understand that this information forms the foundation for detecting and reducing disparities in care.
This paper examines the experience of 14 AF4Q alliances (data from additional alliances that joined the AF4Q initiative in 2009-2010 are not included in this analysis). This study’s target population included AF4Q alliance leaders, project directors, and disparities/equity staff leads; we selected these individuals based upon their overall roles in their alliance, and because they had broad and specific knowledge about challenges and strategies to reduce disparities in their communities. The study protocol was approved by the Office of Research Protections at Pennsylvania State University.
Two senior researchers from the AF4Q evaluation team conducted 1-hour, face-to-face, semi-structured interviews with key informants during 2-day visits to each of the 14 AF4Q communities in 2010. Protocol questions covered a wide range of topics, including questions about disparities-related activities and REL data collection. All informants were asked a set of key questions. The interview protocol was also tailored to ensure that we captured the most relevant information from the best sources.
Alliance visit data were supplemented with data collected from the AF4Q initiative project directors in the summer and winter of 2011 about the alliances’ progress on disparities-related activities; senior researchers collected these data through regular, 6-month, 90-minute, semi-structured telephone interviews to provide ongoing perspective on the alliances’ activities. All interviews were digitally recorded after consent was obtained.
Data were analyzed using the ATLAS.ti (Version 2011;Cologne, Germany) qualitative analysis software program, which assists with organizing and analyzing large quantities of data through creation of analytic units, which contain all the documents, quotations, codes, and associated files for the study. We used deductive and inductive qualitative methods to analyze the key informant interview data.12,13,15 First, data were transcribed from audio recordings and saved in standard word processing document files. The text files were then read and coded using a codebook that was developed a priori, based on theoretical concepts from the extant literature. This codebook served as a tool for organizing segments of similar or related text and provided a framework for subsequent analysis. The coded texts were entered into ATLAS.ti. Code-based queries were used to retrieve responses regarding disparities-reduction activities and race and ethnicity data collection from target subjects. We then reviewed these transcripts to distill emerging themes. Three members of the research team developed a list of themes that focused on strategies that alliances used to collect race and ethnicity data in the community, barriers and facilitators to collecting data, and other disparities-related activities and discussions. The authors jointly extracted the themes discussed by the interviewees using a deductive and inductive analytical process and resolved discrepancies through discussion to attain group consensus.13
We conducted a total of 51 face-to-face interviews across 14 AF4Q alliances with 22 project directors and 29 disparities/equity staff leads. We conducted a total of 28 follow-up telephone interviews with 14 AF4Q project directors (2 rounds of telephone interviews).
Results: Study Cohort
We identified 6 emergent themes. To begin, we present results that describe alliance leaders’ and staffs’ perceptions of alliance members’ sentiments about the AF4Q initiative’s focus on reducing REL disparities, starting with data collection. The first theme is quite broad, and provides a backdrop to interpret the remaining themes, which highlight several more specific challenges and opportunities.
Theme 1: Disparities Are Difficult to Tackle
Overall, disparities in care were not a top-of-mind issue for the majority of AF4Q alliance leaders, given the range of demands of the AF4Q initiative and other competing priorities. Even when alliance directors and key staff acknowledged that disparities in care were a recognized problem within their community, they were unsure how to start tackling them. Key informants from every AF4Q alliance expressed trepidation about how to engage community leaders and partners on the topic of disparities reduction:
“[In] the AF4Q initiative, the focus on disparities is probably the trickiest for us because we don’t have a lot of experience in this area.… Even for our leadership team to have some frank discussions around what type of disparities are occurring [in our community] is challenging because we don’t necessarily feel comfortable talking about inequities in terms of the care that’s being delivered.”
“….There are elements of the AF4Q initiative, like the focus on disparities, that our board (alliance) hasn’t addressed. We will need a level of commitment from their organizations [individual providers] if we’re really going to collect REL data.”
Some alliance directors and staff were enthusiastic about addressing disparities despite challenges. They saw the logic of beginning with one dimension of disparities in care (REL)before addressing other aspects (socioeconomic status or geography) and viewed systematic data collection as a logical entry point. An informant captured these sentiments:
“We (alliance) had a visioning session and started meeting after that to discuss disparities. It is hard, because it is such a large issue. Where do you start in equity work? We agreed that we would start with REL and get grounded in what is happening in our community, which will then help us in other disparities work.”
Theme 2: REL Data Collection Is a Hard Sell
Many alliance directors and staff found the focus on REL data collection misguided; most of this group came from communities where there was minimal REL diversity. Representative comments included:
“The biggest challenge that we confront in the (alliance) is the small number of members that we have in vulnerable populations, if you’re looking at REL as a defining feature. In terms of socioeconomic status, there’s certainly a fairly high level of poverty in the state, but we’re 96% Caucasian.”
“People are thinking, racial and ethnic disparities in our community? Really? Urban and rural disparities, yes, but racial disparities? You’re kidding me. We have to work really thoughtfully to raise awareness.”
These alliances were concerned about getting buy-in for data collection activities from key community partners. Some racially homogeneous communities related that the only reason they were attempting to collect REL data was that it was a requirement for continued participation in the AF4Q initiative. These alliance directors and staff said that their communities had more pressing issues, such as health differences by socioeconomic status. They suggested that REL data collection was an ineffective use of resources and unintentionally stymied the alliance’s ability to achieve its other quality improvement goals by undermining the social capital built with the member organizations of the alliance.
For alliances with racially and ethnically diverse populations, targeting disparities was generally supported by their partners and the broader community, but getting buy-in to specifically get providers to collect REL data as a starting point was still often a hard sell to community organizations representing minorities. In part, these community organizations saw REL data collection as myopic or misguided, since they felt REL disparities in health outcomes stemmed from sources other than differences in physician and hospital care. As an alliance disparities staffer summarized:
“We’re tending to focus on the provider end, but there is a whole community component that’s going on around disparities and we need to look at that. If we started to collect REL data without the larger community understanding why—the leaders of the black community, the church leaders, the politicians—that could be misunderstood. We have to lay groundwork of ‘disparities exist and there are some things we need to do (ie, REL data collection) to get a handle on them.’”
Some AF4Q alliances were able to build constructive partnerships with community organizations on the topic of disparities, and then use this to leverage buy-in on REL data collection activities:
PDF is available on the last page.