Bruce Siegel, director of the Robert Wood Johnson Foundation’s (RWJF) Aligning Forces for Quality (AF4Q) initiative and the RWJF legacy program, Expecting Success: Excellence in Cardiac Care (Expecting Success), writes about collecting race, ethnicity and language data.
In the medical profession, we diagnose problems before we attempt to treat them. It shouldn’t be any different when we try to fix our health care system.
As Congress focuses on expanding health coverage and reducing cost and waste in the health care system, an important question is: how will we make health care not only more efficient, but better for patients and families from all racial and ethnic backgrounds?
It starts with working with the best possible data. We need to know who’s not receiving high quality care so we can target our efforts. Reams of research have shown that racial and ethnic disparities persist despite efforts to reduce them. To address these gaps in care, hospitals, medical practices and health plan members need precise and standardized data on patients’ race, ethnicity and primary language.
Reducing disparities and enhancing data collection have been discussed in many of the health reform proposals in Congress. If included in the final legislation and as health care systems continue to work to close the gap, guidance and best practices on how to do this will be needed.
The Office of Management and Budget (OMB) has a standard set of race and Hispanic ethnicity categories that are widely used (race categories: Black or African American, White, Asian, American Indian or Alaska Native, and Native Hawaiian or Other Pacific Is¬lander; ethnicity categories: yes or no in reference to Hispanic/Latino ethnicity), but that is not enough.
A recent report by the Institute of Medicine (IOM) recommends that health care organizations continue to use the OMB categories and add more detailed ethnicity categories (known as granular ethnicity and based on a patient’s ancestry), which will help organizations better identify disparities and move away from broad categorizations of people. For example, in very diverse communities such as Miami, “black” can mean a lot of things and include many cultures, making more specific categories such as Haitian or Bahamian useful. The IOM also recommends using categories to assess language needs (ratings of spoken English language pro¬ficiency and a patient’s preferred language for health-related encoun¬ters). Understanding a patient’s language needs will increase the ability to communicate in medical settings, which is critical to providing and receiving high-quality care. The report suggests that organizations list location-relevant categories and provide a space for patients to self-identify their ethnicity or language if it is not listed.
With better information, health care organizations can better understand the populations they serve, detect disparities in care, design actual solutions to improve care and evaluate progress. Through our work with two RWJF programs, Expecting Success and Speaking Together: National Language Services Network (Speaking Together), we have done just that. We have found that more detailed data enables us to better develop targeted quality improvement interventions for more specific populations.
As part of the Expecting Success program, we worked with 10 hospitals to develop and share tools for improving cardiac care for African-American and Hispanic patients with acute myocardial infarction or congestive heart failure. Hospital leaders want to believe that their hospitals provide equal care regardless of a patient’s race, ethnicity or primary language, but few know for sure. Without uniform standards for collecting this information, there is no way of knowing if all patients receive the same level of care.
The Expecting Success hospitals established standardized collection of patient race, ethnicity and language data. For the first time ever, the hospitals analyzed 23 cardiac care quality indicators by patient race, ethnicity and language. Although some had to face the reality that there were disparities in care in their hospitals, they were better equipped to address these gaps.
Several questions arose as the hospitals analyzed their race, ethnicity and language data such as “Why are some Hispanic patients consistently not receiving all discharge instructions?” and “Why are readmission rates so much higher for minority patients?”. These questions prompted the hospitals to design interventions to specifically address these issues. As they developed these and other programs, they were able to compare data on core measures before and after implementation to help assess their efficacy and adapt their approach as needed.
In Speaking Together, an initiative modeled after the Expecting Success program, we worked with 10 hospitals to improve the quality and availability of language services. For many patients whose first language is not English, language services are integral to getting the right care at the right time. An organization’s ability to provide appropriate language services depends on its ability to accurately screen for language needs of its patients and to identify patients’ preferred language for health care encounters—data collection is fundamental. Without this information, we are trying to diagnose and treat a problem with a limited exchange of information.
At the beginning of Speaking Together, several hospitals were collecting language data, but there was room for improvement. To increase screening for patient language needs, hospitals used a combination of efforts, including using data to open a discussion with the leaders of registration and scheduling; training staff on screening for language needs; programming reminders in the registration and scheduling screens to prompt staff to complete the language field; using scripts for language screening; and integrating demographic information with other electronic systems in the organization. At one hospital, screening rates went from 60 percent to more than 80 percent. At another hospital, screening rates improved from approximately half of patients screened to nearly all patients screened. The data collected through these screening efforts enabled the hospitals to more accurately identify which languages were spoken by their patients, improve their overall language services and deliver safer, higher quality care.
In both of these projects, the ability to reduce disparities began with the knowledge that we gained through data collection. As we work to reform our health system to improve care for all, we need to ensure that our health care organizations are actively collecting information on our patients’ race, Hispanic and granular ethnicity, and language needs. We know the symptoms of poor-quality, unequal care, but to truly diagnose it and treat it, we need the data.