Listening To The Voice Of All Patients To Help Heal Health Disparities In A Post-COVID-19 World

One year ago this March in a Health Affairs Blog post, leaders of the Patient Experience Policy Forum (PXPF) published a vision for taking patient experience measurement and reporting to a new level. They outlined major improvements needed to modernize national patient experience surveys—to improve survey content, administration, and analysis to capture current, policy-relevant information on patient needs and priorities—and to democratize them, by making the data more accessible and understandable to all health care stakeholders.

During the intervening months, the devastating COVID-19 public health crisis has exposed many strengths, as well as shortcomings, of the nation’s health care system. The alarming toll that COVID-19 has disproportionally taken on Americans of color highlights even more starkly the persistent disparities in the provision of health care in the US.

As representatives of a large and diverse PXPF workgroup focused on measurement improvements, we believe that the fundamental principles articulated a year ago to modernize and democratize patient experience measurement and reporting can and must be rigorously applied to help eliminate racial and ethnic disparities in health care overall and in patient experience in particular. This is a call to action long overdue—and requires the engagement of health care organizations, regulators, policy makers, and all survey sponsors and vendors.

Scope Of Disparities

Disparities in health care quality for underserved populations have been well documented in recent years, prompting increased efforts to close these gaps. These disparities are reflected directly in the opinions of health care consumers as well. In a recent The Beryl Institute—Ipsos PX Pulse survey, US consumers reported very different perceptions of the quality of care they receive based on their racial/ethnic backgrounds. While a total of 4 percent of White people report that they “often” or “sometimes” feel discrimination in health care, 35 percent of Black people and 24 percent of Hispanic/Latino people report this experience. These are all realities we must both acknowledge and act on in our efforts to tackle disparities and inequities in health care.

Exhibit 1: Patient experience of race/ethnicity-based discrimination or prejudice in health care encounters

Source: The Beryl Institute—Ipsos PX Pulse: Consumer Perspectives on Patient Experience in the US, July 2020.

Recommendations For Addressing Racial And Ethnic Disparities In Patient Experience Through Improved Measurement And Reporting

The March 2020 blog post outlined several recommendations for improving the measurement and reporting of patient experience. We revisit and add to these recommendations with a specific focus on how they can help address racial and ethnic disparities in patient experience.

1. Continue To Update Survey Content To Reflect What Is Important To All Patients

Patient reports of experiences and outcomes of care are an important source of information to provide direction for addressing health and health care inequities across population groups. We must ensure that all racial, ethnic, age, and gender groups are well represented in all patient-centered measurement efforts (for example, through patient focus groups and stakeholder input) and that their input, which reflects what is important to them in their health care interactions, is used in developing and testing patient experience survey questions.

We also urge inclusion of specific concepts and questions related to racism and ethnic bias in care experiences. For example, in Canada, the Office of Patient-Centered Measurement within the British Columbia Ministry of Health now asks such questions as, “Did you feel you were treated unfairly due to your race or cultural background?” in standardized provincial surveys. Questions that explicitly address “cultural safety” and unfair treatment and discrimination in care delivery were designed in consultation with an Indigenous Advisory Committee. We strongly support efforts along these lines—especially the inclusion of racial/ethnic groups and organizations in the development and testing of survey questions—to help identify how discrimination is experienced in health care.

In addition to obtaining direct input on survey content from all demographic and socioeconomic groups, the development of survey questionnaires to capture this feedback must be thoroughly assessed through “cognitive testing.” Cognitive testing in the development of the federally sponsored Consumer Assessment of Healthcare Providers and Systems (CAHPS) surveys—widely accepted as the gold standard for patient experience surveys—is one of the steps that sets these questionnaires apart from many others. This step involves interviewing patients about the survey questions that they are asked, to confirm that the questions are clearly understood. The testing of question wording and their underlying concepts is critical to assure that any substantive differences in patient understanding of survey questions across subgroups are addressed and resolved.

2. Include Patient Narratives

In addition to including standardized questions designed to gather feedback relevant to all racial/ethnic groups, we strongly support providing patients the opportunity to have an even more direct voice regarding their experiences through the use of open-ended comments. Research has shown that the use of open-ended survey questions can often capture issues and insights not covered in standardized closed-ended questions. The use of open-ended survey questions may be especially effective to capture the voices of marginalized patient populations, helping to reveal what is particularly salient to them.

During the past decade, the CAHPS Consortium, with funding from the Agency for Healthcare Research and Quality (AHRQ), has developed and tested a Patient Narrative Item Set for the CAHPS Clinician and Group Survey; the Consortium is currently working on narrative questions for other CAHPS surveys. We urge survey sponsors to add these rigorously developed open-ended questions to standardized surveys to gain a deeper understanding of patient experience and preferences, as well as insights into how to make improvements in health care delivery.

3. Expand Survey Administration Modes

We support the development and use of new survey administration methods such as email, web, and smartphone applications to achieve more efficient and timely data collection. Research has shown that each mode has relative advantages and disadvantages, and achieves different response rates from different demographic groups.

While the increased use of digital modes has promise, they are not the silver bullet that some proponents have suggested. By contrast, evidence suggests that the most promising approach to maximizing survey participation is to use sequential mixed modes. This involves using one mode after another to offer different options by which people can reply. Mixed-mode administration promises to more fully represent the experiences of patient populations that have been hard to reach.  

4. Consider Adjusting For Race And Ethnicity While Retaining Transparency In Disparities

An important principle in the use of standardized surveys such as CAHPS for “high stakes” applications, such as public reporting and value-based payments, is the need for adequate case-mix adjustment to assure fair comparisons across health care organizations. Case-mix adjustment is necessary to account for differences in patient characteristics—such as age, education, and health status—that are known to influence survey scores but do not reflect differences in quality of care. The Centers for Medicare and Medicaid (CMS) currently uses sophisticated case-mix adjustment models to “level the playing field,” but these models do not include the race and ethnicity of survey respondents as factors.

Inadequate case-mix adjustment could lead to lower reimbursement from value-based payment programs for safety-net organizations serving vulnerable patient populations. This, in turn, may exacerbate existing disparities experienced by the patients they serve. We believe there is a strong rationale and growing evidence to support adjustments for respondents’ race and ethnicity in the case-mix models used by CMS for public reporting and value-based payments to hospitals, health plans, and medical groups. However, case-mix adjustment should not disguise the existence of racial/ethnic disparities. Therefore, we urge continued and expanded public reporting of these differences. Health care leaders and the public should take advantage of current resources such as the AHRQ’s National Health Care Quality and Disparities Reports and stratified reports published by the CMS Office of Minority Health to frame their understanding and response to disparities.

5. Standardize The Collection Of Demographic Data

Critical to surveying and analyzing results by race and ethnicity is the availability of self-reported race and ethnicity for every patient. Although progress has been made in the systematic collection of race, ethnicity, and language data, more effort is needed to ensure that this information is readily available for every patient and transmitted to survey vendors for use in constructing “sample frames”—the lists of all patients that are used to draw random samples. The questions asked of patients to determine their race and ethnicity should be standardized across all health care providers and systems to ensure comparability, that is, identical questions regarding ethnicity and race, including the use of identical racial and ethnic categories. This standardization of race and ethnicity questions should extend to all CAHPS surveys as well.

The collection of patient background data in health care organizations should also include the collection of language preference. This can be a challenge for many institutions, but it is essential to offer patients a survey they can actually read. CMS provides translations of surveys in a number of languages—but survey vendors need this essential information from their clients so that they can send out language appropriate surveys.

6. Use Other Sampling And Feedback Methods To Complement Standardized Surveys

A critical element in the use of standardized surveys for high-stakes applications is the use of representative samples of patients—in which each person has the same chance of being surveyed—to ensure that survey results are comparable across hospitals, medical groups, and clinicians.

A simple, random sample may not produce a sufficient number of responses from smaller subgroups of patients or groups of patients whose survey response rates are lower than those of others. Some organizations that believe they do not receive enough surveys to reliably analyze the responses of patient subgroups go to the expenses of surveying all patients—a census survey—rather than a sample of patients.

CMS provides another option that could reduce this cost of census surveying while providing more responses from the subgroups of interest. Instead of a census or a simple, random sample, hospitals are allowed to oversample groups of patients—that is, survey more of these patients than you would in a random selection. This can provide hospitals a larger number of responses from these groups for their internal patient experience improvement processes. When a hospital chooses to oversample, CMS then reweights the survey results so that, for public reporting and high-stakes uses, the results are representative. While this option is available, it is not in general use because it requires hospitals to collect complete demographic information about all patients prior to survey administration—underscoring the importance of obtaining this information as noted above.

Finally, given their chief focus on standardized comparisons across organizations, CAHPS surveys were never intended to be the primary inputs driving rapid-cycle quality improvement. While CAHPS data certainly can and do contribute to quality improvement, other patient feedback methods can be used to drive rapid-cycle improvement and obtain input from underrepresented groups—complementing the results obtained from standardized surveys and the oversampling techniques described above. Examples of such feedback methods include short, targeted surveys coupled with rapid-cycle results; focus groups; patient shadowing; and patient and family advisory councils.

7. Enhance Public Reporting

A critical element of democratizing patient experience surveys is making public reports of survey results more meaningful, understandable, and accessible to all consumers and patients. While CMS and other sponsors of public reports have made important strides in making health care quality information more available, little information exists on the extent to which these public reporting websites are actually being used.

To achieve increased use of science-based, standardized survey results in the public domain, the content, design, and capabilities of existing websites must be improved, starting with the following steps:

• Ensure that the publicly reported information is relevant and meaningful to all patients, including racial and ethnic subgroups. This should include the addition of patient narratives to public reports.
• Conduct research and implement methods for publicly reporting reliable and valid patient experience survey results by racial and ethnic subgroup, at the individual, organization, and clinician level.
• Ensure that website content is produced at an eighth-grade reading level or lower, to enable people at various levels of literacy to access it.
• Ensure that navigation of websites is easy and intuitive, based upon user testing.
• Make content available in multiple languages, not just English.

Concomitant with the need to improve content is the need to make public reports more easily accessible to all consumers. For example, those without access to the internet should have a means of accessing these reports (for example, through local libraries). A concerted and funded effort to increase consumer awareness of public reports of patient experience is needed. A relatively simple starting point would be adding a link to relevant websites in all survey cover letters.

Identifying And Spreading Models For Success

Acting on these recommendations for reducing racial and ethnic disparities in patient experience of care will require a significant commitment by policy makers, regulators, and health care organizations. One resource that may be of use is The Beryl Institute’s framework for future action in health care—The New Existence—which includes a segment on policy and systemic issues. The objective of this segment is to advocate for equitable institutional, governmental, and payer policies, incentives, and funding to drive positive change. A central aim of this segment is to “research, measure and dismantle the structures and systems that lead to disparities.” We believe we must do no less.

We also believe, to be successful, all survey sponsors and their vendors will need to comply with the survey content, sampling, and administration recommendations we propose. Health care leaders willing to respond to this challenge will face a rewarding but difficult journey. But starting on this journey does not need to mean stepping into the unknown. We are confident that there are hospitals, health systems, medical groups, and individual providers that are already successfully measuring racial/ethnic disparities and working aggressively to eliminate them. Who are these entities? How are they identifying disparities? What strategies and tactics are they using to reduce disparities?

With this call to action, PXPF and The Beryl Institute commit to organizing a coalition of likeminded organizations to tackle racial and ethnic disparities in patient experience. Working collaboratively, this coalition will: a) identify a group of these pacesetter organizations that are reducing disparities; b) explore the keys to their success; and c) spread their stories widely within the health care field. Ending racial and ethnic disparities in health care is one of our most important challenges. If we have the will and work together, we will succeed.

Authors’ Note

The Patient Experience Policy Forum, supported by The Beryl Institute, is a broad-based and diverse coalition of patients, family members, caregivers, and health care professionals uniquely led by a balanced board of patient/family partners and senior patient experience leaders. Our purpose and commitment are to advocate for and help shape policy at the national and local levels on issues that directly affect patient and family experience and elevate the human experience in health care. We do so through convening policy forums, educating policy makers, sponsoring advocacy events, providing communication updates, and publishing calls to action.