Gender and racial biases in Press Ganey patient satisfaction surveys

^{ILLUSTRATION © IRYNA INSHYNA/SHUTTERSTOCK}

Patient satisfaction questionnaires were developed in the 1980s as part of the movement to better understand the patient’s experience and their perspective of the quality of care. In 1985, the Press Ganey survey—now the most widely used method to assess patient satisfaction—was developed by 2 professors in anthropology and sociology-statistics at Notre Dame. Initially intended for inpatient admissions, the survey was validated based on a few thousand survey results.¹ Given the strong interest in improving patient satisfaction at the time, it became widely adopted and quickly expanded into outpatient encounters and ambulatory surgery settings.

Although other surveys have been developed,² the Press Ganey survey is the most commonly used assessment tool for patient satisfaction metrics in the United States, with approximately 50% of all hospitals and more than 41,000 health care organizations using its services.^3,4 The survey consists of 6 domains related to satisfaction with:

1. the care provider

2. the nurse or assistant

3. personal issues

4. overall assessment

5. access

6. moving through the visit.

Survey items are scored using a 5-point Likert scale, with scores ranging from “very poor” (a score of 1) to “very good” (a score of 5). According to the company, because this format is balanced and parallel (unlike a “poor” to “excellent” format), responses can be quantified and used statistically without violating methodologic assumptions. Also, variability in patients’ responses with this format allows for the identification of opportunities to improve, unlike “yes/no” response formats.¹ There are limitations to this design, however, which can impact data quality,⁵ as we will see.

While the distribution process varies by institution, there is an algorithm laid out by Press Ganey for administering surveys to patients in their preferred language after outpatient visits. Based on recent research into Press Ganey response rates, the typical response rate is estimated to be 16% to 19%.⁶ Although this low response rate is typical of survey data, it inherently introduces the risk of responder bias—meaning results may be skewed by patients who represent the extremes of satisfaction or dissatisfaction.

Adoption of the survey as we move toward value-based care

More recently, patients’ satisfaction with their health care has received increased attention as we move to a patient-centered care model and as health care reimbursement models shift toward value-based care. Current trends in health care policy statements include the importance of raising the standard of care and shifting from a “fee-for-service” to a “pay-for-performance” reimbursement model.^7,8 As a result, hospitals are establishing systems to measure “performance” that are not nationally standardized or extensively studied with objective measures. The changing standard of health care expectations in the United States is a topic of much public debate.⁹ And as expectations and new standards are defined, the impact of implementing novel measures of performance should be evaluated prior to widespread adoption and utilization.

Patient satisfaction also has been identified as a driver for hospital finances through loyalty, described as the “likelihood to return to that system for future medical services.”^10,11 This measure has contributed to policy changes that reinforce prioritization of patient satisfaction. For example, the Affordable Care Act tied Medicare reimbursement and patient satisfaction together in the Hospital Value-Based Purchasing Program. This program uses measures of clinical processes, efficiency, outcomes, and patient experiences to calculate a total score that results in hospital reimbursement and incentives,¹² which creates a direct pathway from patient experience to reimbursement—underscoring hospitals’ desire for ongoing assessment of patient satisfaction.

Continue to: Gender, race, and age bias...