Zusman, Edie E.
Beginning in just a few months, hospital patient satisfaction scores will have a direct impact on the bottom line for health care reimbursement. In October 2012, the Center for Medicare and Medicaid Services (CMS) is reducing by 1% the base operating diagnosis-related group (DRG) payments to hospitals to create an incentive fund, estimated at $850 million.1 How this money is distributed to hospitals will depend on their performance on several “quality” measures, 30% of which will be based on how patients rate their hospital experience on the Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) patient satisfaction survey. The component of payment to hospitals, which is variable and based on performance measures, is expected to gradually increase over the next several years, and private payers are likely to follow suit.1
This so-called value-based purchasing initiative is now required under the Patient Protection and Affordable Care Act, the 2010 national health-care reform legislation. The patient satisfaction survey tool selected for value-based purchasing, HCAHPS, is available for distribution through many authorized vendors including Press Ganey Associates, Inc., which previously distributed its own proprietary survey to 40% of United States hospitals. The HCAHPS survey is designed to allow consumers to rate their inpatient experiences and perception of care and is anticipated to be an improvement over the highly criticized Press Ganey Patient Satisfaction Survey. Because HCAHPS has been selected by CMS as the validated and transparent national survey tool with publicly available results at the Hospital Compare website, (www.hospitalcompare.hhs.gov) these data can then be used to evaluate hospitals, improve patient decision-making and increase incentives for hospitals and providers to deliver what patients perceive as high-quality care.2
The inclusion of patient satisfaction scores in the reimbursement equations, however, presents a number of challenges. Based on concerns about Press Ganey and other available survey tools, researchers have been encouraged to analyze HCAHPS to identify potential biases, and CMS is considering using conversion factors to normalize the national data where biases have been identified.
Aside from reporting bias, the question of whether patient satisfaction surveys are a useful tool upon which to make administrative or medical decisions remains. Among the concerns are that patients may not be in the best position to evaluate their care, and that hospitals may not be able to improve health-care delivery based on results of patient surveys. In addition, some critics argue that the surveys do not adequately take into account variables that can skew results against academic medical centers or hospitals in regions of the country that treat large numbers of patients with mental or other serious illnesses.3
Perhaps most concerning is new data that suggests that higher hospital patient satisfaction scores are actually associated with higher inpatient use, overall health care and prescription drug expenditures and increased mortality. In a recent article published in Archives of Internal Medicine Joshua J. Fenton of the University of California Davis Health System evaluated data from more than 50 000 adult respondents of the Medical Expenditure Panel Survey, a nationally representative survey of the US population that assesses the use and costs of medical services. Respondents completed questionnaires about their health status and experiences with health care, including how often their health-care providers listened carefully, were respectful and spent enough time with them. Participants also were asked to rate their health care on a scale of 0 to 10. The data were linked to the national death certificate registry.
What Fenton and colleagues found was that even after adjusting for numerous variables, patients who were most satisfied had greater chances of being admitted to the hospital and had about 9% higher total health-care costs as well as 9% higher prescription drug expenditures. Most strikingly, death rates were also higher. For every 100 people who died over an average period of nearly 4 years in the least satisfied group, about 126 people died in the most satisfied group. More satisfied patients had better average physical and mental health status at baseline than less satisfied patients. The association between high patient satisfaction and an increased risk of dying was also stronger among healthier patients.4 An alternative explanation for this unexpected finding, likely based on surveys completed by family members of the deceased patients, may reflect the perception that everything possible was done to help their loved ones. As health-care providers, we may be very good at focusing attention on and communicating well with this patient population and their families.
In an accompanying editorial in the same journal, Brenda E. Sirovich, MD, MS, of the Department of Veterans Affairs Medical Center in Virginia, explains that the Fenton research infers that efforts to cater to patient satisfaction may be ill guided because by “implicitly encouraging health care providers to honor requests for (or to explicitly offer) discretionary health care services, such efforts may lead to overutilization, higher costs, and worse outcomes.”5
The author further suggests that while the relationship between customer satisfaction and subsequent consumption in the business world is doctrine, it is not necessarily appropriate in health care. “While most Americans may accurately assess how well their washing machines, their hairdressers, or even their airlines are performing, their evaluations of physicians and health care interventions may have limited validity,” she writes.
These and other pitfalls of patient satisfaction surveys and their increasing use in physician compensation and hospital reimbursement for care are important for neurosurgeons and all physicians to consider as these measurement tools seemingly affect every aspect of the services we provide.
In one online survey posted by Emergency Physician Monthly, 16% of the 717 medical professionals who responded said they had their employment threatened by low patient satisfaction scores, and 27% stated that their income was in some way tied to satisfaction scores.6
In addition to the effect of patient satisfaction scores on physician employment and income, the authors, William Sullivan, DO and JD, and Joe DeLucia, DO, explain that low survey scores can affect medical care, and not always for the benefit of the patient. When asked to rate on a 1 to 10 scale how patient satisfaction scoring affects the amount of testing performed, 41% of respondents said they decreased the amount, while 59% said they increased testing. Additionally, 48% of health-care providers reported altering medical treatment—including providing unnecessary care—in an effort to influence a patient satisfaction survey. Adverse outcomes from treatment rendered due to patient survey concerns included allergic reactions to medications, resistant bacterial infections, kidney damage and medication overdose.6 These authors go on to theorize that hospital liability could increase from the effects of these scores. “If adverse patient outcomes due to unnecessary medical treatment can be tied to pressures that hospitals place on the medical staff to improve patient satisfaction scores, civil liability to the hospital could result.”
By examining the flaws in Press Ganey's approach to patient satisfaction measurement and reporting, physicians and hospitals can be more informed users of HCAHPS. For example, according to Press Ganey, a minimum of 30 survey responses collected over the designated time period is necessary to draw meaningful conclusions of the data for a specific individual, program or hospital. Despite this requirement to achieve statistical significance, Sullivan and DeLucia found that the firm often provides comparative data about hospital departments and individual physicians based on a smaller sample size that may create an unacceptably large margin of error.6 Sullivan, for example, said his department may have 8 to 10 Press Ganey survey responses per month and yet still receives monthly reports from the company analyzing the data. Because of the small sample size, 1 month his department ranked in the first percentile and 2 months later it ranked in the 99th percentile.6
The authors further point out that emergency patients who are admitted—those most in need of emergency care—do not receive Press Ganey surveys about their ER experience of care. These are the patients who likely had the most thorough evaluation and possibly, the most heroic and excellent care. Those who are treated in the ER and discharged, and may have had to wait longer because of triaging protocols, are likely those to have the least satisfactory experiences. But only the individuals who are not admitted to the hospital are evaluated on Press Ganey Surveys for ER.
Whether patient satisfaction surveys actually measure satisfaction has been questioned. Patient motivation to fill out a survey can skew results, since it may be those who had an extreme experience—either superb or terrible—who are likely to bother. How a survey is administered can also affect the patient's evaluation.7 In a study funded by CMS, researchers found that patients randomized to the telephone and active interactive voice response modes provided more positive evaluations than patients randomized to mail and mixed-mailed survey with phone call follow up.7
Patient satisfaction scores are being used to measure quality of care, but patients may not be in the best position to evaluate their care, and hospitals may not be able to improve health-care delivery based on results of patient surveys. Geographical, cultural and racial differences can affect a patient's perspective about their medical or hospital experiences. Research aimed specifically at identifying biases in the HCAHPS reporting system has already found that hospital rankings vary substantially by patient health status and ethnicity/language and moderately by patient education and age (P < .05).8 Because HCAHPS largely rates patients' perception of their hospital care, there is concern regarding higher scores in more affluent communities where supplemental services and philanthropy dollars may be available to improve the patient experience.2 One Cleveland Clinic study evaluating bias in HCAHPS reporting, found that “no hospital in the nation with 500 or more beds has scored in the 90th percentile for such basic measures as physician or nurse communication.”3 Vinski et al identified yet another variable potentially, out of our control to change, which lowers patient satisfaction scores—whether a patient is in isolation for infection.9
Research has described how patients' opinions of the care they receive can be different from the actual quality of the medical care they receive, and that opinions and experiences vary by race.10 In a randomized controlled trial, researchers from Mount Sinai school of Medicine and Columbia University Medical Center surveyed inner-city women with newly diagnosed and surgically treated early-stage breast cancer for their perceived quality of care and the process of getting care including referrals, test results, and treatments. They compared the responses to patient records to determine the actual quality of care.
Of the 374 women who had received treatment for early stage breast cancer, 55% said they received “excellent care,” but most—88%—actually got care that was in line with the best current treatment guidelines. Among the other findings: African-American women were less likely to report excellent care than Caucasian or Hispanic women, less likely to trust their doctor, and more likely to say they experienced bias during the process. However, there was no difference in actual quality of care received in any group.
Geographic differences may also play a role in how patients perceive their care experience. Modern Healthcare reported that HCAHPS survey results indicate patients favor smaller, rural hospitals. One Alabama hospital official interviewed said his facility may have ranked highly because it encourages its employees to practice their faith with patients, which is well-received by and mentioned in survey responses by patients.11 Overall, there is a perception that a typical patient, based on acuity, diagnosis or demographics at an institution will report the highest patient satisfaction scores.
It is not clear yet whether HCAHPS will use a qualitative threshold score to influence funding decisions, or whether they will replicate the flawed statistical process of Press Ganey by using percentiles. Robert C Lloyd, Ph.D, author of Quality Health Care: A Guide to Developing and Using Indicators12 is among the statisticians who helped develop the percentile statistical analysis mapping that, in 1985, was based on a classic bell-shaped distribution of patient satisfaction survey scores.
Lloyd explained that because hospitals, medical groups and physicians have been working hard these past 20 years to achieve higher Press Ganey scores, there is now a significant clustering of raw scores at the high end with a very narrow response range. The data no longer have the bell-shaped distribution, and when this now-condensed data distribution maps to the percentile spectrum, the percentile data are highly inaccurate. A difference of 20 or 30 percentile points may actually be based on raw scores that are not statistically different and simply the result of random variation, yet these percentile results sometimes announced at staff meetings and posted in physician lounges are also used to determine bonus pay for clinicians and promote or replace program managers.13
Physicians and other providers should be aware of the limitations identified in using Press Ganey and the biases inherent with any patient satisfaction survey methodology. With this knowledge, strategies can be developed to avoid the same misleading analyses with the new HCAHPS survey so that physician pay, institutional reimbursement and hiring and firing of managers are based on real and meaningful changes in patients' perception of their health-care experience, and the quality of care provided.
2. Giordano LA, Elliott MN, Goldstein E, Lehrman WG, Spencer PA. Development, implementation, and public reporting of the HCAHPS survey. Med Care Res Rev. 2010;67(1):27–37.
3. Daly R. Unsatisfactory marks. Hospitals question use of HCAHPS in scoring for value-based purchasing. Mod Healthc. 2011;41(33):30.
4. Fenton JJ, Jerant AF, Bertakis KD, Franks P. The cost of satisfaction: a national study of patient satisfaction, health care utilization, expenditures, and mortality. Arch Intern Med. 2012;172(5):405–411.
5. Sirovich BE. How to feed and grow your health care system: comment on The cost of satisfaction. Arch Intern Med. 2012;172(5):411–413.
7. Elliott MN, Zaslavsky AM, Goldstein E, et al.. Effects of survey mode, patient mix, and nonresponse on CAHPS hospital survey scores. Health Serv Res. 2009;44(2 pt 1):501–518.
8. Elliott MN, Lehrman WG, Goldstein E, Hambarsoomian K, Beckett MK, Giordano LA. Do hospitals rank differently on HCAHPS for different patient subgroups? Med Care Res Rev. 2010;67(1):56–73.
9. Vinski J, Bertin M, Sun Z, et al.. Impact of isolation on hospital consumer assessment of healthcare providers and systems scores: is isolation isolating? Infect Control Hosp Epidemiol. 2012;33(5):513–516.
10. Bickell NA, Neuman J, Fei K, Franco R, Joseph KA. Quality of breast cancer care: perception versus practice. J Clin Oncol. 2012;30(15):1791–1795.
11. DerGurahian J. Southern comfort. HCAHPS data favors Ala. hospitals. Mod Healthc. 2008;38(14):10.
12. Lloyd R. Quality Health Care: A Guide to Developing and Using Indicators. Sudbury, Massachusetts: Jones & Bartlet Learning; 2004.
13. Lloyd R. “Statistics for Managers” Sutter Health Management for Clinical Excellence Course. Green Valley, CA: Sutter Health; 2012.