Emergency Department
Patient Satisfaction Surveys

An Information Paper

Created by Members of the ACEP
Emergency Medicine Practice Committee

June 2011

INTRODUCTION

This information paper provides background on patient satisfaction surveys, methodologies utilized, limitations and utilization of survey data for emergency physicians to work with hospital leaders on appropriate interpretation of scores and creating an environment conducive to quality care and patient satisfaction. The American College of Emergency Physicians has a policy statement, “Patient Satisfaction Surveys” adopted in September 2010 that outlines the College’s position.

In a free enterprise, competitive healthcare system such as that of the United States, patient satisfaction has been called “The Indispensable Outcome.”1 The Centers for Medicare & Medicaid Services (CMS) have embraced Value-Based Purchasing (VBP) as a methodology for apportioning entitlement healthcare resources, and identified patient satisfaction (patient experience of care) as a key marker of value. Prior to the VBP initiative implementation, Medicare mandated HCAHPS (Hospital Consumer Assessment of Healthcare Providers and Systems), a highly prescriptive post-discharge survey of consumer perceptions of their inpatient care. The results of these surveys, along with other hospital performance measures, are published on the CMS Hospital Compare (hospitalcompare.hhs.gov) website. Clinician and Group CAHPS (CGCAHPS) is in its early execution phase, and together with the PQRS (Physician Quality Reporting System), it will be used for measurement and by 2015 for pay for performance in the outpatient setting. There is no comparable mandate for emergency services as this is written (early 2011), but hospital administrators are acutely aware that they are being graded according to patient perceptions of their facilities’ care, including the emergency department (ED). There is, however, a strong correlation between patient satisfaction in the ED and inpatient HCAHPS scores. Hospital administrators have come to accept the ED as the “front door to the hospital” and expect their emergency medicine physician groups and ED staff to take the issue of patient satisfaction seriously. Increasingly, emergency medicine group contracts are mandating that certain levels of patient satisfaction be met in order to retain the contract.

Value is not the same thing as quality. Value is a perceptual attribute, subjectively determined by the recipient of the service, whereas the quality of both the science and art of medical practice can be objectively measured according to evidence-based standards.2 Value is typically determined in relationship to price, but in the U.S. healthcare system, where the patient is largely shielded from the price, service expectation predominates. Retrospectively, there has been little evidence in the literature linking greater patient satisfaction perception to quality 3 More recently, published studies have linked higher patient satisfaction with lower hospital readmission rates for pneumonia, heart failure, and acute myocardial infarction,4 and another study has linked the patient rating of physician empathy with better glucose control and lower lipid levels in diabetic patients.5 Emergency physicians are challenged in that they must interact with patients without the benefit of an established physician-patient relationship. They must inspire confidence and communicate their concern for the patient and family in a very short period of time. This may require a different skill set from the clinical practice and is part of the art of emergency medicine. In services industries, including health care, satisfaction surveying has become a common practice. Emergency physicians must become well versed in the process of surveying patients on their expectations and perceptions of care.

SURVEY HISTORY

The origin of patient satisfaction is not a mystery. It has its roots in the human connection that is at the foundation of the doctor-patient relationship. The annals of medicine are replete with stories of the failure of technically excellent care delivered by an unapproachable physician and the success of technically poor care delivered by a charismatic physician. Some might even say that the quality of medical care cannot possibly be considered excellent in the absence of this all-important human connection.

In the eighties, businesses began to use customer satisfaction surveys to monitor the service they provide, and as a tool to gauge customer loyalty.

Some experts in the survey industry feel that SERVQUAL, short for service quality, is the “most complete attempt to ….measure quality.” SERVQUAL was originally measured on 10 aspects of service quality: reliability, responsiveness, competence, access, courtesy, communication, credibility, security, understanding the customer and tangibles. It measures the gap between customer expectations and experience. Generally, healthcare uses the SERVQUAL process in determining patient satisfaction scores. In the early nineties the SERVQUAL tool was refined to 5 aspects; reliability, assurance, tangibles, empathy and responsiveness or RATER. Reliability represents the ability to perform service dependably and accurately. Assurance represents the ability of staff to inspire confidence and trust. Tangibles represent the physical facilities, equipment, and staff appearance. Empathy is the extent to which caring individualized service is given. Responsiveness is the willingness of staff to help and respond to customer need. Source: Wikipedia, SERVQUAL, 2/16/2011 update

The surveys are designed to determine the difference between the customers’ expectations of a service offering and the customers’ perceptions of the offering received. A criticism is that the surveys measure perceived service as opposed to actual service. This makes the SERVQUAL measure an attitude measure that is related to, but not the same as, satisfaction. Source: Theories Used in Research, York University, http://www.istheory.yorku.ca/SERVQUAL.htm

 HCAHPS is a CMS-driven process that was initiated in 2006. It has been described as the first national public and standardized survey that measures the patient’s perspective of healthcare. Prior satisfaction surveys were used internally by hospitals. In contrast, HCAPS is used to compare hospitals publicly in relation to the patient’s perception of care. It is now being extended to include home health care and physician practices.

SURVEY METHODOLOGY

Currently the HCAHPS Method revolves around surveys sent to a random sample of adult patients between 48 hrs and 6 wks of discharge. Hospitals may use a survey vendor to acquire data or perform their own survey. The surveys may be performed by telephone, mail, or interactive voice recognition, and are offered in 5 different languages. There is a target minimum number of 300 returned surveys per year. HCAHPS excludes patients under age 18, patients who died in the hospital, patients discharged to hospice, patients discharged with a primary psychiatric diagnosis, prisoners and patients with international addresses, and “No Contact” patients.

Press Ganey and Associates, the nation’s largest patient satisfaction survey vendor, surveys hospitals, medical groups, and other health care entities. While the surveys are sent to individual patients, the data is collated for the hospital as an entity. The data may be dissected and drilled down to individual physicians and other providers. For this company the stated number of surveys required to make appropriate judgments and comparisons about individual physicians is 30-50. With that said, results are tabulated, returned and compared when greater than 7 surveys are returned. The position of the Press Ganey company is that the data is relevant though statistically insignificant and “belongs” to the survey participants. It is then up to the participants and member hospitals to decide what, if any, useful information can be gleaned from the surveys.

While Press Ganey does not release the internal statistical calculations and goals of its survey tools, independent review reveals that depending on the size of the ED population, for the ED as a whole 30-50 results could provide as low as a 50-55% confidence interval (a flip of a coin to decide whether the results are valid). In order to create the generally accepted 95% confidence interval, 175-225 surveys would be necessary for a scaled (1-5 for example) survey, which is the methodology that the company uses.

SURVEY LIMITATIONS

With the exception of rare survey systems that are performed in person and at the time of discharge, satisfaction survey data is not completely randomized.

• Most health care satisfaction survey companies do not include patients admitted to the hospital or transferred out of the host hospital in the ED data. In most cases, those patients receive an inpatient survey and are counted in the “Inpatient” data. The ED surveys therefore include a higher proportion of low acuity visits, patients with whom the ED personnel spend less time and resources.
• Telephone-based surveys tend to skew the data away from non-English speakers and those who may not have a personal phone.
• Mail-based surveys eliminate those without permanent addresses and the illiterate.
• Pediatric patients have their surveys filled out by parents or caretakers.
• Institutionalized patients may be unable to complete surveys independently and tend to be underrepresented.
• Depending on how surveys are distributed and interpreted, frequent ED utilizers are either over- or underrepresented – most commonly, ED surveys are not re-sent out to patients if they have been seen in the same ED within 90 days.

Most satisfaction surveys are based on a “normal” bell-shaped curve with a central mean exactly in the middle of possible values/results (ie, a 1-5 survey would be assumed to have a mean of 3). Across giant populations such as the “United States” or “Europe,” this may be true, but the mean and standard deviation of increasingly smaller populations (New York, New York City, Queens, Flushing) will vary increasingly. To this extent, absolute results and even trends in results are not statistically comparable across small populations. They are only valid when comparing an individual variable (a practitioner, a hospital, a trait) to itself over time with the same distribution of results. Though we often use results to compare practitioners within a department or departments within a city (most would agree, our ideal use of the data), there is no statistical basis on which to do this. From a statistical standpoint, these survey tools are created only to measure trends over time, not to compare practitioners. Any discussion between the ED leadership and the hospital leadership should recognize this statistical reality. While surveys can be a valuable tool to receive information from patients, they should not be used to drive significant decisions that are not statistically valid.

UTILIZATION OF SURVEY RESULTS

When discussing how best to utilize the results of patient satisfaction survey three general areas should be considered:

Obtaining the best available data

Understanding the target audience

Driving process improvements

Obtaining the best available data

One of the biggest concerns voiced by practitioners is the small number of surveys returned (ie, the “n”). This may not be reflective of a poor tool, but rather of the number of surveys sent out/obtained by phone and the percentage of those sent out/called that are returned/answered. Since survey companies charge for sending/calling whether the survey is successfully completed, hospitals often limit the number of patients to be contacted. Almost every satisfaction review company will agree that more data points will provide more power to the tool, but many facilities choose to limit the number of surveys sent due to financial considerations. Some patient satisfaction survey companies only do 25-30 surveys a month, regardless of the size of the ED. These companies allege that this number is adequate for statistical validity. There is a threshhold at which the data begins to be relevant, but this doesn’t mean it’s the best data available. The ED team must be sure to have the discussion with the hospital leadership on what the desired outcome of the surveys might be.

First, determine what it is that you want. Align the amount of data received with its significance. Most facilities will receive enough group feedback within a given time period to provide sufficient power to the results, and in most cases we can lead customer service improvements with group data. Even a small percentage of responses will provide dozens, if not hundreds, of data points within a month, and thus provide a fairly powerful metric for leading change. Monthly collation of data is the lowest frequency recommended. If individual data is requested the process should ensure that there are at least 30 surveys per provider for any chosen time period. Monthly and quarterly group data can provide great customer feedback on how care is being perceived, and often be one of the initial methods of trending either satisfying or dissatisfying patterns that are developing within a group.

Next, make sure that you understand your peer group. Most satisfaction companies are willing to “match” your ED to its proper cohort usually by patient volume although facilities with the same number of visits can be radically different. A 100k visit community ED is not the same as 100K visit urban ED with a residency program. Factors such as payer mix, patients per room, teaching, psychiatric populations etc. should be factored when looking at “matching” your ED with its cohort. That said, your facility or hospital system may prefer that you be compared to all hospitals in the database, since patients themselves do not differentiate their expectations on the characteristics of the ED. If this is the case consider having data run for both all hospitals and for appropriate peer groups, which can usually be done for little or no additional cost.

Finally, understand which questions are being asked in your survey, and modify them as allowed. Most companies will have a standard group of questions that they feel are necessary to include in a survey - you usually won’t have the ability to alter these. But you may have the ability to add/delete other questions on request, and often even have certain “limits” placed on who will receive questionnaires. For example, many facilities will request patients receive no more than one questionnaire per year, thus limiting the number that may go to “frequent narcotic fliers” who might otherwise express frequent dissatisfaction at not having received desired prescriptions.

Understanding the target audience

Any contemporary hospital leadership will be looking at patient satisfaction data. Your knowledge of the relevance and limitations of the data can help guide performance expectations hospital leadership may have, but that is the bare minimum! More knowledgeable and proactive directors will understand the influence of this metric on hospital administration and leverage that as a tool towards process improvement and potentially even performance incentives. Aligning goals on the surveys between the hospital and the ED is important to maintaining a good relationship. Understand what your organizational goals are and create target incentives for increasing levels of performance (eg, good/great/outstanding). Most hospitals will be very willing to contribute money for a guaranteed performance. It is important that the contribution be substantial enough to drive individual performance, making it an easier process to “obtain buy-in from the ED team. Ultimately, the financial results are a part of achieving a satisfied hospital leadership and solidifying your relationship with them. It’s a “win-win” when your group’s interests are aligned with those of the hospital!

Do not underestimate the value of your public as a target audience. Much of the patient satisfaction data is now reported publicly, and easily accessible. Make sure your group understands this, and understands the effects this has on the public perception of your hospital within your community.

Driving Process Improvement

Driving process improvement is ideally what this data should do. You should have no problem getting quarterly data powerful enough to drive group incentive programs as an achievable goal, but you’ll create a lot of animosity and distrust if you attempt to use data of poor power for individual incentives. However, physicians are data driven individuals, and if you can get a large enough sample size, the data can be very helpful. Without individual data, there may be difficulty in driving individual accountability. At the same time, patient satisfaction data should not be used by itself as a peer review tool. Occasionally there will be complaints that merit further investigation regarding the quality of care provided or the behavior of the provider, but appropriate clinical quality peer review should be separate from the patient perceptions of care. Patient satisfaction data is perceptive, and in almost every case affected by the circumstances unique to that particular patient/family’s moment in time. We must be continuously open to these perceptions, yet be always cognizant to the fact that they may have less capacity to understand the medical decision making behind the scenes. This is where the peer review process becomes so valuable.

Do not under-sell patient perceptions as a tool for driving proper change within our organizations – it may be one of the best tools available to us! Our patients can be one of our most valuable resources in pointing out what is working, and what isn’t, within our EDs. Take the time to look closely at the comments provided by patients – you’ll be amazed at the value offered there. Indeed, many satisfaction companies provide tools that can analyze particular questions within your survey in more detail. Press Ganey and Associates provides the very useful “Priority Index” tool with correlates every question with others that affect it most directly; they also have online “Solutions” resources which gives specific practices to improve performance. If used properly these can provide a virtual road map towards process improvement within your department! Decreasing door to provider times, turn-around-times for lab and radiology studies and time to final disposition have been shown to improve patient satisfaction scores.

IMPROVING PATIENT SATISFACTION

There are a number of reasons why emergency physicians and ED groups should want to improve scores.

• the patient is more likely to be compliant with the care provided
• it reduces malpractice risk
• it improves physician and staff morale
• if patients are satisfied they return for their next episode of care
• it may be an important part of the negotiation if your group contracts with the hospital for provider services

One step in understanding and improving satisfaction is to look at your demographics. Older, insured, sicker patients at community hospitals tend to be more satisfied with their care. The younger, uninsured, low acuity, minority patient at a teaching hospital is the most unsatisfied.6

Look at your ED scores and see which age groups and demographics are consistently lower scoring. Strategies can be developed to improve these scores. For instance, at one facility it was noted that young women with vaginal bleeding and pregnancy had very low scores if they had to wait in the lobby. In an effort to improve their perceptions of care, these patients are now offered a stretcher instead of remaining in the lobby area while they wait.

All staff should be empathetic and professional in appearance. Physicians should wear clean white coats. Scrubs are acceptable, but leave the jeans, sneakers and sandals at home. A nice pair of pants and a collared shirt go a long way in projecting a professional appearance followed by a non-coffee and non-blood stained white coat. Always have a back up white coat or scrubs should your clothing have something splashed on it. Men should shave before coming to work regardless of the shift. Perfumes and colognes should be minimal and unobtrusive.

Patient assessment of satisfaction has strong correlation with the physician’s interpersonal skills and less with whether or not the physician met the expectation of diagnosis and therapy. Part of this interpersonal skill is body language that can be construed as negative, such as:6

Holding objects in front of your body
Checking the time or inspecting your fingernails
Picking lint off of your clothes
Stroking your chin while looking at someone
Narrowing your eyes
Standing too close
Looking down while in the presence of others
Touching your face during a conversation
Faking a smile
Resting hands behind the head or on the hips
Not directly facing the person you’re speaking to
Crossing your arms
Foot and finger tapping 7

Below are some approaches and environmental factors that can result in a positive effect on the patient’s perception of the care provided.

Greet the patient appropriately. The initial interaction is very important in setting the tone for the patient’s perception of their visit. You are now “onstage” with the patient, and any tension you might be feeling must be left “offstage.” Introduce yourself to the patient and acknowledge them. It is also beneficial to address those who might be in the room with the patient. If appropriate, smile and break down barriers by shaking the patient’s hand. Hand out business cards.

Sit down. Taking a seat in the patient’s room gives them a sense that they are important enough to warrant your full attention. Studies have suggested they will also have a perception that you were there longer than you were, by as much as sevenfold. Lower yourself to below the patient’s eye level if possible. Don't forget the importance of touch. That's the magic healing of being a physician or nurse in the first place. Holding a hand or a gentle touch of the shoulder is reassuring.

Use active listening and open body language. The most important thing you can say to a patient or parent is, 'What's your biggest concern?' It may be that their concern is entirely different than you think. Often, the most important thing that can be provided to a patient is reassurance. You may not find a reason for their concern, but you likely can reassure them that it is not anything serious. You can certainly tell them the items you have ruled out, and possibly the minor ones that remain. Explaining this often is enough to alleviate the patient’s fears and make them start to feel better. Be sure you know their concern. For instance the patient with a simple rash may have had a loved one die from skin cancer, and now think they have the cancer as well. Know the questions that are included on the patient satisfaction survey for your ED. Use key words specific to the survey during your encounter with the patient.

Manage expectations. Finding out what the patient or parent wants, and explaining to that person at the beginning what the ED is able to provide will help avoid frustrations later.8 Give them an idea what the duration of time might be to complete your evaluation. Check back on the patient as often as you can and keep them informed as to their progress. Be sure their pain is adequately controlled. If you are discharging the patient, be sure you explain what is wrong, what the treatment is, the expected course and required follow up. Be sure the patient and/or family understand. If an admission is required, explain why and what further testing may be needed during the stay. This helps patient satisfaction scores on the hospital admission side significantly. Ask the patient at the beginning and end of the visit, “Do you have any questions or concerns?”

Establish privacy. Even if there is only a curtain between patients, speak in as soft a voice as necessary to maintain discretion. Close the door in a private room. Draw the curtain. When you do these things, emphasize to the patient that you are doing it to try and maintain their privacy.

Maintain a clean and comfortable environment. A clean work area and patient care area promotes professionalism, means that an organization cares, and medically helps stop the spread of disease. Pick up objects on the floor when you see them and throw them away. If you see a dirty or soiled object removed it from the patient care area to be cleaned or have it cleaned immediately. Providing a blanket when a patient is cold goes a long way to meeting patient expectations and comfort needs.

Provide diversions. Add TV’s to lobby and rooms when possible. Provide items to pass the time. Magazines, books, games for kids, etc 

Call patients after their visit. ED patient callbacks have long been known to improve ED patient satisfaction.9 A more recent study has reaffirmed this.10 The experience of members of the committee has been that patients who receive a follow phone call rate their ED satisfaction 30 to 70 percentile points higher than those who did not receive such a call. This is a truly differentiating factor in patient perception.

CLOSING

The use of patient satisfaction surveys in EDs has been controversial. Critics would suggest that the surveys lack validity and therefore the data they produce is being interpreted incorrectly. Without disputing that concern, there is every indication that patient satisfaction surveys will continue to be used in healthcare, and that the scope of use will increase. Results will be tied to reimbursement. Satisfaction surveys are one of the few tools available to facilities, administrators, regulatory bodies and providers to gauge the perception of the care experienced by the patient. Indeed, the current move in health care is towards more transparency of this information with incentives for high performers (and the subsequent corollary – disincentives for poor performers). Moving forward in this climate, ED leaders need to be armed with all the appropriate tools to use the information received by survey in the most productive manner. With a thorough understanding comes a greater ability to interpret the data, educate those who need the knowledge, and drive to a better performance. In addition, it could also lead to developing better methodology for understanding how our consumers perceive care. In the end, ED providers would and should be appropriately recognized for the highly skilled care they deliver in an incredibly challenging environment.

Created by members of the ACEP Emergency Medicine Practice Committee
June 2011

William P. Jaquis, MD, FACEP, Chair, EMPC
Laurence R. DesRochers, MD, FACEP, Chair, Subcommittee on Patient Satisfaction Surveys
Enrique R. Enguidanos, MD, FACEP
Diana L. Fite, MD, FACEP
Juan Francisco Fitz, MD, FACEP
Daniel Freess, MD
Ronald A. Hellstern, MD, FACEP
William P. Jaquis, MD, FACEP
Carla E. Murphy, DO, FACEP
John Newcomb, MD, FACEP
Mark S. Rosenberg, DO, MBA, FACEP
Adhi N. Sharma, MD, FACEP

References

1. Managed Care. April 1999, © 1999 Stezzi Communications
2. Thomas AM. Patient satisfaction: Measuring the art of medicine. JAMA. 1998;280(24): 2127.
3. Cleary PD, McNeil BJ. Patient satisfaction as an indicator of quality care. Inquiry. 1988;25(1) 25-36.
4. Boulding W, Glickman SW, Manary MP, et al. Relationship between patient satisfaction with inpatient care and hospital readmission within 30 days. Am J Manag Care. 2011 Jan;17(1):41-8.
5. Hojat M, Louis DZ, Markhan FW, et al. Physicians’ empathy and clinical outcomes for diabetic patients. Acad Med. 2011 Mar;86(3):359-64.
6. Toma G, Triner W, McNutt LA. Patient satisfaction as a function of emergency department previsit expectations. Ann Emerg Med. 2009;54(3):360-367.e6. Epub 2009 Mar 12.
7. 25 Acts of Body Language to Avoid. http://www.marcandangel.com/2008/07/07/25-acts-of-body-language-to-avoid/
8. Dan Hurley. Mentor staff to improve patient satisfaction. ACEP News, August 2010.
9. Shesser R, Smith M, Adams S, et al. The effectiveness of an organized emergency department follow-up system. Ann Emerg Med. 1986;15(8):911-915.
10. Dunn L. Four best practices for improving emergency department results. The Studer Group Newsletter, January 25, 2010.