Patient satisfaction: Is your agency measuring and using it effectively?

Use these tips to help your patients help you improve care

When you gauge the success of your agency’s efforts, clinical measures are always of paramount concern: How much do patients’ conditions improve as a result of care? Now, another measure is gaining greater importance: How satisfied are patients with the care and other services you provide?

As agencies seek an answer to that question, they are spending more time and effort on patient satisfaction surveys, tailoring them to give a more complete and specific picture of patients’ wants and needs.

Carol O. Long, PhD, RN, assistant professor in the College of Nursing at Arizona State University in Tempe, says interest in patient satisfaction surveys has skyrocketed in recent years. "I’d say probably six years ago, nobody cared a whole lot about it," says Long, who has studied patient satisfaction for the past decade. "But today, it’s much more important. Now, it’s considered a bona fide outcome measure."

One reason is financial: Insurance companies are increasingly looking to patient satisfaction results as a factor in awarding contracts. As interest evolves, agencies have begun to use good results from such surveys in their marketing efforts.

Bad patient satisfaction can affect referrals and word-of-mouth recommendations, Long says, citing the oft-used consumer statistic that a dissatisfied customer tells 10 people about his or her experience.

But, she says, the best agencies also look to patient surveys to fine-tune their care, identifying areas of weakness and acknowledging the efforts of staff who score well among their clients.

The recent attention to the Patient’s Bill of Rights, requirements of the Medicare Conditions of Participation, and other quality improvement initiatives will only increase the importance of accurate patient satisfaction measurements, she adds.

At the Joint Commission on Accreditation of Healthcare Organizations (JCAHO), the new ORYX initiative requires accredited agencies to begin tracking different performance measurements. Among the measures that agencies can choose to track are those dealing with "patient perception of care," which is commonly gauged through satisfaction surveys.

Those patient perception measures are given equal weight with other measures of agencies’ care, says Sharon Sprenger, RHIA, MPA, project director for core measure identification and evaluation in JCAHO’s division of research.

"If I’m going to look at an organization, I really need to get a balanced picture," Sprenger says. "I need to look at things related to clinical care. I need to look at health status to see if people are functioning better after they get our services. And I’d also want to know something about the patients."

Surveying styles vary

However, as satisfaction surveys become more important, Long says not every agency is using them to their best effect. Poorly written surveys, or those that don’t adequately encourage responses, won’t tell an agency much about its operations. Because agencies use a wide variety of surveys — some building their own questionnaires, some relying on packaged ommercial surveys — it’s hard to draw useful comparisons among them.

Those agencies that choose to use patient perception measures to meet their ORYX requirements must use a patient satisfaction tool offered by their performance measurement system, Sprenger says.

Long surveyed Arizona agencies a year ago and found a wide variety of patient satisfaction survey efforts. "Some agencies do it right on discharge, others while they are still on service," she says. "Some do a sampling, while others survey all patients that have used agency services. Some don’t know whether it’s the client who fills out the questionnaire or someone else in the family. All of those things may make a difference as far as what your final satisfaction scores are."

Home care itself has certain features that make it difficult to compare satisfaction with other health care providers, she says. "There’s a lot of research that shows that continuity of care and interactions with staff are highly valued. In home care, where you are generally having the same person throughout your duration of illness, that can play more of an important role. The proficiency of clinicians can be important."

A patient in the home is expected to participate more in his or her own care, and that, too, can affect the relationships with caregivers and the satisfaction of the patient.

Front-line care staff aren’t the only ones who can have an impact on patient attitudes. Billing, scheduling, and other interactions between the agency and the patient all can lead to client perceptions, both good and bad.

Long says these results must be considered in tandem with other outcomes measures. She notes, for example, that patients who are doing poorly can have high satisfaction with their care, while the opposite might also hold true — an unhappy client may be doing well medically.

"That’s why it is important to examine how utilization rates, access to service, and other key clinical indicators may relate to patient satisfaction," she says.

The response rate on surveys can vary widely depending upon the patient involved. Overall response rates, Long says, can be up to 80% to 85% for some agencies, which is considered quite successful. She says there are steps an agency can take to ensure more patients understand the survey and have an interest in returning it:

The survey’s format. It is important to consider the presentation and content of the form the patient must fill out. Is it too long? Does it adequately address all of the dimensions of care a patient is likely to encounter — nursing care, therapy visits, aides’ services, billing, and scheduling?

Is it confusing? Long says many patients don’t differentiate between the nurses and the other professionals who visit them. When asking about specific care, a survey should describe the care involved, rather than just providing a label.

"Sometimes you have to clarify what those disciplines are," she says. For instance, "A question [should] ask about, the registered nurse — the nurse who came in to provide your teaching, take your vital signs, and report to your doctor.’"

She says a survey that is too long may turn off patients who feel they are too busy to participate, while a short, vague list of questions may lead a patient to believe it’s not important enough to bother with.

Patients can provide a grade for each service — A, B, C, D, F — or can be asked to gauge their satisfaction on a scale from "very good" to "very poor," or "strongly agree" to "strongly disagree."

In addition to scored responses or rating scales, a patient should have the opportunity to discuss the care in his or her own words. Often, Long says, those comments will prove to be more useful than the simple satisfaction scale. The short answers will give you the "context of care."

She says it’s important to give the client the opportunity to remain anonymous if he or she wishes or to leave a name and phone number for follow-up.

Long and Diane Greeneich, DNSc, RN, president and chief executive officer of Healthcare Systems Management Inc. in Goodyear, AZ, have developed a patient satisfaction survey called "HomeSat." (A copy is inserted in this issue.) The company also offers other patient satisfaction management services.

Invitations to participate. It is also important to encourage the client to fill out the survey. That process can start with a self-addressed, stamped envelope, but it shouldn’t end there, Long says. "What is the letter of introduction for the survey? Is there a special letter from the executive director that says, We really value your input?’ People want to know that when they complete the survey, that someone will be reading it."

Long says successful invitation letters point out how the survey can help improve an agency’s services.

Staff enthusiasm. Staff play an important role in helping convince clients fill out the survey and send it in. In some agencies, Long says, the survey is included in the patient’s admission packet. At the end of care, the nurse removes the survey and explains to the patient how important the feedback is to the agency.

"The nurse or therapist can say, Here is the survey. We’re really interested in knowing what you think.’ Bring it to their attention, along with the stamped envelope."

Reinforce to staff the importance of the information, pointing out the effect that good patient satisfaction results can have on contracts and referrals.

Long says it’s also important to provide individual feedback to employees who are praised in the surveys. "Most nurses and therapists like to get the feedback," she says. "When positive feedback comes back, share it. When staff get accolades back to them from the home setting, it’s good to bring that to their attention."

Using the data

Once the responses start coming in, what do you do with them? Long says that process needs to start before the surveys are distributed, recognizing that multiple analyses can and should be done.

"Before you send out any survey, you need to have a database, some kind of statistical tool with which to enter and evaluate the data," she says. "If you’re not set up for that analysis, it’s very difficult. I’ve seen agencies send out the surveys and then have to hand-tabulate the data. They don’t know how to really analyze it once they have the data in place." Some agencies may use Microsoft Excel, SPSS, or SAS, which are more sophisticated data entry and analysis packages.

Most prepackaged surveys will analyze data the agency has gathered. The manager then needs to interpret the data, she says.

Long says agencies should know what kind of information they want — averages or mean scores, the range of scores, and information broken out for each discipline.

When reviewing the surveys, an agency should note any written comments that point to an immediate medical or legal problem and address them right away. Otherwise, Long says, the goal is to find trends, comparing results over time and looking for areas of weakness and strength. When a negative trend shows up, it’s important to carefully analyze what may be causing it. Many agencies use control charts. "Like any other outcome measurement, you go back and look at the processes of care," Long says.

One factor that can affect results is a small sample size, in which outliers are more likely to skew the results. "You can’t talk about 10 surveys when, generally, they say you should have 100 to look at," she says.

Another way of blunting the effect of outliers is to use median measurement, which is the point in the scale that has an equal number of higher and lower results.

Judicious use of both scored results and written comments can help an agency focus on a problem and fix it. At Hospice of the Valley in Phoenix, results from patient satisfaction surveys were used to support a decision to change pharmacy services, says quality/compliance officer Madeline E. Wollmer, RN.

Wollmer says staffers already had been complaining about problems with the pharmacy. Then, she says, complaints started showing up on surveys. "I won’t tell you that it was the only reason we changed pharmacies, but when your customers begin to tell you there’s a problem, then there really needs to be some action taken."

On the clinical side, Long points to a home health agency providing intravenous therapy, an environment in which the proficiency of nursing staff is especially important. She notes that if a nurse comes out to start IV therapy and has to make multiple attempts, it can cause difficulty for the patient.

"The nurse may have to call another nurse out, but by that time, she’s done, say, 10 sticks," Long says. "Patients are pretty good at evaluating technical things. They can tell when a nurse knows what she’s doing or doesn’t know what she’s doing. That can relate to scores that are not very favorable."

She says the agency can review the scores, making note of any written comments that explain the problem in more detail, and use them to make changes. "When you look at your data for the month and see, for example, 10 comments about IV therapy, and it appears that the nurses are not very proficient in administering IV therapy, you’d want to look at that and see what’s going on. Is it the same nurse? Are they patients who are very old and are a difficult stick? What are the issues that seem to be affecting the satisfaction score?"

At Hospice of the Valley, survey results are handled differently depending on their content. A wholly positive review goes directly to the manager of the team caring for the client. Neg- ative results are sent to Wollmer, who reviews them and assigns the team leader to call the family. In the case of hospice, families are usually sent satisfaction surveys after the patient’s death.

Complaints are tracked by the team, with an eye toward emerging trends. As an example, she points to a recent increase in the number of complaints about the hospice’s after-hours service. "The director of that department has had to respond to all of those [individual complaints], but now I have a whole quarter’s worth of results," Wollmer says. "I will say, Here’s the trend for the whole quarter. You need to examine what’s going on in your department.’"

Wollmer says survey results also can serve as a powerful argument to staff that they should improve. "It’s very helpful for the staff member to see the comments on a survey. It means more than anything else. It’s very powerful to be able to say to a staff member, Here’s what your patients have said about you.’"

Outside comparisons

Drawing comparisons with other agencies can be more problematic because of the different types of surveys used, as well as the difficulty of finding similar organizations with which to compare an agency.

One agency may deal with terminally ill patients, a factor that could have a serious effect on satisfaction rates. Another might have a higher population of elderly people or Medicaid patients, both of whom, Long says, are more likely to be satisfied with their care than other groups. Patients with higher incomes are often less likely to be satisfied.

"All of these kinds of things are usually not factored into those surveys," she says. "You wouldn’t ask a person’s income level in a survey. There are so many variables that enter into the process that you can’t control for on a 10-item questionnaire."

What an agency gains in creating its own highly individualized survey, it may lose in benchmarking ability. For example, Hospice of the Valley is switching from a fairly detailed survey to one with fewer questions so it can benchmark its results nationally.

Wollmer admits that it’s a trade-off, but notes that even on a more general survey form, there’s room for customization. "We’ve tweaked it. We’ve added some questions that we wanted to ask. We just won’t be able to benchmark those answers nationally."

Despite the hassle, you can glean useful information from comparing your agency with others, Long says. "There’s some value in benchmarking, but you need to understand there are caveats and variables. Standardization is forthcoming, and benchmarking is inevitable. The greatest challenge, however, is to truly capture the essence of patient satisfaction in your surveys."


1. Long CO. The Arizona Home Health Care and Hospice Survey, 1999. A Final Report. Tempe, AZ: Arizona State University College of Nursing; 1999.

Diane Greeneich, President and CEO, Healthcare Systems Management Inc., 14820 W. Trevino Drive, Goodyear, AZ 85338. Telephone: (623) 536-9744. Fax: (623) 536-9745.

Carol O. Long, Assistant Professor, College of Nursing, Arizona State University, P.O. Box 872602, Tempe, AZ 85287-2602. Telephone: (480) 965-7444. Fax: (480) 965-0212. E-mail:

Sharon Sprenger, Project Director, Core Measure Identification and Evaluation, Division of Research, JCAHO, 1 Renaissance Blvd., Oakbrook Terrace, IL 60181. Telephone: (630) 792-5968. Fax: (630) 792-4968. E-mail:

Madeline E. Wollmer, Quality/Compliance Officer, Hospice of the Valley, 1510 E. Flower St., Phoenix AZ 85014. Telephone: (602) 530-6985. Fax: (602) 530-6904. E-mail: