Cheryl L. Stavins, RN, FACHE, vice president, Texas Children's Hospital, Houston

ORGANIZATIONAL INFORMATION
The institution is a freestanding, private, not-for-profit pediatric hospital founded in 1954. The 456-bed teaching hospital serves as the parent organization for an integrated delivery system that consists of a primary care physician network, health plan, insurance corporation, international division, and home health service. The metropolitan area served includes seven counties and a population of more than four million. The hospital is a tertiary care referral center, receiving patients from all over the United States and other countries, and also serves as the primary pediatric teaching site for its affiliated medical school. The hospital's mission is to provide the finest possible patient care, education, and research. Average occupancy is 89 percent, and 51 percent of the beds consist of special care unit beds. The emergency center treats more than 70,000 patients annually, and the outpatient specialty clinics generate more than 170,000 visits each year. More than $20 million in NIH (National Institutes of Health) grant-funded research is conducted annually within the hospital.

BRIEF STATEMENT OF THE PROBLEM
A major concern for any healthcare organization is patient satisfaction. When consumers have a choice in selection of healthcare providers, quality and service become the selection criteria. Quality of care remains an elusive measure for many, but customer service is readily understood and often becomes the ultimate driver in choosing a provider of care. Therefore, measurement of patient satisfaction and incorporating results to create a culture where service is deemed important should be a strategic goal for all healthcare organizations.

DESCRIPTION OF THE PROBLEM
The organization had used a written satisfaction survey from a major survey company for a number of years. Survey results were distributed to department directors, who were then asked to address issues and develop plans to improve scores. Survey results tended to be relatively good, with little variation on a quarterly basis. A Patient Satisfaction Committee was developed to provide focus and direction to the process.
As the vice president responsible for the patient relations department, I was the chief administrative member of the committee. I appointed the manager of the department, who was most closely responsible for the survey process, as chair of the committee. Other committee members included the director of the department; nursing director of the emergency department and acute care areas; assistant vice president of ambulatory services; representatives from the operating room, home health, the primary care physician corporation, marketing, public affairs, and quality improvement; and a physician with a strong interest in patient satisfaction and a strong background in statistics.
The committee's first task was to review and assess the current process for measuring patient satisfaction. The major issues identified from that review included lack of control over sample size or randomization of sample, length of time between responses and reporting of information, lack of a unified approach to results, generic questions that did not always suit a pediatric setting, and lack of widespread accountability within the organization toward improving service.

ADMINISTRATIVE DECISIONS
Armed with these concerns, the committee explored options available and the pros and cons of each approach. Consensus was quickly reached regarding switching to a telephone survey. Although it is a significantly more costly approach, a telephone survey had the distinct advantages of control to produce a randomized, statistically significant sample and relatively short reporting time. It also provided for an immediate response to clients who desired follow-up on complaints. These strategic advantages were deemed to outweigh the increased cost of the survey.
A well-known national organization was selected as the vendor to conduct the survey. This vendor was chosen because it lent credibility and anonymity to the responders and allowed for a national benchmarking of results, which was desired by our board of trustees. They also allowed us some flexibility in development of survey questions.
The response scale of the survey tool was one of our first challenges. The vendor used a four-point scale, and our physician/statistician committee member felt strongly that a five-point scale was the minimum acceptable for a statistically valid tool. A review of the literature and an opinion from an expert consultant from a nearby school of public health resulted in agreement on a five-point scale. Because this meant that we could no longer benchmark our results, we decided to keep two questions on a four-point scale for benchmarking and use a five-point scale for the rest. "Overall satisfaction" and "likelihood to recommend" were the two questions that remained on a four-point response and would be our benchmark questions.
The committee then reviewed each of the survey questions, a process that is now done on an annual basis. Each question is evaluated for its correlation to overall satisfaction, and those questions with low correlation were dropped from the survey. Because telephone surveys are limited in terms of the time a responder devotes to the process, each question had to achieve maximum impact. This limits department-specific questions, requiring buy-in and understanding of the process from the whole organization. The issue of accountability for results led the committee on a search for best practices. Tying financial rewards to results seemed to achieve the best outcomes. With support from the executive team and financial planning group, a program called P3 (people, performance, and preeminence) was developed. In addition to accountability for customer service, the organization was trying to promote fiscal accountability. P3, as an employee bonus program tied to both efforts, was adopted.
The program is divided into 50 percent financial goals and 50 percent customer service goals. Entry into the program is dependent on meeting budget targets. The financial portion of the bonus pool requires meeting an activity ratio developed by the finance department that measures expenses in relation to activity. The customer service portion was based on 25 percent of the pool paid for meeting or exceeding the national benchmark for overall patient satisfaction and 25 percent on a sliding scale based on the average of seven survey questions. The seven survey questions chosen were those whose results could be affected by the whole organization, such as cleanliness of the hospital and how child-friendly the hospital is. We did not choose any questions about physicians, for example, because the staff could not be directly accountable for those results. One percent of total salary dollars constitutes the pool of dollars to be paid for goal achievement.
The fiscal year for our organization ends on September 30. Timing for the bonus payout is the first week of December. This timing has proven to provide maximum appreciation of the bonus. Continuing debate occurs each year surrounding the methodology of the payout. We opted to pay a bonus based on 1 percent of each person's annual salary. Many such bonus programs pay a fixed amount regardless of a person's salary. The majority of our bonus checks average from $200 to $500, which is well received in December.
The introduction of P3 accomplished its goal of focusing attention on customer service and energized the patient-satisfaction survey process. Employees at all levels began asking about quarterly results and the status of P3 goals. Employee interest fostered management focus on ways to provide training and performance improvement projects to increase patient satisfaction. The committee produced a radar screen showing overall satisfaction, results on the seven P3 questions, and best and worst results. Each of the four areas surveyed-inpatient, outpatient, emergency center, and day surgery-demonstrated problems centered around waits and delays. A campaign centered around communication regarding delays might help to improve satisfaction when delays were unavoidable. Staff were encouraged to tell waiting families how long the wait might be or if the physician was delayed in surgery. "Don't wait-communicate" became a slogan for the organization.
Each year, the committee tries to keep interest and enthusiasm for customer service high by injecting something new into the program. Several years ago, we added a party and a photo banner, with the staff and CEO, for each quarterly winner of the best area results. The party (pizza or subs) was well received, but it was the photo banner that ignited the competitive spirit. The banner was displayed in a prominent public location for three months and then given back to the winning area. Employees were seen taking photos of the photo banner. The measure of the level of success of the banner, however, came from the emergency center. Employees seeking to win this honor started a "BOB" (be on the banner) campaign. They produced BOB buttons for all emergency center employees, had pens made that showed half a dozen customer service clues, and posted BOB banners in the employee lounge. This was a staff-led effort that unified and boosted morale in an area overwhelmed by staggering numbers of patients in a winter exacerbated by a meningitis scare. Unfortunately, winning results are still an elusive goal for the emergency center, but their effort was recognized in the employee newsletter and at a management forum.
Last year, the committee attempted to motivate departments that had little or no direct patient contact and tie them more closely to survey results. An additional requirement was added to the P3 bonus program. Each department had to participate in at least one customer service-related performance improvement project to allow its respective employees to qualify for the 50 percent of the P3 pool relating to customer service. They could partner with a direct patient care area, and the patient relations department would assist in finding a project for any department needing help. Because most direct patient care areas already had ongoing service-oriented performance improvement projects, they merely had to register a project to qualify. The resulting projects from departments such as accounting, materials management, and human resources exceeded our expectations.
Financial planning adopted the infant areas and spent lunch hours sewing isolette covers to create a nonstimulating environment for preemies. One employee brought her sewing machine in and taught others, including male employees, to sew the covers. Parents of these smallest patients deeply appreciated the caring gesture of employees they would never meet. Accounting employees spent time as sitters during lunch hours and in the evenings when patients' parents left for meals. Materials management spent evenings and weekends as greeters in the emergency center and gained a whole new perspective on the patients we serve.

RESULTS
The development of a Patient Satisfaction Committee has created a unified approach to customer service throughout our integrated delivery system. It has served to strengthen the survey process, elevate awareness of the need for customer service throughout the organization, and begin the enculturation of customer service beyond the traditional direct patient care providers. The maturation of the committee has resulted in a group of people with extensive knowledge regarding survey tools, processes, and statistical validity of results and an abiding passion for improving the service the organization provides to patients and families. Spillover from this committee has led to an enhancement of family-centered care; the creation of a parent advisory board; and a joint effort with our affiliated medical school to begin customer service training at all levels, including physicians.
The major lessons learned over the years are (1) training must be an integral part of any plan to improve patient satisfaction; (2) there is no one-time fix or approach to patient satisfaction, and the initiative must be a sustainable, permanent focus constantly refreshed through innovative approaches; and (3) there must be a torch-bearer group dedicated to keeping the effort in the forefront of the organization. A continuous examination of the reliability of data and accountability for results improves an institution's delivery of quality patient service.