Begin with the right foundation

Editor’s note: Patrick Quinlan is principal of Quinlan & Associates, an Adrian, Mich., marketing research firm, and a professor of marketing at Adrian College.

It’s not news that health care providers face increasingly competitive environments. Maintaining, or even increasing, market share can be critical to reaching financial goals. Positive word-of-mouth from satisfied users to non-users, as well as repeat use of the service offerings by satisfied users is vital to achieving this goal.

Patient perception of the quality of individual service offerings undoubtedly affects intention to use other services. For example, patient perception of the quality of an emergency room encounter affects intention to use other services of that hospital. Therefore, measurement of patient satisfaction is essential as hospitals strategically allocate limited resources to result in increased performance.

Satisfaction and service quality foundations

Several paradigms concerning the nature of consumer satisfaction have been generated, but disconfirmation theory, which suggests that satisfaction is a transaction-specific process where consumer expectations are compared to actual experience in specific service dimensions, is the most long-lived. (See graphic on the disconfirmation paradigm.)

In addition to gaining insight by comparing expected performance and perceived performance of individual service dimensions, we can also compare overall expectation of performance with overall perception of performance. An evaluation that measured satisfaction of individual service dimensions and measured overall satisfaction would yield more useful information.

Service quality is a construct resulting from a long-term evaluation of an organization’s service offerings. It is often measured with procedures quite similar to the expectation/performance approach of the disconfirmation theory and, in many instances, with the ServQual instrument developed by Parassuraman, Zeithaml and Berry. Their work identified a global definition of service quality as including tangibles, reliability, responsiveness, assurance and empathy. Service dimensions specific to a health care service encounter have also been discussed in the literature and are, for example: expressive caring, expressive professionalism, expressive competence (Bopp, 1990) and curing, caring and access (Joby 1989), among many others.

Although both consumer satisfaction and service quality issues have been explored in the marketing and health care literature, the nature of the relationship between service quality and satisfaction has been historically undefined. Understanding their combined impact on intention to purchase could provide great insight. Recent research suggests the following:

1. Perceptions of service quality may be the drivers of satisfaction.

2. Overall satisfaction impacts intention to purchase more significantly than does perception of performance in individual service dimensions alone.

3. Individual service dimensions impact satisfaction and dissatisfaction differently. This suggests the importance of developing separate diagnostic models to evaluate both satisfaction and dissatisfaction.

Applying these foundations

A rural Michigan hospital was involved in applying TQM principles to the challenges of servicing non-critical patients in their emergency room setting. Rather than rely on traditionally accepted global measurement scales alone, a series of focus groups with former patients, spouses and parents of minor patients was used to identify determinants of satisfaction applicable to this particular setting. This qualitative phase yielded 20 service quality dimensions and a single measure of the perceived value of an emergency room visit by the non-critical patient.

The survey instrument that was then developed contained two overall measures of satisfaction. One, relying on the disconfirmation theory, asked respondents if their overall expectations of service quality had been met, not met, or exceeded. The other, utilizing a Likert format, provided a 10-point scaling of overall satisfaction.

Despite concerns expressed by hospital employees that it would set the bar too high, the frequency with which the sample provided the highest rating, a 10 on the Likert scale, was selected as the benchmark to which future performance would be compared. The percentage of future respondents providing the highest rating would prove to be a more responsive measurement to changes in overall patient satisfaction.

Utilizing the disconfirmation paradigm as the foundation, both importance and performance on the service dimensions were measured. Demographic and behavioral measurements were also made. A multi-stage mail survey resulted in a 40 percent response rate that provided 350 usable respondents. Non-response error was unknown. Hospital administrators felt the sample fairly represented the population. Legal counsel had been consulted for insight into confidentiality compliance prior to contact with former patients. No confidentiality concerns were registered by sampling frame members receiving any of the mailed materials.

Analysis and results

The hospital administrators required data they could use to properly prioritize desired improvements in service quality. Furthermore, it was essential that the information be understandable to all members of the hospital staff.

Examination of the correlation matrix indicated that the data was appropriate for factor analysis. A principal component procedure with a Varimax rotation was conducted and a six-factor solution was selected. One factor, AIDS Precautions, was selected with an eigenvalue less than one primarily due to the intense discussions in earlier focus groups where participants expressed extreme concern that hospital procedures must minimize the potential exposure of patients in the ER to blood of other patients.

The six-factor solution included Hospital Comfort and Concern, Physician Competence, Responsiveness to Patients, Billing Accuracy and Appropriateness, Time and AIDS Precautions (see Figure 1). While these do not perfectly replicate factors identified in marketing and health care literature, similarities are evident.

Two stepwise discriminant analyses used factor scores as the set of predictor variables and the results of the overall satisfaction variable asking respondents if their expectations of performance had been met, not met or exceeded. Due to the untested nature of the instrument, a surrogate variable procedure was employed to provide factor scores.

Two separate discriminant analyses were performed: expectations met/not met and expectations met/exceeded as the dependent variable. The expectations met/not met analysis has important implications for the hospital since organizations that retain loyal customers may financially outperform organizations with higher market share and/or lower cost structures.

Using a scaling of the F value as a measure of the discrimination power of each factor, dissatisfaction was shown to be driven, in descending order, by patient perceptions of poor performance with Time (54 percent), Billing Accuracy and Appropriateness (23 percent), Hospital Comfort and Concern (17 percent), AIDS Precautions (4 percent) and Physician Competence (1 percent). Classification results indicated statistically significant differences between the groups.

The final step in analyzing dissatisfaction involved separate multiple regression analyses with the factor scores as the dependent variable and the performance rating for each of the variables constituting a factor as independent variables. The resulting Beta coefficients were scaled in response to the hospital’s need for an indication of the percent of impact of each variable to the factor.

For example, in the eliminating dissatisfaction chart (Figure 2), you can see the relative contribution of the three variables comprising the Time factor. Patient perception of Promptness of Physician Arrival, Time from End of Treatment to Final Discharge and Time of Total Encounter contributed fairly equally to patient perception of the hospital’s performance in moving them through a non-critical emergency room visit in a timely manner. This procedure was repeated to determine the relative impact of the individual variables to each of the other factors comprising this model.

Developing action plans aimed solely at minimizing dissatisfaction will not result in the intense loyalty and positive word-of-mouth that are characteristics of organizational excellence. Therefore, a second discriminant analysis of expectations met/exceeded was performed to understand the antecedents of an overall state where patient expectations were exceeded. As Figure 3 shows, this resulted from, in descending order of importance, outstanding performance in these factors: Physician Competence (34 percent), Time (27 percent), Responsiveness to Patient (26 percent), Hospital Comfort & Concern (8 percent), Billing Accuracy and Appropriateness (4 percent) and AIDS Prevention (1 percent).

As with the Eliminating Dissatisfaction model, the contribution of each individual variable to each factor was generated to assist the hospital in prioritizing improvements. As shown in the Increasing Satisfaction chart (Figure 3), the individual variables contributing to the Physician Competence factor included Perceptions of Correct Treatment (27 percent), Putting Patient at Ease (26 percent), Perceptions of Diagnosis Accuracy (24 percent) and Physician Including Others in the Decision Process (23 percent).

Additional analysis including the focus group results revealed significantly more dissatisfaction among parents of patients under the age of 12 years and with certain types of diagnoses.

As you might suspect, there was a great deal of discussion surrounding the set of questions to be included in the physician competence section of the questionnaire. Like many small hospitals, this ER was staffed by physicians provided via a contractual agreement with an outside firm. The desire of the hospital to secure a patient measurement of physician competency ultimately outweighed concerns that patients would not be capable of such an evaluation. It is interesting to note that the hospital’s decision to include this set of questions was influenced by preexisting concerns about the outside physicians’ competence voiced by some areas of the hospital.

Hospital response

Forty percent of respondents indicated that their overall expectation of hospital performance had been met, 20 percent indicated that their expectations had been exceeded and the remaining 40 percent rated hospital performance as not meeting expectations.

The hospital’s top priority became the reduction of the number of patients who claimed their expectations had not been met. They aimed their efforts at improvement in the Time factor. Since that factor also had a significant impact (27 percent) on satisfaction, the hospital also felt that improvement on this factor would increase the number of future patients who report that their expectations had been exceeded. To this end, the hospital invested in a major renovation of the emergency room area. Separate entrances were created for trauma patients and for ambulatory non-critical patients. Separate triage areas were created as well. Finally, they developed a “fast track” system to quickly move less critical patients through the emergency room encounter.

Another area of hospital response centered on the contracted ER physicians. Focus group participants discussed the “good old days” when the hospital emergency room was staffed by local physicians. They characterized the current physicians as “outsiders,” “unresponsive to community needs” and “uncaring.” On the other hand, internal measurements of physician quality pointed to a competent emergency room staff.

Despite the focus groups’ concerns and despite concerns of other factions in the hospital, the chief of staff was unwavering in his support of the contracted physicians. The hospital selected two actions aimed at improving satisfaction through the Physician Competence factor. Physician assistants were increasingly used in the diagnosis and treatment of patients. They also developed a public relations campaign to enhance the image of the ER physicians as being qualified, caring and highly competent.

Through using basic knowledge of customer satisfaction and service quality as a guide to data analysis decisions, this small hospital’s administrative staff felt secure in its decisions to allocate funds and efforts in order to bring about desired quality improvements.