patient satisfaction with health care: recent …

41

Corresponding author: Catherine Worthington, Faculty of Social Work, University of Calgary, 2500 University Drive NW, Calgary, AB T2N 1N4; <[email protected]>

PATIENT SATISFACTION WITH HEALTH CARE: RECENT THEORETICAL DEVELOPMENTS AND IMPLICATIONS FOR EVALUATION PRACTICE

Catherine WorthingtonUniversity of CalgaryCalgary, Alberta

Using client or patient satisfaction instruments to assess the quality of services and programs is an integral component of much program research and evaluation. But however methodologically straightforward and programmatically useful, the constructs and theory underlying patient or client satisfaction measures are not particularly simple, and the informativeness and usefulness of these satisfaction measures have been repeatedly called into question. To help program evaluators more effectively incorporate service users’ perspectives into program evaluations, this article reviews major approaches to, and research on, satisfaction theory, including disconfi rmation, fulfi llment and consumer models, and sociological perspectives, and discusses emerging approaches and implications for evaluation practice.

Utiliser des instruments de mesure de la satisfaction des clients ou des patients pour évaluer la qualité des services et des pro-grammes est une composante intégrale de bon nombre de travaux de recherche et d’évaluation de programme. Mais peu importe à quel point ils sont méthodologiquement simples et utiles du point de vue programmatique, les concepts et la théorie qui sous-tendent les mesures de la satisfaction des clients ou des patients sont complexes, et le caractère informatif ainsi que l’utilité de ces mesures de la satisfaction ont été maintes fois remis en question. Afi n d’aider les évaluateurs de programme à intégrer plus effi -cacement les points de vue des utilisateurs aux évaluations de programme, le présent article examine les principales méthodes utilisées ainsi que les travaux de recherche effectués sur la théorie de la satisfaction, y compris les modèles de réfutation, d’exécu-tion, et de consommation, et les points de vue sociologiques, et discute des nouvelles approches et de ce qu’elles signifi ent pour la pratique de l’évaluation.

The Canadian Journal of Program Evaluation Vol. 20 No. 3 Pages 41–63ISSN 0834-1516 Copyright © 2005 Canadian Evaluation Society

Abstract:

Résumé:

THE CANADIAN JOURNAL OF PROGRAM EVALUATION42

Using client or patient satisfaction instruments to assess the quality of services and programs is an integral component of much program research and evaluation. Philosophically, pragmatically, and intuitively, measuring patient or client evaluations of program elements and processes makes sense: it values the consumer voice and potentially empowers service users; can provide structured data on services in a relatively straightforward, cost-effective, and easily developed and administered manner; and has inherent face validity. What could be more obvious than asking for user opinions on service elements to gauge program performance? But however methodologi-cally straightforward and programmatically useful, the constructs and theory underlying patient satisfaction measures are not par-ticularly simple, and the informativeness and usefulness of patient satisfaction measures have been repeatedly called into question (Avis, Bond, & Arthur, 1997; Williams, Coyle, & Healy, 1998).

To help program evaluators more effectively incorporate service us-ers’ perspectives into program evaluations, this article reviews ma-jor approaches to, and research on, satisfaction theory, including disconfi rmation, fulfi llment and consumer models, and sociological perspectives, and discusses emerging approaches and implications for evaluation practice. Canadian researchers have made a small but signifi cant contribution to this research base, and in keeping with the theme of this special issue of the Canadian Journal of Program Evaluation, these contributions will be noted.

SATISFACTION THEORY

Although patient or client satisfaction research has been performed for several decades and many standardized satisfaction instruments and countless ad hoc measures have been developed (with varying but usually weak validity and reliability [Sitzia, 1999]), most re-search and program evaluation applications have been practical and problem-oriented, and therefore theoretically the area is underdevel-oped (Linder-Pelz, 1982; Locker & Dunt, 1978; Sitzia & Wood, 1997; Williams et al., 1998). However, a small but growing body of work from several disciplines and approaches has attempted to clarify the constructs and measures of both satisfaction and expectations, and identify the personal and service-related factors determining expec-tations and satisfaction. Most theoretical work has been carried out in the health care area, and so this literature will be the focus of the following review.

43LA REVUE CANADIENNE D’ÉVALUATION DE PROGRAMME

Satisfaction and Elements of Care

In the majority of studies, patient satisfaction is defi ned as an evalu-ation or set of evaluations of the medical intervention on a specifi c health problem or issue (Sitzia & Wood, 1997; Williams, 1994). Re-views of the large body of patient satisfaction studies have produced a range of potentially relevant dimensions of satisfaction, more properly labelled components or elements of health care (Sitzia & Wood, 1997), that may be evaluated by patients. Table 1 summarizes three of the most comprehensive reviews from different time periods, and shows the many elements used in patient satisfaction measures.

As is evident from the table, there is considerable variation in the specifi c elements of care that may be considered in studies of patient satisfaction. Although the diverse components of care produced by and used in different studies have been criticized for variability, these may be expected to vary according to the specifi c care context in which sat-isfaction is being measured and also according to the individuals and sub-populations measuring them (Avis, Bond, & Arthur, 1995; Like & Zyzanski, 1987; Sitzia & Wood, 1997). However, as may also be seen from the table, there is also considerable overlap in the components of care when these are grouped into broad categories.

Generally, through adoption of the Donabedian quality assurance framework (Donabedian, 1980, 1988), relevant elements of care for any particular medical care context may be categorized according to whether they are related to structure (facilities, personnel), proc-ess (technical process, interpersonal process), or outcomes (somatic, psychosocial, and fi nancial) of care (Kravitz, 1996; Pascoe, 1983). In research to date, measures of the dimensions of care have tended to focus on process variables, particularly around the interpersonal manner of the health care professional (Kravitz, 1996; Pascoe, 1983). As well, in addition to dimensions of care, an overall measure of sat-isfaction is usually taken in patient and client satisfaction research studies either directly through a Likert-scale type question on overall satisfaction with a service or through weighting and summing various ratings on elements of care.

There are several weaknesses in the current conceptualizations of satisfaction and the dimensions of care that are rated. First, the tra-ditional classifi cations of the components of care related to satisfaction largely have been constructed by service providers and represent serv-ice provider concerns or service provider perceptions of client concerns


Table 1Reviews of the Literature on Satisfaction with General Health/Medical Care

Components of Health Care Ware et al. (1983) Hall & Dornan (1988) Wensing et al. (1998) (number of studies not indicated) (221 studies) (57 studies)

Interpersonal manner of health care provider: concern, friendli-ness, disrespect, rudeness

Technical quality of care: thor-oughness, accuracy, unnecessary risks, making mistakes

Accessibility/convenience: time and effort required to get to an ap-pointment, waiting time at offi ce, ease of reaching care location

Finances: reasonable costs, alternative payment arrangements, comprehensiveness of insurance coverage

Effi cacy/outcomes: helpfulness of medical staff in improving or maintaining health

Continuity: sameness of provider and/or location of care

Physical environment: orderly facilities and equipment, pleasant-ness of atmosphere, clarity of signs and directions

Availability: presence of medical care resources

Overall satisfaction: global item or items or composite

Humaneness: warmth, respect, kindness, willingness to listen, appropriate nonverbal behaviours, and interpersonal skill

Technical competence: technical performance and competence defi nable in traditional medical terms

Outcome

Physical facilities: aesthetic and functional aspects, parking, and adequacy of equipment and laboratories

Continuity of care

Access: convenience, hours, dis-tance, perceived availability, and ease of getting appointments

Amount of information: explana-tions of treatment, procedures, or diagnoses

Cost

Bureaucracy/organization

Attention to psychosocial problems

Availability and accessibility: waiting times, fl exibility, telephone consultations, physical accessibil-ity, fi nancial accessibility

Organization and cooperation: premises, continuity, coopera-tion, effi ciency, special services available

Medical care: effectiveness, burden on the patient, compe-tence/accuracy

Doctor/patient relation: humane-ness, exploring patients’ needs, patients’ involvement in decisions, time for patient care, patients’ privacy

Information and support: infor-mativeness, stimulating self-help, counselling, supporting patients’ relatives

(Avis et al., 1997; Sitzia & Wood, 1997). Until recently, patients have been infrequently asked to defi ne the dimensions of care related to service satisfaction; therefore, these dimensions may be incomplete or mis-specifi ed (Wensing, Grol, & Smits, 1994), particularly since patients and health care providers differ in their perceptions of qual-ity health care dimensions (Jung, Wensing, de Wilt, Olesen, & Grol, 2000; Zandbelt, Smets, Oort, Godfried, & de Haes, 2004).


Second, the validity of a global satisfaction construct is in question. In traditional survey research on satisfaction, overall satisfaction with a service is generally found to be very high, often with 85% or more of respondents claiming to be generally satisfi ed with services. However, if questions are asked on specifi c dimensions of service or occurrences, much lower levels of satisfaction are found (Avis et al., 1997; Sitzia & Wood, 1997). An explanation for this discrepancy could lie in the conceptualization of satisfaction, or may be due to meth-odological weaknesses in satisfaction surveys and methods, including inadequate sampling, particularly of program drop-outs (Blais, 1990); social desirability bias; respondent situational bias; poor question wording; inadequate weighting of the various dimensions of care; and so on (Avis et al., 1997; Locker & Dunt, 1978).

Perhaps most importantly, the generally employed, often implicit, models underlying the determinants of satisfaction are simplistic and inadequate. Traditional patient satisfaction theory rests on social-psychological theory, and assumes that the expression of satisfaction is an expression of an attitude (an affective response) that is related to both the belief (expectation) that the care possesses certain at-tributes (dimensions) and the patient’s evaluation of these attributes (fi rst defi ned in Linder-Pelz, 1982, drawing upon the Fishbein and Ajzen formulation of attitude [Fishbein & Ajzen, 1975]). Measures of satisfaction, then, potentially provide both a measure of care and a measure of the patient or client who provides the rating (Pascoe, 1983; Ware, Snyder, Wright, & Davies, 1983). This general approach has been called the disconfi rmation paradigm (Zegers, 1968). Within this approach, dissatisfaction is hypothesized to increase as the dis-parity between a standard of care and perceived care increases. Other models of satisfaction include fulfi llment theory and consumer or marketing theories. There is also an emerging sociological perspective on satisfaction that places these theories within their interactional contexts. Each of these will be briefl y reviewed, and their contribu-tions to satisfaction theory and defi ciencies will be discussed.

Disconfi rmation Theories of Satisfaction

Theories subsumed under the disconfi rmation paradigm include dis-crepancy theory and equity theory, both of which draw on Lawler’s (1973) work on job satisfaction. Simple discrepancy theory of patient satisfaction is the most widely adopted (Kravitz, 1996; Pascoe, 1983; Sitzia & Wood, 1997) and defi nes satisfaction as the difference be-tween what is desired or thought to be needed (“I want”/ “I need”)


and what is perceived to occur. There is mixed research support for an association between desires and satisfaction (Kravitz, 1996; Sitzia & Wood, 1997). For example, Like and Zyzanski (1987) found that fulfi llment of patient desires by the physician was strongly correlated with visit satisfaction; however, more recent research (Peck et al., 2004) using prospective methods has found no relationship between satisfaction and met or unmet expectations in men visiting their primary health care provider.

Less frequently, discrepancies are measured using perceptions of equity or entitlement (“services should be”/ “I deserve”) or as the difference between what is thought to be likely and what actually occurs, that is, a probabilistic approach rather than a value-based one (Kravitz, 1996). While there is some evidence that these two ap-proaches may be relevant to patient satisfaction, the research in these areas is too limited to allow for any fi rm conclusions (Davis, Albino, Tedesco, Portenoy, & Ortman, 1986; Sitzia & Wood, 1997; Smith & Sanderson, 1992).

In research performed to date, simple discrepancy/disconfi rmation theories have not been found particularly successful in explaining variation in satisfaction. Generally, and irrespective of the type of expectation measured (desires, entitlements, or probabilities), the variation explained by discrepancies between expectations and sat-isfaction, controlling for socio-demographic differences, remains at or under 20% (Williams, 1994).

Fulfi llment Theory of Satisfaction

Fulfi llment theory is often categorized with discrepancy theories (e.g., Kravitz, 1996; Linder-Pelz, 1982), although it comes from an older tradition in job satisfaction research that posits that satisfaction results from the outcomes of an experience regardless of how much one feels he or she should get and/or wants to receive (see Lawler, 1973; Pascoe, 1983). It is assumed that there is a direct relationship between factors such as pay or promotion and satisfaction. In the job satisfaction and consumer satisfaction research areas, it has been found that both discrepancy and fulfi llment factors are predictors of satisfaction (Lawler, 1973; Oliver, 1993). In the health care area, if fulfi llment is related to satisfaction, one would expect a direct relationship between satisfaction and outcomes of care, such as im-proved physical, mental, or social functioning. As noted earlier, most satisfaction research has focused on process measures, and so there is


little information available on the relationship between satisfaction and outcomes of health care. However, one study (Covinsky et al., 1998) of older adult inpatients suggests that there is no association between health status change and satisfaction, once baseline health status is controlled for. Those with higher functioning or outcomes will be more satisfi ed whether there was improvement during the period of care or not.

Consumer Models of Satisfaction

Both discrepancy and fulfi llment approaches to patient satisfaction have been judged logically inadequate (Pascoe, 1983). Fulfi llment theory assumes that objective outcomes alone determine satisfaction, while discrepancy theories assume that any deviation from what is expected will create dissatisfaction, whether the outcome is more negative or more positive than was expected. As early as 1983, Pascoe suggested that consumer models of satisfaction might be incorporated into patient satisfaction modelling to take into account the type and degree of discrepancy. Contrast theory suggests that when consumers perceive a discrepancy between expectations and outcome, they will magnify the difference. (So, for example, a car that does not perform to expectations will be perceived as a “complete lemon.”) Assimilation theory (from Festinger’s [1957] cognitive-dissonance theory) suggests that inconsistencies between expectations and outcomes will be de-creased or assimilated in order to adjust perceptions of outcomes to be consistent with expectations. The combination of these two models — the assimilation-contrast theory — suggests that a consumer’s sat-isfaction response will be nonlinear: assimilation will occur within a certain range of discrepancy between expectations and outcome (“the zone of tolerance”), while outside that range, the contrast effect will occur and inconsistencies will be magnifi ed, leading to either dissat-isfaction or higher satisfaction than within the assimilation range. There appears to be support for the assimilation-contrast model in consumer research (Oliver, 1997; Pascoe, 1983). In the health care area, however, little if any research has been performed on these models, although some research from a more sociological perspective (discussed in the next section) provides some insights into the con-cept of a “zone of tolerance” in health care (Edwards, Staniszweska, & Crichton, 2004).

Finally, one of the most ambitious contributions to social-psychologi-cal satisfaction theory comes from work conducted by Oliver (1993, 1997). His model incorporates fulfi llment theory, discrepancy and


equity disconfi rmation theories, and assimilation-contrast theory in a model of consumer satisfaction. In addition, he adds affect as an explicit component of his satisfaction model. He criticizes most satis-faction theories for being primarily cognitive, when satisfaction is, by defi nition, and according to research results emerging from consumer theory, part cognitive reaction and part affective reaction. Oliver suggests that, based upon research results, affect is composed of two dimensions, positive and negative, which operate simultaneously. He tested the portion of his model related to affect with automobile purchasers and students evaluating a course, and found (a) that af-fect comprised separate positive and negative dimensions, (b) that attribute satisfaction and dissatisfaction were signifi cantly related to positive and negative affect, respectively, and (c) that both were related to overall satisfaction.

The newest research on affect and satisfaction extends Oliver’s model of positive and negative affect to incorporate concepts of “delight” and “outrage” (Arnold, Reynolds, Ponder, & Lueg, 2005; Schneider & Bowen, 1999). Schneider and Bowen (1999) move well beyond con-ventional cognitive modelling to argue that outrage is produced when customers’ fundamental needs of security, justice, and self-esteem are degraded during a shopping experience. Some work in the health care area supports this needs-based approach. Coyle (1999; Coyle & Williams, 2001), in her qualitative work on patient dissatisfaction, has found that patient dissatisfaction results when patient identity is threatened by experiences perceived as dehumanizing, objectifying, disempowering, and devaluing. She proposes the idea of “personal identity threat” as a framework for the development of patient dis-satisfaction instruments (Coyle, 1999; Coyle & Williams, 2001).

Consumer-based models of satisfaction appear promising, as they build upon previous research and acknowledge the complexity of the factors related to or determining satisfaction. However, attempts to transfer or apply consumer standardized measures of satisfac-tion (such as the SERVQUAL instrument) to health care have been only partially successful due to the multiple dimensions of service involved in health care and specifi c service and cultural contexts (Alden, Hoa, & Bhawuk, 2004; Bowers, Swan, & Koehler, 1994; Car-man, 1990; Parasuraman, Zeithaml, & Berry, 1988). As well, there is some philosophical opposition to the notion of health care recipient as a “consumer” in the Canadian context (Deber & Sharpe, 1999; Feldberg & Vipond, 1999). It has also been suggested that while consumer models have begun to address emotional aspects of service


reactions, they neglect the important relationship between health care professional and patient, and other social contexts of health care (Williams, 1994). Sociological perspectives on satisfaction attempt to address these concerns.

Sociological Perspectives on Satisfaction

The fi nal set of emerging theory and research to be reviewed examines the interactional and social aspects of determinants of satisfaction. This sociological research questions the veracity of the direct expecta-tions–satisfaction theoretical link, primarily because social-psycho-logical theory does not recognize the social context of health care or the social infl uences on health care recipients. For example, although “consumer” perspectives on the health care system have been widely sought recently and consumer models have been applied to patient satisfaction, the supposition that the patient/client necessarily adopts a consumerist role in obtaining medical care has been challenged by critical health researchers (Lupton, Donaldson, & Lloyd, 1991). Consumer models are based on a primarily rational view of patient behaviour that assumes that patients have bargaining power, freedom of choice, the knowledge and the motivation to choose a particular option from available services, and the power to challenge medical authority. However, Lupton and colleagues (1991) found that these assumptions do not necessarily hold at the individual choice level. In a survey of 333 patients attending general practitioners in Sydney, Australia, patients surveyed tended not to think of themselves as consumers who should be wary of the quality of the service offered by doctors; rather, they preferred to trust their doctor, and therefore did not devote effort to actively seeking out information about their doctor or evaluating his or her services. In a subsequent study, Lupton (1997) found that individuals’ personal characteristics (particularly age) infl uenced the degree to which they wished to act as consumers (as opposed to more passive patients), but that the degree of control patients wanted in the medical interaction also depended upon the medical context.

Two studies were conducted that question whether patients evaluate services at all when not impelled to do so by forced choice surveys, and if they do, how they evaluate services (Avis et al., 1997; Williams et al., 1998). In the fi rst study, Avis et al. (1997) repeatedly interviewed 81 cardiology or respiratory patients to obtain their changing perspec-tives on satisfaction with their care. Results suggested that patients are uncertain in their expectations; that these may be based on pre-


vious negative experiences, hearsay, and background assumptions about medical care; and that these change over time with expectations becoming more pronounced as experience and knowledge increases. Also, they found that patients tended to excuse poor performance. Avis and colleagues suggest that satisfaction may result from psychosocial aspects of the health care encounter other than expectations, such as relief, gratitude, fear of wasting a physician’s time, or confi dence in the health care professional. They also suggest that satisfaction with the health care experience may not be particularly relevant to patients in terms of desired outcome, particularly in situations where patients may learn “bad news” or must cope with debilitating condi-tions. For example, a recent study demonstrated that more than 75% of cancer patients were satisfi ed or very satisfi ed with their overall pain management, despite almost half reporting moderate to severe pain (Dawson et al., 2002).

In the second qualitative study investigating whether and how pa-tients evaluate services, Williams and colleagues (1998) found that mental health clients provided value descriptions of their experience of services; however, these were not value descriptions or evaluations of the service itself. Williams et al. suggest that there are two parts to the move from valuing/judgement of the experience to valuing/judgement of the service: (a) Consistent with expectations theory, it appears that patients have perceptions of the duties of services or personnel. These obligations of services are perceived as what will happen or what should happen, although patients hold these beliefs or expectations with varying degrees of certainty. Most often, service users expressed their expectations with a large degree of uncertainty. (b) The second step in the valuing/judgement of a service itself is the attribution of culpability or blame for a negative experience to the service. Most individuals in the study did not negatively evaluate the service or personnel even if they had a negative experience and the service had “failed” in its perceived duties. Usually, patients found suffi cient mitigating circumstances to excuse the service or personnel. For example, a long waiting period for service was excused as “the usual government waiting time” (p. 1355), or the lack of resolution to a mental health problem was explained away because the patient concluded that “that’s down to me really to solve the problem. There’s no medication they can give me” (p. 1356). Williams et al. suggest that the generally high ratings of services found on satisfaction surveys relate to the lack of congruence between ratings of the experience and ratings of the service itself.


Edwards and colleagues (2004) recently conducted a follow-up study to the Williams et al. (1998) exploration. They conducted re-peated in-depth interviews with 19 elective orthopaedic patients, and found that only 1 of the 19 did not refl ect on their treatment in a way that shifted their assessment of their care to a positive view. They assessed the patients’ processes of refl ection that infl uence the transformation of negative opinions of experiences to positive evaluations of services, and identifi ed three psychosocial pressures that infl uence patients’ interpretations (or at least expressions of interpretation) of events: the relative dependency of patients within the health care system, social etiquette and their need to maintain constructive working relationships with those providing care, and their general preference for holding a positive outlook. These proc-esses allowed patients to hold seemingly contradictory views. For example, one participant who had a long wait for surgery said she did not mind “waiting her turn,” but in the next breath modifi ed this to, “I do mind waiting my turn! … but I understand where he [the GP] was coming from” (Edwards et al., 2004, p. 172). Edwards and colleagues conclude that some degree of cognitive-dissonance (Festinger, 1957) helps patients cope with the health issue in a positive way, and suggest that if a “zone of tolerance” exists as sug-gested by the assimilation-contrast model from the consumer satis-faction literature, it is infl uenced to a large degree by the amount of choice a patient has, and further, that if a patient has little or limited choice in the care provided, the zone of tolerance will be very wide. In fact, the positive reinterpretations will become the default action. Similar results were found in a qualitative Cana-dian study (Sinding, 2003) conducted with those who had provided care to a relative or friend who had died of breast cancer. In this care context, participants muted their potential expressions of dis-satisfaction, indicating there was no “point” in expressing negative opinions after the death of the cared-for person.

The results of these three studies suggest a further complexity in sat-isfaction theory and measurement. While not necessarily inconsistent with social psychological expectations theory, these results suggest that an individual’s social context and the interpersonal and power relationships between health care professionals and patients must be incorporated into modelling, and that broader determinants of satisfaction (beyond individual perceptions and attitudes) should be considered. As work in medical anthropology and sociology suggests, factors beyond individual perception infl uence the way in which peo-ple understand health care services, including the diverse symbolic


meanings society attaches to these services and the social control inherent in them (Annandale, 1998).

Expectations

As noted, in most models of satisfaction, expectations (whether op-erationalized as measures of values, entitlements, or probabilities) are included as the major determinant. The Avis et al. (1997) study suggests that the formation of these beliefs is complex and infl uenced by a variety of factors, both individual and social. There is little theory or research available that focuses on the determinants of patient expectations (Kravitz, 2001; Thompson & Sunol, 1995). Kravitz and colleagues (1996) investigated infl uences on the development and expression of expectations. Based upon transcript analysis for 88 individuals who identifi ed their internists’ omissions of care, they proposed a preliminary model of patient expectations in the symptom-driven clinical context. Perceived vulnerability to particular health issues (e.g., due to age or family history), past experiences with similar symptoms (either personal or acquired while caring for others) and transmitted knowledge (e.g., from friends, via the media, or through health education) were found to infl uence expectations by affecting the interpretation of symptoms and by establishing an implicit standard of care. Kravitz (2001) posits that in the clinical context, the encoun-ter therefore becomes a negotiation between patient and physician. However, given the sociological research on satisfaction, it is clear that this negotiation is situated within the context of professional-patient relationships, including the power and knowledge imbalance inherent between health care or social service professional and patient, and the resulting uncertainty or tentativeness of expectations.

Patient Characteristics

Patient characteristics comprise the other major category of determi-nants of satisfaction found in the literature (Hall & Dornan, 1988). However, these have been found to be a minor predictor of satisfaction (Sitzia & Wood, 1997), and results to date have been inconsistent, perhaps because the characteristics refl ect complex social factors. The only generally consistent fi nding in the literature has been that age is positively related to satisfaction. In American studies, educational attainment has been found to be negatively related to satisfaction, but this variable may be confounded with income (which is a major factor in health care in the United States). The relationship of satis-


faction to other measures of socio-economic status, sex/gender, racial or ethnocultural group, and psychological state (distress/anxiety) have all been found to be situation and context dependent (Sofaer & Firminger, 2005).

EMERGING APPROACHES AND IMPLICATIONS FOR EVALUATION PRACTICE

Several weaknesses of satisfaction theory and research have been identifi ed. In the majority of published studies, the components of care rated by patients using patient satisfaction instruments or surveys traditionally have been defi ned by health care professionals, leav-ing open the question of whether these dimensions are exhaustive, or are those of most importance to service users. In the relatively small number of studies that explicitly test psychosocial satisfaction theory, the generally accepted and adopted discrepancy theory has been shown to be inadequate. At the psychological level, consumer theory suggests factors such as affect and assimilation-contrast re-sponses should be incorporated into models. More recent work from a sociological perspective suggests that the framing of the construct satisfaction may itself be simplistic, and that the link between ratings of the experience and ratings of the care is mediated by factors such as patient health system expectations and the power relationships inherent in the health care provider-patient interaction, as well as broader social factors such as the patient’s adoption of the role of “consumer” or more passive health care recipient, cultural factors, and so on. These factors also need to be taken into consideration in models of patient satisfaction. Current knowledge about expectations and expectations formation is even more rudimentary. Calnan (1988) suggests that a model of patient (“lay”) evaluation of health care must move away from simplistic satisfaction models toward an attempt to understand the logic, knowledge and beliefs of individuals, including purposes in seeking professional help, their level of previous experi-ence with health services, cultural values placed on health care, and personal images of health.

Given the documented shortcomings of satisfaction theory, what is a program evaluator to do? Obviously, as an accepted (and expected) practice by both program managers and service users, satisfaction measures are not going to simply fade away. Newer research fi nd-ings and approaches to capture users’ views of services provide some suggestions regarding appropriate methods.


There have been several calls for qualitative research into patient satisfaction (e.g., Aharony & Strasser, 1993, Sitzia & Wood, 1997), and indeed, over the past 10 years, there have been a growing number of published studies that use qualitative methods to document experi-ences with services rather than assess satisfaction directly. In areas as diverse as primary care (Anderson et al., 2001; Conacato & Fein-stein, 1997), HIV testing in Canada (Worthington & Myers, 2002), nursing care (Larrabee & Bolden, 2001), and sexual health planning (Baraitser, Pearce, Blake, Collander-Brown, & Ridley, 2005), qualita-tive methods have been used to elicit descriptions and defi nitions of elements of care that demonstrate quality from the patient perspec-tive. These studies show the same range and types of elements seen in more traditional patient satisfaction measures (see Table 1). Sofaer & Firminger (2005) summarize dimensions captured in 12 qualitative studies into seven categories: patient-centred care, access, commu-nication and information, courtesy and emotional support, technical quality, effi ciency of care/organization, and structure and facilities. The advantage of these descriptions is that they are service-specifi c, avoid some of the demonstrated diffi culties with eliciting negative evaluations of services, and provide direct descriptions of service elements which service providers may examine for congruence with quality guidelines or best practices in a service area. However, quali-tative approaches have their own challenges. While they have become widely accepted in evaluation and the social sciences generally over the past several decades (Patton, 2002), they are time-consuming, unstructured endeavours that rely heavily on the skills and integrity of the researcher-evaluator. If undertaken without adequate rigour and attention to both investigator preconceptions and relevant theory, qualitative research can lead to anecdotal, unrefl ective, and ultimate-ly inadequate fi ndings that are not useful for program improvement (Pope, Ziebland, & May, 2000).

Another approach that has been growing in popularity is the use of qualitative methods to assist in the development of satisfaction instruments that are specifi c to different care contexts. Canada has a strong cadre of qualitative health researchers, and many of these studies are Canadian (e.g., Boechler & Neufeld, 2002; Hilton, Budgen, Molzahn, & Attridge, 2001; Hudak, Hogg-Johnson, Bombardier, Mc-Keever, & Wright, 2004; Hudak, McKeever, & Wright, 2004). With these instruments, qualitative methods (e.g., thematic analysis of individual interviews and/or focus group data) are used to identify elements of care that are incorporated into satisfaction instruments. Others have heeded the calls from sociological researchers to develop


measures that focus on service experiences rather than evaluations of service in order to avoid the power and social dynamics inherent in judging services (Webb, Clifford, & Graham, 1999). But early results from one such set of instruments, the Picker surveys in the United Kingdom (“National Patient Experience Survey,” developed based on qualitative studies with patients to focus directly on experiences) show the familiar pattern of positively rated experiences (Edwards et al., 2004). Another approach that has been suggested is to focus on dissatisfaction. Coyle and Williams (2001) constructed an instru-ment based on qualitative fi ndings using a dissatisfaction approach, more particularly the theory of “personal identity threat” (previously described) that they hypothesize leads to dissatisfaction. Using this approach with a small (n = 97) clinical sample in a hospital, they measured dimensions of personalization, empowerment, information, approachability and availability of staff, and respectfulness, and found higher percentages of negative experiences reported with some care dimensions than would be found with a typical satisfaction sur-vey. For example, 44% agreed or strongly agreed with the statement “I do not like to bother the nurses because they are so busy,” but for many other statements a high percentage reported positive experi-ences (e.g., 93% agreed or strongly agreed with the statement “I felt very reassured by the doctors”). Thus, unless we assume superlative service provision in these studies’ service contexts, these approaches to measure care experiences in a standard survey format need more developmental work before they are widely useful in capturing vari-ation in experiences.

CONCLUSION

So where does that leave the program evaluator? At the least, those implementing patient or client satisfaction instruments should be aware of the diffi culties and complexities inherent in their theoreti-cal foundations, and exercise caution and thoughtfulness in their use in a specifi c context. Satisfaction levels of 80% and higher should be expected on surveys and not be considered clear evidence of high performing services or programs without substantial additional in-formation (i.e., triangulation of results from other sources). If time and resources allow, qualitative methods should be considered as an alternative to surveys where possible, or as a complement to them. In addition, the social, interactional, and power dynamics of health and social services should be taken into consideration wherever possible, whether the evaluator chooses to frame this in their work as social desirability bias to be addressed in survey construction, or uses con-


cepts from emerging satisfaction research such as “zone of tolerance” or “identity threat” in the way research questions are framed and data collection is designed. In any case, the implication for evaluat-ing quality service provision is clearly that attention needs to be paid to the context of patient-provider relationships and communication when conducting evaluations.

An emphasis on relationship factors in health is not new, and at present patient-centred care is a prominent model of service being promoted in many areas (Arbuckle & Herrick, 2006; Floyd & Seale, 2002; Gertais, Edgman-Leviatn, Daly, & Delbanco, 1993; Stewart, 2003). Satisfaction measures have long included interpersonal man-ner or provider-client relations as the most extensive area of ques-tioning, and evaluation research using a patient-centred perspective certainly shows that provider behaviours are linked to satisfaction. For example, Kim, Kaplowitz, and Johnston (2004) found that pa-tient-perceived physician empathy signifi cantly infl uenced patient satisfaction (and compliance) via the mediating factors of information exchange, perceived expertise, interpersonal trust, and partnership. Feddock et al. (2005) found that patient dissatisfaction with waiting time was associated with a shorter physician visit (i.e., patients did not mind a longer wait as long as they were able to spend more time with their physician). In their review of patient perceptions of quality using a patient-centred approach, Sofaer and Firminger (2005) have come to the conclusion that patients need to be educated about the pathways through which positive health outcomes are achieved so that they can take a more active role in assessing quality. If adopted within a comprehensive patient-centred approach (where the patient is a partner in planning and monitoring care and outcomes) this method makes sense. However, where there is a more traditional ap-proach to care, less egalitarian power dynamics, and social etiquette that keeps patients from expressing contrary opinions, placing more responsibility on patients to be fully informed about their care and progress is simply placing additional burden on those the service is designed to assist. And given the sociological research fi ndings from the last decade on the social contexts of care that infl uence expression of patient satisfaction, it appears that these social factors are impor-tant, and thus this approach does not appear widely applicable.

Finally, one Canadian researcher (Sinding, 2003) has pointed out that program evaluation generally (unless conducted from a critical or emancipatory perspective) ignores systemic and structural issues, and that these also deserve an evaluator’s attention. The health


care system has been of growing public concern in Canada. When patients excuse service performance on the basis of limited resources, staff pressures, or funding constraints, they are acknowledging and accepting a health care systems reality. However, this masks the need for evaluators and evaluation recommendations to consider “the big picture,” view services within their service system contexts, and include recommendations to promote services development as a component of their broader systems of health care.

ACKNOWLEDGEMENTS

The author would like to acknowledge support provided by the Ca-nadian Institutes for Health Research (CIHR) (New Investigator award), and the Alberta Heritage Foundation for Medical Research (AHFMR) (Population Health Investigator award).

REFERENCES

Aharony, L., & Strasser, S. (1993). Patient satisfaction: What we know and what we still need to explore. Medical Care Review, 50(1), 49–79.

Alden, D.L., Hoa, D.M., & Bhawuk, D. (2004). Client satisfaction with reproductive health-care quality: Integrating business approaches to modeling and measurement. Social Science & Medicine, 59(11), 2219–2232.

Anderson, R., Barbara, A.M., Weisman, C., Scholle, S.H., Binko, J., et al. (2001). A qualitative analysis of women’s satisfaction with primary care from a panel of focus groups in the national centers of excel-lence in women’s health. Journal of Women’s Health & Gender-Based Medicine, 10, 637–747.

Annandale, E. (1998). The sociology of health and medicine. Cambridge: Polity Press.

Arbuckle, M., & Herrick, C. (2006). Child and adolescent mental health: interdisciplinary systems of care. Sudbury, MA: Jones and Bartlett.

Arnold, M J., Reynolds, K.E., Ponder, N., & Lueg, J.E. (2005). Customer de-light in a retail context: Investigating delightful and terrible shopping experiences. Journal of Business Research, 58(8), 1132–1145.


Avis, M., Bond, M., & Arthur, A. (1995). Satisfying solutions? A review of some unresolved issues in the measurement of patient satisfaction. Journal of Advanced Nursing, 22, 316–322.

Avis, M., Bond, M., & Arthur, A. (1997). Questioning patient satisfaction: An empirical investigation in two outpatient clinics. Social Science and Medicine, 44(1), 85–92.

Baraitser, P., Pearce, V., Blake, G., Collander-Brown, K., & Ridley, A. (2005). Involving service users in sexual health service development. Journal of Family Planning and Reproductive Health Care, 31(4), 281–284.

Blais, R. (1990). Assessing patient satisfaction with health care: Did you drop somebody? Canadian Journal of Program Evaluation, 5(2), 1–13.

Boechler, V., & Neufeld, A. (2002). Evaluation of client satisfaction in a community health centre: Selection of a tool. Canadian Journal of Program Evaluation, 17(1), 97–117.

Bowers, M.R., Swan, J.E., & Koehler, W.F. (1994). What attributes deter-mine quality and satisfaction with health care delivery? Health Care Management Review, 19(4), 49–55.

Calnan, M. (1988). Towards a conceptual framework of lay evaluation of health care. Social Science and Medicine, 27(9), 927–933.

Carman, J.E. (1990). Consumer perceptions of service quality: An assess-ment of the SERVQUAL dimensions. Journal of Retailing, 66(1), 33–55.

Conacato, J., & Feinstein, A. R. (1997). Asking patients what they like: Over-looked attributes of patient satisfaction with primary care. American Journal of Medicine, 102(4), 399–406.

Covinsky, K.E., Rosenthal, G.E., Chren, M.-M., Justice, M.C., Fortinsky, R.H., Palmer, R.M., et al. (1998). The relation between health status changes and patient satisfaction in older hospitalized medical pa-tients. Journal of General Internal Medicine, 13, 223–229.

Coyle, J. (1999). Exploring the meaning of ‘dissatisfaction’ with health care: The importance of ‘personal identity threat’. Sociology of Health & Illness, 21(1), 95–123.


Coyle, J., & Williams, B. (2001). Valuing people as individuals: Development of an instrument through a survey of person-centredness in secondary care. Journal of Advanced Nursing, 36(3), 450–459.

Davis, E.L., Albino, J.E., Tedesco, L.A., Portenoy, B.S., & Ortman, L.F. (1986). Expectations and satisfaction of denture patients in a university clinic. Journal of Prosthetic Dentistry, 55(1), 59–63.

Dawson, R., Spross, J.A., Jablonski, E.S., Hoyer, D.R., Sellers, D.E., & Solomon, M.Z. (2002). Probing the paradox of patients’ satisfaction with inadequate pain management. Journal of Pain and Symptom Management, 23(3), 211–220.

Deber, R.B., & Sharpe, N. (1999). One million decisions at the micro-level: Patient choice. In D. Drache & T. Sullivan (Eds.), Market limits in health reform: Public success, private failure (pp. 275–291). London: Routledge.

Donabedian, A. (1980). The defi nition of quality and its assessment. Ann Arbor, MI: Health Administration Press.

Donabedian, A. (1988). The quality of care: How can it be assessed? Journal of the America Medical Association, 260(12), 1743–1748.

Edwards, C., Staniszweska, S., & Crichton, N. (2004). Investigation of the ways in which patients’ reports of their satisfaction with healthcare are constructed. Sociology of Health & Illness, 26(2), 159–183.

Feddock, C.A., Hoellein, A.R., Griffi th, C.H., Wilson, J.R., Bowerman, J.L., Becker, N.S., et al. (2005). Can physicians improve patient satisfac-tion with long waiting times? Evaluation & the Health Professions, 28(1), 40–52.

Feldberg, G., & Vipond, R. (1999). The virus of consumerism. In D. Drache & T. Sullivan (Eds.), Market limits in health reform: Public success, private failure (pp. 48–64). London: Routledge.

Festinger, L. (1957). A theory of cognitive dissonance. New York: Harper and Row.

Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention and behavior: An introduction to theory and research. Reading, MA: Addison-Wesley.


Floyd, M., & Seale, P. (2002). Substance abuse: A patient-centered approach. Abingdon, UK: Radcliffe Medical Press.

Gertais, M., Edgman-Leviatn, S., Daly, S., & Delbanco, T.L. (1993). Through the patient’s eyes: Understanding and promoting patient-centered care. San Francisco: Jossey-Bass.

Hall, J.A., & Dornan, M.C. (1988). What patients like about their medical care and how often they are asked: A meta-analysis of the satisfaction literature. Social Science and Medicine, 27(9), 935–939.

Hilton, B.A., Budgen, C., Molzahn, A.E., & Attridge, C.B. (2001). Developing and testing instruments to measure client outcomes at the Comox Valley Nursing Center. Public Health Nursing, 18(5), 327–339.

Hudak, P.L., Hogg-Johnson, S., Bombardier, C., McKeever, P.D., & Wright, J.G. (2004). Testing a new theory of patient satisfaction with treat-ment outcome. Medical Care, 42(8), 726–739.

Hudak, P.L., McKeever, P.D., & Wright, J.G. (2004). Understanding the meaning of satisfaction with treatment outcome. Medical Care, 42(8), 718–725.

Jung, H.P., Wensing, M., de Wilt, A., Olesen, F., & Grol, R. (2000). Com-parison of patients’ preferences and evaluations regarding aspects of general practice care. Family Practice, 17(3), 236–242.

Kim, S.S., Kaplowitz, S., & Johnston, M.V. (2004). The effects of physician empathy on patient satisfaction and compliance. Evaluation & the Health Professions, 27(3), 237–251.

Kravitz, R.L. (1996). Patients’ expectations for medical care: An expanded formulation based on review of the literature. Medical Care Research and Review, 53(1), 3–27.

Kravitz, R.L. (2001). Measuring patients’ expectations and requests. Annals of Internal Medicine, 134, 881–888.

Kravitz, R.L., Callahan, E.J., Paterniti, D., Antonius, D., Dunham, M., & Lewis, C.E. (1996). Prevalence and sources of patients’ unmet expec-tations for care. Annals of Internal Medicine, 125, 730–737.

Larrabee, J.H., & Bolden, L.V. (2001). Defi ning patient-perceived quality of nursing care. Journal of Nursing Care Quality, 16, 34–60.


Lawler, E.E. (1973). Motivation in work organizations. Monterey, CA: Brooks/Cole.

Like, R., & Zyzanski, S.J. (1987). Patient satisfaction with the clinical encounter: Social psychological determinants. Social Science and Medicine, 24(4), 351–357.

Linder-Pelz, S. (1982). Toward a theory of patient satisfaction. Social Science and Medicine, 16, 577–582.

Locker, D., & Dunt, D. (1978). Theoretical and methodological issues in so-ciological studies of consumer satisfaction with medical care. Social Science and Medicine, 12, 283–292.

Lupton, D. (1997). Consumerism, refl exivity and the medical encounter. Social Science and Medicine, 45(3), 373–381.

Lupton, D., Donaldson, C., & Lloyd, P. (1991). Caveat emptor or blissful ignorance? Patients and the consumerist ethos. Social Science and Medicine, 33(5), 559–568.

Oliver, R.L. (1993). Cognitive, affective and attribute bases of the satisfac-tion response. Journal of Consumer Research, 20, 418–430.

Oliver, R.L. (1997). Satisfaction: A behavioral perspective on the consumer. Boston, MA: Irwin-McGraw Hill.

Parasuraman, A., Zeithaml, V.A., & Berry, L.L. (1988). SERVQUAL: A multiple-item scale for measuring consumer perceptions of service quality. Journal of Retailing, 64(1), 12–40.

Pascoe, G.C. (1983). Patient satisfaction in primary health care: A lit-erature review and analysis. Evaluation and Program Planning, 6, 185–210.

Patton, M.Q. (2002). Qualitative research and evaluation methods (3rd ed.). Thousand Oaks, CA: Sage.

Peck, B.M., Ubel, P.A., Roter, D.L., Goold, S.D., Asch, D.A., Jeffreys, A S., et al. (2004). Do unmet expectations for specifi c tests, referrals, and new medications reduce patients’ satisfaction? Journal of General Internal Medicine, 19(11), 1080–1087.


Pope, C., Ziebland, S., & May, N. (2000). Analysing qualitative data. British Medical Journal, 320, 114–116.

Schneider, B., & Bowen, D.E. (1999). Understanding customer delight and outrage. Sloan Management Review, 41, 35–45.

Sinding, C. (2003). Disarmed complaints: Unpacking satisfaction with end-of-life care. Social Science & Medicine, 57(8), 1375–1385.

Sitzia, J. (1999). How valid and reliable are patient satisfaction data? An analysis of 195 studies. International Journal for Quality in Health Care, 11(4), 319–328.

Sitzia, J., & Wood, N. (1997). Patient satisfaction: A review of issues and concepts. Social Science and Medicine, 45(12), 1829–1843.

Smith, J., & Sanderson, C. (1992). What makes outpatient attendance worthwhile for patients? Quality Assurance in Health Care, 4(2), 125–132.

Sofaer, S., & Firminger, K. (2005). Patient perceptions of the quality of health services. Annual Review of Public Health, 26, 513–559.

Stewart, M. (2003). Patient-centered medicine: Transforming the clinical method. Abingdon, UK: Radcliffe Medical Press.

Thompson, A.G.H., & Sunol, R. (1995). Expectations as determinants of patient satisfaction: Concepts, theory and evidence. International Journal for Quality in Health Care, 7(2), 127–141.

Ware, J.E., Snyder, M.K., Wright, R., & Davies, A.R. (1983). Defi ning and measuring patient satisfaction with medical care. Evaluation and Program Planning, 6, 247–263.

Webb, Y., Clifford, P.I., & Graham, P. (1999). A simple method of evaluat-ing patients’ perceptions of their treatment and care. Bulletin of the Menninger Clinic, 63(3), 401–412.

Wensing, M., Grol, R., & Smits, A. (1994). Quality judgements by patients on general practice care: A literature analysis. Social Science and Medicine, 38, 45–53.


Wensing, M., Jung, H.P., Mainz, J., Olesen, F., & Grol, R. (1998). A system-atic review of the literature on patient priorities for general practice care. Part 1: Description of the research domain. Social Science and Medicine, 47(10), 1573–1588.

Williams, B. (1994). Patient satisfaction: A valid concept? Social Science and Medicine, 38(4), 509–515.

Williams, B., Coyle, J., & Healy, D. (1998). The meaning of patient satis-faction: An explanation of high reported levels. Social Science and Medicine, 47(9), 1351–1359.

Worthington, C., & Myers, T. (2002). Desired elements of HIV testing services: Test recipient perspectives. AIDS Patient Care and STDs, 16(11), 537–548.

Zandbelt, L.C., Smets, E.M.A., Oort, F.J., Godfried, M.H., & de Haes, H. (2004). Satisfaction with the outpatient encounter: A comparison of patients’ and physicians’ views. Journal of General Internal Medicine, 19(11), 1088–1095.

Zegers, R.A. (1968). Expectancy and the effects of confi rmation and discon-fi rmation. Journal of Personality Social Psychology, 9, 67–71.

Catherine Worthington is an assistant professor in the Faculty of Social Work, University of Calgary, where she teaches research and evaluation methods and conducts community-based health and HIV services research. As a doctoral student, she was the recipient of the 1997 Canadian Evaluation Society’s Best Student Essay Award.

patient satisfaction with health care: recent …

Documents