The Canadian Journal of Program Evaluation Vol. 20 No.

3 Pages 41–63
ISSN 0834-1516 Copyright © 2005 Canadian Evaluation Society 41

PATIENT SATISFACTION WITH HEALTH

CARE: RECENT THEORETICAL
DEVELOPMENTS AND IMPLICATIONS FOR
EVALUATION PRACTICE

Catherine Worthington
University of Calgary
Calgary, Alberta

Abstract: 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 disconfirmation, fulfillment and consumer models, and
sociological perspectives, and discusses emerging approaches and
implications for evaluation practice.

Résumé: 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.
Afin 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 signifient pour
la pratique de l’évaluation.

Corresponding author: Catherine Worthington, Faculty of Social Work, University of

Calgary, 2500 University Drive NW, Calgary, AB T2N 1N4; <cworth@ucalgary.ca>
 THE CANADIAN JOURNAL OF PROGRAM EVALUATION
42

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
disconfirmation, fulfillment and consumer models, and sociological
perspectives, and discusses emerging approaches and implications
for evaluation practice. Canadian researchers have made a small but
significant 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.
LA REVUE CANADIENNE D’ÉVALUATION DE PROGRAMME
43

Satisfaction and Elements of Care

In the majority of studies, patient satisfaction is defined as an evalu-

ation or set of evaluations of the medical intervention on a specific
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

specific 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 specific 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 financial) 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 classifications 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
 THE CANADIAN JOURNAL OF PROGRAM EVALUATION
44

(Avis et al., 1997; Sitzia & Wood, 1997). Until recently, patients have
been infrequently asked to define the dimensions of care related to
service satisfaction; therefore, these dimensions may be incomplete
or mis-specified (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).

Table 1
Reviews of the Literature on Satisfaction with General Health/Medical Care
Ware et al. (1983)
(number of studies not indicated)
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 office,
ease of reaching care location
Finances: reasonable costs,
alternative payment arrangements,
comprehensiveness of insurance
coverage
Efficacy/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
Components of Health Care
Hall & Dornan (1988)
(221 studies)
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
definable 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
Wensing et al. (1998)
(57 studies)
Availability and accessibility:
waiting times, flexibility, telephone
consultations, physical accessibil-
ity, financial accessibility
Organization and cooperation:
premises, continuity, coopera-
tion, efficiency, 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

Availability: presence of medical Cost

care resources
Bureaucracy/organization
Attention to psychosocial
problems
LA REVUE CANADIENNE D’ÉVALUATION DE PROGRAMME
45

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 satisfied with services.
However, if questions are asked on specific 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
(first defined 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 disconfirmation 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 fulfillment 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 briefly reviewed, and their contribu-
tions to satisfaction theory and deficiencies will be discussed.

Disconfirmation Theories of Satisfaction

Theories subsumed under the disconfirmation 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 defines satisfaction as the difference be-
tween what is desired or thought to be needed (“I want”/ “I need”)
 THE CANADIAN JOURNAL OF PROGRAM EVALUATION
46

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
fulfillment 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 firm conclusions (Davis, Albino,
Tedesco, Portenoy, & Ortman, 1986; Sitzia & Wood, 1997; Smith &
Sanderson, 1992).

In research performed to date, simple discrepancy/disconfirmation

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).

Fulfillment Theory of Satisfaction

Fulfillment 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 fulfillment factors are predictors
of satisfaction (Lawler, 1973; Oliver, 1993). In the health care area,
if fulfillment 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
LA REVUE CANADIENNE D’ÉVALUATION DE PROGRAMME
47

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 satisfied whether there was improvement during the
period of care or not.

Consumer Models of Satisfaction

Both discrepancy and fulfillment approaches to patient satisfaction

have been judged logically inadequate (Pascoe, 1983). Fulfillment
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 magnified, 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 fulfillment theory, discrepancy and
 THE CANADIAN JOURNAL OF PROGRAM EVALUATION
48

equity disconfirmation 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
definition, 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 significantly 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 specific 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
LA REVUE CANADIENNE D’ÉVALUATION DE PROGRAMME
49

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 final 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 influences 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) influenced 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 first 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-
 THE CANADIAN JOURNAL OF PROGRAM EVALUATION
50

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 confidence
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 satisfied or very satisfied 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
sufficient 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.
LA REVUE CANADIENNE D’ÉVALUATION DE PROGRAMME
51

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 reflect on their treatment in
a way that shifted their assessment of their care to a positive view.
They assessed the patients’ processes of reflection that influence
the transformation of negative opinions of experiences to positive
evaluations of services, and identified three psychosocial pressures
that influence 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 modified
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 influenced 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 influence the way in which peo-
ple understand health care services, including the diverse symbolic
 THE CANADIAN JOURNAL OF PROGRAM EVALUATION
52

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 influenced
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 influences on the development and
expression of expectations. Based upon transcript analysis for 88
individuals who identified 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 influence 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 reflect complex social factors. The
only generally consistent finding 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-
LA REVUE CANADIENNE D’ÉVALUATION DE PROGRAMME
53

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

identified. In the majority of published studies, the components of care
rated by patients using patient satisfaction instruments or surveys
traditionally have been defined 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 find-
ings and approaches to capture users’ views of services provide some
suggestions regarding appropriate methods.
 THE CANADIAN JOURNAL OF PROGRAM EVALUATION
54

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 definitions 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, efficiency of care/organization, and structure and facilities.
The advantage of these descriptions is that they are service-specific,
avoid some of the demonstrated difficulties 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, unreflective, and ultimate-
ly inadequate findings 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 specific 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
LA REVUE CANADIENNE D’ÉVALUATION DE PROGRAMME
55

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 findings 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 difficulties and complexities inherent in their theoreti-
cal foundations, and exercise caution and thoughtfulness in their use
in a specific 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-
 THE CANADIAN JOURNAL OF PROGRAM EVALUATION
56

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 significantly influenced 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 findings from
the last decade on the social contexts of care that influence 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
LA REVUE CANADIENNE D’ÉVALUATION DE PROGRAMME
57

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.
 THE CANADIAN JOURNAL OF PROGRAM EVALUATION
58

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.
LA REVUE CANADIENNE D’ÉVALUATION DE PROGRAMME
59

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 definition 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., Griffith, 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.
 THE CANADIAN JOURNAL OF PROGRAM EVALUATION
60

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). Defining patient-perceived quality of

nursing care. Journal of Nursing Care Quality, 16, 34–60.
LA REVUE CANADIENNE D’ÉVALUATION DE PROGRAMME
61

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, reflexivity 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 specific tests, referrals, and
new medications reduce patients’ satisfaction? Journal of General
Internal Medicine, 19(11), 1080–1087.
 THE CANADIAN JOURNAL OF PROGRAM EVALUATION
62

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). Defining 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.
LA REVUE CANADIENNE D’ÉVALUATION DE PROGRAMME
63

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 confirmation and discon-
firmation. 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.