Content uploaded by Jeffrey L Jackson
Author content
All content in this area was uploaded by Jeffrey L Jackson on Aug 11, 2015
Content may be subject to copyright.
The Effect of Unmet Expectations among Adults Presenting with
Physical Symptoms
Jeffrey L. Jackson, MD, MPH, and Kurt Kroenke, MD
Background: Unmet patient expectations are common and have
been associated with decreased patient satisfaction.
Objective: To assess the prevalence and effect of unmet expec-
tations in patients presenting with physical symptoms.
Design: Prospective cohort study.
Setting: Primary care walk-in clinic. Most patients were seeing a
particular provider for the first time.
Patients: 750 adults whose principal reason for the clinic visit
was a physical symptom.
Measurements: Patients completed previsit questionnaires that
assessed symptom characteristics, the patient’s expectations of the
visit, functional status (Medical Outcomes Study Short Form-6),
and mental disorders (Primary Care Evaluation of Mental Disor-
ders [PRIME-MD]). Patient questionnaires given immediately after
the visit and 2 weeks after the visit assessed patient satisfaction
with the visit and unmet expectations; the 2-week questionnaire
also assessed symptom outcome and functional status. Postvisit
physician questionnaires measured encounter difficulty (Difficult
Doctor Patient Relationship Questionnaire) and what the physi-
cian did in response to the patient’s symptom.
Results: Nearly all patients (98%) had at least one previsit ex-
pectation, including a diagnosis (81%), an estimate of how long
the symptom was likely to last (63%), a prescription (60%), a
diagnostic test (54%), and a subspecialty referral (45%). Imme-
diately after the visit, the most common unmet expectations were
for prognostic information (51%) or diagnostic information
(33%). Only 11% of patients had an unmet expectation of a
diagnostic test, subspecialty referral, prescription, or sick slip. Un-
met patient expectations were more common after encounters
experienced as difficult by the clinician and in patients with un-
derlying mental disorders. Patients with no unmet expectations
had less worry about serious illness (54% vs. 27%;
P
< 0.001)
and greater satisfaction (59% vs. 19%;
P
< 0.001), and patients
who reported receiving diagnostic or prognostic information were
more likely to have symptom alleviation (relative risk, 1.2 [95%
CI, 1.02 to 1.3]) and functional improvement (functional status
score, 25 vs. 23;
P
ⴝ 0.01) at 2 weeks.
Conclusions: Patients who seek care for physical symptoms and
do not leave the encounter with an unmet expectation are more
likely to be satisfied with their care and to have less worry about
serious illness. Diagnostic and prognostic information are partic-
ularly valued by patients and may be associated with greater
improvement in symptoms and functional status 2 weeks after the
visit.
Ann Intern Med. 2001;134:889-897. www.annals.org
For author affiliations and current addresses, see end of text.
A
lmost 20 years ago, Barsky (1) wrote of “hidden”
reasons why patients seek medical care, and he sug-
gested that patient dissatisfaction should trigger explora-
tion for unmet expectations. Subsequent research has
shown that patient expectations, distinct from requests
(2), are ubiquitous. Broad categories include expecta-
tions of information, support, and medical diagnosis or
treatment (3, 4). Unfortunately, physicians often under-
value or do not recognize patient expectations (5–7),
and expectations are therefore often unmet (8 –14). Un-
met expectations have been associated with decreased
patient satisfaction (8, 15–21), no adherence (15, 22–24),
and possibly worse health-related outcomes (18, 25).
Our goal was to determine the frequency of symp-
tom-related patient concerns and expectations and to
assess the relationship between expectations and patient-
centered outcomes.
METHODS
Adults presenting to the general medicine walk-in
clinic at Walter Reed Army Medical Center, Washing-
ton, D.C., with a chief complaint of a physical symptom
were eligible to participate. The demographic character-
istics, medical and psychiatric comorbid conditions, and
satisfaction with care of patients seen in a military gen-
eral medicine clinic are similar to those seen in civilian
settings (26, 27).
These protocols were approved by our institutional
human use committee.
Previsit Patient Questionnaire
Immediately before seeing a physician, all patients
completed a questionnaire on their presenting symptom
(“What problem brings you to the clinic today?”), symp-
The Physician–Patient Relationship
© 2001 American College of Physicians–American Society of Internal Medicine 889
Downloaded From: http://annals.org/ by a Med College of Wisconsin User on 08/11/2015
tom severity (ranked from 0 to 10 on a visual analogue
scale), symptom duration (in days), previous visits for
the symptom (yes or no), worry about serious illness (yes
or no), stress in the previous week (yes or no), and
presence of common symptom-related expectations.
The expectations included expectations of a diagnosis
(an explanation of the symptom’s cause), prognostic in-
formation (an estimate of how long the symptom was
likely to last), a prescription, a diagnostic test, a referral
to another clinic, or another physician action (8, 13). In
addition, all patients completed the Medical Outcomes
Study Short Form-6, a six-item scale that measures func-
tional status in six domains: general health, role func-
tion, physical function, social function, emotional
health, and physical pain (28). Patients were also evalu-
ated for depressive and anxiety disorders by using the
Primary Care Evaluation of Mental Disorders (PRIME-
MD) (29).
Postvisit Patient Questionnaire
Immediately after the visit, patients completed the
Medical Outcomes Study nine-item satisfaction survey
(30), which asks about overall satisfaction and eight
domains of satisfaction. Additional questions assessed
residual worry about serious illness and unmet expecta-
tions with respect to a diagnosis, prognostic informa-
tion, a prescription, a diagnostic test, or a referral. Pa-
tients were invited to list any other unmet expectations.
Two-Week Patient Questionnaire
Two weeks after the visit, patients were mailed a
questionnaire that assessed symptom outcome and se-
verity, residual worry about serious illness, unmet expec-
tations, functional status (Medical Outcomes Study
Short Form-6), and a single question on satisfaction:
“Overall, how do you feel about the care you received
for this problem from your doctor?” Patients were also
asked whether they had had or had anticipated having
another physician visit for the original symptom and
whether the symptom had lasted longer than expected.
Physician Variables
After each visit, physicians completed the 10-item
Difficult Doctor Patient Relationship Questionnaire
(31) to assess clinician-perceived difficulty of the en-
counter. This questionnaire was previously shown to be
reliable, with scores greater than 30 points (on a scale of
10 to 60 points) indicating a “difficult” encounter (32).
Physicians also indicated whether the symptom had led
them to order a prescription, diagnostic test, or referral
and whether they had told the patient what the problem
was and how long it would probably last.
Statistical Analysis
Our primary analysis assessed the presence of unmet
expectations immediately after the visit and 2 weeks af-
ter the visit in relation to other variables, using the chi-
square test or the Student t-test. The McNemar test was
used to compare the proportion of patients who had an
unmet expectation immediately after the visit with the
proportion of patients who had an unmet expectation at
2 weeks. To assess functional status, overall scores were
created by summing scores for each of the individual
domains. Logistic regression techniques were used to de-
termine independent correlates of unmet expectations
and satisfaction. Unmet expectations were dichotomized
into any unmet expectation or no unmet expectation;
overall satisfaction was dichotomized into fully satisfied
or less than fully satisfied. Overall satisfaction was used
at both time points for two reasons. First, each of the
eight domains of satisfaction of the Medical Outcomes
Study survey correlated strongly with overall satisfaction
(r ⬎ 0.85). Second, we thought that at 2 weeks, patients
would recall overall satisfaction more accurately than
they would recall specific aspects of the encounter.
Data were collected as part of two clinical trials in
which previsit information on patients’ symptom-related
expectations and mental disorders was given to clini-
cians. Both trials were done in the same clinic, had the
same inclusion criteria (walk-in patients presenting with
a physical symptom), and included the same survey in-
struments. The first study, a pre–post trial involving 500
patients, was conducted from November 1994 to Janu-
ary 1996 (33). During the intervention period, physi-
cians were given previsit information about their pa-
tients’ worry over serious illness, expectations of care,
and mental disorders. The second study, a randomized
trial involving 250 patients, was done from August to
September 1998. Patients were randomly assigned to
one of three groups. One group’s clinicians were given
no information, the second group’s clinicians were given
previsit information about worry over serious illness and
expectations, and the third group’s clinicians received
information about mental disorders.
The Physician–Patient Relationship Unmet Expectations in Primary Care
890 1 May 2001 Annals of Internal Medicine Volume 134 • Number 9 (Part 2) www.annals.org
Downloaded From: http://annals.org/ by a Med College of Wisconsin User on 08/11/2015
The two cohorts were similar in terms of patient
sex, race, educational status, worry over serious illness,
symptom duration, symptom type or severity, number
of previsit expectations, functional status, recent stress,
prevalence of mental disorders, satisfaction, and 2-week
improvement rates. Patients in the first cohort were
slightly older than those in the second (55 years vs. 51
years; P ⫽ 0.007). Because the two cohorts were similar
in most variables and visited the same clinic, they were
combined into one group for purposes of our analysis.
Although no differences were seen in any outcome mea-
sured in the second trial, the pre–post trial modestly
reduced unmet expectations and patient difficulty (33).
We explored for potential confounding of our outcomes
by this intervention; because we found none, we report
all data unadjusted. All analyses were done by using
Stata 6.0 (Stata Corp., College Station, Texas).
RESULTS
Baseline Data
Table 1 presents the baseline characteristics of the
study sample. The 750 participating patients averaged
55 years of age; 52% were women, 49% were white, and
46% were African American. Most patients (93%) were
seeing the examining physician for the first time. Thirty
percent of patients had a depressive or anxiety disorder.
Patients presented with numerous problems, which we
collapsed into 15 broad categories. One hundred forty-
three patients (19%) answered the question “What
problem brings you to the clinic?” by stating that they
had more than one physical symptom, 97 (13%) noted
two symptoms, 14 (2%) had three symptoms, and 3
(0.4%) listed four symptoms. The most common type
of symptom was pain (53%), and the second most com-
mon was symptoms suggestive of an upper respiratory
tract infection (congestion or cough), present in 21%.
In addition to listing their presenting symptoms, pa-
tients were asked whether they had been “often both-
ered” by 15 common symptoms on the PRIME-MD.
Patients had had a mean (⫾SD)of4⫾ 2.8 of these
symptoms during the past month (median, 4; range, 0
to 14). The median duration of the presenting physical
symptom was 14 days (range, 12 hours to 13 years).
Sixty-three percent of patients were worried that their
symptom might represent a serious illness.
Nearly all patients (98%) reported at least one pre-
visit expectation (Table 2). Eighty-one percent hoped
for a diagnosis (an explanation of the symptom’s cause),
and 63% desired prognostic information (an estimate of
how long the symptom would last). Sixty-six percent
hoped for a prescription, 54% a diagnostic test, 45% a
subspecialty referral, and 7% an excuse from work. The
250 patients in the randomized trial were also asked
about expectations for counseling or referral for specific
issues, but these expectations were held by only small
proportions of patients. The specific issues included to-
bacco use (0.43%), nutrition (1.7%), alcohol use (0.43%),
obesity (1.7%), exercise (1.3%), cancer screening (0.9%),
cholesterol levels (1.3%), sexual function (0.43%), stress
management (2.2%), and domestic violence (0%).
Most patients had more than one expectation for
Table 1. Characteristics of the Study Sample
Characteristic Value
Demographic characteristics
Mean age ⫾ SD, y 54.6 ⫾ 18.4
Female sex, % 52
Ethnicity
White 49
African American 46
Other 6
Symptom characteristics
Median duration (range) 14 d (12 h–13 y)
Mean symptom severity ⫾ SD
(on a scale of 1 to 10) 5.6 ⫾ 2.6
Type of presenting symptom, %
Back pain 7
Chest pain 4
Cough, congestion 21
Dermatologic 12
Dizziness 5
Ear, nose, throat 6
Fatigue 2
Gastrointestinal 11
Genitourinary 5
Headache 4
Musculoskeletal 30
Neurologic 2
Ophthalmologic 4
Any pain 53
Psychiatric 0.5
Previous visit with any physician for
this problem, % 45
Worry about serious illness, % 63
Recent stress, % 43
Mental disorder, %
Any 30
More than one 15
Depression
Major 8
Minor 12
Anxiety disorder
Generalized anxiety disorder 2
Panic disorder 1
Anxiety not otherwise specified 12
The Physician–Patient RelationshipUnmet Expectations in Primary Care
www.annals.org 1 May 2001 Annals of Internal Medicine Volume 134 • Number 9 (Part 2) 891
Downloaded From: http://annals.org/ by a Med College of Wisconsin User on 08/11/2015
the visit, and the mean number of expectations was
3 ⫾ 1.3 (Figure). Patients who desired a diagnosis were
more likely to want prognostic information (relative risk
[RR], 1.6 [95% CI, 1.5 to 1.8]), but neither of these
expectations clustered with desires for a prescription, test,
or referral. In addition, no evidence was seen of cluster-
ing of desires for prescriptions, referrals, or diagnostic
testing. Neither the type nor the number of previsit ex-
pectations was associated with demographic characteris-
tics; presence of a mental disorder; functional status; or
duration, type, or severity of the presenting symptom.
There was also no relationship between the type or
number of previsit expectations and the number of
physical symptoms, whether volunteered by the patient
as a presenting symptom or noted on the PRIME-MD.
Immediately after the Visit
Immediately after the visit, the proportion of pa-
tients worrying about serious illness was substantially
reduced, from 63% to 30%. One third of patients had
an unmet expectation of a diagnosis, and half had an
unmet expectation of prognostic information (Table 2);
only 12% had any other unmet expectation (5% had
desired a diagnostic test, 4% a referral, 2% a prescrip-
tion, and 1% a sick slip).
On multivariate analysis, continuing worry about
serious illness, not receiving a diagnosis, being consid-
ered difficult by the physician, and having an underlying
mental disorder independently increased the likelihood
of having an unmet expectation immediately after the
visit (Table 3). In contrast, patient demographic char-
acteristics, type or duration of the presenting symptom,
recent stress, and functional status were not associated
with unmet expectations. Patients with an underlying
mental disorder had more symptoms, both volunteered
as presenting symptoms and noted on the PRIME-MD
(P ⬍ 0.001 for both). However, neither measure of symp-
tom count was associated with a greater likelihood of
unmet expectations after adjustment for mental disorders.
Most patients who had an unmet expectation after
their visit had had a previsit desire for the same item,
including 76% of patients with a residual desire for a
prescription, 68% of those still wanting a diagnostic
test, and 62% of those still desiring a subspecialty refer-
ral. Patients who expressed previsit expectations for a
prescription (RR, 1.5 [CI, 1.3 to 1.8]), a diagnostic test
(RR, 1.5 [CI, 1.3 to 1.8]), or a referral (RR, 1.6 [CI, 1.3
to 1.9]) were more likely to receive them. This was not
the case for those desiring a diagnosis or prognostic in-
formation.
Immediately after the visit, 55% of patients were
fully satisfied with the care they had received. Patients
were more likely to be fully satisfied if they had received
a diagnosis (RR, 1.4 [CI, 1.2 to 1.6]) or prognostic
information (RR, 1.3 [CI, 1.2 to 1.5]), and patients
were less likely to be fully satisfied if they had an unmet
expectation for a test or referral (RR, 0.4 [CI, 0.2 to
0.6]). Unmet expectations for prescriptions were not as-
sociated with satisfaction.
Two Weeks after the Visit
We obtained follow-up information on 632 patients
(84%) at 2 weeks. Nonrespondents differed from re-
spondents only in being younger (43 years vs. 55 years;
P ⬍ 0.001); no differences were seen in previsit expec-
Figure. Distribution of previsit expectations among
patients presenting with physical symptoms.
Table 2. Prevalence of Symptom-Related Expectations
Symptom-Related
Expectation
Before
the Visit
(
n
ⴝ 750)
Immediately
after the Visit
(
n
ⴝ 750)
At 2-Week
Follow-up
(
n
ⴝ 632)
4
OOOOOOOO % OOOOOOOO
3
Physician communication
Diagnostic (causal explana-
tion of symptom) 81 33 Not measured
Prognostic (likely duration
of symptom) 63 51 Not measured
Physician action
Diagnostic test 54 5 14
Prescription for medication 66 2 6
Subspecialty referral 45 4 16
Work excuse 7 0 0
The Physician–Patient Relationship Unmet Expectations in Primary Care
892 1 May 2001 Annals of Internal Medicine Volume 134 • Number 9 (Part 2) www.annals.org
Downloaded From: http://annals.org/ by a Med College of Wisconsin User on 08/11/2015
tations, unmet expectations or satisfaction after the visit,
or likelihood of receiving a diagnosis or prognostic in-
formation. By 2 weeks, 71% of patients had symptom
alleviation. Patients who reported immediately after the
visit that they had received a diagnosis (RR, 1.2 [CI,
1.02 to 1.3]) or prognostic information (RR, 1.2 [CI,
1.04 to 1.3]) were more likely to have symptom im-
provement, even after adjustment for patient demo-
graphic characteristics (age and sex), the presence of
mental disorders, and symptom characteristics (dura-
tion, severity, and type of symptom).
Worry about serious illness rebounded slightly; 30%
of patients had it immediately after the visit, and 38%
had it by 2 weeks (Table 2). Two thirds of patients
thought that their symptom had lasted longer than ex-
pected, and 44% either had had or had anticipated an-
other visit for the same symptom. Actual or anticipated
revisits were less likely in patients who were given a
diagnosis (RR, 0.88 [CI, 0.78 to 0.99]) or prognostic
information (RR, 0.79 [CI, 0.69 to 0.91]).
Unmet expectations were more prevalent at 2 weeks
than immediately after the visit (27% vs. 11%; P ⬍
0.001). Among patients reporting complete resolution
of their symptom, 14% reported an unmet expectation.
In contrast, an unmet expectation was reported by 28%
of patients reporting their symptom as “better,” 48% of
those reporting “no change,” and 64% of those report-
ing “worsening.” Overall, 16% of patients at 2 weeks
wished that they had received a referral, 14% wished
that they had received a diagnostic test, and 6% wished
that they had received a prescription (Table 2). No re-
lationship was seen between previsit expectations and
unmet expectations at 2 weeks. Whereas patients report-
ing an unmet expectation immediately after the visit
were more likely to still have unmet expectations at 2
weeks (RR, 2.1 [CI, 1.6 to 2.8]), 78% of patients with
unmet expectations at 2 weeks had had no unmet ex-
pectations immediately after the visit. On multivariate
analysis, persistent worry over serious illness, lack of
alleviation of the symptom, a longer duration of the
symptom than expected, and an encounter rated as dif-
ficult by the physician correlated with the presence of an
unmet expectation at 2 weeks (Table 3).
By 2 weeks, 60% of patients reported that they were
fully satisfied with the care they had received. The stron-
gest correlate of satisfaction at 2 weeks was the absence
of any unmet expectation (OR, 15 [CI, 8.7 to 26.1]).
Other correlates included symptom alleviation (OR, 3.0
[CI, 1.7 to 5.1]), better functional status (OR, 1.1 [CI,
1.04 to 1.2]), the symptom not lasting longer than ex-
pected (OR, 2.0 [CI, 1.2 to 3.4]), and patient age
greater than 65 years (OR, 1.9 [CI, 1.1 to 3.1]).
Physician–Patient Interactions
Physicians rated 100 encounters (14%) as difficult.
Patients involved in difficult encounters were less likely
to report receiving prognostic information (RR, 0.68
[CI, 0.5 to 0.92]) and more likely to have unmet expec-
tations both immediately and at 2 weeks (Table 2).
Patients and physicians had good agreement on cer-
tain aspects of the visit, including whether a prescription
(85% agreement;
⫽ 0.63), a diagnostic test (83%
agreement;
⫽ 0.64), or a referral (84% agreement;
⫽ 0.67) was provided. However, patient–physician
agreement was poor on whether a diagnosis (72% agree-
ment;
⫽ 0.17) or prognostic information (67% agree-
ment;
⫽ 0.36) was provided.
DISCUSSION
Among patients presenting with a physical symp-
tom, symptom-specific expectations for care are com-
mon and include desires for a diagnosis and prognostic
information as well as diagnostic testing, prescriptions,
and referrals. Other common aspects of routine health
care seem not to be expected in visits triggered by phys-
ical symptoms. Patients who reported receiving prognos-
tic and diagnostic information were more likely to have
symptom alleviation and improvement in functional sta-
tus 2 weeks after the visit. Both immediately and 2
Table 3. Independent Correlates of Unmet Patient
Expectations*
Correlate Odds Ratio for Unmet Expectation
(95% CI)
Immediately
after Visit
2 Weeks
after Visit
Worry about serious illness 2.4 (1.5–4.0) 2.5 (1.6–3.8)
Provider considered encounter
difficult 1.8 (1.1–3.0) 2.1 (1.2–3.7)
Presence of a depressive or anxiety
disorder 1.8 (1.1–3.0) 1.4 (0.9–2.4)
Receiving a diagnosis (causal
explanation of symptom) 0.5 (0.3–0.9) 0.9 (0.6–1.4)
Symptom same or worse NA 1.9 (1.2–3.0)
Symptom lasted longer than expected NA 2.3 (1.4–3.8)
* NA ⫽ not applicable.
The Physician–Patient RelationshipUnmet Expectations in Primary Care
www.annals.org 1 May 2001 Annals of Internal Medicine Volume 134 • Number 9 (Part 2) 893
Downloaded From: http://annals.org/ by a Med College of Wisconsin User on 08/11/2015
weeks after the visit, patients who reported any unmet
expectation were less satisfied with their care.
Previous studies have found that unmet expecta-
tions are common (8, 12, 20) and that patients partic-
ularly value diagnostic and prognostic information (20,
35–37). Similarly, we found that receiving either diag-
nostic or prognostic information was associated with a
reduction in worry about serious illness immediately af-
ter the visit, fewer actual or anticipated return clinic
visits at 2 weeks, increased satisfaction both immediately
after the visit and at 2 weeks, and a greater likelihood of
symptom improvement at 2 weeks. Other trials have
shown that better health outcomes (blood pressure,
blood glucose level, or functional status) are consistently
related to aspects of physician–patient communication
(18), but our study links health outcomes specifically to
the provision of diagnostic and prognostic information.
One possible explanation is that patients with simple,
self-limiting disorders are both easier to diagnose and
more likely to improve; consequently, their clinicians
may be more apt to give them specific diagnostic and
prognostic information. However, this does not seem to
entirely explain our findings for several reasons.
First, although patients presenting with symptoms
lasting 72 hours or less were more likely to be given
diagnostic (RR, 1.1 [CI, 1.03 to 1.3]) or prognostic
(RR, 1.5 [CI, 1.3 to 1.8]) information, receiving such
information predicted improved symptom outcomes
even after adjustment for symptom duration. Second,
patients presenting with specific categories of symptoms
suggesting self-limited illnesses, such as a urinary tract
infection, were not more likely to improve than were
patients with other categories of somatic problems.
Third, no relationship was seen between the physician’s
report of providing prognostic (P ⬎ 0.2) or diagnostic
(P ⫽ 0.13) information and the likelihood of symptom
alleviation at 2 weeks. If patients with self-limited,
straightforward symptoms were more likely to receive a
diagnosis or to be given prognostic information, one
would expect this to be reflected by physician reports of
providing such information. Instead, only the patient’s
perception of receiving this information was associated
with symptom alleviation.
Although patients and physicians agreed rather well
about concrete visit events, such as whether a prescrip-
tion, a diagnostic test, or a subspecialty referral had been
ordered, patients frequently disagreed with physicians
about whether diagnostic and prognostic information
had been provided. Although disturbing, this discor-
dance has several possible explanations. The physician–
patient discussions could be implicit rather than explicit.
Physicians may use language not readily understood by
the patient, or the patient may reject the physician’s
prognostic or diagnostic statements. The interactive na-
ture of patient–physician communication is exemplified
by the fact that patients from encounters experienced by
clinicians as difficult were less likely to report receiving
diagnostic information and more likely to have unmet
expectations.
Because of limited symptoms-based research, pro-
viders themselves may be uncertain about the cause of a
symptom in a particular patient. A clear-cut diagnosis
cannot be established for physical symptoms at least one
third of the time (38). Diagnostic uncertainty may in
turn foster hesitancy about discussing prognosis. Prelim-
inary research does suggest that, regardless of symptom
type, improvement occurs in most primary care patients
within a few weeks to several months of the index visit,
and occult, serious causes seldom become manifest at
follow-up (39). Clearly, additional research on specific
symptoms is needed to strengthen the ability of clini-
cians to confidently provide diagnostic and prognostic
information. Meanwhile, even generic “positive reassur-
ance” strategies may improve symptom outcomes (40).
Our study focused on unscheduled visits for somatic
symptoms. Expectations are also common in patients
presenting in continuity settings and include desires for
physical examination, prescription refills, blood testing,
prognostic counseling, and discussion of the patient’s
own ideas about management (10, 17, 41). Expectations
may also vary with the type of visit. In our sample of
symptomatic walk-in patients, few were interested in
components of health delivery considered part of rou-
tine health care, such as counseling on tobacco use, nu-
trition, weight control, cholesterol levels, cancer screen-
ing, and stress reduction. Other studies have suggested
that among patients with symptoms, receipt of diagnos-
tic information is more valued than being included in
the management decision-making process (42). Clini-
cians need to consider the nature and reason for the
patient’s visit to anticipate and elicit particular expecta-
tions for the encounter.
Since unmet expectations have been found to be
associated with lower satisfaction (8, 10, 16 –21), several
The Physician–Patient Relationship Unmet Expectations in Primary Care
894 1 May 2001 Annals of Internal Medicine Volume 134 • Number 9 (Part 2) www.annals.org
Downloaded From: http://annals.org/ by a Med College of Wisconsin User on 08/11/2015
investigators have tried to improve management of pa-
tient expectations. One study found that giving infor-
mation to clinicians on patient stress and functioning
improved patient satisfaction (47). Other studies, focus-
ing on patient worry about serious illness (48) and pro-
viding previsit feedback on patient expectations (49),
had no effect. Two trials, using pre–post designs and
coupling workshops on eliciting and managing expecta-
tions with previsit information on patient desires, re-
duced the number of unmet expectations without im-
proving satisfaction (33, 50). One of these trials found
no improvement in physician communication on diag-
nosis or prognosis (33); the other trial did not report on
this variable. If diagnostic and prognostic information
are particularly important, these trials may not have pro-
duced changes necessary to improve patient outcomes.
On the other hand, although it is tempting to conclude
that unmet expectations trigger patient dissatisfaction,
the direction of the relationship is not clear. It may be
that patients finding themselves less than fully satisfied
are more likely to report an unmet expectation immedi-
ately after the visit.
Our study has several important limitations. First,
the relationship between receipt of diagnostic and prog-
nostic information and symptom alleviation may be due
to self-selection. Patients with easily diagnosable, self-
limited illness may be more likely to receive diagnostic
and prognostic information. Although we adjusted for
symptom type and duration as well as for various patient
and encounter characteristics, these may not be optimal
surrogate markers. However, we also asked physicians
whether they had given the patient a diagnostic expla-
nation and discussed prognosis. Notably, physician re-
ports of giving such information failed to predict symp-
tom alleviation, whereas patient reports of receiving this
information did.
A second limitation is that the relationship between
receiving a diagnosis or prognosis and symptom allevia-
tion was modest (RR, 1.2 for both). Only 10% of the
variance in symptom severity at 2 weeks was explained
by these two variables. Third, without direct observa-
tion, it is difficult to determine precisely the source of
discordance between patient and physician reports of
what occurred during the encounter. Future studies may
wish to include audiotape or videotape. Fourth, our en-
counters involved almost exclusively new patient–physi-
cian interactions. Although one must be cautious about
generalizing these findings to continuity settings, 7% of
encounters in our study involved established patient–
physician relationships, and these did not differ from
other encounters in the number of previsit and postvisit
expectations or the rate of symptom alleviation. More-
over, previous studies of established relationships have
found the type and number of expectations to be similar
to those that we report (9).
Almost all patients who present with physical symp-
toms have one or more symptom-specific expectations
of what is desired from the physician. Unmet expecta-
tions after the visit, although less prevalent, may have
significant effects on numerous outcomes, including sat-
isfaction, symptom resolution, functional status, and
health care utilization. The most important expectations
relate to diagnostic information (a causal explanation for
the symptom) and prognostic information (an estimate
of symptom duration). Unfortunately, this information
is often either not provided or not clearly grasped by
patients. Clinicians could endeavor to improve this as-
pect of physician–patient communication by explicitly
addressing the diagnostic and prognostic concerns of
patients who present with physical symptoms.
From the Uniformed Services University of the Health Sciences, Be-
thesda, Maryland; and the Regenstrief Institute for Health Care and
Indiana University School of Medicine, Indianapolis, Indiana.
Grant Support: In part by an intramural grant from the Uniformed
Services University of the Health Sciences.
Requests for Single Reprints: Jeffrey L. Jackson, MD, MPH, Depart-
ment of Medicine-EDP, Uniformed Services University of the Health
Sciences, 4301 Jones Bridge Road, Bethesda, MD 20814; e-mail,
jejackson@usuhs.mil.
Current Author Addresses: Dr. Jackson: Department of Medicine-EDP,
Uniformed Services University, 4301 Jones Bridge Road, Bethesda, MD
20814.
Dr. Kroenke: Regenstrief Institute for Health Care, 1001 West 10th
Street, Indianapolis, IN 46202.
References
1. Barsky AJ 3rd. Hidden reasons some patients visit doctors. Ann Intern Med.
1981;94:492-8. [PMID: 0007212508]
2. Uhlmann RF, Inui TS, Carter WB. Patient requests and expectations. Defi-
nitions and clinical applications. Med Care. 1984;22:681-5. [PMID: 0006748787]
3. Valori R, Woloshynowych M, Bellenger N, Aluvihare V, Salmon P. The
Patient Requests Form: a way of measuring what patients want from their general
practitioner. J Psychosom Res. 1996;40:87-94. [PMID: 0008730648]
4. Salmon P. Medical care appraisal. Hospitals. 1968;42:103-4 passim. [PMID:
0004868679]
The Physician–Patient RelationshipUnmet Expectations in Primary Care
www.annals.org 1 May 2001 Annals of Internal Medicine Volume 134 • Number 9 (Part 2) 895
Downloaded From: http://annals.org/ by a Med College of Wisconsin User on 08/11/2015
5. Burgoyne RW, Staples FR, Yamamoto J, Wolkon GH, Kline F. Patients’
requests of an outpatient clinic. Arch Gen Psychiatry. 1979;36:400-3. [PMID:
0000426606]
6. van Dulmen AM, Fennis JF, Mokkink HG, van der Velden HG, Bleijenberg
G. Doctors’ perception of patients’ cognitions and complaints in irritable bowel
syndrome at an out-patient clinic. J Psychosom Res. 1994;38:581-90. [PMID:
0007990066]
7. Gillette RD, Kues J, Harrigan JA, Franklin L. Does the family physician’s role
correspond to the patient’s expectations? Fam Med. 1986;18:68-72. [PMID:
0003556853]
8. Marple RL, Kroenke K, Lucey CR, Wilder J, Lucas CA. Concerns and
expectations in patients presenting with physical complaints. Frequency, physi-
cian perceptions and actions, and 2-week outcome. Arch Intern Med. 1997;157:
1482-8. [PMID: 0009224227]
9. Kravitz RL, Callahan EJ, Paterniti D, Antonius D, Dunham M, Lewis CE.
Prevalence and sources of patients’ unmet expectations for care. Ann Intern Med.
1996;125:730-7. [PMID: 0008929006]
10. Kravitz RL, Cope DW, Bhrany V, Leake B. Internal medicine patients’
expectations for care during office visits. J Gen Intern Med. 1994;9:75-81. [PMID:
0008164081]
11. Starfield B, Steinwachs D, Morris I, Bause G, Siebert S, Westin C. Patient-
doctor agreement about problems needing follow-up visit. JAMA. 1979;242:344-6.
[PMID: 0000556592]
12. Sanchez-Menegay C, Stalder H. Do physicians take into account patients’
expectations? J Gen Intern Med. 1994;9:404-6. [PMID: 0007931752]
13. Good MD, Good BJ, Nassi AJ. Patient requests in primary health care
settings: development and validation of a research instrument. J Behav Med.
1983;6:151-68. [PMID: 0006620371]
14. Uhlmann RF, Carter WB, Inui TS. Fulfillment of patient requests in a
general medicine clinic. Am J Public Health. 1984;74:257-8. [PMID: 0006696158]
15. Barsky AJ, Borus JF. Somatization and medicalization in the era of managed
care. JAMA. 1995;274:1931-4. [PMID: 0008568987]
16. Sanchez-Menegay C, Hudes ES, Cummings SR. Patient expectations and
satisfaction with medical care for upper respiratory infections. J Gen Intern Med.
1992;7:432-4. [PMID: 0001506950]
17. Eisenthal S, Lazare A. Evaluation of the initial interview in a walk-in clinic.
The patient’s perspective on a “customer approach.” J Nerv Ment Dis. 1976;162:
169-76. [PMID: 0001255147]
18. Kaplan SH, Greenfield S, Ware JE Jr. Assessing the effects of physician-
patient interactions on the outcomes of chronic disease. Med Care. 1989;27:
S110-27. [PMID: 0002646486]
19. Uhlmann RF, Inui TS, Pecoraro RE, Carter WB. Relationship of patient
request fulfillment to compliance, glycemic control, and other health care out-
comes in insulin-dependent diabetes. J Gen Intern Med. 1988;3:458-63. [PMID:
0003171726]
20. Korsch BM, Gozzi EK, Francis V. Gaps in doctor-patient communication.
1. Doctor-patient interaction and patient satisfaction. Pediatrics. 1968;42:855-
71. [PMID: 0005685370]
21. Brody DS, Miller SM, Lerman CE, Smith DG, Lazaro CG, Blum MJ. The
relationship between patients’ satisfaction with their physicians and perceptions
about interventions they desired and received. Med Care. 1989;27:1027-35.
[PMID: 0002586185]
22. Francis V, Korsch BM, Morris MJ. Gaps in doctor-patient communication.
Patients’ response to medical advice. N Engl J Med. 1969;280:535-40. [PMID:
0005764453]
23. Eisenthal S, Emery R, Lazare A, Udin H. “Adherence” and the negotiated
approach to patienthood. Arch Gen Psychiatry. 1979;36:393-8. [PMID:
0000426605]
24. Linn MW, Linn BS, Stein SR. Satisfaction with ambulatory care and com-
pliance in older patients. Med Care. 1982;20:606-14. [PMID: 0007109743]
25. Brody DS, Miller SM. Illness concerns and recovery from a URI. Med Care.
1986;24:742-8. [PMID: 0003736145]
26. Jackson JL, Strong J, Cheng EY, Meyer G. Patients, diagnoses, and proce-
dures in a military internal medicine clinic: comparison with civilian practices.
Mil Med. 1999;164:194-7. [PMID: 0010091492]
27. Jackson JL, O’Malley PG, Kroenke K. A psychometric comparison of
military and civilian medical practices. Mil Med. 1999;164:112-5. [PMID:
0010050567]
28. Ware JE, Nelson EC, Sherbourne CD, Stewart AL. Preliminary tests of a
6-item general health survey: a patient application. In: Ware JE, Stewart EL, eds.
Measuring Functioning and Well-Being: The Medical Outcomes Study
Approach. Durham, NC: Duke Univ Pr; 1992:291-308.
29. Spitzer RL, Williams JB, Kroenke K, Linzer M, deGruy FV 3rd, Hahn SR,
et al. Utility of a new procedure for diagnosing mental disorders in primary care.
The PRIME-MD 1000 study. JAMA. 1994;272:1749-56. [PMID: 0007966923]
30. Rubin HR, Gandek B, Rogers WH, Kosinski M, McHorney CA, Ware JE
Jr. Patients’ ratings of outpatient visits in different practice settings. Results from
the Medical Outcomes Study. JAMA. 1993;270:835-40. [PMID: 0008340982]
31. Hahn SR, Thompson KS, Wills TA, Stern V, Budner NS. The difficult
doctor-patient relationship: somatization, personality and psychopathology.
J Clin Epidemiol. 1994;47:647-57. [PMID: 0007722577]
32. Hahn SR, Kroenke K, Spitzer RL, Brody D, Williams JB, Linzer M, et al.
The difficult patient: prevalence, psychopathology, and functional impairment.
J Gen Intern Med. 1996;11:1-8. [PMID: 0008691281]
33. Jackson JL, Kroenke K, Chamberlin J. Effects of physician awareness of
symptom-related expectations and mental disorders. A controlled trial. Arch Fam
Med. 1999;8:135-42. [PMID: 0010101984]
34. Sanchez-Menegay C, Hudes ES, Cummings SR. Patient expectations and
satisfaction with medical care for upper respiratory infections. J Gen Intern Med.
1992;7:432-4. [PMID: 0001506950]
35. Deyo RA, Diehl AK. Patient satisfaction with medical care for low-back pain.
Spine. 1986;11:28-30. [PMID: 0002939566]
36. Laine C, Davidoff F, Lewis CE, Nelson EC, Nelson E, Kessler RC, et al.
Important elements of outpatient care: a comparison of patients’ and physicians’
opinions. Ann Intern Med. 1996;125:640-5. [PMID: 0008849148]
37. Kroenke K. Symptoms and science: the frontiers of primary care research
[Editorial]. J Gen Intern Med. 1997;12:509-10. [PMID: 0009276658]
38. Kroenke K, Jackson JL. Outcome in general medical patients presenting with
common symptoms: a prospective study with a 2-week and a 3-month follow-up.
Fam Pract. 1998;15:398-403. [PMID: 0009848423]
40. Joos SK, Hickam DH, Borders LM. Patients’ desires and satisfaction in
general medicine clinics. Public Health Rep. 1993;108:751-9. [PMID: 0008265760]
41. Strull WM, Lo B, Charles G. Do patients want to participate in medical
decision making? JAMA. 1984;252:2990-4. [PMID: 0006502860]
42. Sox HC Jr, Margulies I, Sox CH. Psychologically mediated effects of diag-
nostic tests. Ann Intern Med. 1981;95:680-5. [PMID: 0007305144]
43. Macfarlane J, Holmes W, Macfarlane R, Britten N. Influence of patients’
expectations on antibiotic management of acute lower respiratory tract illness in
general practice: questionnaire study. BMJ. 1997;315:1211-4. [PMID: 0009393228]
44. Froehlich GW, Welch HG. Meeting walk-in patients’ expectations for test-
ing. Effects on satisfaction. J Gen Intern Med. 1996;11:470-4. [PMID: 0008872784]
45. Hamm RM, Hicks RJ, Bemben DA. Antibiotics and respiratory infections:
are patients more satisfied when expectations are met? J Fam Pract. 1996;43:56-
62. [PMID: 0008691181]
46. Brody DS, Lerman CE, Wolfson HG, Caputo GC. Improvement in phy-
The Physician–Patient Relationship Unmet Expectations in Primary Care
896 1 May 2001 Annals of Internal Medicine Volume 134 • Number 9 (Part 2) www.annals.org
Downloaded From: http://annals.org/ by a Med College of Wisconsin User on 08/11/2015
sicians’ counseling of patients with mental health problems. Arch Intern Med.
1990;150:993-8. [PMID: 0002331204]
47. Hornberger J, Thom D, MaCurdy T. Effects of a self-administered previsit
questionnaire to enhance awareness of patients’ concerns in primary care. J Gen
Intern Med. 1997;12:597-606. [PMID: 0009346455]
48. Sanchez-Menegay C, Stalder H. Do physicians take into account patients’
expectations? J Gen Intern Med. 1994;9:404-6. [PMID: 0007931752]
49. Joos SK, Hickam DH, Gordon GH, Baker LH. Effects of a physician
communication intervention on patient care outcomes. J Gen Intern Med. 1996;
11:147-55. [PMID: 0008667091]
50. Jackson JL, Kroenke K. Patient satisfaction and quality of care. Mil Med.
1997;162:273-7. [PMID: 0009110554]
The Physician–Patient RelationshipUnmet Expectations in Primary Care
© 2001 American College of Physicians–American Society of Internal Medicine 897
Downloaded From: http://annals.org/ by a Med College of Wisconsin User on 08/11/2015
