Foundations and Essentials for the Doctor of Nursing Practice
RESEARCH ARTICLE
Practice variation and practice guidelines:
Attitudes of generalist and specialist
physicians, nurse practitioners, and physician
assistants
David A. Cook1,2,3*, Laurie J. Pencille1,4, Denise M. Dupras5, Jane A. Linderbaum1,6, V.
Shane Pankratz7, John M. Wilkinson8
1 Knowledge Delivery Center, Mayo Clinic, Rochester, Minnesota, United States of America, 2 Mayo Clinic
Online Learning, Mayo Clinic College of Medicine, Rochester, Minnesota, United States of America,
3 Division of General Internal Medicine, Mayo Clinic, Rochester, Minnesota, United States of America,
4 Center for the Science of Health Care Delivery, Mayo Clinic, Rochester, Minnesota, United States of
America, 5 Division of Primary Care Internal Medicine, Mayo Clinic, Rochester, Minnesota, United States
of America, 6 Department of Cardiovascular Diseases, Mayo Clinic, Rochester, Minnesota, United States
of America, 7 Department of Internal Medicine, University of New Mexico Health Sciences Center,
Albuquerque, New Mexico, United States of America, 8 Department of Family Medicine, Mayo Clinic,
Rochester, Minnesota, United States of America
* cook.david33@mayo.edu
Abstract
Objective
To understand clinicians’ beliefs about practice variation and how variation might be
reduced.
Methods
We surveyed board-certified physicians (N = 178), nurse practitioners (N = 60), and physi-
cian assistants (N = 12) at an academic medical center and two community clinics, repre-
senting family medicine, general internal medicine, and cardiology, from February—April
2016. The Internet-based questionnaire ascertained clinicians’ beliefs regarding practice
variation, clinical practice guidelines, and costs.
Results
Respondents agreed that practice variation should be reduced (mean [SD] 4.5 [1.1]; 1 =
strongly disagree, 6 = strongly agree), but agreed less strongly (4.1 [1.0]) that it can realisti-
cally be reduced. They moderately agreed that variation is justified by situational differences
(3.9 [1.2]). They strongly agreed (5.2 [0.8]) that clinicians should help reduce healthcare
costs, but agreed less strongly (4.4 [1.1]) that reducing practice variation would reduce
costs. Nearly all respondents (234/249 [94%]) currently depend on practice guidelines. Clini-
cians rated differences in clinician style and experience as most influencing practice varia-
tion, and inaccessibility of guidelines as least influential. Time to apply standards, and
patient decision aids, were rated most likely to help standardize practice. Nurse practitioners
PLOS ONE | https://doi.org/10.1371/journal.pone.0191943 January 31, 2018 1 / 12
a1111111111
a1111111111
a1111111111
a1111111111
a1111111111
OPENACCESS
Citation: Cook DA, Pencille LJ, Dupras DM,
Linderbaum JA, Pankratz VS, Wilkinson JM (2018)
Practice variation and practice guidelines: Attitudes
of generalist and specialist physicians, nurse
practitioners, and physician assistants. PLoS ONE
13(1): e0191943. https://doi.org/10.1371/journal.
pone.0191943
Editor: Kimon Bekelis, Dartmouth-Hitchcock
Medical Center, UNITED STATES
Received: April 4, 2017
Accepted: January 13, 2018
Published: January 31, 2018
Copyright: © 2018 Cook et al. This is an open
access article distributed under the terms of the
Creative Commons Attribution License, which
permits unrestricted use, distribution, and
reproduction in any medium, provided the original
author and source are credited.
Data Availability Statement: All relevant data are
within the paper and its Supporting Information
files.
Funding: The authors received no specific funding
for this work.
Competing interests: The authors have declared
that no competing interests exist.
and physicians assistants (vs physicians) and less experienced (vs senior) clinicians rated
more favorably several factors that might help to standardize practice. Differences by spe-
cialty and academic vs community practice were small.
Conclusions
Clinicians believe that practice variation should be reduced, but are less certain that this can
be achieved. Accessibility of guidelines is not a significant barrier to practice standardiza-
tion, whereas more time to apply standards is viewed as potentially helpful.
Introduction
Practice variation is an important target for clinical systems improvement. Although some var-
iation in clinical practice is justified, unwarranted variation–"variation that is not explained on
the basis of illness, patient risk factors or patient preferences"[1]–is common. This can lead to
underuse of effective care, overuse of non-beneficial services, and emphasis on physician opin-
ions rather than patient preferences.[2, 3] Unwarranted variation has been linked to subopti-
mal outcomes[4–8] and to increased cost for the same outcome (i.e., inefficient care).[9, 10]
While clinical practice guidelines offer one potential solution to unwarranted practice
variation,[5, 11, 12] implementation of guidelines encounters resistance, including both unin-
tentional and intentional non-adherence.[13–17] Studies exploring clinicians’ reasons for
resistance to guidelines have identified concerns about decreased autonomy, oversimplifica-
tion of medicine, uncertainty regarding the evidence base, financial conflicts of interest, and
potential litigation.[18–23] Interventions to improve guideline adherence have met with only
modest success.[24] Other approaches to reducing practice variation include patient decision
aids,[25, 26] clinical decision support,[27–29] provider performance feedback,[29, 30] pro-
vider financial incentives,[31–33] and regulatory changes,[34] yet these too have had limited
impact. Given these suboptimal results, we believe that before further trying to solve the prob-
lem of practice variation, we ought to better understand the nature of practice variation itself,
including clinician beliefs about the issue.
Qualitative studies have empirically identified sources of practice variation, including the
clinician (including personal experiences and attitudes), the patient and family, and the work
environment.[35–37] Conceptual models have added social values and regulatory bodies to
the list.[38] Yet we found no quantitative studies that directly explored the prevalence of clini-
cian attitudes about practice variation. Moreover, studies exploring attitudes about clinical
practice guidelines[13, 19, 20] and evidence-based medicine[39, 40] have incompletely
explored how attitudes vary across clinician characteristics such as time in practice, provider
type, and specialty. To address these gaps, we surveyed generalist and specialist clinicians to
answer the following research questions:
1. What do clinicians believe about the sources and impact of practice variation?
2. What do clinicians believe about the credibility and utility of clinical guidelines and other
approaches to reduce variation?
3. How do these beliefs vary by provider type, time in practice, specialty, and practice
location?
Attitudes about practice variation and guidelines
PLOS ONE | https://doi.org/10.1371/journal.pone.0191943 January 31, 2018 2 / 12
We anticipated that nurse practitioners / physician assistants (given the pragmatic nature of
their training and practice), clinicians early in their career (who may rely on external supports
as they continue to learn), and generalists (who must treat a wide variety of often-unfamiliar
conditions), would have more favorable beliefs about guidelines than physicians, senior clini-
cians, and subspecialists, respectively.
Methods
Overview
Clinicians completed a 27-item survey regarding their beliefs about clinical guidelines and
practice variation, as part of a larger study exploring clinical decision-making in the context of
written clinical vignettes (manuscript submitted).
Questionnaire development
An extensive search of PubMed did not reveal an existing questionnaire that specifically
addressed our questions regarding clinician attitudes about practice variation, so we created a
new questionnaire. We used prior surveys and review articles[19, 20, 40–43] to generate a list
of relevant themes. From these we developed 14 Likert-type items regarding clinical guidelines
and practice variation generally (response options, 1 = strongly disagree, 6 = strongly agree),
and 13 items exploring the impact on practice variation of six specific situational or personal
factors and seven potential practice changes (response options, 1 = not at all, 4 = very).
Authors DAC and DMD, both of whom have specialized training and extensive experience in
survey item development, supervised the development process and examined the wording of
each item. We iteratively reviewed and revised all items, including incorporation of findings
from pilot-testing by two internal medicine physicians. Table 1 contains the final full text of all
items. Respondents self-reported demographic information on their provider type, time in
practice, specialty, and practice location. The survey also asked respondents to describe how
they would manage four hypothetical patients (written clinical vignettes, not reported in this
manuscript) presenting for routine management of hyperlipidemia or syncope. Two of the
Likert-type items and all 13 impact items appeared after the vignettes; since not all respondents
completed the vignettes, the response rate for these "second half" items is lower than for the 12
"first half" items.
Sampling, participants, and survey administration
We sent email invitations to all 617 clinicians (458 physicians, 123 nurse practitioners, and 36
physician assistants) in the Mayo-Rochester, MN Divisions of General Internal Medicine
(N = 123), Primary Care Internal Medicine (N = 86), and Cardiology (N = 228), the Mayo-
Rochester Department of Family Medicine (N = 92), and the Mayo Clinic Health System com-
munity sites in Austin, MN (N = 19) and La Crosse, WI (N = 69). Participants were offered a
monetary incentive. The study was judged exempt by the Mayo Clinic Institutional Review
Board.
We used SurveyMonkey to administer the questionnaire from 26 February to 20 April,
2016. The invitation email indicated that this was a "research study to understand how clini-
cians make decisions at the point of care." All data were anonymous although we did track
responses. Invitees received up to two reminders.
Our original goal was to obtain data from 120 participants, and we over-enrolled partici-
pants in order to allow for those who might start but not finish the survey. The response to our
invitations and the volunteer rate were higher than expected, such that we exceeded our target
Attitudes about practice variation and guidelines
PLOS ONE | https://doi.org/10.1371/journal.pone.0191943 January 31, 2018 3 / 12
and had to close the survey before all those invited had a chance to respond, due to limited
funds for participant remuneration.
Statistical analyses
We report means, median, and standard deviation for each item. We determined the fre-
quency of favorable responses (e.g., "agree" or "important") by dichotomizing results as above
or below the midpoint of the respective scale. We classified time in practice as early (�10
years), mid (11–20 years), and late (>20 years) career. We used Kruskal-Wallis tests to
Table 1. Main survey results.
Statement or question Mean (SD),
mediana No./N (%) agree or moderate/
higha
Indicate your level of agreement with the following statements (range 1–6a):
Clinicians should always apply the latest research findings to each patient. 4.5 (1.0), 5 214/250 (86%)
Clinicians have a responsibility to help reduce the overall cost of medical care.b 5.2 (0.8), 5 245/250 (98%)
Reducing variation in clinical practice would benefit most patients.b 4.6 (1.1), 5 216/249 (87%)
It is hard to find and quickly comprehend state-of-the-art practice standards when I need them. 3.5 (1.3), 4 130/250 (52%)
Reducing variation in clinical practice would reduce costs. 4.4 (1.1), 5 201/249 (81%)
I trust the findings in most research studies and systematic reviews. 4.2 (0.9), 4 200/248 (81%)
I depend on practice guidelines to help me provide optimal care for my patients.b 4.9 (0.9), 5 234/249 (94%)
It is easy to apply practice guidelines to most of my patients. 4.3 (1.0), 4 197/247 (80%)
Most practice variation among clinicians is justified by relevant differences in clinical situations.b 3.9 (1.2), 4 161/250 (64%)
I am quick to adapt my clinical practice to align with new practice guidelines. 4.5 (0.9), 5 223/249 (90%)
Clinicians should encourage patients to follow guideline recommendations for diagnosis and treatment. 4.9 (0.8), 5 239/248 (96%)
Clinicians should resist patient requests that are not grounded in solid evidence of benefit. 4.4 (1.1), 4 202/250 (81%)
Between-clinician practice variation should be substantially reduced.c 4.5 (1.1), 5 133/152 (88%)
Between-clinician practice variation can realistically be substantially reduced.c 4.1 (1.0), 4 121/153 (79%)
How much impact do the following factors have on between-clinician variation in practice? (range 1–4a)
Lack of access to needed evidence and guidelines.c 2.6 (0.9), 3 87/153 (57%)
Lack of awareness of existing evidence and guidelines.c 3.0 (0.7), 3 115/153 (75%)
Differences in practice context and patient population.c 3.0 (0.7), 3 118/152 (78%)
Differences in clinician experience and training.c 3.2 (0.6), 3 134/153 (88%)
Differences in clinician style and preferences.c 3.3 (0.6), 3 136/153 (89%)
Individual patient preferences.c 2.9 (0.8), 3 99/153 (65%)
How helpful would each of the following be in helping you to standardize your clinical practice with other
clinicians? (range 1–4a)
Better access to guidelines and synthesized evidence.c 3.0 (0.9), 3 107/153 (70%)
More time to look up, appraise, and apply available practice standards.c 3.5 (0.7), 4 135/153 (88%)
Clearly stated institution-wide standard practices.c 3.1 (0.9), 3 116/153 (76%)
Standardized order sets.c 3.1 (0.9), 3 114/153 (75%)
Decision aids to help with patient counseling.c 3.2 (0.8), 3 124/153 (81%)
More frequent feedback on how my practice compares with that of others.c 3.0 (0.9), 3 103/152 (68%)
Having someone else order common / straightforward tests.c 2.3 (1.1), 2 58/153 (38%)
a For the first 14 items the response options were: 1 = strongly disagree, 2 = disagree, 3 = slightly disagree, 4 = slightly agree, 5 = agree, 6 = strongly agree; we
dichotomized 4–6 as "agree." For the remaining items the response options were: 1 = not at all, 2 = slight, 3 = moderate, 4 = very; we dichotomized 3–4 as "moderate/
high." b Key item (used to compare respondents who completed the second half of the survey and those who completed only the first half). c Item from the second half of the survey, which participants completed after they responded to four written clinical vignettes.
https://doi.org/10.1371/journal.pone.0191943.t001
Attitudes about practice variation and guidelines
PLOS ONE | https://doi.org/10.1371/journal.pone.0191943 January 31, 2018 4 / 12
compare responses across demographic subgroups. We used Cohen’s d to calculate a standard-
ized mean difference for statistically significant models, and classified these differences as
small (d = 0.2–0.49), moderate (d = 0.5–0.79), or large (d� 0.8). To compare those who com-
pleted the entire survey vs those who completed only the first half we used chi-squared tests to
compare demographic features, and used Kruskal-Wallis tests to compare responses for four a
priori "key items" (regarding responsibility to reduce cost, benefit of reducing variation,
dependence on guidelines, and justification of variation). We used Wilcoxon signed rank tests
to compare item responses within individuals. To explore the possibility that nonrespondents
were systematically different from respondents we compared available demographic informa-
tion (provider type, specialty, and practice location) among respondents and non-respondents
using the chi-squared test. Because we stopped enrollment before all those invited could
respond, we modeled the probability of participation as a function of the time allowed to reply
to the survey invitation. From this model, we predicted the probability of participation for all
individuals according to the larger of either 14 days (the per-protocol minimum) or the actual
time allowed to respond to the invitation. The "expected" response rate was the average of
these individual-level predictions that accounted for the early withdrawal of some invitations.
Given the large number of comparisons, we used a two-tailed alpha level of 0.01 as the thresh-
old of statistical significance. We used SAS version 9.4 (SAS Institute Inc., Cary, North Caro-
lina). We powered the study for the analysis of the vignette management (to be reported
elsewhere), and estimated the required sample size at 120. We did not separately power this
survey study.
Results
Two hundred fifty clinicians completed at least one questionnaire item (response rate 41%).
We found no statistically significant differences in provider type, specialty, or practice location
for respondents vs nonrespondents. As a secondary estimate of response, we determined the
response rate expected if all those invited had had an equal chance of responding (i.e., if none
had been uninvited) at 45%.
The sample included 178 physicians, 60 nurse practitioners, and 12 physician assistants.
Respondents had been in clinical practice a mean (SD) of 15.1 (11.3) years. Table 2 lists addi-
tional demographics. We found no statistically significant differences between the 153 respon-
dents (61%) who completed the second half of the survey and those who completed only the
first half, in demographic features (provider type, time in practice, specialty, or practice loca-
tion; p>.15) or in responses to the key questionnaire items (listed in Methods; p>.06).
Main survey findings
Impact and sources of variation in care. Clinicians generally agreed that practice varia-
tion should be reduced (mean [standard deviation], 4.5 [1.1]; 1 = strongly disagree, 6 = strongly
agree), but 65/152 (43%) respondents agreed less strongly that it can realistically be reduced
(4.1 [1.0], p< .001). Clinicians believed that reducing variation would benefit most patients
(4.6 [1.1]), and only slightly agreed that variation is justified by differences in clinical situations
(3.9 [1.2]).
When asked to rate factors that influence practice variation (1 = no impact, 4 = high
impact), respondents perceived differences in clinician style (3.3 [0.6]) and experience (3.2
[0.6]) as most influential, while patient differences (3.0 [0.7]) and preferences (2.9 [0.8]) were
moderately influential. Lack of access to guidelines was only slightly/moderately influential
(2.6 [0.9]).
Attitudes about practice variation and guidelines
PLOS ONE | https://doi.org/10.1371/journal.pone.0191943 January 31, 2018 5 / 12
When asked to rate factors that might help to standardized practice (1 = not at all helpful,
4 = very helpful), respondents rated time to identify and apply standards the highest (3.5
[0.7]), followed by patient decision aids (3.2 [0.8]), institution-wide standard practices (3.1
[0.9]), and standardized order sets (3.1 [0.9]). Having someone else order common tests was
rated lowest (2.3 [1.1]).
Cost. Clinicians strongly agreed (5.2 [0.8], range 1–6) that clinicians have a responsibility
to reduce the cost of care. They agreed less strongly (4.4 [1.1]) that reducing practice variation
would reduce costs.
Guidelines and evidence. Nearly all respondents (234/249 [94%]) agreed that they depend
on practice guidelines to help them provide patient care (mean 4.9 [0.9], range 1–6). They
reported that they are quick to adapt their own practice to align with new guidelines (4.5 [0.9])
and agreed that clinicians should apply new research findings (4.5 [1.0]). Repondents were
slightly less favorable regarding the ease of applying guidelines to their patients (4.3 [1.0]) and
their trust of research (4.2 [0.9]). Only 130/250 (52%) agreed that practice standards are hard
to find and apply (mean 3.5 [1.3]).
Nearly all (239/248 [96%]) agreed that clinicians should encourage patients to follow guide-
line recommendations (4.9 [0.8]). The also agreed, but less strongly, that clinicians should
resist patient requests that are not evidence-based (4.4 [1.1]).
Analyses across clinician demographic subgroups
Differences across provider type, time in practice, specialty, and practice location were gener-
ally small and not statistically significant (see Table 3, and S1 eTable). The statistically signifi-
cant (at alpha 0.01) comparisons across provider type were that nurse practitioners and
physician assistants rated more highly than physicians their trust in research (Cohen’s
d = 0.42), dependence on practice guidelines (d = 0.35), quickness to align with practice guide-
lines (d = 0.32), and five actions that might help standardize their clinical practice (better
access to guidelines and evidence, d = 0.65; more time to look up, appraise, and apply stan-
dards, d = 0.70; institution-wide standard practices, d = 0.87; decision aids for patient counsel-
ing, d = 0.56; and practice feedback, d = 0.47); all p�.008.
We found statistically significant differences by time in practice for four items. Early- or
mid-career clinicians reported higher agreement than late-career clinicians in dependence on
Table 2. Characteristics of invitees and respondents.
Demographic Feature Invited
No. (%)
Respondents
No. (%)
Provider type Physician 458 (74%) 178 (71%)
Nurse practitioner 123 (20%) 60 (24%)
Physician assistant 36 (6%) 12 (5%)
Years in practice �10 - - 109 (45%)
11–20 - - 66 (27%)
>20 - - 69 (28%)
Specialty Cardiology 227 (37%) 95 (38%)
Family medicine 181 (29%) 76 (30%)
Internal medicine 209 (34%) 79 (32%)
Practice location Academic 528 (86%) 213 (85%)
Community 89 (14%) 37 (15%)
N = 250 respondents. Years in practice information not available for invited cohort, and not reported by 6 respondents.
https://doi.org/10.1371/journal.pone.0191943.t002
Attitudes about practice variation and guidelines
PLOS ONE | https://doi.org/10.1371/journal.pone.0191943 January 31, 2018 6 / 12
practice guidelines (d = 0.42, p = .01); and rated time, institution-wide standard practices, and
decision aids as more helpful in standardizing practice (d = 0.47–0.60, p�.006).
Only one analysis by specialty reached statistical significance, namely that cardiologists
agreed more than internal medicine or family medicine clinicians that practice variation
should be reduced (d = 0.29, p = .01; see S1 eTable). None of the comparisons between aca-
demic vs community practice location were statistically significant (p>.01; see S1 eTable).
Table 3. Survey results: Subgroup analyses by provider type and years in practice.
Provider type Years in practice
Question / statement Physicians
N = 178
Mean (SD)
NP/PA
N = 72
Mean (SD)
P �10 yrs
N = 109
Mean (SD)
11–20 yrs
N = 66
Mean (SD)
>20 yrs
N = 69
Mean (SD)
P
Agree / disagree (range 1–6)
Apply latest research findings 4.4 (1.0) 4.6 (1.1) .08 4.6 (1.0) 4.5 (1.0) 4.2 (1.2) .12
Responsibility to reduce costs 5.2 (0.8) 5.1 (0.8) .30 5.2 (0.8) 5.3 (0.7) 5.1 (0.9) .81
Reducing variation would benefit patientsa 4.6 (1.0) 4.4 (1.2) .68 4.6 (1.1) 4.6 (1.1) 4.5 (1.2) .96
Hard to find & understand practice standards 3.5 (1.3) 3.3 (1.3) .26 3.7 (1.3) 3.3 (1.3) 3.2 (1.2) .02
Reducing variation would reduce costsa 4.4 (1.1) 4.3 (1.1) .30 4.4 (1.0) 4.4 (1.1) 4.4 (1.3) .62
Trust research & systematic reviewsc 4.1 (0.9) 4.5 (1.0) .001 4.2 (1.0) 4.2 (0.9) 4.1 (0.9) .74
Depend on practice guidelinesa 4.8 (0.9) 5.1 (0.9) .002 5.0 (0.9) 5.0 (0.8) 4.6 (0.9) .004
Easy to apply practice guidelinesd 4.2 (1.0) 4.5 (1.0) .02 4.4 (1.0) 4.3 (1.0) 4.1 (1.0) .31
Variation justified by clinical situations 3.8 (1.2) 4.2 (1.1) .03 4.1 (1.1) 4.0 (1.2) 3.6 (1.2) .03
Quick to adapt to align with guidelinesa 4.4 (0.8) 4.7 (0.9) .008 4.6 (0.9) 4.4 (0.8) 4.4 (0.9) .22
Encourage patients to follow guideline recommendationsc 4.8 (0.7) 4.9 (0.8) .18 4.9 (0.7) 4.8 (0.8) 4.8 (0.8) .48
Resist patient requests not grounded in evidence 4.4 (1.0) 4.3 (1.2) .91 4.4 (1.0) 4.3 (1.2) 4.3 (1.1) .74
Variation should be reduceda,b 4.5 (1.1) 4.6 (0.9) .64 4.5 (1.1) 4.7 (1.0) 4.4 (1.1) .32
Variation can realistically be reducedb 4.1 (1.0) 4.0 (0.9) .44 4.1 (0.8) 4.2 (1.2) 3.9 (0.9) .25
How much impact on between-clinician variation? (range 1–4)
Lack access to evidence & guidelinesb 2.6 (0.8) 2.8 (0.9) .12 2.6 (0.8) 2.6 (0.9) 2.6 (0.9) .91
Lack awareness evidence & guidelinesb 2.9 (0.7) 3.0 (0.7) .39 3.0 (0.7) 3.1 (0.9) 2.8 (0.7) .16
Differences in context & patient populationa,b 3.0 (0.7) 3.0 (0.7) .90 3.1 (0.7) 3.0 (0.6) 2.9 (0.7) .53
Differences in clinician experience & trainingb 3.2 (0.7) 3.3 (0.6) .22 3.4 (0.6) 3.1 (0.7) 3.1 (0.6) .04
Differences in clinician style & preferencesb 3.2 (0.7) 3.3 (0.6) .75 3.2 (0.6) 3.3 (0.7) 3.2 (0.7) .72
Individual patient preferencesb 2.8 (0.8) 3.0 (0.8) .08 2.8 (0.8) 2.9 (0.9) 2.9 (0.7) .89
How helpful to standardize your practice? (range 1–4)
Access to guidelines & evidenceb 2.8 (0.9) 3.4 (0.8) < .001 3.1 (0.9) 3.1 (0.9) 2.7 (1.0) .11
Time to look up, appraise, & apply standardsb 3.3 (0.7) 3.8 (0.5) < .001 3.6 (0.7) 3.5 (0.7) 3.1 (0.8) < .001
Clear institution-wide standard practicesb 2.9 (0.9) 3.6 (0.6) < .001 3.3 (0.7) 3.1 (1.0) 2.7 (0.9) .008
Standardized order setsb 3.0 (0.9) 3.3 (0.9) .07 3.3 (0.8) 3.1 (1.0) 2.9 (0.9) .12
Decision aids for patient counselingb 3.1 (0.9) 3.6 (0.6) .001 3.4 (0.8) 3.4 (0.8) 2.9 (0.9) .005
Feedback comparing my practice with othersa,b 2.8 (1.0) 3.3 (0.8) .009 3.0 (0.9) 3.1 (1.0) 2.6 (0.9) .03
Someone else order straightforward testsb 2.2 (1.1) 2.3 (1.1) .62 2.2 (1.1) 2.2 (1.1) 2.4 (1.0) .71
NP/PA = nurse practitioner / physician assistant, SD = standard deviation, yrs = years. See Table 1 for exact wording and response options for each item. Six
respondents did not report time in practice. a Missing 1 data point (N = item total-1). b Sample size N = 153 for this item unless otherwise noted by additional footnote; provider type N = 105 physicians, N = 48 NP/PAs; time in practice N = 70 at�10
years, N = 39 at 11–20 years, N = 39 at >20 years. c Missing 2 data points (N = item total-2). d Missing 3 data points (N = item total-3).
https://doi.org/10.1371/journal.pone.0191943.t003
Attitudes about practice variation and guidelines
PLOS ONE | https://doi.org/10.1371/journal.pone.0191943 January 31, 2018 7 / 12
Discussion
The physicians, nurse practitioners, and physician assistants responding to this survey agreed
that practice variation should be reduced, but many seemed to have reservations about the fea-
sibility of this aim. Time to appraise and apply practice standards was rated as most helpful in
standardizing practice. By contrast, access to guidelines was not perceived to be an important
factor in creating or resolving practice variation. Clinicians perceived that differences in clini-
cian style and experience have more influence on variation than patient differences or prefer-
ences. In subgroup analyses, nurse practitioners and physician assistants had more favorable
beliefs than physicians about practice guidelines and the potential value of several other
changes to help standardize practice. Younger clinicians rated nearly all proposed approaches
to standardization as more helpful. These differences were, with one exception, small or mod-
erate in magnitude (Cohen’s d<0.8). Generalists and specialists, and those at academic vs
community sites, had overall similar beliefs.
Limitations
The generalizability of our findings is limited by the response rate and the sampling from a sin-
gle geographic region and health system. While it is possible that those completing the survey
had stronger feelings about the topic than those who did not, the invitation email did not men-
tion practice variation or clinical practice guidelines, and respondents were similar to nonre-
spondents for available demographics. We acknowledge the large number of statistical tests
and suggest that statistically significant findings should be interpreted cautiously, highlighting
potentially interesting relationships that may merit further study. Finally, it is possible that
those responding to the second-half items were systematically different from those who
responded only to the first-half items (although we found no differences between these groups
in demographic features or responses to the key items), or that that the clinical vignettes could
have shaped their responses to the second-half items.
Strengths include the sample that encompassed multiple provider types and specialties, and
both academic and community sites; the carefully developed survey; and the sample size that
exceeded our planned enrollment.
Integration with prior work
Our findings extend those of earlier qualitative studies[35, 36, 40] by prioritizing the factors
that influence variation. Our findings also parallel prior work exploring the adoption of or
resistance to clinical practice guidelines, namely that physicians have favorable attitudes about
practice guidelines in general yet remain somewhat skeptical of specific recommendations,
have difficulty applying them to specific patients, and lack time to do so.[18–20] A recent
review of studies exploring non-adherence to guidelines found that much non-adherence is
intentional, and at least partly warranted by patient decisions or case-specific contraindica-
tions.[13] Similar to our results, at least one US study has found more favorable attitudes
among younger physicians.[44] By contrast with our findings, a study in Norway found signifi-
cantly greater challenges for generalists compared with non-generalist physicians.[43]
Implications
The clinicians in this sample did not express strong concerns about practice variation: about
two-thirds agreed that variation is justified by relevant clinical differences, and the vast
majority indicated that they already follow guidelines themselves. While some variation is
indeed justified,[13] studies suggest that non-adherence to guidelines is widespread.[1, 16,
Attitudes about practice variation and guidelines
PLOS ONE | https://doi.org/10.1371/journal.pone.0191943 January 31, 2018 8 / 12
45] Moreover, clinicians’ perceptions of their own guideline adherence is typically more
favorable than reality.[46–48] Yet even if clinicians are mistaken in their beliefs, their per-
ceptions must be taken seriously and at face value. It will be difficult to fix the problem if
those involved fail to recognize its magnitude, importance, or potential for correction.
Changing attitudes will require that beliefs be acknowledged, needs understood, evidence
sought, and misperceptions corrected. Systems-level solutions will play an essential role in
standardizing practice, but attempts to circumvent clinicians, i.e., through automated imple-
mentation of guidelines, may backfire. We note, for example, that the lowest-rated item on
our survey was a potential practice change in which someone else would order common or
straightforward tests.
Physicians seem to have lingering doubts that practice variation can be reduced. This may
reflect low "outcome expectancy"–the belief that certain outcomes will result from given
actions.[49] Before clinicians will invest personal effort, they must believe that their efforts will
result in desired outcomes.[18] Low outcome expectancy may reflect concerns about the effec-
tiveness of currently-available solutions including guidelines,[18, 19, 21, 22] decision aids,[25]
clinical decision support tools,[27] and individual performance feedback.[30] Research to
identify new solutions, and how to more effectively implement existing ideas, remains a top
priority. Alternatively, doubts about reducing practice variation may reflect a lack of personal
self-efficacy—the capability to perform at a certain level in a specific task and context.[50]
Research exploring how to influence clinicians’ self-efficacy and other motivations may prove
insightful.[49, 51, 52]
Although younger clinicians and nurse practitioners and physician assistants rated several
items higher than senior clinicians and physicians, relative prioritizations (rankings) were gen-
erally uniform across subgroups. This suggests that solutions need not target specific groups
during development, although testing during implementation would be warranted.
Perhaps most importantly, our respondents indicated that the factor anticipated to most
help standardize their clinical practice is not new standards or better access to guidelines, but
is rather time to apply existing standards. Others have reported similar beliefs.[18, 19] This
suggests that instead of focusing attention exclusively on developing and implementing guide-
lines, we might have more success by freeing up clinicians’ time or reducing the time needed
to apply patient-tailored guideline recommendations. Since decreasing clinicians’ time burden
is likely unrealistic (and, even if achieved, might not actually be used to study and apply stan-
dards), we propose to focus on how to efficiently integrate practice standards into clinicians’
workflow.
Conclusions
Clinicians believe that practice variation should be reduced, but are less certain that this can be
achieved. Accessibility of guidelines is not a significant barrier to practice standardization,
whereas more time to apply standards is viewed as potentially helpful.
Supporting information
S1 eTable. Survey results: Subgroup analyses by specialty and practice location.
(DOCX)
Acknowledgments
We thank Phil Einspahr for his support in administering the survey.
Attitudes about practice variation and guidelines
PLOS ONE | https://doi.org/10.1371/journal.pone.0191943 January 31, 2018 9 / 12
Author Contributions
Conceptualization: David A. Cook, Laurie J. Pencille, John M. Wilkinson.
Data curation: David A. Cook, Laurie J. Pencille, Denise M. Dupras, Jane A. Linderbaum,
John M. Wilkinson.
Formal analysis: David A. Cook, V. Shane Pankratz.
Investigation: David A. Cook, Laurie J. Pencille, Denise M. Dupras, Jane A. Linderbaum, V.
Shane Pankratz, John M. Wilkinson.
Methodology: David A. Cook, Denise M. Dupras, V. Shane Pankratz.
Project administration: David A. Cook, Laurie J. Pencille.
Resources: David A. Cook.
Supervision: David A. Cook.
Validation: David A. Cook, Laurie J. Pencille, Denise M. Dupras.
Writing – original draft: David A. Cook.
Writing – review & editing: David A. Cook, Laurie J. Pencille, Denise M. Dupras, Jane A. Lin-
derbaum, V. Shane Pankratz, John M. Wilkinson.
References 1. Corallo AN, Croxford R, Goodman DC, Bryan EL, Srivastava D, Stukel TA. A systematic review of medi-
cal practice variation in OECD countries. Health Policy 2014; 114:5–14. https://doi.org/10.1016/j.
healthpol.2013.08.002 PMID: 24054709
2. Wennberg JE. Unwarranted variations in healthcare delivery: implications for academic medical cen-
tres. BMJ 2002; 325:961–964. PMID: 12399352
3. Wennberg JE. Time to tackle unwarranted variations in practice. BMJ 2011; 342:687–690.
4. McCabe C, Kirchner C, Zhang H, Daley J, Fisman DN. Guideline-concordant therapy and reduced mor-
tality and length of stay in adults with community-acquired pneumonia: playing by the rules. Arch Intern
Med 2009; 169:1525–1531. https://doi.org/10.1001/archinternmed.2009.259 PMID: 19752411
5. Patel J, Ahmed K, Guru KA, Khan F, Marsh H, Shamim Khan M, et al. An overview of the use and imple-
mentation of checklists in surgical specialities—a systematic review. Int J Surg 2014; 12:1317–1323.
https://doi.org/10.1016/j.ijsu.2014.10.031 PMID: 25448652
6. van Diepen S, Bakal JA, Lin M, Kaul P, McAlister FA, Ezekowitz JA. Variation in critical care unit admis-
sion rates and outcomes for patients with acute coronary syndromes or heart failure among high- and
low-volume cardiac hospitals. J Am Heart Assoc 2015; 4:e001708. https://doi.org/10.1161/JAHA.114.
001708 PMID: 25725089
7. Adam MA, Goffredo P, Youngwirth L, Scheri RP, Roman SA, Sosa JA. Same thyroid cancer, different
national practice guidelines: When discordant American Thyroid Association and National Comprehen-
sive Cancer Network surgery recommendations are associated with compromised patient outcome.
Surgery 2016; 159:41–50. https://doi.org/10.1016/j.surg.2015.04.056 PMID: 26435426
8. Boland GM, Chang GJ, Haynes AB, Chiang YJ, Chagpar R, Xing Y, et al. Association between adher-
ence to National Comprehensive Cancer Network treatment guidelines and improved survival in
patients with colon cancer. Cancer 2013; 119:1593–1601. https://doi.org/10.1002/cncr.27935 PMID:
23280510
9. Clough JD, Rajkumar R, Crim MT, Ott LS, Desai NR, Conway PH, et al. Practice-Level Variation in Out-
patient Cardiac Care and Association With Outcomes. J Am Heart Assoc 2016; 5.
10. Clough JD, Patel K, Shrank WH. Variation in Specialty Outpatient Care Patterns in the Medicare Popu-
lation. J Gen Intern Med 2016; 31:1278–1286. https://doi.org/10.1007/s11606-016-3745-8 PMID:
27259290
11. Woolf SH, Grol R, Hutchinson A, Eccles M, Grimshaw J. Clinical guidelines: potential benefits, limita-
tions, and harms of clinical guidelines. BMJ 1999; 318:527–530. PMID: 10024268
12. Worrall G, Chaulk P, Freake D. The effects of clinical practice guidelines on patient outcomes in primary
care: a systematic review. CMAJ 1997; 156:1705–1712. PMID: 9220922
Attitudes about practice variation and guidelines
PLOS ONE | https://doi.org/10.1371/journal.pone.0191943 January 31, 2018 10 / 12
13. Arts DL, Voncken AG, Medlock S, Abu-Hanna A, van Weert HC. Reasons for intentional guideline non-
adherence: A systematic review. Int J Med Inform 2016; 89:55–62. https://doi.org/10.1016/j.ijmedinf.
2016.02.009 PMID: 26980359
14. Urkin J, Allenbogen M, Friger M, Vinker S, Reuveni H, Elahayani A. Acute pharyngitis: low adherence to
guidelines highlights need for greater flexibility in managing paediatric cases. Acta Paediatr 2013;
102:1075–1080. https://doi.org/10.1111/apa.12364 PMID: 23879261
15. Patel N, Tong L, Ahn C, Singal AG, Gupta S. Post-polypectomy Guideline Adherence: Importance of
Belief in Guidelines, Not Guideline Knowledge or Fear of Missed Cancer. Dig Dis Sci 2015; 60:2937–
2945. https://doi.org/10.1007/s10620-015-3685-x PMID: 25947332
16. Carpentier MY, Vernon SW, Bartholomew LK, Murphy CC, Bluethmann SM. Receipt of recommended
surveillance among colorectal cancer survivors: a systematic review. J Cancer Surviv 2013; 7:464–
483. https://doi.org/10.1007/s11764-013-0290-x PMID: 23677524
17. Sivananthan SN, Puyat JH, McGrail KM. Variations in self-reported practice of physicians providing clin-
ical care to individuals with dementia: a systematic review. J Am Geriatr Soc 2013; 61:1277–1285.
https://doi.org/10.1111/jgs.12368 PMID: 23889524
18. Cabana MD, Rand CS, Powe NR, Wu AW, Wilson MH, Abboud PA, et al. Why don’t physicians follow
clinical practice guidelines? A framework for improvement. JAMA 1999; 282:1458–1465. PMID:
10535437
19. Carlsen B, Glenton C, Pope C. Thou shalt versus thou shalt not: a meta-synthesis of GPs’ attitudes to
clinical practice guidelines. Br J Gen Pract 2007; 57:971–978. https://doi.org/10.3399/
096016407782604820 PMID: 18252073
20. Farquhar CM, Kofa EW, Slutsky JR. Clinicians’ attitudes to clinical practice guidelines: a systematic
review. Med J Aust 2002; 177:502–506. PMID: 12405894
21. Fischer F, Lange K, Klose K, Greiner W, Kraemer A. Barriers and Strategies in Guideline Implementa-
tion-A Scoping Review. Healthcare (Basel) 2016; 4(3).
22. Kastner M, Bhattacharyya O, Hayden L, Makarski J, Estey E, Durocher L, et al. Guideline uptake is
influenced by six implementability domains for creating and communicating guidelines: a realist review.
J Clin Epidemiol 2015; 68:498–509. https://doi.org/10.1016/j.jclinepi.2014.12.013 PMID: 25684154
23. Nuckols TK, Lim YW, Wynn BO, Mattke S, MacLean CH, Harber P, et al. Rigorous development does
not ensure that guidelines are acceptable to a panel of knowledgeable providers. J Gen Intern Med
2008; 23:37–44.
24. Flodgren G, Hall AM, Goulding L, Eccles MP, Grimshaw JM, Leng GC, et al. Tools developed and dis-
seminated by guideline producers to promote the uptake of their guidelines. Cochrane Database Syst
Rev 2016:CD010669. https://doi.org/10.1002/14651858.CD010669.pub2 PMID: 27546228
25. O’Connor AM, Llewellyn-Thomas HA, Flood AB. Modifying unwarranted variations in health care:
shared decision making using patient decision aids. Health Aff (Millwood) 2004; Suppl Variation:
VAR63–72.
26. Reames BN, Shubeck SP, Birkmeyer JD. Strategies for reducing regional variation in the use of sur-
gery: a systematic review. Ann Surg 2014; 259:616–627. https://doi.org/10.1097/SLA.
0000000000000248 PMID: 24240626
27. Fiander M, McGowan J, Grad R, Pluye P, Hannes K, Labrecque M, et al. Interventions to increase the
use of electronic health information by healthcare practitioners to improve clinical practice and patient
outcomes. Cochrane Database Syst Rev 2015:CD004749. https://doi.org/10.1002/14651858.
CD004749.pub3 PMID: 25770311
28. Prevedello LM, Raja AS, Ip IK, Sodickson A, Khorasani R. Does clinical decision support reduce unwar-
ranted variation in yield of CT pulmonary angiogram? Am J Med 2013; 126:975–981. https://doi.org/10.
1016/j.amjmed.2013.04.018 PMID: 24157288
29. Forrest CB, Fiks AG, Bailey LC, Localio R, Grundmeier RW, Richards T, et al. Improving adherence to
otitis media guidelines with clinical decision support and physician feedback. Pediatrics 2013; 131:
e1071–1081. https://doi.org/10.1542/peds.2012-1988 PMID: 23478860
30. Ivers N, Jamtvedt G, Flottorp S, Young JM, Odgaard-Jensen J, French SD, et al. Audit and feedback:
effects on professional practice and healthcare outcomes. Cochrane Database Syst Rev 2012; 6:
CD000259.
31. Scott A, Sivey P, Ait Ouakrim D, Willenberg L, Naccarella L, Furler J, et al. The effect of financial incen-
tives on the quality of health care provided by primary care physicians. Cochrane Database Syst Rev
2011:CD008451. https://doi.org/10.1002/14651858.CD008451.pub2 PMID: 21901722
32. Houle SK, McAlister FA, Jackevicius CA, Chuck AW, Tsuyuki RT. Does performance-based remunera-
tion for individual health care practitioners affect patient care?: a systematic review. Ann Intern Med
2012; 157:889–899. https://doi.org/10.7326/0003-4819-157-12-201212180-00009 PMID: 23247940
Attitudes about practice variation and guidelines
PLOS ONE | https://doi.org/10.1371/journal.pone.0191943 January 31, 2018 11 / 12
33. Baxter PE, Hewko SJ, Pfaff KA, Cleghorn L, Cunningham BJ, Elston D, et al. Leaders’ experiences and
perceptions implementing activity-based funding and pay-for-performance hospital funding models: A
systematic review. Health Policy 2015; 119:1096–1110. https://doi.org/10.1016/j.healthpol.2015.05.
003 PMID: 26004845
34. McCulloch P, Nagendran M, Campbell WB, Price A, Jani A, Birkmeyer JD, et al. Strategies to reduce
variation in the use of surgery. Lancet 2013; 382:1130–1139. https://doi.org/10.1016/S0140-6736(13)
61216-7 PMID: 24075053
35. Wilson ME, Rhudy LM, Ballinger BA, Tescher AN, Pickering BW, Gajic O. Factors that contribute to
physician variability in decisions to limit life support in the ICU: a qualitative study. Intensive Care Med
2013; 39:1009–1018. https://doi.org/10.1007/s00134-013-2896-x PMID: 23559079
36. Brouwers MC, Makarski J, Garcia K, Akram S, Darling GE, Ellis PM, et al. A mixed methods approach
to understand variation in lung cancer practice and the role of guidelines. Implement Sci 2014; 9:36.
https://doi.org/10.1186/1748-5908-9-36 PMID: 24655753
37. Grant A, Sullivan F, Dowell J. An ethnographic exploration of influences on prescribing in general prac-
tice: why is there variation in prescribing practices? Implement Sci 2013; 8:72. https://doi.org/10.1186/
1748-5908-8-72 PMID: 23799906
38. Reschovsky JD, Rich EC, Lake TK. Factors Contributing to Variations in Physicians’ Use of Evidence at
The Point of Care: A Conceptual Model. J Gen Intern Med 2015; 30(Suppl 3):S555–561.
39. McColl A, Smith H, White P, Field J. General practitioner’s perceptions of the route to evidence based
medicine: a questionnaire survey. BMJ 1998; 316:361–365. PMID: 9487174
40. Zwolsman S, te Pas E, Hooft L, Wieringa-de Waard M, van Dijk N. Barriers to GPs’ use of evidence-
based medicine: a systematic review. Br J Gen Pract 2012; 62:e511–521. https://doi.org/10.3399/
bjgp12X652382 PMID: 22781999
41. Hajjaj FM, Salek MS, Basra MK, Finlay AY. Non-clinical influences on clinical decision-making: a major
challenge to evidence-based practice. J R Soc Med 2010; 103:178–187. https://doi.org/10.1258/jrsm.
2010.100104 PMID: 20436026
42. Mercuri M, Gafni A. Medical practice variations: what the literature tells us (or does not) about what are
warranted and unwarranted variations. J Eval Clin Pract 2011; 17:671–677. https://doi.org/10.1111/j.
1365-2753.2011.01689.x PMID: 21501341
43. Carlsen B, Bringedal B. Attitudes to clinical guidelines—do GPs differ from other medical doctors? BMJ
Qual Saf 2011; 20:158–162. https://doi.org/10.1136/bmjqs.2009.034249 PMID: 21209148
44. Tunis SR, Hayward RS, Wilson MC, Rubin HR, Bass EB, Johnston M, et al. Internists’ attitudes about
clinical practice guidelines. Ann Intern Med 1994; 120:956–963. PMID: 8172440
45. Gouvea M, Novaes Cde O, Pereira DM, Iglesias AC. Adherence to guidelines for surgical antibiotic pro-
phylaxis: a review. Braz J Infect Dis 2015; 19:517–524. https://doi.org/10.1016/j.bjid.2015.06.004
PMID: 26254691
46. Steinman MA, Fischer MA, Shlipak MG, Bosworth HB, Oddone EZ, Hoffman BB, et al. Clinician aware-
ness of adherence to hypertension guidelines. Am J Med 2004; 117:747–754. https://doi.org/10.1016/j.
amjmed.2004.03.035 PMID: 15541324
47. Linder JA, Schnipper JL, Tsurikova R, Volk LA, Middleton B. Self-reported familiarity with acute respira-
tory infection guidelines and antibiotic prescribing in primary care. Int J Qual Health Care 2010;
22:469–475. https://doi.org/10.1093/intqhc/mzq052 PMID: 20935008
48. Oderda GM, Shiozawa A, Walsh M, Hess K, Brixner DI, Feehan M, et al. Physician adherence to ACR
gout treatment guidelines: perception versus practice. Postgrad Med 2014; 126:257–267. https://doi.
org/10.3810/pgm.2014.05.2774 PMID: 24918810
49. Cook DA, Artino AR Jr. Motivation to learn: an overview of contemporary theories. Med Educ 2016;
50:997–1014. https://doi.org/10.1111/medu.13074 PMID: 27628718
50. Bandura A: Self-efficacy. In: Encyclopedia of human behavior. Volume 4. Edited by Ramachaudran
VS. New York: Academic Press; 1994: 71–81
51. Brydges R, Butler D. A reflective analysis of medical education research on self-regulation in learning
and practice. Med Educ 2012; 46:71–79. https://doi.org/10.1111/j.1365-2923.2011.04100.x PMID:
22150198
52. Ng JYY, Ntoumanis N, Thøgersen-Ntoumani C, Deci EL, Ryan RM, Duda JL, et al. Self-Determination
Theory Applied to Health Contexts: A Meta-Analysis. Perspectives on Psychological Science 2012;
7:325–340. https://doi.org/10.1177/1745691612447309 PMID: 26168470
Attitudes about practice variation and guidelines
PLOS ONE | https://doi.org/10.1371/journal.pone.0191943 January 31, 2018 12 / 12
Copyright of PLoS ONE is the property of Public Library of Science and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.