Content uploaded by Cheryl Lynn Eschbach
Author content
All content in this area was uploaded by Cheryl Lynn Eschbach on Jun 02, 2020
Content may be subject to copyright.
Health Promotion Practice
March 2020 Vol. 21, No. (2) 175 –180
DOI:https://doi.org/10.1177/1524839919868980
Article reuse guidelines: sagepub.com/journals-permissions
© 2019 Society for Public Health Education
175
Research Brief
Clinical–community linkages enhance health care
delivery and enable physician–patient partnerships to
achieve better health. The Michigan State University
(MSU) Model of Health Extension includes a strategy
for forming these linkages by focusing on increasing
primary care patient referrals and enrollment in health
programs. This article shares the results of a survey of
Michigan internal medicine and family medicine phy-
sicians (n = 323) to better understand attitudes toward
and familiarity with community-based education (CBE)
programs and to assess the logistical requirements to
make CBE referrals efficient and sustainable. Survey
results showed that at most, 55% of respondents were
aware of at least one CBE program implemented by
Cooperative Extension. Of those who were aware, over
85% agreed that the programs have positive benefits for
patients. Thirty-five percent reported at least one refer-
ral barrier, and familiarity with the CBE programs was
a significant predictor for reporting all referral barriers.
The results suggest that increasing physicians’ familiar-
ity of CBE health programs is a key first step in identify-
ing ideal strategies to overcome referral barriers. Data
from this study may help determine scalable state level
models for increasing awareness of chronic disease
prevention and other CBE programs in efforts to improve
the health of the nation.
Keywords: community-based education; health exten-
sion; patient barriers; physician referrals
>
INTRODUCTION
Chronic conditions are leading causes of death
and disability in the United States (National Center
for Health Statistics, 2016), and are the most com-
mon, costly, and preventable health problems. Sixty
percent of Americans have a chronic health condition
(Buttorff, Ruder, & Bauman, 2017), and fewer than
half of these patients receive health education in out-
patient settings (Ritsema, Bingenheimer, Scholting, &
Cawley, 2014).
Key to improving the nation’s health are strategies
that prevent and manage chronic diseases by linking
clinical care teams and community programs, yet each
face different sets of challenges. Clinical care teams
tend to lack time to effectively deliver the intensive
counseling needed to address health behavior changes
and perceive that social determinants of health fall
outside their scope of influence (Fenton, 2011; Ono et
al., 2018; Torres, 2015). Community programs are chal-
lenged with recruiting the appropriate audiences to fill
classes and a scope of influence that falls outside of
868980HPPXXX10.1177/1524839919868980Health Promotion PracticeKhan et al. / Connecting Primary Care To Cbe
research-article2019
1American Medical Association, Chicago, IL, USA
2Michigan State University, East Lansing, MI, USA
Authors’ Note: Address correspondence to Tamkeen Khan,
American Medical Association, 330 North Wabash Avenue, Suite
39300, Chicago, IL 60611, USA; e-mail: tamkeen.khan@ama-
assn.org.
Connecting Primary Care to Community-Based
Education: Michigan Physicians’ Familiarity With
Extension Programs
Tamkeen Khan, PhD1
Cheryl Eschbach, PhD2
Courtney A. Cuthbertson, PhD2
Cathy Newkirk, MS2
Dawn Contreras, PhD2
Kate Kirley, MD1
176 HEALTH PROMOTION PRACTICE / March 2020
clinical practice (Braun et al., 2014; Centers for Disease
Control and Prevention [CDC], 2015). Establishing clin-
ical–community linkages enhances health care deliv-
ery and enables physicians to partner with patients to
achieve better health (Torres, 2015). Bridging this gap
requires identifying community resources in advance,
developing referral mechanisms, and engaging external
intermediaries to address the challenges.
Health extension is a national movement to improve
health outcomes by linking patients, clinical and pri-
mary care practices, community resources, support,
and health education (Dwyer et al., 2017; Grumbach &
Mold, 2009). The Michigan State University (MSU)
Model of Health Extension includes a strategy for form-
ing clinical–community linkages by focusing on
increasing primary care patient referrals and enroll-
ment in health programs currently provided by
Cooperative Extension, such as those for diabetes pre-
vention (Dwyer et al., 2017).
Our study examines linkages between primary care
and community-based education (CBE) resources such
as those offered by Cooperative Extension. The goal is
to measure physicians’ awareness of and familiarity
with CBE programs provided by Cooperative Extension
in Michigan (Michigan State University [MSU] Exte nsion),
assess logistical requirements to make CBE referrals
efficient and sustainable, and increase physicians’ patient
referrals to CBE programs.
>
METHOD
Data Source and Study Population
The survey sample came from the 2016 American
Medical Association (AMA) Physician Masterfile.
Eligibility criteria included licensed physicians in direct
patient care within Michigan in internal medicine or
family medicine and having an e-mail address. Data
weights were generated using an iterative raking method
to account for gender, age-group, and degree (Izrael,
Hoaglin, & Battaglia, 2000) so the results are representa-
tive of the state. Population distribution information was
obtained from the 2016 AMA Physician Characteristics
and Distribution in the United States (AMA, 2014), and
sample distributions were derived from the Michigan
physicians listed in the AMA Masterfile.
Study Variables
AMA and MSU Extension co-developed a survey
instrument to assess primary care physicians’ familiar-
ity with 10 CBE programs implemented by MSU
Extension, examine the level of agreement that CBE
programs help patients in various ways, and measure
barriers or facilitators they may encounter when refer-
ring to these programs. The survey collected self-
reported data on number of patients seen weekly,
number of weekly CBE referrals, and provider race or
ethnicity. The survey was administered using an online
survey tool and was open for 6 weeks in 2017.
Physician characteristics, including age, gender,
medical degree, medical school, and primary office
location were obtained from the AMA Masterfile.
Urbanicity of physicians’ primary office location was
derived by geocoding at the street level and cross walk-
ing county Federal Information Processing Standards
Codes with 2013 USDA Rural Urban Continuum Codes
(USDA, 2013) composed of the Office of Management
and Budget metro and nonmetro categories.
Data Analysis
Primary outcome variables are presented by physi-
cian characteristics. For univariate analyses, we used
Pearson’s chi-square tests to identify significant rela-
tionships (Bland & Altman, 1995). Weighted multivari-
ate logistic regression analyses are presented, which
model the effects of the various physician characteris-
tics on reasons they are less likely to refer patients to
CBE programs. We present adjusted odds ratios for
each explanatory variable. Statistical significance was
defined as p < .05. All data analyses for this project
were conducted using SAS Version 9.4 and STATA
Version 13.0 and accounted for sampling weights. The
institutional review boards at Michigan State University
and the University of Illinois at Chicago reviewed the
protocol and ruled it exempt.
>
RESULTS
A total of 4,640 physicians met the inclusion crite-
ria and were invited to participate. In total, 323 returned
surveys were eligible, yielding a response rate of 7.3%.
Details on the response rate calculation are available on
request.
Table 1 includes weighted characteristics of survey
respondents. Respondents’ familiarity with individual
CBE programs offered by MSU Extension was low.
Respondents were most familiar with CDC-recognized
lifestyle change programs and programs about living
well with diabetes, and least familiar with programs
related to managing money, home ownership, stress
and anger management, parenting education, and pre-
venting foodborne illness. The respondents who were
familiar with MSU Extension CBE programs felt they
had positive benefits for their patients. For instance,
91.5% believed programs help patients increase aware-
ness of the importance of health condition–related testing
Khan et al. / CONNECTING PRIMARY CARE TO CBE 177
(e.g., cholesterol), 91.0% acknowledged programs help
patients gain access to health literacy or health promotion
tools, and 90.6% agreed programs help patients gain
positive attitudes about self-management. Additional
benefits of CBE programs, as reported by percentage of
physicians who agreed, included patients’ gaining
knowledge (89.9%), making informed decisions (88.1%),
learning skills to manage symptoms (88.0%), and build-
ing a supportive network (85.9%).
Table 2 presents the univariate analyses of the
respondents’ characteristics by reasons physicians are
less likely to refer their patients to CBE programs, con-
trolling for the provider level characteristics. Overall,
52.9% reported limited knowledge and evidence,
43.0% reported inadequate availability, 42.1% reported
patient-level barriers to participation, and 35.3% report ed
inadequate resources and time to make referrals as the
reasons they are less likely to refer patients to CBE pro-
grams. There were no significant differences for the
referral barriers by age, gender, or medical degree.
However, a significantly higher percentage of physi-
cians who graduated from an MSU medical school
reported patient-level barriers to participation (55.7%
vs. 38.9%, p < .05). Physicians who reported referring
to CBE programs were more likely to acknowledge bar-
riers to referrals, where 56.8% (vs. 34.0%, p < .001)
reported inadequate availability, 66.7% (vs. 44.5%,
p < .001) reported limited knowledge, 44.0% (vs.
30.9%, p < .01) reported inadequate resources such as
time, and 53.0% (vs. 35.6%, p < .01) reported patient-
level barriers. Among physicians who were familiar
with the CBE programs, 55.6% (vs. 33.0%, p < .001)
reported inadequate availability, 66.7% (vs. 41.9%,
p < .001) reported limited knowledge and evidence,
47.2% (vs. 25.7%, p < .001) reported inadequate
resources and time to make referrals, and 56.3% (vs.
30.7%, p < .001) reported patient-level barriers to
participation. Furthermore, physicians practicing in
nonurban areas were more likely to report inadequate
availability (66.7% vs. 39.0%, p < .001) and patient-
level participation barriers (57.8% vs. 39.4%, p < .05)
as reasons they were less likely to refer.
Table 3 shows results of four separate logistic regres-
sion models to examine independent predictors of rea-
sons physicians were less likely to refer patients to CBE
programs. Familiarity with the CBE programs was a
significant predictor in all four models. Those who
were familiar with at least one of MSU Extension’s CBE
programs had more than twofold greater odds of report-
ing all four reasons they are less likely to refer their
patients to these programs, ranging from inadequate
availability (adjusted odds ratio [aOR] = 1.852, p < .01)
to patient-level participation barriers (aOR = 2.860,
p < .001). Providers who were making patient referrals
to CBE programs also noted significant barriers includ-
ing patient-level barriers to participation (aOR = 1.529,
p < .05), limited knowledge and evidence (aOR = 2.151,
p < .001), and inadequate availability (aOR = 2.293,
p < .001). Those who attended medical school at MSU
were more likely to report patient-level barriers to par-
ticipation (aOR = 2.359, p < .01), and those who prac-
ticed in nonurban settings were more likely to report
inadequate availability (aOR = 2.875, p < .01) as the
reasons they are less likely to refer.
Among those who reported barriers to referring
patients to CBE programs, we asked what would facilitate
TABLE 1
Characteristics of Survey Respondents
Respondent Characteristic All (n) Weighteda (%)
Age in years
≤44 71 22.02
45-65 192 59.42
66+60 18.56
Gender
Male 210 65.02
Female 113 34.98
Degree
MD 251 77.69
DO 72 22.31
Medical school
MSU 61 18.89
Other 262 81.11
Familiar with programsb
Yes 144 55.42
No 179 44.58
Urbanc
Yes 277 85.76
No 45 13.93
Refer to CBE programsb
Yes 132 40.87
No/don’t know 191 59.13
NOTE: N = 323. MSU = Michigan State University; CDC =
Centers for Disease Control and Prevention.
aWeighting accounts for race/ethnicity, gender, and age-group.
bSelf-reported: binary variable to measure overall familiarity as
somewhat to very familiar with any of the listed programs. Ten
types of programs related to eating healthy and being active, par-
ticipating in CDC-recognized lifestyle change programs, living
well with diabetes, parenting education, managing money, home
ownership education, dealing with stress and anger management,
cooking for health, healthy aging, and preventing foodborne ill-
ness. c“Yes” includes all counties in metropolitan areas.
178 HEALTH PROMOTION PRACTICE / March 2020
overcoming these barriers. Regarding the challenge of
inadequate availability, 78.4% stated increased availa-
bility of CBE programs (in-person and online). For those
who selected limited knowledge and evidence about
the CBE programs as a referral barrier, 55.6% stated
Continuing Medical Education credits about referring to
CBE programs would be a potential solution and 55.0%
suggested meeting with CBE staff or learning more about
the content to better understand what patients will
experience. Regarding the barrier of inadequate resources
and time to make referrals, 75.4% suggested increased
staff resources (program coordinator, community navi-
gators) to facilitate referrals as well as communication
between their offices and CBE programs. Finally, for
those who reported facing patient-level barriers to pro-
gram participation, 72.7% cited expanding insurance
coverage would be a viable solution to promote CBE
program participation.
>
DISCUSSION
Our study shows low awareness among physicians
about CBE programs offered through MSU Extension,
especially those that address social determinants of
health. The results suggest that increasing physician
familiarity of CBE health programs is a key first step in
identifying ideal strategies to overcome referral barriers.
Given that physicians reported very positive attitudes
TABLE 2
Characteristics of Respondents by Reasons Physicians Are Less Likely to Refer Patients
to Community-Based Education (CBE) Programs
Respondent
Characteristic All (%)
Inadequate
Availability (%)
Limited Knowledge
and Evidence (%)
Inadequate Resources
and Time to Make
Referrals (%)
Patient level
Barriers to
Participation (%)
Percentage of total
respondents
n = 323 43.03, n = 139 52.94, n = 171 35.30, n = 114 42.10, n = 136
Age in years
≤44 22.02 39.44 46.48 38.02 42.25
45-65 59.42 45.31 55.21 36.98 41.14
66+18.56 39.99 53.33 26.66 45.00
Gender
Male 65.02 42.38 55.24 38.81 43.36
Female 34.98 44.24 48.67 38.06 41.43
Degree
MD 77.69 43.42 51.39 35.06 43.03
DO 22.31 41.67 58.34 36.11 38.89
Medical school
MSU 18.89 52.46 57.37 42.62 55.73
Other 81.11 40.84 51.91 33.59 38.93*
Familiar with programsa
Yes 55.42 55.56 66.67 47.23 56.26
No 44.58 32.96*** 41.89*** 25.70*** 30.72***
Urbanb
Yes 85.76 38.99 50.90 35.38 39.35
No 13.93 66.67*** 64.45 33.34 57.78*
Refer to CBE programsa
Yes 40.87 56.82 66.67 43.95 53.04
No 59.13 34.02*** 44.50*** 30.89** 35.59**
NOTE: Boldface indicates statistical significance using Pearson’s chi-square test of difference between each category that reported the
referral barrier versus who did not report the referral barrier. MSU = Michigan State University.
aSelf-reported. b“Yes,” includes all counties in metropolitan areas.
*p < .05. **p < .01. ***p < .001.
Khan et al. / CONNECTING PRIMARY CARE TO CBE 179
toward CBE programs, awareness and outreach efforts
are likely to be well-received by physicians. In this
study, 45.2% of the survey respondents (n = 146) were
interested in learning more about establishing an easy
and efficient way to refer patients to MSU Extension
for health-related education and provided their e-mail
address so that MSU Extension staff could reach them.
This reinforces findings from other studies that have
shown the effectiveness of physician outreach activi-
ties in increasing referrals into CBE programs (Paige,
Stellefson, & Singh, 2016).
Our results also indicate that the physicians who
had some familiarity with CBE programs were more
likely to report potential barriers that could interfere
with referrals. This confirms that outreach and com-
munication to physicians to increase their awareness of
CBE programs must also include information and
resources to help them overcome barriers to making
referrals. Through such communication, Cooperative
Extension and other CBE providers should share infor-
mation about the evidence base and effectiveness of
programs to increase physician confidence in and refer-
rals to CBE programs. Additionally, CBE staff should
consider how they might lower barriers to referrals by
ensuring a steady availability of programs and meeting
with primary care regularly to increase the familiarity
of existing CBE. Physicians and Cooperative Extension
in other states may be able to work together to address
or assist in removing the patient-level barriers to par-
ticipation in CBE programs identified in this study.
Limitations of this study include the relatively low
response rate, although the number of responses was
sufficient for exploring the topic. Sampling weights
were applied to ensure controlling for sample bias of
TABLE 3
Adjusted Odds Ratios for Reasons Physicians Are Less Likely to Refer Patients
to Community-Based Education (CBE) Programs
Respondent
Characteristic
Inadequate Availability,
OR [95% CI]
Limited Knowledge and
Evidence, OR [95% CI]
Inadequate Resources
and Time to Make
Referrals, OR [95% CI]
Patient-Level Barriers
to Participation,
OR [95% CI]
Age in years
≤44 (Reference) 1.00 1.00 1.00 1.00
45-65 1.354 [0.752, 2.437] 1.270 [0.728, 2.216] 0.930 [0.529, 1.635] 0.960 [0.546, 1.689]
66+1.180 [0.551, 2.531] 1.238 [0.604, 2.537] 0.673 [0.309, 1.463] 0.1254 [0.621, 2.531]
Gender
Male (Reference) 1.00 1.00 1.00 1.00
Female 0.968 [0.580, 0.617] 0.828 [0.501, 0.369] 1.022 [0.614, 1.701] 1.041 [0.625, 1.733]
Degree
MD (Reference) 1.00 1.00 1.00 1.00
DO 0.799 [0.412, 1.551] 1.461 [0.801, 2.667] 1.049 [0.542, 2.033] 0.664 [0.333, 1.324]
Medical school
MSU 1.550 [0.765, 3.141] 0.898 [0.467, 1.726] 1.322 [0.668, 2.616] 2.359**[1.155, 4.821]
Other (Reference) 1.00 1.00 1.00 1.00
Familiar with programsa
Yes 1.852** [1.132, 3.031] 2.047*** [1.259, 3.330] 2.345*** [1.409, 3.902) 2.860*** [1.737, 4.710]
No (Reference) 1.00 1.00 1.00 1.00
Urbanb
Yes (Reference) 1.00 1.00 1.00 1.00
No 2.875** [1.240, 6.666] 1.680 [0.750, 3.766] 0.585 [0.260, 1.317] 1.186 [0.518, 2.712]
Refer to CBE programsa
Yes 2.293*** [1.388, 3.779] 2.151*** [1.311, 3.530] 1.352 [0.819, 2.233] 1.529* [0.930, 2.515]
No (Reference) 1.00 1.00 1.00 1.00
NOTE: Boldface indicates statistical significance. OR = odds ratio; CI = confidence interval.
aSelf-reported. b“Yes,” includes all counties in metropolitan areas.
*p < .05. **p < .01. ***p < .001.
180 HEALTH PROMOTION PRACTICE / March 2020
race, ethnicity, gender, or age. Physicians who elected
to participate may have been more interested in this
topic or viewed CBE programs more favorably, so some
of the attitudinal responses may be biased toward posi-
tive perceptions. Additionally, the survey items regard-
ing barriers and facilitators were somewhat hypothetical
in nature: respondents were asked whether certain
barriers or facilitators “would make them more/less
likely” to refer patients to CBE programs, not whether
they had encountered specific barriers or facilitators in
practice.
>
CONCLUSIONS
Implications of the current study show the need for
health extension models that connect primary care to
CBE. Engaging an established educational infrastructure
such as Cooperative Extension as the program provider
can facilitate practice transformation and contribute to
improved health outcomes for patients. To that end,
Cooperative Extension is trying recruitment strategies
like prescription pad referrals from health care providers
to patients and working on efficient ways to incorporate
outcomes of CBE into patient electronic health records.
Public health professionals can use the results from
this study to propose scalable state level models for
improving physicians’ familiarity with programs and
ultimately increasing referrals to CBE from primary
care. Furthermore, results of this study encourage a
model of health extension that connects the existing
infrastructure of Cooperative Extension to primary care
in community settings. Overall, this study may help
determine scalable state level models for increasing
awareness of chronic disease prevention and other CBE
programs in efforts to improve the health of the nation.
ORCID iDs
Tamkeen Khan https://orcid.org/0000-0002-4522-0811
Cathy Newkirk https://orcid.org/0000-0001-8836-0737
REFERENCES
American Medical Association. (2014). AMA physician charac-
teristics and distribution in the U.S. 2015. Available from https://
commerce.ama-assn.org/store/catalog/productDetail.jsp?product
_id=prod2530012
Bland, J. M., & Altman, D. G. (1995). Multiple significance tests:
The Bonferroni method. British Medical Journal, 310, 170.
Braun, B., Bruns, K., Cronk, L., Kirk Fox, L., Koukel, S., Le
Menestrel, S., . . . Warrant, T. (2014). Cooperative Extension’s
National Framework for Health and Wellness. Retrieved from
http://nifa.usda.gov/national-framework-health-and-wellness
Buttorff, C., Ruder, T., & Bauman, M. (2017). Multiple chronic
conditions in the United States. Santa Monica, CA: RAND
Corporation. Retrieved from https://www.rand.org/pubs/tools
/TL221.html
Centers for Disease Control and Prevention. (2015). 1-2-3
Approach to provider outreach marketing chronic disease inter-
vention to primary care practices. Retrieved from https://www
.cdc.gov/arthritis/interventions/marketing-support/1-2-3
-approach/docs/pdf/arthritis-marketing-guide_introduction.pdf
Dwyer, J., Contreras, D., Eschbach, C., Tiret, H., Newkirk, C.,
Carter, E., & Cronk, L. (2017). Cooperative extension as a frame-
work for health extension: The Michigan State University model.
Academic Medicine, 92, 1416-1420.
Fenton, M. (2011). Health care’s blind side: The overlooked con-
nection between social needs and good health. Princeton, NJ:
Robert Wood Johnson Foundation. Retrieved from http://health
-equity.lib.umd.edu/3643/
Grumbach, K., & Mold, J. W. (2009). A health care cooperative
extension service: Transforming primary care and community
health. JAMA Journal of American Medical Association, 301,
2589-2591.
Izrael, D., Hoaglin, D., & Battaglia, M. (2000). A SAS macro for
balancing a weighted sample. Proceedings of the Twenty-Fifth
Annual SAS Users Group International Conference. Retrieved
from http://www2.sas.com/proceedings/sugi25/25/st/25p258.pdf
National Center for Health Statistics. (2016). Health, United
States, 2015: With special feature on racial and ethnic health
disparities. Hyattsville, MD: Author. Retrieved from https://www
.ncbi.nlm.nih.gov/books/NBK367640/
Ono, S. S., Crabtree, B. F., Hemler, J. R., Balasubramanian, B. A.,
Edwards, S. T., Green, L. A., . . . Cohen, D. J. (2018). Taking inno-
vation to scale in primary care practices: The functions of health
care extension. Health Affairs, 37, 222-230.
Paige, S. R., Stellefson, M., & Singh, B. (2016). Patient perspec-
tives on factors associated with enrollment and retention in
chronic disease self-management programs: A systematic review.
Patients Intelligence, 1, 935-943.
Ritsema, T., Bingenheimer, J., Scholting, P., & Cawley, J. (2014).
Differences in the delivery of health education to patients with
chronic disease by provider type, 2005–2009. Preventing Chronic
Disease, 11, 130-175. doi:10.5888/pcd11.130175
Torres, J. (2015). Community connections: Linking primary care
patients to local resources for better management of obesity.
Health Promotion Practice, 16, 313-315.
U.S. Department of Agriculture. (2013). Rural-urban continuum
codes. Retrieved from https://www.ers.usda.gov/data-products
/rural-urban-continuum-codes/documentation
