An Electronic Linkage System for Health
Effect on Delivery of the 5A’s
Alex H. Krist, MD, MPH, Steven H. Woolf, MD, MPH, Charles O. Frazier, MD, Robert E. Johnson, PhD,
Stephen F. Rothemich, MD, MS, Diane B. Wilson, EdD, RD, Kelly J. Devers, PhD, J. William Kerns, MD
A variety of factors limit the ability of clinicians to offer intensive counseling to patients
with unhealthy behaviors, and few patients (2%–5%) are referred to the community
counseling resources that do offer such assistance. A system that could increase referrals
through an efficient collaborative partnership between community programs and clini-
cians could have major public health implications; such was the subject of this feasibility
At nine primary care practices, an electronic linkage system (eLinkS) was instituted to
promote health behavior counseling and to automate patient referrals to community
counseling services. Patients were offered 9 months of free counseling for weight loss,
smoking cessation, and problem drinking at a choice of venues: group counseling,
telephone counseling, computer care, and usual care. The delivery of behavioral counsel-
ing, measured by the 5A’s (ask, address, advise, assess, agree, arrange) and patients’
reported experiences with eLinkS, was examined.
For 5 weeks eLinkS was used, until high referral volumes depleted counseling funds. Of the
5679 patients visiting the practices, 71% had an unhealthy behavior. Of these patients, 10%
were referred for intensive counseling from a community program, most often for weight
loss. Counseling and referrals occurred regardless of visit type—wellness, acute, or chronic
care. eLinkS was used more often for middle-aged adults and women and by more-
The intervention increased the rate at which patients were referred for intensive behavioral
counseling compared to current practice norms. Given the evidence that intensive
counseling is more effective in promoting behavior change, implementing eLinkS could
have substantial public health benefits.
(Am J Prev Med 2008;35(5S):S350–S358) © 2008 American Journal of Preventive Medicine
the U.S.1Addressing these behaviors could help stem
the rising prevalence and cost of chronic diseases,2–4
and clinicians can play a pivotal role. Americans have
healthy diet, physical inactivity, and risky alcohol
use—account for approximately 37% of deaths in
regular contact with clinicians and value their advice. A
clinician’s recommendation to change behavior is
widely cited as a motivating factor.5,6Guidelines recom-
mend that clinicians use the 5A’s (assess [A1]; advise
[A2]; agree [A3]; assist [A4]; and arrange [A5]) to
promote healthy behaviors (Table 1).6–11
Following such guidelines is difficult for clinicians due
to inadequate time, staff, reimbursement, and familiarity
with counseling techniques.12Few patients report being
asked regularly by clinicians whether they engage in
unhealthy behaviors.13Intensive counseling, often a ne-
cessity to help patients adopt and maintain healthy behav-
iors, is rarely possible in primary care settings; the per-
centage of patients who receive intensive counseling is
probably less than 2%–5%.14,15Increasing this percentage,
even modestly, could have substantial public health implica-
tions, given the population-attributable risk of unhealthy
behaviors and the benefits of intensive counseling.9,16–18
From the Department of Family Medicine (Krist, Kerns, Frazier), the
Departments of Family Medicine, Epidemiology and Community
Health (Woolf), the Departments of Biostatistics and Family Medi-
cine (Johnson), the Departments of Internal Medicine and Family
Medicine, Massey Cancer Center, (Wilson), and the Departments of
Health Administration and Family Medicine (Devers), Virginia Com-
monwealth University, Richmond, Virginia; the Fairfax Family Prac-
tice Residency (Krist), Fairfax; the Riverside Family Practice Resi-
dency (Frazier), Newport News; and the Shenandoah Family Practice
Residency (Kerns), Front Royal, Virginia
Address correspondence and reprint requests to: Alex H. Krist,
MD, MPH, P.O. Box 980251, Richmond VA 23298-0251. E-mail:
S350 Am J Prev Med 2008;35(5S)
© 2008 American Journal of Preventive Medicine • Published by Elsevier Inc.
0749-3797/08/$–see front matter
Few practices can undertake redesign efforts to system-
atically offer high-quality behavioral counseling.19,20Ex-
ceptional practices and health systems have the infrastruc-
ture support, multidisciplinary team members, and staff
to follow up with patients and motivate health behavior
change,12,21but these conditions are atypical.22Ironically,
to programs and services that do offer this level of
counseling. Such programs commonly report that local
physicians refer few patients.
This study tested the feasibility of an electronic
linkage system (eLinkS) to help connect these entities
to support behavioral counseling. Utilizing the elec-
tronic medical record (EMR) as a platform, eLinkS was
designed to (1) help clinicians systematically perform
elements of the 5A’s that are feasible in busy practice
settings (i.e., asking about health behaviors, offering
brief advice, and agreeing on next steps); (2) make it
fast and easy to refer patients to intensive counseling
outside the office; and (3) establish bidirectional com-
munication between practices and community counsel-
ors.23,24This article examines the effects of eLinkS on
Nine primary care practices in the Tidewater region of
Virginia were recruited.25The practices, members of a single
medical group and of the Virginia Ambulatory Care Out-
comes Research Network (ACORN), share a common type of
EMR (GE Centricity Physician Office©) that is managed by a
central informatics staff. The practices have used the EMR for
3 to 10 years. Practice size ranged from 1 to 30 clinicians
(median?3), and 48 (87%) clinicians participated in the
study. Two sites were solo practices, five had three clinicians,
one had eight clinicians, and one (a family medicine resi-
dency program) had 30 part-time clinicians and residents.
The design of eLinkS reflects early input from clinicians at
the study sites, solicited through site visits. The final version
functioned as follows:
Practices could distribute a flyer in the waiting room that
informed patients about the following four referral options,
which were available for free for up to 9 months:
● Group counseling was offered in community locations: for
weight loss by Weight Watchers; for tobacco use, by the
local hospital’s wellness center26; and for risky alcohol use,
by Alcoholics Anonymous (AA).
● Telephone counseling, offered for smoking cessation and
weight loss, was delivered by trained counselors at the
University of Kentucky Health & Wellness Program Behav-
ioral Health Improvement Program (BeHIP).27
● Computer care provided patients access both to an infor-
mative website developed previously by ACORN28–30and
to an e-counseling service that ACORN and BeHIP de-
signed for this project. The e-counseling program followed
the BeHIP telephone counseling protocol but was used
● Usual care included options other than the above (e.g.,
physician counseling, pharmacotherapy, or even inaction).
When staff who take patients to the examination rooms
(“rooming” staff) obtained patients’ vital signs and entered
them into the medical record, the EMR displayed a screen
prompt to remind the staff to enter height, weight, and
tobacco-use status (A1). If a patient was overweight or obese
(BMI ?25 kg/m2); smoked; or had an EMR entry of risky
alcohol use, a prompt appeared when the clinician opened
the patient’s record. The prompt asked if the clinician wanted
to address the flagged behavior(s) at that visit. If the clinician
clicked yes—a step that was classified as Address (Ad)—the
EMR would open a form (Figure 1). The form included
checkboxes wherein clinicians could document whether they
gave brief advice to change behavior (A2); whether the
patient was ready to improve and engage in the process (A3);
whether the patient wanted to be referred outside the prac-
tice to one of the intensive-counseling options noted above
(A4); and by what means (telephone, appointment, or e-
mail) the patient sought follow-up by the practice (A5).
To allow the clinician to view all of the patient’s unhealthy
behaviors, automatic prompts would not display until the
rooming staff entered both weight and tobacco-use status. A
current or past height measurement also was required for
BMI calculation. At any point in any encounter, clinicians
could also load the form manually.
The screen displays and the EMR programming were
designed to make the interface with clinicians easy and fast, to
automate the referral process electronically, and to facilitate
proactive counseling. When telephone or group counseling
was selected through eLinkS, the EMR automatically e-mailed
contact information to the intensive-counseling program
staff, who then would contact the patient (rather than having
the patient call the intensive-counseling program [reactive
counseling]). When clinicians selected computer care, the EMR
forwarded an e-mail to the patient with a link to the educa-
tional website and instructions for e-counseling. AA referrals
were reactive; the EMR would print a list of AA meetings for
clinicians to hand to patients.
Training sessions for eLinkS of 1-hour duration were held at
all practices prior to launch. The intervention went live on
Table 1. The 5A’s applied to health behaviors8
Assess (A1)Assess health behaviors and factors affecting
Advise (A2)Provide clear, specific, personalized advice,
including harms and benefits associated
Agree (A3)Select treatment goals and approach based
on patient’s needs
Assist (A4) Aid patient in achieving agreed-upon goals
through self-help, counseling, and
adjunctive medical treatments as
Arrange (A5)Arrange follow-up contacts to provide
November 2008Am J Prev Med 2008;35(5S)
April 16, 2006, and was discontinued 5 weeks later (May 22,
2006) when an overwhelming surge in referrals for intensive
counseling exceeded available funds. Patients referred to
intensive counseling during the exposure period were
eligible to receive services for up to 9 months (through
February 2007). Weekly feedback reports notified partici-
pating clinicians of the number of counseling referrals they
The delivery of the 5A’s (this report’s main outcome vari-
ables) was measured by clinicians’ entries in the EMR dia-
logue box. EMR data were collected for all adults (aged ?18
years) visiting the practice from 2 years prior to the encounter
to 1 year afterward. Dependent variables were either patient-
specific (age, gender, and ICD-9 codes for comorbid condi-
tions) or encounter-specific (weight, tobacco-use status, doc-
umentation of any alcohol disorder, patient’s stated reason
for visit, and Current Procedural Terminology [CPT] codes).
The reason for the visit was classified as acute, wellness, or
selected chronic conditions, based on the criteria in Table 2.
Postal surveys were mailed 2
weeks after the encounter to all
counselees (patients referred to
an intensive-counseling option
or receiving usual care), and
asked respondents to describe
the encounter. An established
mailing protocol (modified
Dillman technique31,32) was
used to optimize response ra-
tes. Surveys completed by office
managers provided informa-
tion about clinicians, practices,
and whether only clinicians
or clinician–nurse teams used
Statistical calculations were per-
formed in SAS version 9.1.3. Dif-
ferences in percentages were
tested for significance using
Fisher’s exact test. A logistic re-
of the 5A’s. The initial regressor
variables were those that could
theoretically influence eLinkS
usage: the patient’s age, gender,
health behaviors, and comor-
bidities; encounter type and
complexity; and the physician’s
age, gender, training, years in
practice, and practice site. Vari-
ables were retained in the
model in a stepwise method that
used p?0.25 as the entry crite-
rion and p?0.10 as the reten-
tion criterion. The patient was
the unit of analysis. Practices
were selected purposively, and
fixed-practice effects were con-
sidered. This study was approved by the IRBs of Virginia
Commonwealth University and Riverside Health System.
A total of 5679 adult patients visited the practices during
the 5-week exposure period. Their ages (median?53
years); gender (64% female); and frequency of preven-
tion visits (14%) were typical of adult primary care popu-
lations (Table 2).33The frequency of chronic-care visits at
other places was lower than published norms (9% vs
44%)33because only visits for selected reasons met this
study’s definition of relevant chronic illnesses. The prev-
alence of circulatory diseases was similar to the general
population’s (34% vs 36%); the prevalence of neoplasms
(9% vs 7%) was slightly higher, as was the prevalence of
diabetes (15% vs 7%).34The prevalence of multiple
unhealthy behaviors resembled national norms.35Rates
Figure 1. The electronic linkage system (eLinkS) dialogue box (form). The eLinkS form
would appear after the clinician was prompted that the patient had an unhealthy behavior and
clicked yes to an invitation to open the form. In the example portrayed in the figure, the
clinician chose to address the patient’s obesity. The form provides the current BMI and
response options to document the advice given to the patient, the patient’s choice of
intensive-counseling options, and the arrangements for follow-up. In this example, the
clinician has indicated that the patient was advised to modify diet and/or increase physical
activity, was ready to change behavior and engage in next steps, and desired computer care for
intensive counseling and follow-up by telephone. View Patient Counseling Script is a button that
clinicians clicked to review potential scripts for brief advice. By clicking Add translation to note,
the clinician could transfer the response options to the encounter note to document health
behavior counseling in the patient’s formal medical record. The response options clicked by
the clinicians participating in this study were captured for analysis in this study. See text for
further details about group counseling, computer care, telephone counseling, and usual care.
S352 American Journal of Preventive Medicine, Volume 35, Number 5Swww.ajpm-online.net
for overweight/obesity, tobacco use, and risky alcohol use
were consistent with local norms.36–38
The postal survey was mailed to 583 counselees and was
returned by 358 (61%). The age, gender, weight, and
comorbid conditions of survey respondents did not differ
significantly from those of nonrespondents. Participating
clinicians were not representative of primary care clini-
cians generally, because a large percentage (44%) were
family medicine residents, a group that is younger and
comprises more women than nonresidents.39
Delivery of the 5A’s
The delivery of the 5A’s is depicted in Figure 2.
Assess (A1). Rooming staff collected both BMI and
tobacco-use information for 2117 patients, 37% of the
5679 patients who visited the practices (Figure 2). Data
were missing more often for tobacco-use status than for
BMI. When BMI data were incomplete, the missing
information was more often height than weight (97% vs
Address (Ad). eLinkS prompts appeared for 1860 pa-
tients, 1801 occurring automatically and 59 initiated by
the clinician. When confronted with these prompts,
clinicians elected to open the dialogue box form (ad-
dress the behavior) for 912 (49%) patients.
Advise (A2). Clinicians advised 537 patients to change
their health behavior, 13% of patients with an un-
healthy behavior (n?4030) and 29% of patients for
whom a prompt appeared (n?1860; Figure 2). A2 was
documented for 20% of smokers, 7% of overweight
patients, 17% of obese patients, and 13% of problem
Respondents to the postal survey reported higher
advice rates. Among those referred for weight loss, 75%
reported receiving advice on this topic, and more than
half reported clinician advice about diet or exercise.
Among those referred for smoking cessation, 97%
recalled clinician advice about smoking. Fully 86% of
counselees considered it appropriate for the clinician
to discuss health behaviors, and 54% were unsure
whether they would have mentioned the topic if their
clinician had not done so.
Agree (A3). Clinicians recorded that 461 patients were
ready to improve the targeted behavior, and 441 were
engaged in what to do next. This represents 86% and
82%, respectively, of patients who received clinician
Assist (A4). Fully 407 patients were referred for inten-
sive counseling, 10% of patients with an unhealthy
behavior or 76% of patients who received A2 (Figure
2). The population that received A4 included 12% of
obese patients, 3% of overweight patients, and 8% of
smokers; no risky drinkers were referred for intensive
counseling. Two thirds of patients who sought intensive
Table 2. Patient and clinician demographics (n/% unless
Median age, years (range)
Gender (% women)
Reason for visit
Selected chronic illnessesc
Neoplasm (any type)d
Behavioral risk factorse
BMI ?25–29 kg/m2
BMI ?30 kg/m2
Total number of risk factors
Median age, years (range)
Gender (% women)
Median years in practice (range)
Average number of full-time
equivalent clinicians in the office
Complete use of eLinkS
aBased on electronic medical record (EMR) data
bAs determined from surveys administered to office managers and
participating clinicians (response rate?100%)
cReason for visit classification: Acute illness/injury included any self-
limited condition (lasting days or weeks). Selected chronic illnesses
included chronic disorders for which counseling regarding the three
target risk factors might be particularly relevant, including chronic
cardiovascular disease, respiratory disease, or diabetes. Wellness/
prevention included visits focused on wellness or prevention, but it
excluded visits focused on specific wellness screening tests. Other
included visits not encompassed by the above categories and those for
which the focus could not be determined accurately.
dComorbidities were defined by examining the ICD-9 codes in a
patient’s EMR database for the index and all prior visits. Patients were
classified as having circulatory disease, diabetes, neoplasm, or respi-
ratory disease if their records included an ICD-9 code of 390–459,
250, 140–239, or 460–519, respectively.
eThe number of behavioral risk factors refers to how many of the
three target risk factors (tobacco use, overweight/obesity, risky
alcohol use) were documented in the EMR database.
fIn clinician only practices, the only involvement of rooming staff with
the electronic linkage system (eLinkS) was to collect and record vital
signs; the prompts and forms appeared only for the clinician. In team
approach practices, both rooming staff and clinicians could access and
use the forms.
November 2008Am J Prev Med 2008;35(5S)
counseling chose group counseling as the venue, pri-
marily Weight Watchers. Usual care was requested by
Only 64% of postal survey respondents, including the
usual-care group, recalled being invited by the clinician
to try an intensive-counseling option. Higher percent-
ages reported that the clinician was helpful and pro-
vided enough information to make a choice (80% and
87%, respectively). One third cited the clinician’s rec-
ommendation over other factors (e.g., convenience) as
the primary consideration in selecting an A4 option.
When surveyed ?2 weeks after the visit, 65% of referred
patients indicated that they were still planning to
pursue intensive counseling.
Arrange (A5). Arrangements for follow-up with the
practice were made for 306 patients. Office visits, or
nurses’ telephone calls to monitor progress, were the
preferred arrangements (Figure 2).
Predictors of Use
While all of the practices
used eLinkS, significant vari-
ation was observed by prac-
tice and by individual patient
and clinician characteristics.
For example, the crude rate
for A2 was greater for female
patients than for male pa-
tients (33% vs 23%, respec-
tively). Even after adjustment
(Table 3), ORs for delivering
most of the 5A’s were gener-
ally greater when patients
had unhealthy behaviors and
were either female or middle-
aged (aged 45–60 years)
and when clinicians had had
longer tenures in practice.
Most of the 5A’s occurred
less frequently at acute-care
visits, but patients were re-
ferred to intensive counsel-
ing at all types of visits: 23%
at acute-care visits, 26% at
chronic-care visits, and 34%
at wellness visits. Unexpect-
edly, the presence of com-
orbid diseases that benefit
from behavioral counseling
(e.g., diabetes) was not pre-
dictive of eLinkS use.
This study was designed to
observe whether clinicians
would use eLinkS, what op-
tions the patients would self-select, and what effect the
intervention would have on counseling practices and
patients’ behaviors. The results are encouraging. The
prompts appeared at more than one third of the
encounters (due to the prevalence of overweight/
obesity). The use of eLinks was steady throughout its 5
weeks of availability, and occurred at all manner of
office visits, not just those devoted to prevention. In
semistructured interviews conducted for a qualitative
study, clinicians and office managers commented on
the usefulness and feasibility of the system, and patients
welcomed the selection and quality of the counseling
Public Health Implications
Evaluating eLinkS’ effect on health behaviors requires
a randomized trial, but this report’s data seem encour-
Figure 2. Exposure to eLinkS and delivery of the 5A’s to the study population. The above
figure depicts the delivery of the 5A’s (ask [A1]; address [Ad]; advise [A2]; assess [A3]; agree
[A4]; and arrange [A5]) for health behavior counseling as recorded by eLinkS for all adult
patients seen in the study practices during the exposure period (April 16, 2006–May 22,
2006).aThis value (4030) represents the number of patients who had (1) either BMI or
tobacco use status assessed (n?4923) and (2) abnormal results from this assessment. A
denominator of 4030, or 70% of patients seen (N?5679), is used in the text when reporting
the number of patients with an unhealthy behavior.
(n?1801) only when (1) both the BMI and tobacco-use status were assessed and (2) the
patient had evidence of an abnormality. Clinicians could also manually load the form, if BMI
and/or tobacco use status were not assessed, and this occurred for 59 patients. A denominator
of 1860 is used in the text when reporting the percentage of patients who were identified by
prompts and received counseling.cSee text for description of counseling options.
bPrompts appeared automatically
S354 American Journal of Preventive Medicine, Volume 35, Number 5Swww.ajpm-online.net
aging on several grounds. Brief clinician advice (A2) is
itself effective in promoting smoking cessation and,
potentially, weight loss,4,9and clinicians reported giv-
ing such advice at one third of the encounters in which
the eLinkS prompt appeared. Patients referred for
intensive counseling reported that the topic, if not
initiated by the clinician, might not have come up.
According to Health Plan Employer Data and Informa-
tion Set (HEDIS) data,4074% of patients who smoke
recall receiving a clinician’s advice to quit smoking in
the past year, a larger percentage than this study
reports. However, clinician counseling was examined
for only 5 weeks, rather than 1 year, and A2 was
monitored as documented in the patient’s record. Brief
advice by clinicians is underreported in medical records
compared to patient recall.41
Intensive counseling is substantially more effective than
A2,9,16–18and eLinkS referred patients for up to 9 months
of intensive counseling. As a subsequent article will de-
scribe, patients referred to these programs reported sig-
nificant improvements in health behaviors. Controlled
trials document the effectiveness of proactive telephone
and group counseling.42–44Group weight-loss programs,
the most commonly selected intensive-counseling option
in this study, report reductions of 3 to 6 kg over 6 to 12
The percentage of patients with unhealthy behaviors
who received intensive counseling through eLinkS
(10%) appears to exceed practice norms. If approxi-
mately 62% of patients are overweight/obese,46a clini-
cian who sees 30 patients daily would have to refer 2
patients for intensive weight-loss counseling to achieve
a comparable referral rate as observed with using
eLinkS. Data on the existing rate at which clinicians
refer patients for intensive counseling are sparse.47
Clinicians refer approximately 1%–5% of smokers to
quitlines,15,48,49percentages lower than the referral
rate observed in this study (8%). Some studies report
that A4 is delivered to 13% to 43% of smokers.40,50–54
However, what constitutes A4 in such studies is variable,
and may not equate with intensive counseling as de-
fined here. For example, HEDIS reports a high rate for
A4, but any discussion either of medications or strate-
gies to quit can qualify.40In contrast, patients receiving
A4 through eLinkS participated over 9 months in as
many as 70 sessions of up to 120 minutes each. Pub-
lished A4 rates for weight loss appear significantly lower
than for smoking cessation, and equally brief.55–59The
Table 3. Regression results—predictors for delivery of the 5A’s
AOR (95% CI) for delivery of the 5A’s
Ask (A1) Address (Ad)Advise (A2)Agree (A3)Assist (A4)Arrange (A5)
18–29 (vs ?75)
30–44 (vs ?75)
45–59 (vs ?75)
60–75 (vs ?75)
Gender, female (vs male)
Overweight (BMI ?25–29
Obese (BMI ?30 kg/m2)
Acute visit (vs wellness)
Chronic visit (vs wellness)
Moderate visit CPT (vs
Complex visit CPT (vs simple)
Wellness visit CPT
Gender, female (vs male)
Years in practice (1 year)
Trained clinician (vs resident)
Note: Data fields with no values (—) represent variables not retained in the logistic regression model. Bolded numbers represent significant
differences. See Table 2 for definitions of cancer comorbidity, respiratory comorbidity, and diabetes comorbidity. Simple visits, moderate visits,
complex visits, and wellness visits were those receiving CPT codes of 99212/99201, 99213/99202, 99214–99215/99203–99205, 99239x/99238x,
aEach age group compared to the ?75 age group
CPT, Current Procedural Terminology
November 2008Am J Prev Med 2008;35(5S)
more-extensive counseling offered in this study would
be expected to produce substantially greater effects on
behavior than seems currently possible in ordinary
Caveats and Confounding Variables
Several elements of this intervention may have en-
hanced the referral rate:
● Intensive counseling was free to patients, eliminat-
ing cost as a barrier.
● Participating practices were experienced with EMRs.
● The intervention employed active prompts; EMR
vendors prefer passive prompts, which are less effec-
tive but also interfere less with workflow.60–62
Conversely, several factors may have attenuated the
● Displaying prompts only after rooming staff had en-
tered both BMI and tobacco-use data restricted the
number of patients receiving the full intervention.
● eLinkS was operational for only 5 weeks. The early
shutdown limited patients’ exposure to the interven-
tion over multiple visits, and gave little time for
clinicians to acclimate to eLinkS and community
● Participating clinicians were younger and more in-
experienced than practicing physicians, an artifact
of including the residency program.
● eLinkS was used more for middle-aged patients and
women, perhaps because the available services (e.g.,
Weight Watchers) appealed to this group.
● Problem drinkers received little benefit from eLinkS.
By protocol, rooming staff assessed BMI and tobacco-
use status but not alcohol use. Fewer intensive-
counseling options were available for risky alcohol
use, and referrals were reactive rather than proactive.
Other study limitations include the nonrandomized,
pre–post design. While this design was appropriate for
an initial evaluation of feasibility, comparisons among
groups are subject to biases and confounding. Second,
the study involved only nine practices, and the fidelity
of the intervention varied among sites. Third, EMR data
were used to document the 5A’s; clicking an onscreen
response option does not clarify what actually occurred
during the encounter. Finally, the external validity of
the counselees’ survey responses is limited because they
represent a subset of patients.
Collaboration between clinicians and community re-
sources, as occurred here, presents a win–win scenario
for patients, clinicians, and community programs. Pa-
tients obtain more-intensive assistance. Clinicians, who
frequently cannot provide intensive counseling them-
selves,63welcome an easy means to connect patients
with the help they need. Community programs, which
often struggle to attract clients through media and
advertising, appreciate the influx of referrals from the
This process has implications that extend beyond
behavioral counseling. A system like eLinkS that facili-
tates systematic screening and referrals could, with
some modification, help clinicians arrange screening
tests and chronic disease management—all with the
click of a mouse. Clinician–community collaboration
long has been advocated for these purposes,64but
applying modern technology to make collaboration
easy holds considerable promise.
This work was funded under grant #53769 from the Robert
Wood Johnson Foundation and the Agency for Healthcare
Research and Quality under the Prescription for Health
initiative. The authors thank the practices in Riverside Med-
ical Group that participated in this study: Bruton Avenue
Family Medicine, Eagle Harbor Primary Care, Elizabeth Lakes
Family Practice, Hilton Family Practice, Mathews Medical
Center, Mercury West Medical Center, Patriot Primary Care,
Riverside Family Medicine, and Williamsburg Medical Arts
Family Practice. They also thank the staff of Weight Watchers
(Stephanie Schoemer, MS, RD); Riverside Wellness Center
(Holly Hicks); and BeHIP (Tammy Akins), for their assis-
tance both in arranging intensive-counseling services for the
study subjects and in data collection. They thank Sharon
Flores, MS, for coordinating the project; Amy Burgett, RN, for
interfacing with the practices; Steven Mitchell for database
management; Tina Cunningham, MS, for statistical analysis;
and Kristen Mertz for the EMR programming that underlies
eLinkS. They received invaluable advice and assistance on the
design and conduct of the study and on draft manuscripts
from Carole Hale, Anton Kuzel, MD, MPHE, Daniel Longo,
ScD, David Marsland, MD, and Paul Mazmanian, PhD, and
from the following expert consultants: Richard Botehlo, MD,
Russell Glasgow, PhD, John Hickner, MD, MSc, Abby King,
PhD, Jesse Crosson, PhD, and Atif Zafar, MD. They thank
Michael Fleming, MD, MPH, for his attempts to identify
intensive-counseling options for problem drinkers. Finally,
the study received extensive support from the staff of the
Prescription for Health National Program Office (led by
Larry A. Green, MD, and Maribel Cifuentes, RN, BSN); the
Prescription for Health evaluation team at the University of
Medicine and Dentistry of New Jersey (led by Benjamin
Crabtree, PhD, MA, and Deborah Cohen, PhD); the National
Advisory Committee for Prescription for Health; and the
project officers (Sue Hasmiller, PhD, RN, and Laura Leviton,
PhD) of the Robert Wood Johnson Foundation.
No financial disclosures were reported by the authors of
1. Mokdad AH, Marks JS, Stroup DF, Gerberding JL. Actual causes of death in
the U.S., 2000. JAMA 2004;291:1238–45.
2. Agency for Healthcare Research and Quality. Counseling to prevent
tobacco use and tobacco-caused disease: recommendation statement. U.S.
S356 American Journal of Preventive Medicine, Volume 35, Number 5Swww.ajpm-online.net
Preventive Services Task Force. AHRQ Pub. No. 04-0526. Rockville MD:
Agency for Healthcare Research and Quality, 2003.
3. Eden KB, Orleans CT, Mulrow CD, Pender NJ, Teutsch SM. Does counsel-
ing by clinicians improve physical activity? A summary of the evidence for
the U.S. Preventive Services Task Force. Ann Intern Med 2002;137:208–15.
4. McTigue KM, Harris R, Hemphill B, et al. Screening and interventions for
obesity in adults: summary of the evidence for the U.S. Preventive Services
Task Force. Ann Intern Med 2003;139:933–49.
5. Starfield B, Shi L, Macinko J. Contribution of primary care to health
systems and health. Milbank Q 2005;83:457–502.
6. Goldstein MG, Whitlock EP, DePue J. Multiple behavioral risk factor
interventions in primary care. Summary of research evidence. Am J Prev
7. Whitlock EP, Orleans CT, Pender N, Allan J. Evaluating primary care
behavioral counseling interventions: an evidence-based approach. Am J
Prev Med 2002;22:267–84.
8. Ranney L, Melvin C, Lux L, McClain E, Morgan L, Lohr K. Tobacco use:
prevention, cessation, and control. Evidence Report/Technology Assess-
ment No. 140. Rockville MD: Prepared by the RTI International–University
of North Carolina evidence-based practice center under Contract No.
9. Fiore MC, Bailey WC, Cohen SJ. Treating tobacco use and dependence.
Clinical practice guideline. Rockville MD: USDHHS, Public Health Service,
10. Ockene JK, Adams A, Hurley TG, Wheeler EV, Hebert JR. Brief physician-
and nurse practitioner-delivered counseling for high-risk drinkers: does it
work? Arch Intern Med 1999;159:2198–205.
11. Goldstein MG, DePue J, Kazura A, Niaura R. Models for provider-patient
interaction: applications to health behavior change. In: Shumaker SA,
Schron EB, Ockene JK, eds. The handbook of health behavior change. New
York: Springer, 1998.
12. Rosal MC, Ockene JK, Luckmann R, et al. Coronary heart disease multiple
risk factor reduction: providers’ perspectives. Am J Prev Med 2004;27(2S):
13. Coups EJ, Gaba A, Orleans CT. Physician screening for multiple behavioral
health risk factors. Am J Prev Med 2004;27(2S):34–41.
14. Rothemich SF, Woolf SH, Johnson RE, et al. Effect on cessation counseling
of documenting smoking status as a routine vital sign: an ACORN study.
Ann Fam Med 2008;6:60–8.
15. Cummins SE, Bailey L, Campbell S, Koon-Kirby C, Zhu SH. Tobacco
cessation quitlines in North America: a descriptive study. Tob Control
16. Heshka S, Anderson JW, Atkinson RL, et al. Weight loss with self-help
compared with a structured commercial program: a randomized trial.
17. Truby H, Baic S, deLooy A, et al. Randomised controlled trial of four
commercial weight loss programmes in the UK: initial findings from the
BBC “diet trials”. BMJ 2006;332:1309–14.
18. Dansinger ML, Gleason JA, Griffith JL, Selker HP, Schaefer EJ. Comparison
of the Atkins, Ornish, Weight Watchers, and Zone diets for weight loss and
heart disease risk reduction: a randomized trial. JAMA 2005;293:43–53.
19. Solberg LI, Brekke ML, Fazio CJ, et al. Lessons from experienced guideline
implementers: attend to many factors and use multiple strategies. Jt Comm
J Qual Improv 2000;26:171–88.
20. Solberg LI, Asche SE, Boyle RG, Boucher JL, Pronk NP. Frequency of
physician-directed assistance for smoking cessation in patients receiving
cessation medications. Arch Intern Med 2005;165:656–60.
21. Bodenheimer T, Wagner EH, Grumbach K. Improving primary care for
patients with chronic illness. JAMA 2002;288:1775–9.
22. Crabtree BF, Miller WL, Aita VA, Flocke SA, Stange KC. Primary care
practice organization and preventive services delivery: a qualitative analysis.
J Fam Pract 1998;46:403–9.
23. Woolf SH, Glasgow RE, Krist A, et al. Putting it together: finding success in
behavior change through integration of services. Ann Fam Med 2005;3(2S):
24. Ockene JK, Edgerton EA, Teutsch SM, et al. Integrating evidence-based
clinical and community strategies to improve health. Am J Prev Med
25. Riverside Health System. Riverside online: Riverside medical group. www.
26. Riverside Health System. Riverside online: Riverside wellness and fitness
27. University of Kentucky Health & Wellness Program. BeHIP: overview.
28. Virginia Commonwealth University Department of Family Medicine. My
healthy living. www.myhealthyliving.net.
29. Krist AH, Woolf SH, Rothemich SF, Johnson RE, Wilson DB. It takes a
partnership: the value of collaboration in developing and promoting a Web
site for primary care patients. Ann Fam Med 2005;3(2S):S47–9.
30. Woolf SH, Krist AH, Johnson RE, Wilson DB, Rothemich SF, Norman GJ.
A practice-sponsored website to help patients pursue healthy behaviors: an
ACORN study. Proceedings of the 33rd Annual Meeting of the North
American Primary Care Research Group; 2005 Oct 15–18; Quebec City,
31. Dillman D. Mail and Internet surveys: the total design method. 2nd ed.
Hoboken NJ: John Wiley and Sons, Inc., 1999.
32. Edwards P, Roberts I, Clarke M, et al. Increasing response rates to postal
questionnaires: systematic review. BMJ 2002;324:1183.
33. Cherry DK, Woodwell DA, Rechtsteiner EA. National ambulatory medical
care survey: 2005 summary. Adv Data 2007:1–39.
34. Pleis JR, Lethbridge-Cejku M. Summary health statistics for U.S. adults:
National Health Interview Survey, 2005. Vital Health Stat 10 2006:1–153.
35. Fine LJ, Philogene GS, Gramling R, Coups EJ, Sinha S. Prevalence of
multiple chronic disease risk factors. 2001 National Health Interview
Survey. Am J Prev Med 2004;27(2S):18–24.
36. Behavioral Risk Factor Surveillance System. SMART: selected metropoli-
tan/micropolitam area risk trends. 2006–Virginia Beach–Norfolk–Newport
News, VA–NC metropolitan statistical area. Overweight and obesity (BMI):
Weight classification by body mass index (BMI). apps.nccd.cdc.gov/
37. Behavioral Risk Factor Surveillance System. SMART: selected metropoli-
tan/micropolitan area risk trends. 2006–Virginia Beach–Norfolk–Newport
News, VA–NC metropolitan statistical area. Tobacco use: adults who are
current smokers. apps.nccd.cdc.gov/brfss-smart/MMSARiskChart.asp?yr?
38. Behavioral Risk Factor Surveillance System. SMART: selected metropoli-
tan/micropolitan area risk trends. 2006–Virginia Beach-Norfolk-Newport
News, VA–NC metropolitan statistical area. Alcohol consumption: heavy
drinkers (adult men having more than two drinks per day and adult women
having more than one drink per day). apps.nccd.cdc.gov/brfss-smart/
39. American Academy of Family Physicians. Table 3. Percentage of AAFP
members by selected characteristics, January 1, 2007. www.aafp.org/online/
40. National Committee for Quality Assurance. The state of health care quality
41. Stange KC, Zyzanski SJ, Smith TF, et al. How valid are medical records and
patient questionnaires for physician profiling and health services research?
A comparison with direct observation of patients visits. Med Care
42. Stead LF, Perera R, Lancaster T. Telephone counselling for smoking
cessation. Cochrane Database Syst Rev 2006;3:CD002850.
43. Stead LF, Lancaster T. Group behaviour therapy programmes for smoking
cessation. Cochrane Database Syst Rev 2005;2:CD001007.
44. McAfee TA. Quitlines a tool for research and dissemination of evidence-
based cessation practices. Am J Prev Med 2007;33(6S):S357–67.
45. Rock CL, Pakiz B, Flatt SW, Quintana EL. Randomized trial of a multifac-
eted commercial weight loss program. Obesity (Silver Spring) 2007;15:
46. Behavioral Risk Factor Surveillance System. Weight classification by Body
Mass Index (BMI). 2006. apps.nccd.cdc.gov/brfss/list.asp?cat?OB&yr?
47. Harvey EL, Glenny A, Kirk SF, Summerbell CD. Improving health profes-
sionals’ management and the organisation of care for overweight and
obese people. Cochrane Database Syst Rev 2001;2:CD000984.
48. Bentz CJ, Bayley KB, Bonin KE, Fleming L, Hollis JF, McAfee T. The
feasibility of connecting physician offices to a state-level tobacco quit line.
Am J Prev Med 2006;30:31–7.
49. Unrod M, Smith M, Spring B, DePue J, Redd W, Winkel G. Randomized
controlled trial of a computer-based, tailored intervention to increase
smoking cessation counseling by primary care physicians. J Gen Intern Med
50. Hazlehurst B, Sittig DF, Stevens VJ, et al. Natural language processing in
the electronic medical record: assessing clinician adherence to tobacco
treatment guidelines. Am J Prev Med 2005;29:434–9.
51. Katz DA, Muehlenbruch DR, Brown RL, Fiore MC, Baker TB. Effectiveness
of implementing the agency for healthcare research and quality smoking
November 2008 Am J Prev Med 2008;35(5S)
cessation clinical practice guideline: a randomized, controlled trial. J Natl
Cancer Inst 2004;96:594–603.
52. Solberg LI, Davidson G, Alesci NL, Boyle RG, Magnan S. Physician
smoking-cessation actions: are they dependent on insurance coverage or
on patients? Am J Prev Med 2002;23:160–5.
53. Wadland WC, Holtrop JS, Weismantel D, Pathak PK, Fadel H, Powell J.
Practice-based referrals to a tobacco cessation quit line: assessing the
impact of comparative feedback vs general reminders. Ann Fam Med
54. Ellerbeck EF, Ahluwalia JS, Jolicoeur DG, Gladden J, Mosier MC. Direct
observation of smoking cessation activities in primary care practice. J Fam
55. Nawaz H, Adams ML, Katz DL. Weight loss counseling by health care
providers. Am J Public Health 1999;89:764–7.
56. Jackson JE, Doescher MP, Saver BG, Hart LG. Trends in professional advice
to lose weight among obese adults, 1994 to 2000. J Gen Intern Med
57. Galuska DA, Will JC, Serdula MK, Ford ES. Are health care professionals
advising obese patients to lose weight? JAMA 1999;282:1576–8.
58. Abid A, Galuska D, Khan LK, Gillespie C, Ford ES, Serdula MK. Are
healthcare professionals advising obese patients to lose weight? A trend
analysis. MedGenMed 2005;7:10.
59. Podl TR, Goodwin MA, Kikano GE, Stange KC. Direct observation of
exercise counseling in community family practice. Am J Prev Med
60. Shea S, DuMouchel W, Bahamonde L. A meta-analysis of 16 randomized
controlled trials to evaluate computer-based clinical reminder systems for
preventive care in the ambulatory setting. J Am Med Inform Assoc
61. Hunt DL, Haynes RB, Hanna SE, Smith K. Effects of computer-based
clinical decision support systems on physician performance and patient
outcomes: a systematic review. JAMA 1998;280:1339–46.
62. Weingart SN, Toth M, Sands DZ, Aronson MD, Davis RB, Phillips RS.
Physicians’ decisions to override computerized drug alerts in primary care.
Arch Intern Med 2003;163:2625–31.
63. Yarnall KS, Pollak KI, Ostbye T, Krause KM, Michener JL. Primary care: is
there enough time for prevention? Am J Public Health 2003;93:635–41.
64. Wagner EH, Austin BT, Von Korff M. Organizing care for patients with
chronic illness. Milbank Q 1996;74:511–44.
Did you know?
You can stay updated on key articles with saved searches and table of content e-mail
Visit www.ajpm-online.net today to see what else is new online!
S358 American Journal of Preventive Medicine, Volume 35, Number 5Swww.ajpm-online.net