Adoption of the Healthy Heart Kit by Alberta family physicians.

Bize, Raphael ; Plotnikoff, Ronald C. ; Scott, Shannon D. 等

Cardiovascular disease (CVD) is Canada's leading cause of mortality and morbidity, with estimated direct and indirect costs of $7.2 and $12.4 billion respectively. (1) Further, physicians report difficulty in meeting recommended standards for clinical practice guidelines related to CVD (and other chronic diseases) prevention and treatment. For example, only about half of all physicians routinely advise people who smoke, to quit. (2) The multiplicity of risk factors and the heterogeneity of preventive guidelines are often cited as barriers in achieving such recommendations. (3)

Further, the adoption of a new behaviour of using new innovations--such as employing resource materials related to clinical practice guidelines--often requires a complex cognitive process. Although reports of various resource kits developed for physicians exist, little evaluative research has been conducted regarding physician use of such materials and individual-level determinants that shape their use. One such resource, the "Healthy Heart Kit" (HHK), was developed by Health Canada, the College of Family Physicians, and the Heart and Stroke Foundation. (4) The HHK was launched in 1999 to provide physicians with practical guidelines and tools for the prevention of CVD.

The Kit includes training materials for the systematic assessment and management of six modifiable CVD risk factors (smoking, high blood pressure, high cholesterol, overweight, sedentary lifestyle, and diabetes), charts stickers and paper-based reminders, as well as patients' information sheets. Integrated approaches for the management of multiple CVD risk factors have indeed been found to be superior in achieving recommended goals than non-integrated approaches. (5) In 2001, the intersectoral partnership "Achieving Cardiovascular Health in Canada" (ACHIC) endorsed the HHK as a practical tool for the prevention of CVD. (6,7) The HHK has demonstrated content validity and practical utility, (6) however, no published study to date has examined its actual use or the determinants of its use by family physicians. Based on the literature on the uptake of new information by physicians (8,9) potential predictors of HHK use include: the degree of agreement with the new clinical tool, confidence in using the new tool, the practice setting, academic affiliation (i.e., whether physicians hold a part-time appointment at a university hospital or not), time since graduation, number of hours spent in patient care, average duration of patient visits, and physicians' health behaviours. Our research objective was to investigate the association of the above listed characteristics with the frequency of HHK use by physicians.

METHODS

Study design

A one-group, cross-sectional design was employed to examine our research objective. A questionnaire was completed by family physicians in the province of Alberta after using the HHK in their practice for a duration of two months. The questionnaire collected data on the frequency of using the HHK on patients with at least one cardiovascular risk factor. Physician's socio-demographic, cognitive and behavioural data were also collected. Participating physicians were asked to return the completed questionnaire in a provided self-addressed, stamped envelope. A post-card reminder was mailed to all non-responding physicians two weeks after the initial mailing. A final reminder letter, containing a new copy of the questionnaire and a self-addressed stamped envelope, was sent two weeks later. The study protocol was approved by the Health Research Ethics Board of the University of Alberta. All data were treated in a confidential manner.

Study population

All registered family physicians in Alberta (n=3068) were invited to participate in the study. Physicians were contacted for study recruitment through direct mailing from the Alberta College of Family Physicians (ACFP). This initial mailing included an information letter inviting physicians to take part in the study, and a written consent form to be returned to ACFP in a provided self-addressed, stamped envelope. To be eligible, physicians had to be registered within the ACFP. Participants were blinded to the study objective to limit the risk of information bias.

Intervention

Physicians who took part in our study were sent a Kit between July and September 2006, with the "Guidelines for Management of Modifiable Risk Factors in Adults at High Risk for Cardiovascular Events" published by the Alberta Medical Association. (10) Physicians were asked to test the Kit in their practice for two months.

Data collection: Measures

The frequency with which physicians used the Kit with appropriate patients (those with at least one cardiovascular risk factor) was conceived as the dependent variable and was assessed and reported by the participating physicians on a visual analogue scale ranging from 0 (almost never) to 100 (almost always). Visual Analogue Scales have been used extensively to assess a variety of constructs in health and medical fields, and have demonstrated substantial correlations with Likert-type scales. (11)

Socio-demographic variables collected in the questionnaire were gender, year of graduation from medical school, practice setting, number of hours per week spent in patient care, average visit duration with patients, and whether affiliated with an academic institution. Behavioural variables consisted of the physicians' smoking status, diet and physical activity (PA) habits. In addition, physicians' preventive practices with their patients prior to the study were assessed by asking them to rate on a four-point scale ranging from "never=1" to "frequently=4" the frequency with which they deliver the following services to their patients: weigh patients; calculate BMI; calculate coronary heart disease risk; counsel to cease smoking; counsel to increase PA; and, counsel to improve diet. An overall percentage score of preventive practices was created using the above six individual ratings. Table 1 details the socio-demographic and behavioural characteristics.

Cognitive variables included: an 11-item scale assessed the overall agreement with the Kit with response options ranging from "strongly disagree=1" to "strongly agree=5". The 11 positively framed statements assessed the HHK's usefulness, effectiveness, relevance, credibility, ease of use, understandability, compatibility with physician beliefs, benefits and adaptability. The scale score was obtained by adding the individual scores of agreement, and then transforming the resulting sum to a 0 to 100 scale. A similar measure assessed agreement with the Guidelines for Management of Modifiable Risk Factors in Adults at High Risk for Cardiovascular Events. Confidence in being able to use the Kit was assessed on a 9-point Likert-type scale (e.g., 1=not at all confident; 9=completely confident). The degree of control while using the Kit was assessed on a similar scale (e.g., 1=very little control; 9=complete control).

[FIGURE 1 OMITTED]

The content validity of the measures assessing the agreement of the Kit were based on the Diffusion of Innovation Theory which considers factors such as its relative advantage, consistency with values of the adaptor, complexity, the degree to which it may be experimented (trialability), and its visibility of its results to others. (12) The degree of control and confidence of using the Kit were based on the Theory of Planned Behavior (TPB) (13) and Self-Efficacy Theory (SET). (14) Perceived behavioural control (i.e., the perceived ease or difficulty of performing the behaviour) and self-efficacy (i.e., perceived confidence in performing the recommended behaviour) are core tenants of TPB and SET respectively.

Statistical analysis

The nonparametric test for trend (15) was used to assess the statistical significance of trends across practice settings. Multiple regression models were built using a purposeful selection method. (16) Univariate linear regression was conducted with each potential predictor of Kit use. All variables significant at a p-value < 0.2 were then selected as candidates for three multiple regression models (i.e., socio-demographic, cognitive, and behavioural *). A combined model was also tested in which all eligible variables from the single variable regression analyses (with p<0.2) were simultaneously entered. In the final model, those variables that were not significant (at the p<0.05 level) in the first combined model, were removed.

RESULTS

153 physicians agreed to participate in the study and received the HHK. 115 survey questionnaires were returned at the 2-month follow-up (follow-up rate = 75%). Figure 1 shows the flow of participants through the inclusion and follow-up process. Participating physicians were predominantly male (53%), worked in group practices or clinics (86%), and graduated before 1990 (61%). Table 1 displays the detailed baseline characteristics of respondents. The mean score of Kit use (dependent measure) was 61 [SD=26].

Socio-demographic model: Single and multiple variable associations between the score of Kit use and the socio-demographic variables (gender, year of graduation, practice setting, academic affiliation, and visit duration) were not statistically significant (Table 2, Model 1). There was a statistically significant trend for smaller practices to be associated with lower scores of Kit use with means of 51 [SD=29] for solo practice, 60 [SD=23] for group practice, and 70 [SD=32] for clinical settings; p-value for the nonparametric test for trend was 0.018 (z=2.37).

Cognitive model: The scale employed to assess agreement with the Kit demonstrated strong inter-item correlation (reliability), with a Cronbach's alpha coefficient of 0.92. The mean score of agreement was 77 [SD=14]. This measure, together with the other cognitive variables (agreement with the guidelines, control in using the Kit, and confidence in using the Kit) were strongly associated (p<0.001) in single variable analyses with the kit use score. "Agreement with the Kit" and "confidence in using the Kit" remained as strong independent cognitive predictors of Kit use in the multiple linear regression model, and explained 46% of its variability (Table 2, Model 2).

Behavioural model: Behavioural variables, (i.e., initial preventive practices with patients, and physicians' own health behaviours) were not associated with Kit use either in single or in multiple variable analyses (Table 2, Model 3).

Combined models: Finally, in the combined models, year of graduation was a significant predictor of Kit use in addition to the cognitive variables "agreement with the Kit" and "confidence in using the Kit". Compared with those who graduated before 1970, those who graduated between 1970-1989 scored 9 points higher on Kit use, whereas those who graduated since 1990 scored 14 points higher on Kit use (Table 2, Combined Model, second column). A one-point increase in the agreement with the Kit and the degree of confidence in using the Kit resulted respectively in a 1 and 6 points increase on Kit use.

DISCUSSION

Practice-based research is advocated to better understand which factors influence how evidence-based guidelines can best be translated into actual care delivery. (17) In this study assessing physician practice, a theory-driven score of agreement with a CV risk management Kit and the degree of confidence in using such tools were shown to be strongly associated with its reported use. Results also suggest that a more recent graduation year is also a significant predictor of Kit use. A trend for smaller practices to be associated with lower scores of Kit use was also shown.

Similar to our findings, other studies have found physicians in solo practices to be less prone to adopt new clinical practices compared to physicians in group practices or working in outpatient clinics. (8,9) This may be partly attributed to greater opportunities for collegial input, influence of respected opinion leaders, and exchange of knowledge in the latter. (8,9) As physicians in solo practices represent about one quarter of all family physicians in Alberta, this trend should raise the question of the need to adopt specific implementation strategies for physicians in this particular setting. For example, solo practitioners could be encouraged to build or engage in primary-care networks. (18)

Our findings also show that time since graduation was inversely associated with Kit use. Interestingly, a recent systematic review found an inverse relationship between clinical experience and quality of health care. (19) Of 62 published studies that measured physician knowledge or quality of care (and described time since medical school graduation or age), more than half of the studies suggested that physician performance declined over time for all measured outcomes, with only one study demonstrating improved performance for all assessed outcomes. (19)

Low participation rates when recruiting primary care physicians as participants is a well-recognized barrier to practice-based research. (20,21) For example, Sin18 obtained a 7.1% mail-out recruitment rate in an asthma study which fell to 5.1% when Alberta physicians were contacted for access to their patient charts; this is comparable to the 5.0% recruitment rate in our study.

Self-selection of physicians (e.g., with a particular interest in preventive practices) cannot be excluded as a potential study limitation and may have limited the external validity of our findings. Our sample, however, appears to reflect demographic and behavioural practice profiles reported in other studies involving Alberta family physicians. In the 2001 National Family Physicians Workforce Survey (NFPWS), (22) the 2,274 Alberta family physicians who completed a mail survey had very similar age group and practice setting proportions to our study. Further, physicians in our sample reported relatively similar rates of smoking cessation counselling (97.4% vs. 90.5% in the NFPWS survey) and PA counselling (91.2% vs. 89.6% in the NFPWS survey).

A study limitation was that there were no objective measures regarding the frequency of HHK use. As data collection was based on self-report, a social desirability bias cannot be excluded. Physicians may have over-reported professionally-valued opinions or behaviours. (21) This phenomenon would however not affect our findings, as long as such over-reporting was uniformly distributed across the ranges of the variables retained in our final model. Standardized patients would also have provided a valuable additional source of information to assess whether and how the HHK was actually used. Physicians participating in the study did not receive any specific training on how to use the Kit. Such training may have significantly improved the adoption of the Kit. Future studies on this topic should consider combining qualitative and quantitative approaches in order to better understand how physicians' characteristics influence the adoption of new clinical tools.

In summary, this study found that being a younger physician, working in a group practice or a clinic, reporting a high degree of agreement with positive statements about the Kit, and a high confidence level in using the Kit were all associated with a higher score of Kit use. Future research should explore whether agreement with a new clinical tool might be influenced by gaining support and involvement of practitioners (including older physicians and solo practitioners) at the development stages, and whether increasing confidence in using this tool might be influenced by choosing an implementation strategy allowing physicians to familiarize themselves and have personalized feedback on how to best use such a tool.

Received: January 30, 2008

Accepted: September 25, 2008

REFERENCES

(1.) Health Canada. The Growing Burden of Heart Disease and Stroke in Canada. 2003. Ottawa: Health Canada, 2003. Available online at: http://www.cvdinfobase.ca/cvdbook/ (Accessed December 14, 2007).

(2.) Smith HE, Herbert CP. Preventive practice among primary care physicians in British Columbia: Relation to recommendations of the Canadian Task Force on the Periodic Health Examination. CMAJ 1993;149:1795-800.

(3.) Lomas J, Anderson GM, Domnick-Pierre K, Vayda E, Enkin MW, Hannah WJ. Do practice guidelines guide practice? The effect of a consensus statement on the practice of physicians. N Engl J Med 1989;321:1306-11.

(4.) Smith ER. The Healthy Heart Kit. Can J Cardiol 1999;15(3):349,351-52.

(5.) Hyman DJ, Pavlik VN, Taylor WC, Goodrick GK, Moye L. Simultaneous vs. sequential counseling for multiple behavior change. Arch Intern Med 2007;167:1152-58.

(6.) McClaran J, Kaufman D, Toombs M, Beardall S, Levy I, Chockalingam A. From death and disability to patient empowerment: An interprofessional partnership to achieve cardiovascular health in Canada. Can J Public Health 2001;92:I3-I9.

(7.) Kaufman DM, McClaran J, Toombs M, Beardall S, Levy I, Chockalingam A. Achieving cardiovascular health through continuing interprofessional development. Can J Public Health 2001;92:I10-I16.

(8.) Lopez-de-Munain J, Torcal J, Lopez V, Garay J. Prevention in routine general practice: Activity patterns and potential promoting factors. Prev Med 2001;32:13-22.

(9.) Kopelow ML, Schnabl GK, Hassard TH, Tamblyn RM, Klass DJ, Beazley G, et al. Assessing practicing physicians in two settings using standardized patients. Acad Med 1992;67:S19-S21.

(10.) Tsuyuki RT, Koshman S, Pearson GJ. Guidelines for Management of Modifiable Risk Factors in Adults at High Risk for Cardiovascular Events. Alberta Medical Association, 2005. Available online at: http://www.topalbertadoctors.org/ NR/rdonlyres/A548C784-8630-45C8-9220-30D5C1926875/0/ cardiovascular_guideline.pdf (Accessed August 16, 2007).

(11.) Streiner DL, Norman GR. Health Measurement Scales. A Practical Guide to their Development and Use. Oxford University Press, 1996.

(12.) Rogers EM. Diffusion of Innovations, 5th ed. New York, NY: Free Press, 2003.

(13.) Ajzen I. From intentions to actions: A theory of planned behavior. In: Kuhl J, Beckman J (Eds.), Action-control: From Cognition to Behavior. Heidelberg: Springer, 1985;11-39.

(14.) Bandura A. Self-Efficacy Theory. Social Foundations of Thought and Action: A Social Cognitive Theory. Englewood Cliffs, NJ: Prentice-Hall, 1986.

(15.) Cuzick J. A Wilcoxon-type test for trend. Stat Med 1985;4:87-90.

(16.) Hosmer DW, Jr, Lemeshow S. Stepwise Logistic Regression. Applied Logistic Regression, 2nd ed. Hoboken: John Wiley & Sons, 2000;116-21.

(17.) Pierce M. Doing research in general practice: Advice for the uninitiated. Diabet Med 1998;15:S25-S28.

(18.) Sin DD, Man SF, Cowie RL, Sharpe HM, Andrews EM, Bell NK, et al. Recruitment for a provincial asthma study. Participation of network and nonnetwork primary care physicians. Can Fam Phys 2004;50:1251-54.

(19.) Choudhry NK, Fletcher RH, Soumerai SB. Systematic review: The relationship between clinical experience and quality of health care. Ann Intern Med 2005;142:260-73.

(20.) Levinson W, Dull VT, Roter DL, Chaumeton N, Frankel R. Recruiting physicians for office-based research. Med Care 1998;36:934-37.

(21.) Asch S, Connor SE, Hamilton EG, Fox SA. Problems in recruiting community-based physicians for health services research. J Gen Intern Med 2000;15:591-99.

(22.) The College of Family Physicians of Canada. The National Family Physician Workforce Survey (NFPWS) 2001; 2001. Available online at: http://www.cfpc.ca/English/cfpc/research/janus%20project/nfpws/default.asp ?s=1 (Accessed December 14, 2007).

* socio-demographic (i.e., gender, year of graduation, practice setting, academic affiliation, and visit duration), cognitive (i.e., degree of agreement with the guidelines, degree of agreement with the Kit, degree of control in using the Kit, and degree of confidence in using the Kit), and behavioural (i.e., previous initial preventive practice score, and own health behaviours).

Raphael Bize, MD, MPH, [1,2] Ronald C. Plotnikoff, PhD, [1,3] Shannon D. Scott, PhD, [1,4] Nandini Karunamuni, MSc, [1] Wendy Rodgers, PhD [3]

Author Affiliations

[1.] Centre for Health Promotion Studies, School of Public Health, University of Alberta, Edmonton, AB

[2.] Department of Ambulatory Care and Community Medicine, University of Lausanne, Switzerland

[3.] Faculty of Physical Education and Recreation, University of Alberta, Edmonton, AB

[4.] Department of Pediatrics, Faculty of Medicine and Dentistry, University of Alberta, Edmonton, AB

Correspondence and reprint requests: Ronald C. Plotnikoff, Centre for Health Promotion Studies, School of Public Health, University of Alberta, 5-10 University Extension Centre, 8303--112 Street, Edmonton, AB T6G 2T4, Tel: 780-492-4372, Fax: 780-492-9579, E-mail: ron.plotnikoff@ualberta.ca

Acknowledgements: Dr. Bize holds salary support from The Swiss National Science Foundation. Dr. Plotnikoff is supported from Salary Awards from the Canadian Institutes of Health Research (Applied Public Health Chair) and the Alberta Heritage Foundation for Medical Research (Health Scholar). Dr. Scott received funding from the Canadian Institutes of Health Research, Alberta Heritage Foundation for Medical Research and the Canadian Child Health Clinician Scientist program to support this work. We would like to acknowledge the Alberta College of Family Physicians for their assistance and provision of the family physicians contact information, and the University of Alberta for funding this project.

Table 1. Sample Characteristics (n=114 unless otherwise specified)

Variables                                            % (n)

Sex
  Male                                               52.6 (60)
  Female                                             47.4 (54)
Year of graduation
  [less than or equal to] 1969                       11.4 (13)
  1970-1989                                          50.0 (57)
  [greater than or equal to] 1990                    38.6 (44)
Practice setting
  Solo practice                                      14.0 (16)
  Group practice                                     71.1 (81)
  Outpatient clinic                                  14.9 (17)
Academic affiliation
  Yes                                                36.0 (41)
  No                                                 64.0 (73)
Time spent in patient care (hours)
  <20                                                4.4 (5)
  20-40                                              33.3 (38)
  >40                                                62.3 (71)
Average duration of visits (minutes)
  0-10                                               25.4 (29)
  11-20                                              70.2 (80)
  >20                                                4.4 (5)
Smoking status (n=112)
  Current smoker                                     1.8 (2)
  Former smoker                                      17.0 (19)
  Never smoker                                       80.2 (91)
Moderate intensity physical activity level (n=112)
  [less than or equal to] 1d/wk with 30 min. of      10.7 (12)
  mod. intensity PA
  2-4 d/wk with 30 min. of mod. intensity PA         54.5 (61)
  [greater than or equal to] 5 d/wk with 30 min.     34.8 (39)
  of mod. intensity PA
Buy low-fat food
  Never/seldom                                       8.8 (10)
  Occasionally                                       14.9 (17)
  Often/Very often                                   76.3 (87)

Table 2. Single and Multiple Linear Regression Results Showing the
Associations between the Score of Kit Use and the Independent
Variables (Standardized [beta]-coefficients reported)

                       Model 1                   Model 2

Socio--          Single       Multiple     Single       Multiple
demographic      Variable     Variable     Variable     Variable
Factors          Regression   Regression   Regression   Regression

Graduation       .30 *        .31 *
  year between
  1970 and
  1989
  (reference
  category:
  <1970)
Graduation       .29 *        .30 *
  year since
  1990
Female gender    .07
  (ref. cat.:
  male)
Group practice   .17 *        .16
  (ref. cat.:
  solo)
Outpatient       .27 *        .20
  clinic
Visit duration   -.04 *       -.05 *
  11-20 min
  (ref. cat.:
  0-10 min)
Visit duration   .17 *        .15 *
  >20 min
No academic      .01
  affiliation
  (ref. cat.:
  academic
  affiliation)
20-40 hrs in     .07
  patient care
  (ref. cat.:
  <20 hrs)
>40 hrs in       .10
  patient care

Cognitive
Factors

Agreement with                             .47 **       -.01
  the
  guidelines
  (0-100)
Agreement with                             .62 **       .42 **
  the Kit
  (0-100)

Degree of                                  .32 **       -.01
  control in
  using the
  Kit (1-9)
Degree of                                  .60 **       .39 **
  confidence
  in using
  the Kit (1-9)

Behavioural
Factors

Score of
  previous
  preventive
  practice
  with
  patients
  (0-100)
Former smoker
  (ref. cat.:
  current
  smoker)
Never smoker
Buy low-fat
  food
  occasionally
  (ref. cat.:
  never/seldom)
Buy low-fat
  food often/
  very often
2-4 days/week
  with 30 min
  of moderate
  PA (ref.
  cat.: [less
  than or equal
  to] 1 day/
  week)
[greater than
  or equal to]
  5 days/week
  with 30 min
  of moderate
  PA
Adjusted                      .05                       .46
  [R.sup.2]
(SE)                          (25.11)                   (18.74)

                       Model 3              Combined Models ([dagger])

Socio--          Single       Multiple     Multiple     Final Model
demographic      Variable     Variable     Variable     ([dagger])
Factors          Regression   Regression   Regression   (standardized/
                                           ([dagger])   unstandardized
                                                        coefficients)

Graduation                                 .19 **       .18 **
  year between                                          (9.45)
  1970 and
  1989
  (reference
  category:
  <1970)
Graduation                                 .28 **       .26 **
  year since
  1990                                                  (13.94)
Female gender
  (ref. cat.:
  male)
Group practice                             .01
  (ref. cat.:
  solo)
Outpatient                                 .09
  clinic
Visit duration                             -.01
  11-20 min
  (ref. cat.:
  0-10 min)
Visit duration                             .11
  >20 min
No academic
  affiliation
  (ref. cat.:
  academic
  affiliation)
20-40 hrs in
  patient care
  (ref. cat.:
  <20 hrs)
>40 hrs in
  patient care

Cognitive
Factors

Agreement with                             -.01
  the
  guidelines
  (0-100)
Agreement with                             .42 **       .42 **
  the Kit
  (0-100)
                                                        (0.79)
Degree of                                  -.03
  control in
  using the
  Kit (1-9)
Degree of                                  .38 **       .38 **
  confidence
  in using
  the Kit (1-9)
                                                        (5.82)

Behavioural
Factors

Score of         .15 *                     .00
  previous
  preventive
  practice
  with
  patients
  (0-100)
Former smoker    -.28
  (ref. cat.:
  current
  smoker)
Never smoker     -.28
Buy low-fat      .14
  food
  occasionally
  (ref. cat.:
  never/seldom)
Buy low-fat      .18
  food often/
  very often
2-4 days/week    -.07
  with 30 min
  of moderate
  PA (ref.
  cat.: [less
  than or equal
  to] 1 day/
  week)
[greater than    .09
  or equal to]
  5 days/week
  with 30 min
  of moderate
  PA
Adjusted                      .01          .49          .50
  [R.sup.2]
(SE)                          (25.58)      (18.33)      (18.27)

([dagger]) All elgible variables from the single variable multiple
regression analysis (with p<0.2) were entered together to the first
combined model. In the final model, those variables that were not
significant at the p<0.05 in the first combined model were removed.

* 0.05 [less than or equal to] p-value <0.02; ** p-value <0.05