Abstract
Background: Discussing self-care activities like exercise is challenging with the many competing demands during primary care encounters. Our objective was to study the relationship between the Chronic Care Model (CCM) score in the clinic and time spent by the physicians discussing exercise during encounters with diabetic patients.
Methods: Consecutive patients with diabetes across 20 primary care clinics in South Texas were included. Time spent discussing exercise was determined using the Davis Observation Code on audio recordings of the visits. Clinicians completed the Assessment of Chronic Illness Care survey, a validated measure of the extent to which care delivered is consistent with the CCM. Data were analyzed using hierarchical linear models.
Results: A total of 162 transcribed recordings were analyzed. Age, the number of problems addressed, stage of change (SOC), and overall length of the visit were associated with time spent discussing exercise. There was a significant relationship between clinic CCM score and time spent by physicians advising about exercise, independent of SOC for exercise (P < .01). Also, a discussion about exercise was more likely to occur with patients who were in the contemplation SOC for exercise.
Conclusions: Discussions of exercise may be 18 to 33 seconds longer in clinics with full implementation of the CCM compared with those with basic implementation. Facilitating more complete CCM implementation in clinics with a basic level of CCM that serve a population of patients who are sedentary may realize the most benefit.
Primary care physicians can play an important role in supporting self-care activities, such as exercise among patients with a chronic illness.1 Despite the evidence of effectiveness of preventive services and supporting self-management behaviors,2,3 primary care encounters are bounded by time constraints within which multiple diagnoses, problems, and patient concerns compete with each other for a place on the agenda.4 These multiple competing demands often result in health behavior counseling taking a back seat to issues and concerns considered more pressing or important.2
The Chronic Care Model (CCM) suggests that interactions between informed activated patients and prepared, proactive care teams can produce better care and improved outcomes.1,5 The CCM was developed by identifying 6 evidence-based elements associated with improved chronic illness outcomes. It is based on the concept that when proactive patient care teams provide care to informed, activated patients, clinical quality and outcomes will improve. For example, for patients to engage in proactive care (delivery system design), practices need to be able to view all the patients in their panels (clinical information systems) who need certain guideline-based treatments (decision support), and patients must agree to any changes in their care and integrate them into their lives (self-management support).1
Considerable experience using the CCM to improve the quality of chronic illness care has accumulated over the past decade.6 Evidence suggests that practices redesigned in accordance with the CCM generally improve the quality of care and the outcomes for patients with various chronic illnesses.7,8 This finding seems to be consistent in both US and international settings.9,10
Of particular interest is evidence suggesting that the CCM is associated with clinical outcomes such as control of glycosylated hemoglobin. This relationship seems to be mediated by patients’ self-care activities, such as diet and exercise.5 The Transtheoretical Model11,12 asserts that there are distinct stages in the adoption of health behavior change, which individuals cycle through en route to consistent adoption of the health behavior in question. In previous studies we found a relationship between the stage of change (SOC) for exercise and the CCM. We found that patients in clinics with high CCM scores were more likely to be in the maintenance SOC for exercise and other self-care behaviors.3,5,7,8,13–17 One possible explanation for this relationship is that in clinics with higher CCM scores patients receive more advice and support to improve their level of exercise.
The purpose of this study was to expand our understanding of the relationship between the CCM and exercise SOC. We hypothesize that presence of the CCM in primary care clinics will be associated with the amount of time spent discussing exercise during the visit.
Methods
Setting
The results reported here are from data collected in the Direct Observation of Diabetes Care study.5,13,17 Details of the methods have been explained in previous studies.7,8 The data were collected as part of an in-depth examination of the of quality of care delivered to patients with type 2 diabetes across a wide variety of primary care settings in 2002 to 2003. The study was cross-sectional and was conducted in 20 primary care clinics with 45 primary care physicians from across South Texas: 11 solo physician clinics (11 physicians); 3 group practice settings (10 physicians); 1 community health center (1 physician); 2 Veterans Affairs primary care clinics (11 physicians); and 3 city/county health clinics for uninsured patients (12 physicians).
Patients and Data Collection
Within each clinic, 8 to 10 consecutive patients who presented with an established diagnosis of type 2 diabetes were recruited to participate in the study and had their visit audio-recorded. None of the patients who were approached declined participation. After the visit patients completed an exit survey and were asked about their SOC for self-care behaviors for exercise. The SOCs were adopted from the Transtheoretical Model: precontemplation, contemplation, preparation, action, and maintenance.11,12,16 In addition to describing specific SOC in the analysis, we also constructed an SOC variable as a dichotomous outcome: yes, the patient is in the maintenance stage of change for the self-care behavior, or no, the patient is not in the maintenance SOC for the self-care behavior. Patients in the maintenance SOC reported that they have been adherent to the behavior for at least the past 6 months. Additional patient characteristics included in the analysis were age, sex, race/ethnicity, and self-reported health status (excellent, very good, good, fair, poor).
Measurements
Chronic Care Model
We used the Assessment of Chronic Illness Care survey (ACIC) to determine the degree to which care in each clinic was consistent with the CCM.1 This 25-item survey, which measures the presence of the elements of the CCM, was completed by all clinicians in each clinic (physicians, nurse practitioners, and physician assistants). Each item is scored on a 0 to 11 scale and provides subscale scores for each of the 6 CCM components. Several studies support the validity of this instrument. For example, all 6 subscales were responsive to process of care improvement in a study of an intervention for diabetes and congestive heart failure,17–20 whereas in a collaborative intervention study overseen by the Institute for Health Care Improvement, ratings by an external team on the depth of implementation of the elements of the CCM were significantly associated with the overall ACIC score for 5 of the 6 elements.6
Content of Encounter (Exercise Discussion)
Direct observation and audio recordings of the visit allowed a trained coder to discriminate the presence or absence of discrete behaviors that have been operationally defined, systematically minimizing experimenter basis. The Davis Observation Code is a reliable and valid interactional analysis system that has been used to characterize differences in physician practice styles in a variety of previous studies.7,21–23 The coder recorded the occurrence or nonoccurrence of each of 20 clinically significant behaviors during successive 15-second observation intervals of the medical encounter, including a code for the discussion of exercise. For purposes of this analysis, the number of 15-second intervals devoted to a discussion of exercise were converted into minutes by multiplying by 4.
Data Analyses
A mean total CCM score was calculated for each clinic. In addition to characterizing individual SOC, we constructed a SOC variable as a dichotomous outcome for the patient self-care behavior for exercise: the patient is in one of the 5 SOCs (precontemplation, contemplation, preparation, action, or maintenance). Additional patient and visit characteristics such as age, sex, length of visit, reason for visit, maintenance SOC, and SOC for exercise were controlled for in the final model. To account for clustering of patients within clinics, a 2-level hierarchical regression model was used, with patient- and visit-level predictors entered at level 1 and clinic ACIC scores entered at level 2. The number of 15-second intervals spent discussing exercise was the dependent or outcome variable. Patient-level predictors included age, sex, whether or not the visit was an acute visit, whether or not the patient was in the maintenance SOC for exercise, and the length of visit (dependent variable: time spent discussing exercise). We also controlled for precontemplation SOC for exercise in the final model.
Results
A total of 188 visits were audio recorded, but 26 visits were in Spanish, so 162 visits were coded using the Davis observation codes in the year 2003. Patient and visit characteristics are present in Table 1. The average age of patients was 58 years (range, 35–71 years). More than half (54%) were Hispanic, consistent with the demographics of the South Texas region. A majority were high-school graduates. Less than half rated their health as fair/poor. In addition to diabetes, the average number of chronic illness diagnoses for each patient was 5, ranging from 2 to 7 comorbid conditions. Patients were taking an average of 7 medications (range, 4–10). The purpose of most visits was for a routine checkup; only 23% were acute visits. The average duration of a visit was approximately 18 minutes (range, 10–26 minutes). The average time discussing exercise was 22 seconds (range, 8–36 seconds).
Descriptive Statistics of Patients and Visits (n = 162)
Table 2 presents the predictors for the time spent discussing exercise. The number of problems addressed during the visit, along with the overall length of the visit, was positively associated with time spent discussing exercise; however, age and the number of symptoms or problems raised by the patient were not. SOC for exercise was also associated with the likelihood that exercise was discussed (Table 3). For patients who were in the contemplation, preparation, and action SOCs, the likelihood of discussing exercise was higher than among those in the maintenance and precontemplation SOCs (P = .005).
Bivariate Relationships between Each Predictor and Time Spent* Discussing Exercise (n = 162)
Stage of Change (SOC) of Exercise and If Time Was Spent Discussing Exercise
In the hierarchical regression model (Table 4), the CCM score was associated with the time spent during the encounter discussing exercise, after adjusting for patient and visit characteristics. For each 1-point increment in the CCM score, the time spent discussing exercise increased by approximately 3 seconds.
Multivariable Random Effects Model Predicting Time Spent Discussing Exercise
Discussion
In clinics where the CCM is more fully implemented, primary care physicians spend more time discussing exercise during encounters with patients who have type 2 diabetes. This is true even after controlling for the reason for the visit, the length of visit, and the individual patient's exercise SOC. What does this mean in more practical terms? The CCM score ranges from 0 to 11, with scores from 9 to 11 representing full CCM implementation and scores of 0 to 3 (a range of difference from 6–11 points) representing basic implementation. The regression coefficient suggests that each 1-point increase is associated with an additional 3 seconds of exercise discussion. Therefore, discussion of exercise may be 18 to 33 seconds longer in clinics with full implementation of the CCM compared with those with basic implementation.
Although overall time spent counseling on exercise-related matters was statistically correlated with CCM scores, the time differential was very small and the total range of time spent in this activity was very limited across the entire visit cohort. Behavioral change “counseling” for this brief period of time is similar to other studies in which providers incorporated brief physical activity advice into routine primary care visits with little disruption. The response to physician advice protocol was positive and participation in the study was viewed as beneficial.24,25 Primary care settings provide an important opportunity to promote physical activity for adults. Although those most at risk are more likely to receive such advice, there are many more that may benefit.26 Evidence-based primary care physical activity counseling protocols can be translated into routine practice.27 Opportunistic strategies show promise for a higher yield of identification of patients at risk and leading to advice.28
In addition, patient SOC for exercise seems to predict the likelihood that a discussion about exercise will occur. It is less likely to occur with patients who are in the precontemplation and maintenance SOCs and more likely to occur with patients in the contemplation, preparation, and action SOCs. Patient-initiated health behavior discussions are more likely to receive advice if they explicitly indicated readiness to change. A discussion about exercise occurred in only 9% of encounters with patients who were in the precontemplation SOC, but a discussion about exercise occurred in 69% of encounters when the patient was in contemplation stage (see Tables 2 and 3). It is possible that primary care physicians are aware of the SOC and tailor the time during the encounter to match the SOC for exercise. Amid complex visits with a high level of competing demands, primary care physicians are somehow aware of the SOC of their patients, as reflected in varying percentage of exercise discussions. Future studies will need to explore this finding in more detail.
Our findings are consistent with other studies suggesting that the CCM is related to clinician behavior. For example, implementation or presence of the CCM was associated with traditional process of quality of care indicators dependent on clinician behavior, such as performance of a foot or eye examination for patients with diabetes.17,19,29,30 In other studies where CCM scores were higher, providers were more likely to use recommended therapies such as lipid-lowering and angiotensin-converting enzyme inhibition therapy.
Other studies have found a relationship between the CCM and clinical outcomes, such as glycosylated hemoglobin.5,31,32 It is possible that the pathway to better clinical outcomes for patients seen in clinics with higher CCM scores is through improved patient self-care behavior such as exercise. For example, in settings with higher CCM scores, patients with asthma were more likely to monitor their peak flows and have a written action plan, and their quality of life improved.33
Patient age was found not to be an important predictor of issues addressed during the medical visit. This is contrary to studies that have shown that health promotion discussions, asking patients to make changes in their behavior to increase or promote health, occurred more frequently with younger patients.28,34–38 Irrespective of patient age, exercise counseling was significantly more likely to take place with patients who presented with more problems and who were in for a routine visit. Alternatively, physicians may perceive that a discussion of exercise is needed with patients who have more problems and more medications, and that these patients were more suited to discussions of health promotion. Moreover, the competing demands of managing chronic conditions in older patients during the scheduled visit time may preclude addressing other issues such as health promotion.4,13,39
Why should the presence of the CCM be related to length of discussion about exercise? It is possible that higher CCM scores reflect a more proactive approach to improving exercise. Perhaps patients seen in clinics with higher CCM scores are more likely to be actively involved in their own care and ask questions related to exercise, thus prompting a discussion of exercise. Also, clinics with higher CCM scores may have clinic structures and processes such as reminders and resources that support clinicians and encourage them to discuss exercise. For example, computer prompts, encounter templates, or a flow chart may increase the time devoted to a discussion of exercise. It is also possible that clinics with a higher CCM score may have longer visits, but in further analysis of the data this did not prove to be true (data not shown). Developed more than a decade ago, the CCM is a widely adopted approach to improving ambulatory care that has guided clinical quality initiatives in the United States and around the world. Here we examined the evidence of the CCM's effectiveness by reviewing articles published since 2000 that used one of 5 key CCM articles as a reference. Accumulated evidence seems to support the CCM as an integrated framework to guide practice redesign. Although work remains to be done in areas like cost-effectiveness, these studies suggest that redesigning care using the CCM leads to improved patient care and better health outcomes.6
One limitation of this study is the cross-sectional nature of the data. For example, it is possible that some physician characteristics would result in both the presence of the CCM and a longer discussion of exercise. Another limitation of this study is an inability to draw any conclusion or causality or the direction of observed relationships because of the cross-sectional nature of the data. We do not know whether the SOC was influencing the time spent discussing exercise or if visit time spent discussing exercise was influencing exercise SOC, as reported after the visit by the patients on the administered survey. It is possible that some as-yet unrecognized factor may influence both the CCM score and the reason for increased time spent by physicians advising about exercise. Prospective studies are needed to further evaluate the effect of CCM on the time spent discussing exercise and whether this time equates to patient behavior change.
Conclusions
In primary care clinics where care is more consistent with the CCM, more time is spent discussing exercise during a routine visit encounter. These clinics may have prepared proactive teams interacting with informed, activated patients, resulting in better support for self-care behaviors. Self-management programs for chronic diseases such as diabetes, hypertension, and heart disease probably produce clinically important benefits. Because primary care clinics are the principal source of the tools for self-care and because they support patients who need to learn to manage complex chronic diseases, a greater focus on strengthening the presence of the CCM may be needed in these settings.
Notes
This article was externally peer reviewed.
Funding: The research reported was supported by the Agency for Healthcare Research and Quality (AHRQ K08 HS013008-03) and the Bureau of Health Professions; Health Resources and Services Administration Department of Health and Human Services grant D01 HP08795; The John A Hartford Foundation Center for Excellence in Geriatrics Education; and by the Department of Veterans Affairs.
Conflict of interest: none declared.
Disclaimer: The views expressed in this article are those of the authors and do not necessarily represent the views of the Department of Veterans Affairs.
- Received for publication June 15, 2010.
- Revision received September 20, 2010.
- Accepted for publication October 4, 2010.
