Cardiovascular Risk Factors, Type 2 Diabetes & Primary Care Clinic Structure
Michael L. Parchman, MD1 Amer Kassai, PhD2 Jacqueline A. Pugh, MD1 Raquel L. Romero, MD1
1University
of Texas Health Science Center, San Antonio, Texas
2Trinity
University, San Antonio, Texas
Cardiovascular Disease (CVD) Risk Factors
Glucose Control
Hemoglobin
A1c Goal: <= 7.0%
Blood Pressure
Goal:
<= 130/80
Lipids
LDL
Cholesterol Goal: <= 100 mg/dl (if no CAD)
Self-Care Activities
Diet, Exercise, Glucose Monitoring, Medication Adherence 5 Stages of Change:
Pre-contemplation Contemplation Preparation Action
Maintenance:
adherence for 6 months or
more
The Chronic Care Model (CCM)
Purpose
Examine
the relationship between control of CVD risk factors, patient self-care behaviors, and the presence of the CCM model elements across a diverse group of primary care clinic settings.
Methods
20 small autonomous primary care clinics
Solo
practice physicians (n=11) Small group practices (n=3) Community Health Clinic (n=1) VHA Primary Care OPC (n=2) City/County Indigent Health Clinics (n=3)
Recruited from a Primary Care Practice Based Research Network (PBRN)
Subjects and Data Collection
Patients
30
consecutive presenting pts with an established dx of type 2 DM Exit survey: demographics, stage of change for selfcare behaviors, health status (excellent, v. good, good, fair, poor) Chart Abstraction: most recent values of A1c, BP and LDL-cholesterol
Clinicians
Assessment
of Chronic Illness Care (ACIC) Survey. (Bonomi, Wagner et al 2002) (25 items)
ACIC Survey: Sub-Scales
Organizational Leadership Community Linkages Self-Management Support Decision Support Delivery System Design Clinical Information Systems
Analysis
Outcome: All 3 risk factors well controlled (Y/N) Hierarchical Logistic Model (Random Effects Model)
Patients
clustered within clinic
Predictors:
Patient:
Age (years) Hispanic ethnicity (Y/N) Female gender Maintenance Stage of Change for all 4 behaviors (Y/N) Sub-scale scores from ACIC survey
Clinic
Results: Patient Characteristics
Age Female Hispanic Maintenance Stage of change for all 4 self-care behaviors? 58.6 (12.93) 51% 57% 25%
Results: CVD Risk Factors
Risk Factor
A1c <= 7.0% BP <= 130/80 LDL <= 100
Percent of total (range by clinic) 43% (20 to 69.7)
49% (0 to 72.7) 50% (0 to 73.3)
All 3 well controlled
13% (0 to 31.3)
ACIC Sub-scale Scores
Mean (S.D.)
Orgnzn Leadership Comm Linkage Self-Care Support Decision Support 6.5 (2.3) 7.1 (1.7) 6.9 (1.9) 6.0 (1.8)
Range*
2.5 – 10.0 4.3 – 10.7 2.8 – 10.3 2.7 – 9.0
Delivery System
Clinical Info System
6.7 (2.2)
5.2 (2.4)
3.4 – 11.0
0.6 – 10.2
*Potential Range of each sub-scale: 0 to 11
HLM Model: No Clinic-level Predictors
Patient Characteristic
Age Odds Ratio 1.01 95% C.I. 1.00, 1.02
Female
Hispanic
0.66*
0.86
0.48, 0.92
0.62, 1.19
All Maintenance
1.55*
1.09, 2.21
HLM: No Patient-level predictors
CCM component
Org Leader Comm Linkage Self-Care Support Decision Support
O.R.
0.89 1.65* 0.97 1.10
95% C.I.
0.72, 1.11 1.31, 2.09 0.78, 1.21 0.75, 1.63
Delivery System
Clin Info System
1.38*
0.58*
1.40, 1.67
0.42, 0.81
HLM Final Model
Predictor
Female All Maintenance Comm Linkages Delivery System
O.R.
0.59 1.82 1.56 1.47
95%C.I.
0.36, 0.98 1.08, 4.07 1.23, 1.98 1.17, 1.86
Clin Info System
0.58
0.44, 0.73
Conclusions
Control of CVD risk factors among patients with T2DM is associated with structural characteristics of primary care clinic:
Community
Linkages Delivery System Design Clinical Information Systems
Community Linkages
Linking clinicians to diabetes specialists and educators Patient diabetes education resources Coordinates implementation of diabetes care guidelines with assessment/treatment by specialists
Delivery System Design
Practice Team Functioning Practice Team Leadership Appointment System Follow-up Planned Visits for diabetes care Continuity and Coordination of Care
Clinical Information Systems
Inversely associated with CVD risk factor:
Diabetes
registry Reminders to providers Feedback on performance Identification of patients needing attention Patient treatment plans
CIS may improve measurement of risk factors but not efforts to control Implementation of CIS may distract from risk factor control
Limitations
Small number of primary care clinics Cross-sectional data Selection bias of consecutive patients
Bias
toward worse control of CVD risks Greater burden of illness Worse overall health status
Current/Future Research*
Organizational Intervention in Primary Care Clinics to improve risk factor control
Primary
care clinics are complex adaptive systems with non-linear dynamic behavior No “one-size-fits-all” approach to improving risk factors Facilitation of organizational change with a focus on inter-dependence among agents See Poster by Leykum et al this afternoon
*Funded by NIH/NIDDK 1 R34 DK067300-01
Acknowledgements
Supported by:
Agency
for Healthcare Research and Quality (Grant #K08 HS013008) South Texas Health Research Center Office of Research and Development, Health Services Research and Development Service, Department of Veterans Affairs. The views expressed are those of the authors and do not necessarily represent the views of the Department of Veterans Affairs