A Positive Deviance Approach to Understanding Key Features to Improving Diabetes Care in the Medical Home

Robert A. Gabbay; Mark W. Friedberg; Michelle Miller-Day; Peter F. Cronholm; Alan Adelman; Eric C. Schneider

doi:10.1370/afm.1473

Article Figures & Data

Tables

Additional Files

View popup

Table 1

Diabetes Measures, Demographics, and Survey Response Rates in Higher- and Lower-Performing Practices

Quintile and Practice	Absolute % Change at 18 Months			Improvement Index^a at 18 Months	First NCQA Level^b (2008–2009)	Type of Practice	Size of Practice^c	Adaptive Reserve/Burnout Survey Response Rate, %

	HbA_1c <7%	BP <130/80 mm Hg	LDL-C <100 mg/dL
Higher performing
Practice A	15.8	35.1	14.4	21.7	2	Private	Small	72
Practice B	13.5	20.7	20.3	18.2	1	FQHC	Small	75
Practice C	12.5	12.6	10.2	11.8	3	Private	Small	71
Practice D	0.8	11.9	20.3	11.0	3	Private	Medium	97
Practice E	1.5	17.3	9.3	9.4	3	Private	Medium	12
Average	8.8	19.5	14.9	14.4	2.4	–	–	61
Lower performing
Practice U	−12.1	−4.1	−8.7	−8.3	1	FQHC	Small	63
Practice V	−10.0	−6.4	−10.8	−9.0	1	Health system	Medium	18
Practice W	−9.2	−17.7^d	−7.7	−11.6	3	Private	Solo/partner	58
Practice X	−9.6	−11.2	−14.2	−11.7	2	Private	Medium	39
Practice Y	−18.1	−2.1	−24.7	−15.0	1	Private	Solo/partner	73
Average	−11.8	−8.3	−13.2	−11.1	1.6	–	–	44

HbA_1c = glycated hemoglobin; BP = blood pressure; LDL-C = low-density lipoprotein cholesterol; NCQA = National Committee for Quality Assurance; FQHC = Federally Qualified Health Center.
Note: Practices were classified as higher or lower performing at 18 months (December 2009) as measured by the improvement index. Source: Clinical and NCQA data submitted by practices to the Improving Performance in Practice program.
↵a Calculated for each practice as the arithmetic mean of the absolute percent improvement in the 3 outcomes from baseline to 18 months.
↵b First NCQA Physician Practice Connections Patient-Centered Medical Home (PPC-PCMH) recognition level; possible levels range from 1 (lowest) to 3 (highest).
↵c Practice size categories were based on the number full-time equivalent (FTE) clinicians as solo/partner (1–2 FTE clinicians), small (3–4), or medium (5–9).
↵d Calculated from baseline to January 2010 because of an obvious data error in December 2009.

View popup

Table 2

Structural and Organizational Characteristics of Higher- and Lower-Performing Practices

	Number Higher Performing (5 Practices)^a		Number Lower Performing (4 Practices)^a

Characteristic	Preintervention	Postintervention	Preintervention	Postintervention
Practices in which clinicians use a shared communication system to contact diabetic patients who...
...are due for HbA_1c testing	3	4	0	3
...are due for cholesterol testing	3	4	0	3
...are due for eye examination	2	4	0	3
...are due for nephropathy monitoring	2	4	0	3
...have not had an appointment in the practice for an extended period (longer than clinically appropriate)	1	4	1	2
EHR use
Number of EHR features present, median (range)^b	11 (0–18)	16 (14–19)	5 (0–7)	14 (9–17)
Using an EHR	4	5	2	5
Have staff to support diabetic patients^c	3	5	1	3

HbA_1c=glycated hemoglobin; EHR=electronic health record.
Source: Survey of practice leaders.
↵a Preintervention survey responses reflect practice characteristics in 2008. Postintervention survey responses reflect practice characteristics in 2011.
↵b Out of 20 possible features.
↵c Presence of specially trained nonphysician staff who help patients better manage their diabetes.

View popup

Table 3

Key Elements Distinguishing Higher- and Lower-Performing Practices

Element	Higher-Performing Practices	Lower-Performing Practices
Managing competing demands to medical home implementation
Technology	Most had existing EHRs	Most installed new EHRs during medical home implementation
Finances	Had stable financial systems and processes	Had less stable financial systems and processes
Leadership and vision
Shared vision and buy-in	Champions emphasized the need for all practice members to be on board with the initiative Careful articulation and reinforcement of how the medical home will help patients and the practice and the need for changes	Little to no dissemination of information about the motivations for joining the initiative Confusion about changing roles, uncertainty about processes and expected outcomes
Deliberate planning and testing of changes	Careful, deliberate plan of action, starting slowly with diabetic patients only and with 1 clinician and 1 office staff trying out novel methods and working out the kinks before implementing across the practice	Inconsistent roll-out of methods
Building teams and resource capacity
Sense of team	Collective problem solving and shared decision making High levels of trust, respect, and collaboration Regular multidisciplinary meetings and communication	Top-down approach to decision making Less clarity on roles and responsibilities Noninclusive approach to meetings and communication
Cultivating human resources	Strategic development of team in terms of composition and education/training Expansion of the role of the medical assistant Relatively stable staffing	Less effort to form an integrated team and insufficient education/training for staff Role of the medical assistant remains more limited Moderate to high staff turnover
Monitoring progress and obtaining feedback
Feedback systems	Systematic ongoing processes to solicit and share feedback	Feedback was not systemic; lack of opportunity to provide feedback; little dissemination of feedback
Benchmarking	Data shared across practice regularly; stimulates changes and healthy competition among clinicians	Data not shared regularly or widely
Planning and implementation of changes	Shared planning and decision making regarding changes	Unclear processes in terms of who is involved and what procedures in place to implement changes

EHR=electronic health record.
Source: Site visit observations and semistructured interviews.

View popup

Table 4

Key Factors Supporting Higher-Performing Practices

Key Factor	Description
Health information technology	Early adoption of EHRs (4 of 5 higher-performing practices had EHRs in place =2 years before PCMH implementation)
Administrative leadership	Highly engaged practice administrators who championed the PCMH transformation
Clinician leadership	Regular clinician meetings to discuss performance, agree on clinical guidelines, and establish standards of care
Shared vision and buy-in	Careful articulation and reinforcement of how the medical home will help patients and the practice and the need for changes
Staff development	Team orientation and early development of medical assistant role
Focus on improvement	Meetings revolve around PCMH and clinical quality improvement
Shared decision making	Feedback from practice consistently sought on changes before, during, and after implementation
Accountability	Clear roles and responsibilities and accountability to these roles and responsibilities
Finances	Stable billing and administrative systems
Financial autonomy	Direct receipt of and ability to invest PCMH financial incentives
Benchmarking	Monthly clinician-specific benchmarking to identify best practices and breakdowns in PCMH processes
Reporting and documentation	Careful attention to data reporting and documentation of PCMH changes
Inclusivity	Collective problem solving and open communication
Staff stability	Minimal staff turnover

EHR = electronic health record; PCMH = patient-centered medical home.
Note: A variety of factors supported PCMH implementation in the higher-performing practices.
Source: Site visit observations and semistructured interviews in the higher-performing practices.

Additional Files

Tables

Supplemental Appendix

Supplemental Appendix. Key Contextual Factors and Noteworthy Contextual Changes

Files in this Data Supplement:

Supplemental data: Appendix - PDF file, 3 pages, 139 KB