Use of Quality Improvement Strategies Among Small to Medium-Size US Primary Care Practices

Bijal A. Balasubramanian; Miguel Marino; Deborah J. Cohen; Rikki L. Ward; Alex Preston; Rachel J. Springer; Stephan R. Lindner; Samuel Edwards; K. John McConnell; Benjamin F. Crabtree; William L. Miller; Kurt C. Stange; Leif I. Solberg

doi:10.1370/afm.2172

Article Figures & Data

Figures

Tables
Additional Files

Download figure
Open in new tab
Download powerpoint
Figure 1
Distribution of the CPCQ strategies scores among 1,181 EvidenceNOW practices.
CPCQ = Change Process Capability Questionnaire.

View popup

Table 1

Characteristics of Participating EvidenceNOW Study Practices (N = 1,181)

Characteristic	No. (%)
Practice factor
Practice size
Solo practice	241 (20.4)
2–5 clinicians	558 (47.2)
6–10 clinicians	174 (14.7)
≥11 clinicians	143 (12.1)
Practice ownership
Clinician owned	436 (36.9)
Hospital, health system	275 (23.3)
FQHC, RHC, IHS, federal^a	286 (24.2)
Other, none^b	133 (11.3)
Location^c
Urban core	733 (62.1)
Large town	171 (14.5)
Rural area	191 (16.2)
Suburban	86 (7.3)
Medically underserved area classification	410 (34.7)
External factor
Received external incentives in prior 12 mo	602 (51.0)
Participation in other demonstration projects⁴	369 (31.2)
Patient-centered medical home recognition	481 (40.7)
Part of an accountable care organization	452 (38.3)
Internal factor
Major disruption(s) in prior 12 mo⁵
No major disruption	433 (36.7)
One major disruption	412 (34.9)
More than one major disruption	246 (20.8)
Health information technology characteristics
Completely electronic health record⁶	1,030 (87.2)
Time on EHR
≤3 y	324 (27.4)
4–6 y	344 (29.1)
>6 y	365 (30.9)
ONC-certified EHR	972 (82.3)
Participation in Meaningful Use
Not participating	189 (16.0)
Stage 1 only	139 (11.8)
Stage 1 and stage 2	722 (61.1)
Physical location of data
Server in practice	436 (36.9)
In cloud with vendor	246 (20.8)
Health system data warehouse	229 (19.4)
Do not know where data reside	111 (9.4)
Shares patient health information electronically	762 (64.5)
Clinical laboratories incorporated as structured data	965 (81.7)
Use of EHR data for quality improvement, reporting
Produced quality report(s) in prior 6 mo	741 (62.7)
Report CQMs
Practice level	867 (73.4)
Clinician level	881 (74.6)
EHR vendor helps produce CQM reports
Yes, with restrictions	226 (19.1)
Yes, with no restrictions	287 (24.3)
No	163 (13.8)
Do not know	211 (17.9)
Outside organization helps EHR data capture	731 (61.9)
Ability to create CQM reports
No ability	206 (17.4)
In-house clinician or other staff member writes reports	305 (25.8)
Outside service/consultant writes reports	438 (37.1)
Routinely discuss clinical quality data
Never	115 (9.7)
Infrequently	294 (24.9)
Often	425 (36.0)
Not applicable, solo practice	83 (7.0)
Do not know	98 (8.3)
Use of at least 1 registry	781 (66.1)
Guidelines for CVD⁹ prevention
Not used or clinician agreement to use	390 (33.0)
Included in EHR prompts or standing orders	719 (60.9)
Guidelines for CVD management
Not used or clinician agreement to use	436 (36.9)
Included in EHR prompts or standing orders	673 (57.0)

CPCQ = Change Process Capability Questionnaire, CQM = clinical quality measure, CVD = cardiovascular disease, EHR = electronic health record, FQHC = federally qualified health center, IHS = Indian Health Services clinic, ONC = Office of the National Coordinator for Health Information Technology, RHC = rural health clinic.
Note: Percentages may not add up to 100% because of missing data. Only practices with complete CPCQ strategies scores are included. Additional characteristics of the practices involved in EvidenceNOW are located in the Supplemental Appendix, at http://www.annfammed.org/content/16/Suppl_1/S35/suppl/DC1/.
↵a Federally qualified health centers, rural health clinics, Indian Health Services clinics and Veterans Affairs, military, Department of Defense, or other federally owned practices.
↵b Practices with nonfederal, private/nonclinician, academic, or tribal ownership, those indicating “other” without specifying an ownership type, and practices responding no to every ownership type.
↵c Determined using rural-urban commuting area codes.
d State Innovation Models Initiative, Comprehensive Primary Care Initiative, Transforming Clinical Practice Initiative, community health worker training program, Blue Cross Blue Shield Patient-Centered Medical Home program, and Million Hearts.
e New EHR, new billing system, moved locations, staff turnover, and purchased by or affiliated with larger organization.

View popup

Table 2

Components of the Change Process Capability Questionnaire Strategies Scale (N = 1,181 practices)

The practice has used the following strategies to improve cardiovascular preventive care	Agree or Strongly Agree No. (%)
Changing or creating systems in the practice that make it easier to provide high-quality care	935 (79.2)
Providing information and skills training	891 (75.4)
Removal or reduction of barriers to better quality of care	880 (74.5)
Periodic measurement of care quality for assessing compliance with any new approach to care	816 (69.1)
Using teams focused on accomplishing the change process for improved care	808 (68.4)
Deliberately designing care improvements to make the care process more beneficial to the patient	799 (67.7)
Using opinion leaders, role modeling, or other vehicles to encourage support for changes	757 (64.1)
Providing to those who are charged with implementing improved care the power to authorize and make the desired changes	753 (63.8)
Reporting measurements of practice performance on CVD prevention measures for comparison with their peers	740 (62.7)
Setting goals and benchmarking rates of performance quality on CVD prevention measures at least yearly	683 (57.8)
Customizing the implementation of CVD prevention care changes to the practice	632 (53.5)
Deliberately designing care improvements so as to make clinician participation less work than before	623 (52.8)
Delegating to nonclinician staff the responsibility to carry out aspects of care that are normally the responsibility of physicians	548 (46.4)
Using rapid cycling, piloting, pretesting, or other vehicles for reducing the risk of negative results for introducing organization-wide change in care	492 (41.7)

CVD=cardiovascular disease.

View popup

Table 3

Adjusted Mean Difference in Change Process Capability Questionnaire Strategies Scores by Practice Characteristics Among 1,181 Small to Medium-Size Primary Care Practices

Characteristic	Multivariable Final Model R²= 22.4% Adjusted Mean Difference
Practice characteristics
Practice ownership (Ref: clinician-owned)
Hospital/health system	−3.32^a
FQHC, RHC, IHS, federal^c	−1.49
Other, none^d	0.62
Missing	0.96
Location^e [Ref: urban core]
Large town	1.29
Rural area	0.65
Suburban	−2.13
Medically underserved area classification (Ref: No)
Yes	−1.52
Missing	1.46
External factors
Received external incentives in prior
12 mo (Ref: No)
Yes	0.96
Missing	−1.43
Participation in other demonstration projects^f (Ref: No)
Yes	1.24
Missing	3.10
Part of an accountable care organization (Ref: No)
Yes	2.74^a
Missing	2.23
Internal factors
Major disruption(s) in prior 12 mo^g (Ref: no disruption)
1 major disruption	−0.22
>1 major disruption	−2.68^b
Missing	0.33
Completely electronic EHR (Ref: No, part paper)
Yes	−3.59^b
Missing	−1.54
Produced quality report(s) in prior 6 mo (Ref: No)
Yes	5.09^a
Missing	2.96
Report CQMs at practice level (Ref: No)
Yes	2.21
Missing	5.09^b
Ability to create CQM reports from EHR (Ref: No ability)
In-house clinician or other staff	2.26^b
Outside service/consultant	0.76
Missing	0.18
Routinely discuss clinical quality data (Ref: Never)
Infrequently	1.87
Often	5.56^a
Not available, solo practice	4.40^b
Don’t know	3.81^b
Missing	4.21^b
Use of at least 1 registry (Ref: No)
Yes	3.74^a
Missing	2.22
Guidelines for CVD prevention or management^h (Ref: Not used or clinican agreement to use)
Included in EHR prompts or standing orders	4.45^a
Missing	0.41

AIC=Akaike Information Criterion; CQM=Clinical Quality Measure; CVD=cardiovascular disease; EHR = electronic health record; FQHC = federally qualified health center, IHS = Indian health services clinics; Ref = reference; RHC = rural health clinic; VA = Veterans Affairs.
Note: Final multivariable model was selected using backward model selection by the AIC. The model with the lowest AIC was preferred regardless of statistical significance for individual covariates.
↵a P value <.001.
↵b P value <.05.
↵c Includes FQHCs, RHCs, IHS, and VA, military, Department of Defense or other federally owned practices.
↵d Includes practices with nonfederal, private/nonclinician, academic, or tribal ownership, those indicating “other” without specifying an ownership type, and practices responding “no” to every ownership type.
↵e Location categories determined using rural-urban commuting area (RUCA) codes.
↵f Other demonstrations programs include State Innovation Models initiative, Comprehensive Primary Care Initiative, Transforming Clinical Practice Initiative, community health worker training program, Blue Cross Blue Shield Patient-Centered Medical Home program, and Million Hearts.
↵g Major disruptions include new EHR, new billing system, moved locations, staff turnover, and purchased/affiliated with larger organization.
↵h Guidelines for CVD prevention and management were combined to avoid multicollinearity. A practice was categorized as “Included in EHR prompts or standing orders” if they denoted that in either CVD prevention or CVD management or both.