Academy Health-Annual Research Meeting Presentation

Massachusetts Patient-Centered Medical Home Initiative (MA PCMHI): Impact on Clinical Quality at 30 Months

Judith Steinberg, M.D., M.P.H.Sai Cherala, M.D., M.P.H. Ann Lawthers, S.M., Sc.D. Christine Johnson, Ph.D.

Commonwealth Medicine UMass Medical School

Introduction The Patient‐Centered Medical Home (PCMH) offers an

innovative model of care: comprehensive primary care, quality improvement, care management, and enhanced access in a patient centered environment

PCMH evaluations have shown variable impact

Aims:

To assess data trends of clinical quality measures from participating practices in the Massachusetts Patient-Centered Medical Home Initiative

To evaluate practice and staff level factors that may impact clinical quality performance

Background: Massachusetts Patient Centered Medical Home Initiative

Multi-payer, statewide initiative

Sponsored by Massachusetts Health & Human Services; legislatively mandated

46 participating practices

3-year demonstration: March, 2011 − March, 2014

Includes payment reform and technical assistance

MA PCMHI Evaluation Questions

Question 1:

To what extent and how do practices become medical homes?

• Extent

• Patient-family centeredness

• Care management

• Care coordination

• Access

• Teamwork

• Information technology

• Leadership

• Barriers and Facilitators

Question 2:

To what extent do patients become partners in their health care?

•Perceived self-management efficacy

•Patient-family centeredness by chronic and non-chronic

Question 3:

What is the initiative’s impact on utilization, cost, clinical quality, patient and provider outcomes?

•Emergency Department use

•Hospitalizations

•Cost

•Clinical quality measures

•Staff satisfaction

•Patient satisfaction

Methods

Design: Quality improvement study using practices’ self-reported monthly data on 22 clinical quality measures from June 2011 through February 2014

Methods

Linear Mixed Model

Analysis

Data were divided into three-month periods:

Time 1 (June – August, 2011) ... to Time 11 (December, 2013 – February, 2014)

Analysis of Change over Time: Baseline (Time 1 or Time 2 or Time 6) vs. Time 11

Clinical Quality MeasuresAdult Diabetes

HbgA1c Control (<8%) HbgA1c Control (>9%) BP < 140/90 mmHg LDL Control < 100mg/dL Screened for Depression Self-Management Goal

Adult Prevention Adult Weight Screening and

Follow-up Tobacco Use Assessment Tobacco Cessation Intervention

Other Adult Target Blood Pressure Control Hypertension with Documented

Self-Management Goal Depression with Documented

PHQ-9 Score Depression with Documented Self-

Management Goal

Childhood Prevention Immunization Status Multiple

vaccines Weight Assessment and Counseling

for Children and Adolescents

Pediatric Asthma Use of Appropriate Medications for

Asthma Persistent Asthma Patients with

Action Plan

Other Pediatric Target Follow-up Care for Children

Prescribed ADHD Medication Management Plan for Children

Prescribed ADHD Medication

Care Coordination/ Care Management Follow-up after Hospital Discharge Highest Risk Patients with Care Plan

Results: Study Participants Practice Characteristics Percentage

Geography

Rural (<10,000 town population) 9%

Suburban (Between 10,000 and 50,000) 20%

Urban (>= 50,000) 71%

Practice Size (Based on Number of Full Time Practitioners)

Small (< 6 FTE practitioners) 31%

Medium (Between 6 and 11 FTE practitioners) 29%

Large (> 11 FTE practitioners) 40%

Type of Practice

Community Health Center 56%

Residency or Academic Practice 11%

Group Practice 29%

Solo Practice 4%

Payer Mix (Practices with Financial Incentives N=31)

Commercial 12%

Health Safety Net 15%

Medicaid 72%

Medicare 1%

Clinical Quality Measures that Showed Significant Improvement in Change over Time

25.2 23.8

37.1

82.4

46.5

16.7 17.3

11.5

18.6

46.4

22.3

36.1

48.7

32.0

47.6

90.5

51.3

25.321.4

19.3

62.7 63.161.2

64.7

0

10

20

30

40

50

60

70

80

90

100

Screened forDepression

Self-Management

Goal

Adult WeightScreening &Follow-Up

Tobacco UseAssessment

TobaccoCessation

Intervention

HypertensionSelf-

ManagementGoal

DepressionPHQ-9 Score

DepressionSelf-

ManagementGoal

Patients WithAction Plan

ImmunizationStatus

MultipleVaccines 1

ImmunizationStatus

MultipleVaccines 2

Care Plans forHighest Risk

Patients

Pe

rce

nt

Baseline Time 11

11 of 22 measures showed statistically significant improvement

Adult Diabetes Adult Prevention Other Adult Measures Pediatric

Asthma

Childhood

Prevention

Care

Management

Values met the study’s definition of statistical significance p<.05.

Care Coordination/Care Management Measures: Change over Time

63.3

36.1

66.5 64.7

0

10

20

30

40

50

60

70

Follow-Up After Hospital Discharge Care Plans for Highest Risk Patients

Ave

rage

Rat

e

Measures

Baseline Time 11

Drilling Deeper on Change: Methods

Correlation analysis

Variables included: Performance on clinical measures in the last three months, change in clinical performance over the demonstration, practice characteristics and staff perceptions/attitudes towards the change

Data sources: Clinical data submission, Medical Home Implementation Quotient (MHIQ), staff survey

Results of Correlation Analysis: Care Plan for Highest Risk Patients

Change over Three Years Performance in Last Three Months

Leadership at Baseline(staff survey) (r=0.42, p=0.01)

Leadership at Baseline(staff survey) (r=0.45, p=0.009)

Strong team(staff survey) (r=0.41, p=0.01)

Quality improvement culture at Baseline(staff survey) (r=0.39, p=0.02)

Quality improvement culture at Baseline(staff survey) (r=0.36, p=0.04)

Strong team(staff survey) (r=0.32, p=0.05)

Processes and Practices Characteristics Associated with Clinical Outcomes

HbA1c < 8% BP <140/90mm of Hg

Comfort with HIT (staff survey) (r=0.61, p <0.0001)

Improved care planning for high risk patients (MHIQ) (r=0.50, p=0.005)

QI culture(staff survey) (r=0.51, p <0012)

Strong leadership at baseline(staff survey) (r=0.44, p=0.009)

Strong teamwork (staff survey) (0.50, p=0.002)

Strong teamwork(staff survey) (r=0.41, p=0.01)

Leadership (staff survey) (r=0.48, p=0.002)

QI culture(staff survey) (r=0.38, p=0.02)

Quality Improvement Study

No Comparison Group

Small Sample Size

Correlation Analysis

Limitations

At the close of the MA PCMHI initiative (3 years), 11 of 22 clinical measures showed statistically significant improvement

Measures that showed significant improvement: Process measures

New or newly documented processes

A solid practice QI culture, leadership and strong team functioning were positively correlated with performance and improvement in high risk care planning

Factors that correlated with performance on clinical outcome measures were: QI culture, strong leadership and teamwork, comfort with HIT

Summary

Conclusion and Implications for Policy and Practice

Quality of care in the management of chronic diseases, prevention and screening, and high risk care management was significantly improved in this PCMH demonstration that had a preponderance of safety net practices

Implementation of foundational elements of the PCMH − QI, leadership engagement, teamwork and HIT − may foster improvement in clinical quality

Understanding factors that are correlated with clinical performance can focus transformation efforts

Acknowledgments

We would like to acknowledge the Massachusetts Executive Office of Health and Human Services (EOHHS), the MA PCMHI Leadership and Medical Home Facilitator Teams, as well as MA PCMHI participating practices without whom this work would not be possible.

Contact Information:

Judith Steinberg, M.D., M.P.H.Deputy Chief Medical OfficerCommonwealth MedicineUMass Medical [email protected]

Sai Cherala, M.D., M.P.H.Senior Clinical AnalystCommonwealth MedicineUMass Medical [email protected]