1. October 29, 20121 Thomas T.H. Wan, Ph.D., MHS Maysoun Masri Dimachkie, Ph.D., MPH Judith Ortiz, Ph.D., MBA Blossom Yen Ju Lin, Ph.D. Jeffrey Harrison, Ph.D., MHA

2.  Health Care Reform  Benefits & Barriers in ACO Participation  Research Questions  Research Design  Propensity Score Matching & Analysis  Measurements  Path Analysis  Major Findings  Implications and Conclusions October 29, 20122

3. 3 Is Health Care Reform Needed?

4. 1) Ensure that all Americans have access to quality, affordable health care. 2) Create a new, regulated marketplace where consumers can purchase affordable health care. 3) Extend much needed relief to small business. 4) Improve Medicare by helping seniors and people with disabilities afford their prescription drugs. 5) Prohibit denials of coverage based on pre-existing conditions. 6) Limit out-of-pocket costs so that Americans have security and peace of mind. 7) Help young adults by requiring insurers to allowable dependents to remain on their parents’ plan until age 26. 8) Expand Medicaid to millions of low-income Americans. 9) Provide sliding –scale subsidies to make insurance premiums affordable. 10) Hold insurance companies accountable for how our health care dollars are spent. 11) Clamp down on insurance company abuses. 12) Invest in preventive care. October 29, 20124

5. 5 Focus on unit cost Focus on total cost Patient Centric Care

6.  Establishes partnerships among health care providers to coordinate and deliver efficient care  Assumes joint accountability for improving quality and slowing cost growth  Emphasizes patient-centered care to achieve optimal population health  Adopt innovative delivery systems October 29, 20126

7.  Benefits  Barriers October 29, 20127 Improvement in coordinated health care services Improvement in the patient care outcomes Reduction in the hospital's total costs. Increase in the hospital's profit margins Increase physicians and other healthcare providers' participation to improve health Other benefit Lack of leadership support or commitment Inadequate financial incentives Legal and regulatory barriers Fear of losing autonomy Lack of adequate capital for HIT infrastructure Inadequate number of PCPs and other specialists Inadequate number of patient population served by the facility Not having affiliated with any healthcare network

8.  How do we apply propensity score matching and analysis in health care innovation and diffusion research?  Does the benefit-barrier gap influence the plan for developing a strategic initiative in ACOs (Plan_ACOs) and the willingness to participate in ACOs (WP_ACOs)?  What are the personal, organizational and contextual factors influencing the Plan_ACOs and WP_ACOs? October 29, 20128

9.  Conventional PSM&A Approach:  Latent Variable Approach October 29, 20129

10. MeasureItems/ScaleMeanSD Cronback’s Alpha Coefficient Intra-Class Correlation Coefficient (Average) Organizational Social Capital 6/ Likert scale (0 to 24)18.4402.9390.8490.839 Electronic Medical Record Uses 5/ Yes-No (0 to 5)3.4601.6390.785 Systems Integration 5/ yes-no (0 to 5)3.4411.8190.847 Perceived Benefits 6/ yes-no (0 to 6)3.6551.5330.611 Perceived Barriers 8/ yes-no (0 to 8)1.7382.1060.800 Benefit-Barrier Gap Total benefits- total barriers 1.9172.406 Knowledge About ACOs 1/ Likert scale (1 to 4)2.6310.929 Health Networks 8/ yes-no (0 to 8)2.4172.169 Plan_ACOs 1/ Likert scale (1 to 4)2.7141.013 WP_ACOs 1/ Analog scale (0 to10)5.0483.523 October 29, 201210

11. Perceived Benefits Mean WP_ACOs With a Without benefit a benefit SD WP_ACOs With a Without benefit a benefit F-value Improvement in coordinated health care services 4.904 3.076 5.281 4.1990.225 (NS) Improvement in the patient care outcomes 5.010 3.054 5.107 4.8330.914 (NS) Reduction in the hospital's total costs. 4.794 3.117 5.810 4.5460.255 (NS) Increase in the hospital's profit margins 5.145 3.387 4.773 3.9630.174 (NS) Increase physicians and other healthcare providers' participation to improve health 5.032 3.212 5.095 4.4360.005 (NS) Other benefit 5.909 4.721 4.918 3.3320.753 (NS) Perceived Barriers Mean WP_ACOs With a Without barrier a barrier SD WP_ACOs With a Without barrier a barrier F-value Lack of leadership support or commitment 2.615 5.493 2.399 3.5297.930* Inadequate financial incentives 3.581 5.906 2.953 3.5749.362* Legal and regulatory barriers 3.636 5.548 3.215 3.5195.006* Fear of losing autonomy 4.000 5.206 3.098 3.5781.119 (NS) Lack of adequate capital for HIT infrastructure 4.313 5.221 2.960 3.6440.858 (NS) Inadequate number of PCPs and other specialists 5.136 5.016 2.916 3.7390.019 (NS) Inadequate patient population 4.474 5.215 3.356.5820.648 (NS) No affiliation with a healthcare network 4.750 5.097 2.896 3.6350.099 (NS) October 29, 201211 ANOVA of Mean WP_ACOs by Specific Benefits or Barriers

12. Predictors Standardized Regression Coefficient Unstandardized Regression Coefficient Standard Error T-value Plan_ACOs predicted by EMR uses -0.014-0.0090.067-0.129 Systems integration -0.042-0.0240.063-0.378 Org. social capital -0.053-0.0180.034-0.546 No. of health networks 0.188 0.0880.050 1.743 Knowledge about ACOs 0.055 0.0600.107 0.563 Benefit-barrier gap 0.381 0.381 0.161 0.1610.042 3.867* 3.867* Rural location -0.198-0.4050.211-1.917 Propensity score -0.159-0.6790.418-1.625 R-square value 0.261 October 29, 201212

13. WP_ACOs predicted by Standardized regression coefficient Unstandardized regression coefficient Standard error T-vale Benefit-barrier gap 0.042 0.0620.165 0.376 Plan_ACOs 0.343 0.343 1.192 1.1920.395 3.020* 3.020* Knowledge about ACOs -0.164-0.6220.387-1.608 Propensity score -0.040-0.5941.508-0.394 R-square value 0.152 October 29, 201213.

14. 1) Chi-square= 30.870 with 27 degrees of freedom and P=0.277; 2) GFI = 0.930, AGFI =0. 858, and CFI = 0.951; and 3) RMESEA = 0.042. October 29, 201214

15.  Benefit-barrier gap directly influences Plan_ACOs, but it indirectly influences WP_ACOs via Plan_ACOs.  Three major hurdles for ACOs:  Lack of leadership support or commitment  Inadequate financial incentives  Legal and regulatory barriers  Personal and organizational factors are more influential than contextual factors in predicting Plan_ACOs and WP_ACOs.  Limitations of the study: Response rate, sampling, and longitudinal study design.  Future research: PSM&A could be useful for evaluating the impact of ACO participation (see next diagram). October 29, 201215

16. October 29, 201216