Medical Care blog

Comments from the scientific community regarding Medical Care.

Friday, April 3, 2015

Evaluating the ability of patient-centered medical homes to improve clinical outcomes

By Damika W. Barr, JD and Lisa M. Lines, PhD, MPH


Support for patient-centered medical homes (PCMH) has been growing in recent years.  A typical PCMH model includes a primary care physician supported by a multidisciplinary team, coordinated care, measurements of quality and safety, and enhanced access to same-day appointments. Preliminary program evaluations have shown that PCMHs can improve access and quality, but to date, there has been little evidence that the model improves clinical outcomes.  A new Medical Care paper by Shi and colleagues is one of the first to evaluate the relationship between PCMH model adoption in health centers and clinical process and outcome measures.


Using health center-level data from HRSA’s 2009 Uniform Data System (UDS) and the 2009 Commonwealth Fund National Survey of Federally Qualified Health Centers, researchers generated an analytic sample of 795 health centers, each serving an average of more than 18,000 patients. The population served included minority patients (49%), uninsured patients (40%), Medicaid patients (32%), managed-care patients (17%), and patients with chronic conditions (24%).

Clinical performance measures generated from UDS included two process outcomes: children receiving required vaccines by their 2nd birthday, and females between the ages of 24-64 receiving a Pap test in the last 3 years.  The outcome measures were the percent of adult patients with Type 1 or Type 2 diabetes who had their blood sugar levels under control and percent of adult patients with hypertension who had their blood pressure under control.


The results were mixed. As shown in the figure below, total PCMH score was associated with an increase in the percentage of diabetics with HbA1c of 7% or less. Improvements in access and communication were associated with increases in the percentage of diabetics with HbA1c of 7% or less and HbA1c of 9% or less. Improvements in care management led to improvements in the percent of women receiving Pap tests. Improvements in coordination of care were associated with improvements in the percent of children receiving immunizations. None of the PCMH improvements were associated with improvements in hypertension (HTN) control.



Oddly, in multivariate analyses, practices that increased their use of patient tracking and registries had significantly worse results on three of the outcome measures - a discouraging finding for proponents of health IT. Less surprisingly, practices with larger proportions of uninsured patients, Medicaid patients, and patients with chronic conditions had worse results on some of the performance measures, highlighting the need for risk adjustment in performance measurement.  


The authors point to the findings related to access and communication on influencing clinical outcomes to highlight importance of keeping the focus on the patient and the need for improved targeting of important subpopulations.


A key limitation of this study is data.  The study used 2009 data, when there were fewer PCMHs in the US.  The limited findings may suggest that more time is needed to assess the impact of PCMH on performance. Unfortunately, the Commonwealth Fund survey provides only cross-sectional data. Future evidence from ongoing efforts, such as the Multi-payer Advanced Primary Care Practice demonstration, will allow researchers to track the longitudinal effects of primary care improvements on clinical outcomes.   


This study provides new evidence about PCMHs, highlighting both the difficulty of practice transformation and the promise of better approaches to primary care.


Read the full article to learn more about the methods, results, and future research needs.


Damika Barr is a Senior Policy Analyst at AmeriHealth Caritas, and Lisa Lines is a health services researcher at RTI International. The views expressed on this post are those of the authors and do not necessarily reflect the views of AmeriHealth Caritas or RTI International.