|Getting to first base|
The Population Health Blog likes it too. But first, let's look at the study itself.
The Taconic Independent Practice Association participated in a multi-commercial insurer medical home program. For each patient enrolled in an NCQA Level 3 primary care home, the IPA was offered $2 to $10 per patient per month.
Since not all of the 675 IPA practices created medical homes in the 2008-2010 time frame, researchers were able to compare the quality of care for the approximately 27,000 patients attributed to the medical homes versus the approximately 64,000 patients attributed to usual care settings. While they were at it, they also evaluated the quality impact of having an electronic health record (all medical home practices had one, but not all users of electronic health records were medical homes) vs. having a paper record. Quality was determined by examining a subset of NCQA-based measures of recommended routine screening tests and care of persons with diabetes, asthma and (for children) pharyngitis.
While the impact of being in a medical home was not a home run, it definitely got on base.
Over the three years of the study, the medical home practices' measures diverged from the other two groups by an average of a 7% vs. paper and 6% vs. EHR. While none of the practices hit 100%, and not all of the differences were statistically significant, if an EHR-based clinic was able to hit (for example) a "72%" quality measure, the medical home was"79%" (mammogram screening).
As one more gauge of its impact, as time went on, the spread of the data grew over time as medical homes outpaced the usual care practices.
The authors correctly point out that the study was observational and that there could be hidden factors other than the presence or absence of a medical home that could be biasing the data. While the authors did everything they could to statistically "neutralize" the impact of unequally distributed patient or practice factors, the study process is not perfect.
This was also in a commercial insurance setting. We don't know if Medicare patients would gain the same benefit.
The PHB will also point out that the NCQA's one-size-fits-all measures are intermediate in nature and are imperfectly tied to long term outcomes. They also fail to account for patient preferences.
On the other hand, the PHB likes this because it was a huge "real world" study involving hundreds of clinical practices taking care of tens of thousands of patients over three years. It helps the PHB put things in perspective by showing that the team-based care management of the medical home can have a discernible impact on quality.
Unfortunately, there is nothing on the cost of care. The PCMH is still struggling to prove that it "saves" money; the PHB says health care consumers will get what they pay for and - aside from a selected high-risk subpopulation - the medical home offers high value at a reasonable additional cost. The commercial insurers got their money's worth from the Taconic IPA.
While the authors don't specifically call it out, there doesn't appear to have been a large impact by the EHR on the quality of care in the real world. Compared to practices with paper records, the impact of the EHR seemed to be scant.