“Look before you leap” might be the saddest phrase one can hear in free-fall, but it seems to be the advice given by the Agency for Healthcare Research and Quality regarding the patient-centered medical home.
“Primary care clinicians, health care systems, insurers, state governments, families, and communities” see PCMHs as a solution to many of the problems besetting the system, the AHRQ says in a report from Mathematica Policy Research, “Improving Evaluations of the Medical Home.” (http://www.mathematica-mpr.com/publications/PDFs/health/medicalhome_brief.pdf). So much so that many stakeholders have already begun undertaking PCMH programs.
AHRQ reminds us that as implementers move forward, they must collect good data to direct future investments. “Strong evaluations are critical in determining whether the PCMH model works and for finding ways to refine, improve, customize, and disseminate the model if it does.”
Deborah Peikes, senior health researcher at Mathematica, says, “With such a wide variety of patient-centered medical home models, gathering and assessing the evidence on what’s working and where challenges arise is critical. The [report] describes why and how to commission effective evaluations. It also discusses which outcomes to assess, why to include comparison practices, and the importance of accounting for clustering of patients within practices.”
MANAGED CARE has long noted both the excitement and skepticism surrounding the PCMH model (/archives/2010/8/lessons-learned-building-patient-centered-medical-home). We’ve also reported on how some health plans have not hesitated to take up the challenge of designating a primary care doctor as chief care coordinator (/archives/2011/2/medical-home-common-ground-bcbsnc-and-provider-group).
The AHRQ and Mathematica Policy Research say that studies should account for clustering and focus more on increasing the number of practices rather than the number of patients.
“So that estimates of the effectiveness of interventions that alter the entire practice such as PCMH are not inflated, statistical corrections must be made for the degree to which patients in a practice tend to be more similar to each other than to patients in other practices (clustering),” states the report “Building the Evidence Base for the Medical Home: What Sample and Sample Size Do Studies Need?” (http://www.mathematica-mpr.com/publications/PDFs/health/evidencebase_medicalhome.pdf).
The white paper also says that having more patients per practice does little for the overall study and that “it is better to have many practices with few patients per practice than few practices with many patients in each. For example, a study with 100 practices and 20 patients per practice has much greater power than a study with 20 practices with 100 patients each.”