Team-based care yields better outcomes for patients with chronic disease
Researchers used de-identified athenahealth data from primary care practices to answer both questions. Their findings have important implications for clinical staffing and for scope-of-practice changes, which are on the minds of policymakers who modified these rules last year to care for surges of COVID-19 patients.
Harvard study looked at athenahealth data from more than 2,000 providers
With support from the Robert Wood Johnson Foundation's Health Data for Action program, researchers at Harvard Medical School and Harvard Business School partnered with athenahealth to analyze data from a national sample of electronic health records on the athenahealth network from 2013 to 2018. Included were data from 2,028 providers who worked at 250 practices belonging to 175 organizations. “The athenahealth data is quite unique in that it covers a lot of practices across the country with a high level of clinical detail,” said Maximilian Pany, a co-leader of the study and MD-PhD candidate at Harvard Medical School.
The study, published in the March edition of Health Affairs, analyzed data for individual patients with new-onset type 2 diabetes, hypertension, or hyperlipidemia. The researchers analyzed data on sociodemographics, care processes such as diagnostics and prescribing, and the biomarker outcomes in terms of hemoglobin A1c, blood pressure, and LDL cholesterol.
The investigators sought to answer two questions about quality of care in chronic disease management. First, do care teams provide higher- or lower-quality care than solo providers? Second, do teams composed entirely of one level of provider – physicians versus midlevel practitioners (nurse practitioners or physician assistants) — yield different outcomes than teams that include a mix of these professions?
How the dataset improves the precision of this study
The Harvard study overcame obstacles inherent in other data sources like billing claims data, which often did not fully capture visits conducted by midlevel providers.
Using athenahealth EHR data, the researchers were able to consistently identify the profession of each service provider – physician, nurse practitioner, or physician assistant. The EHR data also included the very biomarkers by which outcomes one year after disease onset are judged: blood pressure, A1c, and LDL. The study also used EHR data on core care processes for disease management, such as diagnosis, prescribing of pharmaceuticals, and patient monitoring via labs.
Study findings reveal that teams of providers produce better outcomes
The research produced two particularly interesting and statistically significant results. For one, the study showed that for the three conditions studied, patients seen by teams of providers were more likely to bring their disease under control than were patients seen by solo practitioners. Some 63.9 percent of patients seen by teams brought their diabetes under control, versus 54.7 percent for diabetes patients seeing solo providers. For hypertension, 71.0 percent of patients seen by teams had a favorable outcome, versus 64.9 percent for patients of solo providers.
A second important finding: In terms of quality of care, it didn’t make much of a difference whether patients were treated by physicians, midlevel practitioners, or an interprofessional mix.
The results highlight the need for quality improvement
“Chronic disease care – even though the principles are simple – is really hard to do,” said Lucy Chen, a study co-leader and MD-PhD candidate at Harvard Medical School. This was confirmed by the “low absolute performance” of providers in the study, in line with long-established findings that frustrate stakeholders all around.
From their study, the researchers have some hypotheses about why quality improvements have been hard to come by. One is that team providers perform better simply because they have more sets of eyes on the case. “Blood pressure that’s a little high might go unnoticed because there are other things that the patient or provider wants to talk about,” said Chen.
Can changes in health policy promote teamwork to boost outcomes?
These research findings point to a challenge for both clinician executives and policymakers: “How can we encourage team-based care with mixed teams that might be able to deliver really good care cost-effectively?” as Pany puts it.
The researchers say that policymakers should consider interventions to encourage the formation of provider teams, such as new care delivery models and financial incentives for collaboration. The costs of innovation and new incentives might be partly offset by potential savings in provider compensation by staffing with more nurse practitioners and physician assistants and fewer physicians.
What’s next for Chen and Pany? “We’re planning to look under the hood of teams and try to understand mechanisms like team structure and communication among team members,” says Chen.