Care teams with nurse practitioners and physician assistants achieved similar outcomes on 13 of 20 measures to physician-only teams when caring for patients with diabetes whose conditions were not highly complex, study findings indicate.

However, diabetes patients with more complex conditions tended to experience worse outcomes in several areas when care was supplemented by NPs and PAs versus physicians only, Christine Everett, PhD, PA-C, MPH, professor in the physician assistant program at Duke University School of Medicine in Durham, N.C., and colleagues reported in Health Affairs.

For example, the rate of ED visits among diabetes patients without highly complex conditions was 30% lower for the NP and PA care teams compared with physician-only teams (odds ratio [OR] for emergency visits 0.70; 95% CI: 0.56-0.93). But among highly complex diabetes patients, ED visits were 50% higher for NP/PA teams compared with physician-only teams (OR 1.5; 95% CI: 1.06-2.03).

Continue Reading

Although previous studies have shown that NPs and PAs can control diabetes just as well as physicians alone, an understanding of the roles of particular providers in team-based care models is not yet well established. To get a better idea of this aspect of care, Everett and colleagues examined data from 2,500 Medicare patients aged 23 to 102 years with diabetes who were treated at a single mulch-speciality physician group in 2008. 

Of the 261 primary care teams studied, 55% had team care with NPs and PAs. Another 39% had only a physician, whereas 5% consisted solely of an NP or PA. The Johns Hopkins Ambulatory Care Group System Predictive Model was used to define highly complex patients.

The researchers found differences between outcomes for highly complex patients if they were treated by teams with NPs and PAs compared with physicians alone.

Noncomplex patients treated by teams that included NPs and PAs were 54% less likely to have poor versus good glycemic control (OR 0.46; 95% CI: 0.22-0.97). Yet among those with highly complex conditions, the odds for poor versus good glycemic control were 1.8 times greater for those treated by NP and PA care teams than physician-only care teams.

“This raises the question of whether a team approach that divides primary care delivery between clinicians would work for all patient populations, particularly the most clinically complex patients,” the researchers wrote. “Such patients may be best served through a continuous relationship with a single primary care clinician.”

They added that determining the optimal role for clinicians within primary care teams “will require an even more nuanced approach than that taken in the current analysis,” and should take into account patient characteristics.


  1. Everett C et al. Health Aff. 2013; doi:10.1377/hlthaff.2013.0506.