NEJM: EHRs associated with high diabetes care
In research published in a special article in the New England Journal of Medicine, EHR sites were associated with higher levels of achievement and improvement in regionally vetted standards for diabetes care and outcomes.
Randall D. Cebul, MD, from the departments of medicine and epidemiology and biostatistics, and the Center for HealthCare Research and Policy, Case Western Reserve University at MetroHealth Medical Center in Cleveland, and colleagues analyzed data from a retrospective cohort of primary care practices from seven healthcare organizations that publicly reported quality standard achievements for adults with diabetes between July 2007 and June 2010. The research examined the independent association of EHR use with quality standard achievements in the care of patients with diabetes.
Clinical practices or healthcare organizations submitted a unique study code for each patient and healthcare provider into a data management center. EHR-based organizations provided data on all eligible patients. The data from paper-based organizations were gathered by trained chart abstractors for a random sample of patients selected by the data management center.
More than 95 percent of patients in the sample had charts available for review. Each site reviewed summary results for accuracy before publicly reporting the data.
"Our primary analyses examined cross-sectional practice-level achievements for the period from July 2009 through June 2010 and longitudinal changes since the July 2007–June 2008 period for each participating practice (in EHR-based organizations) or paper-based organization,” the authors wrote. “Unadjusted comparisons of EHR-based organizations with paper-based organizations were made with the use of the weighted least-squares method and practice-level data. For multivariate modeling, practice data were disaggregated into four insurance types (e.g., Medicare, commercial, Medicaid and uninsured).”
From July 2009 through June 2010, 27,207 adults with diabetes received care from 569 primary care providers in 46 practices at seven healthcare organizations. “In unadjusted analyses, between July 2009 and June 2010, 50.9 percent of patients at EHR sites, as compared with 6.6 percent of patients at paper-based sites, received care for diabetes that met all four standards, representing a difference of 44.3 percentage points,” the authors found. EHR sites showed higher achievement on all components of the composite standard for care.
For diabetes outcomes, the authors noted that 43.7 percent of patients at EHR sites and 15.7 percent of those at paper-based sites had outcomes that met at least four of the five standards, a difference of 28 percentage points. “Achievement was higher for EHR sites on all but one outcome standard. These findings were similar but somewhat blunted in analyses that adjusted for insurance type, age, sex, race or ethnic group, language preference, estimated household income, and educational level,” researchers said.
For diabetes care, the adjusted difference between EHR and paper-based sites was 35.1 percentage points across all practices and 29.8 percentage points at safety-net sites. For diabetes outcomes at all sites, the adjusted difference between EHR and paper-based sites was 15.2 percentage points; for safety-net sites, the difference was 9.7 percentage points.
“Our findings focus on composite standards, although the results were similar for virtually all component standards," the authors concluded. “Because the study was observational, it may be subject to selection bias, although our results were similar after adjustment for the more favorable socioeconomic profiles of patients cared for in organizations with EHRs. The association of type of medical record with quality standards was significant across all insurance types.”
Randall D. Cebul, MD, from the departments of medicine and epidemiology and biostatistics, and the Center for HealthCare Research and Policy, Case Western Reserve University at MetroHealth Medical Center in Cleveland, and colleagues analyzed data from a retrospective cohort of primary care practices from seven healthcare organizations that publicly reported quality standard achievements for adults with diabetes between July 2007 and June 2010. The research examined the independent association of EHR use with quality standard achievements in the care of patients with diabetes.
Clinical practices or healthcare organizations submitted a unique study code for each patient and healthcare provider into a data management center. EHR-based organizations provided data on all eligible patients. The data from paper-based organizations were gathered by trained chart abstractors for a random sample of patients selected by the data management center.
More than 95 percent of patients in the sample had charts available for review. Each site reviewed summary results for accuracy before publicly reporting the data.
"Our primary analyses examined cross-sectional practice-level achievements for the period from July 2009 through June 2010 and longitudinal changes since the July 2007–June 2008 period for each participating practice (in EHR-based organizations) or paper-based organization,” the authors wrote. “Unadjusted comparisons of EHR-based organizations with paper-based organizations were made with the use of the weighted least-squares method and practice-level data. For multivariate modeling, practice data were disaggregated into four insurance types (e.g., Medicare, commercial, Medicaid and uninsured).”
From July 2009 through June 2010, 27,207 adults with diabetes received care from 569 primary care providers in 46 practices at seven healthcare organizations. “In unadjusted analyses, between July 2009 and June 2010, 50.9 percent of patients at EHR sites, as compared with 6.6 percent of patients at paper-based sites, received care for diabetes that met all four standards, representing a difference of 44.3 percentage points,” the authors found. EHR sites showed higher achievement on all components of the composite standard for care.
For diabetes outcomes, the authors noted that 43.7 percent of patients at EHR sites and 15.7 percent of those at paper-based sites had outcomes that met at least four of the five standards, a difference of 28 percentage points. “Achievement was higher for EHR sites on all but one outcome standard. These findings were similar but somewhat blunted in analyses that adjusted for insurance type, age, sex, race or ethnic group, language preference, estimated household income, and educational level,” researchers said.
For diabetes care, the adjusted difference between EHR and paper-based sites was 35.1 percentage points across all practices and 29.8 percentage points at safety-net sites. For diabetes outcomes at all sites, the adjusted difference between EHR and paper-based sites was 15.2 percentage points; for safety-net sites, the difference was 9.7 percentage points.
“Our findings focus on composite standards, although the results were similar for virtually all component standards," the authors concluded. “Because the study was observational, it may be subject to selection bias, although our results were similar after adjustment for the more favorable socioeconomic profiles of patients cared for in organizations with EHRs. The association of type of medical record with quality standards was significant across all insurance types.”