Diabetes Decision Tool Yields Benefit in Low-Resource Clinics

Diabetes Decision Tool Yields Benefit in Low-Resource Clinics

Adding a clinical decision support system (CDSS) to team-based diabetes care only modestly improved patients’ cardiovascular risk factors over team-based care alone, a randomized trial in China showed.

The tool required clinicians to enter patient data into a computer in order to generate individualized treatment recommendations, adding to their administrative burdens. It also couldn’t tackle patients’ problems with access and affordability of medications.

Nevertheless, the model could curtail physician burnout and improve the quality of care in primary care clinics with limited resources, the researchers said in a paper published in the Annals of Internal Medicine.

They concluded that the findings support “widespread adoption” of the model in China and other low- or middle-income countries where diabetes is on the rise.

Co–principal investigator Jiang He, MD, PhD, chair of epidemiology at Tulane University, New Orleans, said the findings could apply to federally qualified health care (FQHC) clinics that treat underserved patients in the United States.

“At many FQHC clinics, nurse practitioners have to take care of patients with multiple chronic disease conditions. Team-based care with a computerized clinical decision support system will help them and improve patient care,” He said.

Small Improvements

To conduct the trial, called Diabetes Complication Control in Community Clinics (D4C), He and colleagues randomly assigned 19 out of the 38 community health centers in Xiamen, China, to have a clinical decision support tool installed on the computers of primary care physicians and health coaches.

Starting in October 2016 the researchers recruited 11,132 patients aged 50 and older with uncontrolled diabetes and at least one comorbid condition, with 5,475 patients receiving team-based care with the CDSS and the remainder receiving team-based care alone.

The CDSS generated individualized risk factor summaries and treatment recommendations, including prescriptions based on Chinese and U.S. clinical guidelines. It incorporated data on patients’ insurance plans and local availability of drugs.

At all centers, primary care physicians received training in managing glycemia, blood pressure, and lipids. Nurses were certified as health coaches after receiving training on nutrition, lifestyle changes, and medication adherence. Patients met with their coaches for half an hour every 3 months, and diabetes specialists visited each clinic monthly for team meetings and consultations.

After 18 months, patients undergoing team-based care alone lowered their hemoglobin A1c by 0.6 percentage points (95% confidence interval, –0.7 to –0.5 percentage points), LDL cholesterol by 12.5 mg/dL (95% CI, –13.6 to –11.3 mg/dL), and systolic blood pressure by 7.5 mm Hg (95% CI, –8.4 to –6.6 mm Hg).

The group whose care teams used the CDSS further reduced A1c by 0.2 percentage points (95% CI, –0.3 to –0.1 percentage points), LDL cholesterol by 6.5 mg/dL (95% CI, –8.3 to -4.6 mg/dL), and blood pressure by 1.5 mm Hg (95% CI, –2.8 to –0.3 mm Hg).

All-cause mortality did not differ between the groups. Serious adverse events occurred in 9.1% of the CDSS group, compared with 10.9% of the group whose care team did not use the CDSS.

Addressing Social Needs

Experts who were not involved in the trial said the marginal impact of the CDSS was no surprise given the mixed results of such tools in previous studies.

However, the lackluster result “might be a shock to people investing a lot in clinical decision support,” said Elbert Huang, MD, MPH, director of the Center for Chronic Disease Research and Policy at the University of Chicago.

Anne Peters, MD, a professor of medicine at the University of Southern California, Los Angeles, said the administrative burden of entering each patient’s data into the system would slow down care and frustrate clinicians. “The system has to be smarter than this.”

On the other hand, the findings of the D4C trial align with other research showing that team-based care strategies are effective for diabetes management.

Huang noted that there is a “well-established history” of diabetes quality improvement programs, health coaches, buddy programs, and community health worker programs. He added that the new findings “might help to remind everyone of the importance of these programs, which are not always well supported.”

“The bottom line of the paper might be that investing in patient engagement programs might get us 90% of the way to our goal of improving diabetes care,” Huang said.

Still, Peters said the portion of patients in the trial who benefited from team-based care seemed “disturbingly low.” Just 16.9% of patients who received team-based care and CDSS and 13% of those who received team-based care alone improved in all three measures. “This system doesn’t get you to where you want to be by a long shot.”

She added that a team-based approach, particularly the use of health coaches, would be a “huge improvement” over fragmented care provided in much of the U.S. safety-net system.

Another Team Approach

Many systems are striving to improve diabetes management in response to payment incentives, Huang said.

In a separate retrospective analysis, published in Annals of Family Medicine, researchers at the Mayo Clinic, Rochester, Minn., reported quality improvement gains among primary care practices that adopted a team-based model called Enhanced Primary Care Diabetes (EPCD). The model deployed a range of strategies, such as empowering nurses to engage with patients outside of scheduled office visits and including pharmacists on care teams.

Mayo’s approach did not specifically target underserved populations. Rather, researchers evaluated the model’s impact on about 17,000 patients treated at 32 Mayo internal medicine and family medicine practices of varying sizes, resources, and community settings.

Among staff clinician practices using the EPCD model improved patients’ scores on a composite quality measure called D5, which incorporates glycemic control, blood pressure control, low-density lipoprotein control, tobacco abstinence, and aspirin use.

Following implementation, the portion of patients in those practices meeting the D5 indicator increased from 42.9% to 45.0% (incident rate ratio, 1.005; P = .001).

Meanwhile, the portion of patients meeting the indicator increased from 38.9% to 42.0% (IRR, 1.011; P = .003) at resident physician practices that used the EPCD model and decreased from 36.2% to 35.5% (IRR, 0.994; P < .001) at staff clinician practices that did not use the model.

In contrast to the team-based approach used in China, the EPCD protocol “is very complex, and it will be difficult to implement in low-resource settings,” He said.

The D4C trial was funded by the Xiamen Municipal Health Commission. The Mayo study was funded by a National Institutes of Diabetes and Digestive and Kidney Diseases grant. He, Peters, and Huang reported no relevant financial interests.

This article originally appeared on MDedge.com, part of the Medscape Professional Network.

 

Source: Read Full Article