EHR-Based Research Yields Insights on CKD

Patients with chronic kidney disease who also have chronic obstructive pulmonary disease (COPD) have a 41% increased risk of death, according to a recently published study that relied on electronic health records (EHRs) (Navaneethan SD, et al. Am J Nephrol 2016; 43:39–46).

The finding is part of a growing body of evidence demonstrating the power of EHR-based studies to help elucidate the many factors that contribute to poor outcomes for patients with CKD. The technology is also being used to help test ways to improve their care.

Joseph Nally, MD, a coauthor of the COPD study and Director of the Center for Chronic Kidney Disease at the Cleveland Clinic, and his colleagues began exploring electronic data for CKD research nearly a decade ago. There were a few “false starts” using billing data before the multidisciplinary team eventually created a CKD registry using EHRs, Nally said. Their registry now includes 110,000 patients, he said.

One of the reasons EHRs are particularly useful for CKD research is that objective laboratory values found in EHRs can identify patients even if they haven’t been diagnosed, said Jesse D. Schold, PhD, of the Cleveland Clinic’s Quantitative Health Sciences Department.

“That allows you to identify patients who haven’t been recognized through other traditional means,” Schold said.

This is particularly important in CKD patients because only about 10% have CKD listed as a condition in their records (Jolly SE, et al. Am J Nephrol 2014; 39:288–96). Patients who aren’t listed as having CKD are less likely to see a nephrologist and receive recommended testing, and are more likely to receive contraindicated medications, Nally noted.

Another factor that aids EHR-based kidney disease research is the ability to access data from the United States Renal Data System, which provides a census of all patients with end stage renal disease in the US, Schold said. Nally noted that his team also taps data from the Ohio Death Index, which can provide cause-specific deaths for CKD patients (Navaneethan SD, et al. J Am Soc Nephrol 2015; 26:2512–2520).

Having a CKD registry also makes study recruitment more efficient. “The CKD registry is a spectacular vehicle for recruitment into research studies and randomized controlled trials,” Nally said.

Currently, Nally and his team are using their registry for continuous quality improvement efforts (CQI). For example, they have compared the outcomes of 500 CKD patients who receive care in a CKD clinic, which uses nurse practitioners and algorithm-based care, with 1500 matched CKD patients receiving standard nephrology care. So far, the CKD clinics have performed better on patient processes of care and patient education, but whether it translates into better outcomes is unknown, Nally noted.

The team is also exploring the use of EHR-linked technologies for patient engagement through an ongoing randomized trial funded by the National Institute of Diabetes and Digestive and Kidney Diseases. The Cleveland Clinic’s CKD-enhanced patient portal allows patients to access their lab results, request appointments or prescriptions, or send messages to their clinicians. It also provides CKD-stage–specific information. For example, stage 4 patients may receive information about managing anemia or dialysis options.

“There is momentum in health care to facilitate transparency, which allows for joint decision-making [among] patients, caregivers, and providers,” Schold said. “The proliferation of [electronic] data and technology will facilitate that and [help us] understand best practices.”