Analysis of Merged Electronic Health Records Aims to Improve CKD Patient Care

Analysis of Merged Electronic Health Records Aims to Improve CKD Patient Care

Electronic health records (EHR) can minimize medical errors, duplication of tests, and delays in treatment. But problems arise when healthcare providers use different EHR systems, and can be especially serious for patients with multiple chronic conditions.

For this reason, Christiana Care Health System physicians and University of Delaware computer scientists are using merged EHR from different institutes to improve care management and clinical outcomes for patients with chronic kidney disease (CKD).

Claudine Jurkovitz, a nephrologist and researcher in the Value Institute at Christiana Care, is leading the project with Hagit Shatkay, an associate professor in the Department of Computer and Information Sciences at UD.

Researchers in the study, supported by the Delaware ACCEL program, are using longitudinal data gathered from a large cohort of CKD patients to make predictions about patterns of hospitalization, and later plan to investigate trends in CKD.

“Hospitalizations increase in frequency as kidney function declines and are mostly due to cardiovascular events or infections.  We’d like to be able to predict the risk of hospitalization within a defined time period following an office visit,” Jurkovitz said in a news release. “Our hypothesis is that, most often, catastrophic events leading to hospitalization are preceded by the convergence of trends such as increases in blood pressure, fluid overload, and the frequency of skin or respiratory infections, as well as uncontrolled high blood sugar for diabetic patients and declines in natremia, which is a drop in sodium levels in the blood.”

According to Shatkay, a hospitalization model based on patient data includes a large number of variables, such as trends in laboratory results, blood pressure, medication changes, the frequency of calls made to physicians, and outpatient visits.

“These rich and diverse data require that we develop and examine machine-learning based methods for representation of, and prediction from, such data,” she said.

Shatkay and doctoral student Moumita Bhattacharya are pinpointing the characteristics that carry the most information about hospitalization, and aim to represent that data more compactly. The researchers are also working on making predictions more efficiently by looking at inter-dependencies in the different measurements.

A next step is to investigate clinical care provided to a cohort of children, who moved from pediatric care to adult care at Nemours/Alfred I. duPont Hospital for Children over a seven-year interval. They will evaluate the control of risk factors for cardiovascular disease, long-term clinical outcomes, and medication adherence. Their assumption is that medical adherence is a critical determinant of clinical outcomes in young CKD patients transitioning to adult care.

As Shatkay stressed, while this study is focused on the efficacy of the healthcare delivery system for CKD patients in Delaware, its approach will have wider applicability — medically, geographically, and demographically.

“The number of patients is in the thousands, and each one is associated with about 100 pieces of information,” she said. “We’re figuring out ways to sift through these massive amounts of data and determine what’s relevant and what’s not.”

Jurkovitz hopes the research will improve the coordination of care for all CKD patients, reducing hospitalization rates.

“The rate of hospitalization for patients with CKD is as high as 430 per 1,000 patient-years,” she said. “If we can identify the patients most at risk of being hospitalized, we can bring this number down. But first we have to figure out which factors are the most predictive of the decline that leads to hospitalization.”

Leave a Comment