|
||||||
|
|
|
|
|||||||||||
|
|||||||||||||
| ||||||||||||||||||||||||||||||||||||
Avera Health in Sioux Falls, S.D., has achieved 65% participation in disease management among a pilot group of diabetes patients, which is at least twice the national average, according to Chief Medical Officer Richard Miller, DO.
After just 1 year, the return on investment (ROI) in the health system's diabetes management program was 4.4 to 1. "I thought that was too good to be true," Miller says, so he reanalyzed the numbers. At the end of year 2, ROI was 5.5 to 1, though it was based on a pilot sample of just 55 patients.
Among the program's small population, the average weight loss was a staggering 15%, Miller reports.
Avera realized such positive results by turning to a rich storehouse of electronic data, going beyond diagnosis when targeting patients for diabetes management, since a diagnosis alone can be misleading.
Instead of just looking for a diabetes billing code, Avera focuses on "episode-of-treatment groups" and specific bundles of services, explains Miller. A patient with a history of abnormal glycated hemoglobin (A1C) and blood glucose readings who has been on diabetes medications is a better candidate than someone with just a diabetes-related code. "You know the majority of the services, but you don't know the outcomes of the services," Miller says.
Avera's health plan uses nurses as health coaches, but its disease management program includes dietitians, physical therapists, and medical professionals. "It's basically a behavioral intervention," Miller says. "Most of the diabetic care occurs in patients' kitchens," and focusing on comorbidities during the coaching process has helped, he says.
The information now comes from a claims database, but Avera is converting to electronic medical records for ambulatory and inpatient care, a move that should enhance clinical knowledge capabilities.
But extracting clinical knowledge only works if you know what you're trying to solve. "If you don't get the question right, it's most certain that you won't get the answer right," Miller says. "The right question is: `Which people in your diabetes population need the most help?'"
|
SIMPLE TRACKING SYSTEMS CAN HELP SMALL PRACTICES
Avera is an integrated delivery network, but small practices just need simple disease-based tracking systems to accomplish the same goals, says James Veline, senior vice president and chief information officer for Avera. "Using care protocols really makes a huge difference," he says.
Miller says the greatest challenge is persuading physicians. "It is an educational process to convince docs that this will make them more efficient." Data collection cannot be a burden or doctors will not do it, he says. "Paper documentation templates can work, but it's burdensome, especially if patients have multiple conditions."
The health care industry at large has sophisticated support systems for financial decisions. "We have been remiss in our duties by not developing these systems on the clinical side," Veline says, adding that health care leaders are "shirking their responsibilities" if they aren't taking advantage of clinical-decision support tools.
THE RIGHT PLACES TO START
Jon Hultman, DPM, a Los Angeles–based health care technology consultant, says diabetes, obesity, and chronic heart disease are "perfect" places to start digging into clinical information.
"All three respond very favorably to walking," he notes. But with today's health care payment system geared toward episodic treatment of symptoms rather than education and prevention, physicians typically don't spend even 5 minutes telling patients how walking can benefit them.
"Right now, doctors won't take those 5 minutes, but those 5 minutes might save an hour someplace else," Hultman argues. This is where data mining excels, by helping pinpoint candidates who would most benefit from a little extra cautionary attention.
Clinical information systems also can aid disease management in medication reconciliation and compliance. Memorial Hospital in Gulfport, Miss., is pushing forward with an aggressive information technology strategy that includes electronic links to pharmacy benefits so clinicians can track whether patients are following through on their care plans.
"If I know that the patient hasn't filled the prescription, then I know he's noncompliant, unless he's gotten samples from the physician," explains Memorial's medical information officer, internist Richard Ferrans, MD.
"The underlying problem of noncompliance hasn't been
addressed," Ferrans says. "Information systems will help us be
able to do that." 
Footnotes
You can read more about data mining's utility in areas such as the evaluation of treatment effectiveness, management of health care, customer relationship management, and the detection of fraud and abuse in the following article: Koh HC, Tan G: Data mining applications in healthcare. J Healthc Inf Manag 19: 64–72, 2005[Medline] .
CiteULike 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||
|
||||||
|
| DOC News | Diabetes | Diabetes Care | Clinical Diabetes | Diabetes Spectrum |
