equipment. Then you go to those in the most expensive quartile and
try to pull those people down to the mean, and ultimately you try to
pull the mean down to the most frugal group. Of course the most
expensive group will say their patients are sicker. Oh, gee, I hadn't
thought about that. Let's take a look. No, it turns out they're not. So
they'll say, well, my outcomes are better. You diplomatically say:
Gee, I hadn't thought about that either. Let's take a look. And no,
5 2
O U T P A T I E N T S U R G E R Y M A G A Z I N E O N L I N E | F E B R U A R Y 2 0 1 5
T
here's virtually no limit to the amount of information that can be collected
and stored in electronic health records, and researchers are increasingly find-
ing ways to unlock the secrets stored in that shared data.
For example, researchers at Chicago's Northwestern Medicine have developed an
algorithm that identifies patients with previously undiagnosed hypertension
(tinyurl.com/qgkbq3x), with a goal of creating a surveillance system that notifies
staff and primary care physicians any time a high-risk patient arrives in the office.
Harvard researchers, meanwhile, are using a surveillance algorithm to more
efficiently detect and classify type 1 and type 2 diabetes
(tinyurl.com/n8o7dj8). Using EHR data they're able to detect more cases and
accurately distinguish between type 1 and type 2.
Stanford (Calif.) Hospital physicians are using an algorithm to dramatically
increase the efficiency of catheter-associated urinary tract infection (CAUTI) surveil-
lance (tinyurl.com/mcjshhj). In a study of 6,379 positive urine cultures, the
algorithm identified 95.6% as not CAUTIs, 3.0% as possible CAUTIs and 1.4% as
definite CAUTIs, reducing overall surveillance requirements by 97%.
And most recently, pathologists at Dartmouth-Hitchcock, in Lebanon, N.H., have
been able to reduce unnecessary transfusions by adding a "best-practice alert" to
patients' EMRs (tinyurl.com/phn36o4). As a result, the proportion of expensive
and potentially hazardous two-unit transfusions has decreased from 47% to 15%.
— Jim Burger
BETTER CARE
Unlocking the Secrets in EHR Data