With outpatient care becoming more popular, managed-care organizations are trying to find an appropriate tool to measure medical group performance. Episode treatment groups may be a good option, says Henry Dove, principal at Hamden, Conn.-based Casemix Consulting.
Dove will discuss the pros and cons of ETGs in measuring physician performance and efficiency in the Medical Group Management Association session "Episode Treatment Groups: A New Tool for Measuring Physician Productivity and Profiling Physicians' Performance" on Monday, Oct. 5, at 10: 30 a.m.
Episode treatment groups link patient visits to form an episode and can be used to compare the treatments of the same health problem by different physicians and even different medical groups. Managed-care organizations can see how each physician manages each ETG-for example, urinary tract infections-as well as average treatment costs, tests administered and drugs prescribed.
"The difference in treatment costs can be great from one physician to another," says Dove. "That's the kind of thing managed-care groups are interested in."
Reports are based on medical claims submitted by physicians. Currently, there are 588 ETGs.
By comparing ETGs among physician groups, managed-care organizations can make decisions on whom they contract with, Dove says. ETGs can also help physicians improve efficiency and affect physician compensation, he adds.
One strength of the ETG classification system is that it looks at aggregate data from the entire period a patient is treated instead of data from just one visit, Dove says.
In addition, ETGs compare physicians in a specialty to one another rather than to other types of specialists.
"(ETG) takes a group of dermatologists and compares what they do against each other, rather than comparing a dermatologist and an orthopedic surgeon," says Dove. "It's very reasonable and lends itself to specialty profiling."
However, ETGs pose some disadvantages because larger data sets are needed. To be effective, the system needs to analyze a large number of episodes over a long period of time.
"It wouldn't make much sense if you analyzed two cases of urinary tract infection," Dove says. "You want at least 20 different cases for a given provider. Also, the data need to be over an extended time period. Since the time frame of analyses is longer than one visit, you preferably need two years' worth of data."