Article
By comparing your coding habits against these national norms, you can learn whether you're overcoding or undercoding.
By comparing your coding habits against these national norms, you can learn whether you're overcoding or undercoding.
When you circle codes on a charge ticket, do you have the queasy feeling that you're off the mark?
Theres good reason to feel that way, especially when it comes to the complicated CPT codes for evaluation and management. Overcode, and you could get in trouble with Medicare and other payers. But undercode which is what far more doctors doand you leave thousands of dollars on the table. Medicare, for example, pays on average $29 more for a 99214 office visit than for a 99213.
One way to find out whether you're circling the right numbers is to compare yourself to fellow doctors nationwide. To help you do that, we've reproduced a treasure trove of coding data from the Medical Group Management Association. These benchmarks are compiled from more than 100 million procedure records involving 20,000 physicians and other providers. With this data, you'll learn that FPs code an office visit with an established patient as a 99214 almost 14 percent of the time. If you're coding only 6 percent of these as 99214, maybe you're shortchanging yourself.
You can also get comparative data from Medicare, which has published E&M coding statistics by specialty for years. MGMA numbers, however, are broader and deeper because they cover patients in various age groups, not just those eligible for Medicare. They also break down key CPT codes by diagnosishow doctors code visits for hypertensive patients, for example. Experts say the ability to drill down to this level is especially valuable. "A lot of doctors who overcode say they have sicker patients," says MGMA coding consultant Lisa Stavrakas in Dillon, CO. Coding profiles based on diagnosis put that argument to the test, she says.
To take advantage of the MGMA benchmarks, first you'll need to profile your own coding habits. Most practice management software programs are capable of generating reports that match up with MGMA data.
Focus your attention on your E&M codes because these represent the lion's share of a primary care doctor's procedures. Then, to get a true picture of how your numbers compare to the national norms, you must factor in your clinical expertise and patient mix. For instance, "some internists have a strong subspecialty in endocrinology, so they might attract more diabetic patients, and consequently code higher," says Curtis Mayse, a St. Louis consultant with LarsonAllen Health Care Group. This doctor would need to examine the E&M codes for diabetic patients.
If your numbers and the benchmark numbers vary by only 1 or 2 percentage points, that usually doesn't warrant further investigation, especially among the most widely used codes, says David Gans, the MGMA's director of practice management resources. Larger variances must be analyzed in the context of the bell curve for a particular visit type. "If a doctor is an outlier for an established-patient code like 99213, he's going to be an outlier somewhere else, maybe on his 99212s or 99214s," he says. "You want to look for patterns."
What if you discover that you exceed the national averages for the most lucrative E&M codes? Yes, you might be overcoding, but don't necessarily jump to that conclusion. After all, national averages don't necessarily represent correct coding. They may, in fact, set the bar too low, since experts say that physicians tend to undercode to avoid the wrath of Medicare. So if an internist codes 12 percent of new-patient office visits as 99205, and the norm for his specialty is not quite 8 percent, that internist is an outlier. But he also may be choosing his codes according to the book.
Whatever variances you uncover, you won't know what they mean until you audit representative medical charts for the codes in question. Does the documentation support the CPT code on the charge ticket? Record checks can reveal three possible problems:
1) You documented that you performed all the services necessary for a 99214, for example, but you chose 99213 to play it safe.
2) You did, indeed, provide 99214 services, but documentation was sketchy.
3) Your work simply didn't warrant a 92214, no matter what you wrote in the chart.
"You have to keep medical necessity in mind," adds Mayse. "You can code a 99214 for a hangnail, talk to the patient 30 minutes, and document the visit perfectly, but it won't fly."
A benchmark analysis can spotlight other charge-ticket miscues that keep you from receiving your rightful compensation. For instance, blood draws (36415*) and immunization administration (90471) rank in the top 10 procedure codes for FPs who don't deliver babies, but doctors don't always record them, say consultants. One possible reasonthe codes aren't printed on the charge ticket.
You also may identify opportunities to add or expand clinical services. Complete electrocardiograms (93000), for example, are the seventh most common procedure internists do for their hypertensive patients. If you're an internist, and 93000s are No. 20 on your list, maybe you should be doing more. Maybe you have an ECG machine, but with the number of hypertensives in your practice, you could use another.
Physicians who measure their practice's coding against national averages shouldn't demand overnight changes of themselves. "Benchmarking is a slow, gradual process," says William Tyson, who manages a six-physician urgent care clinic in Farmington, NM, that subscribes to MGMA coding data online.
If you're like most physicians, though, merely seeing your numbers alongside the benchmarks will provide enough motivation to fine-tune your charge-ticket skills. "Doctors don't like to be outliers," says Tyson. "They like to be part of the crowd."
The tables and charts accompanying this story come from the 2003 Coding Profile Sourcebook for Primary Care Specialties published by the Medical Group Management Association. There are two other editions: one for medical specialists such as cardiologists, gastroenterologists, and emergency physicians, and another for specialists in surgery, pathology, and radiology. Each book costs $395 for MGMA members and $635 for nonmembers. A set of three is $995 and $1,600, respectively (plus shipping and handling). To order, go to www.mgma.com, click on "Store" at the top of the page, and use the search engine to find the Coding Profile Sourcebook you want.
More coding data are available online from Physcape, a for-profit MGMA subsidiary. Physcape takes coding data from a doctor's practice-management system and automatically compares it to the coding patterns of other Physcape customers in minute detail. You can go side by side with fellow specialists in groups that match your practice size, region, and payer mix. You also can compare one doctor to others within your group.
Physcape goes a step further and calculates how much your practice could be losing in gross billings by not coding as many 99214s or 99215s, say, for diabetic patients as your peers. Company CEO Luke McCain is quick to call this figure merely a "revenue opportunity" because it's possible the practice is correctly coding these visits, and has good reasons to be an outlier. "It's up to the doctors to study the medical records and determine whether the variance is legitimate," he says. But chart audits usually reveal that doctors who are below national norms are undercoding.
The MGMA Coding Profile Sourcebooks represent a yearly snapshot of the Physcape database, which contains more than 140 million procedure records. The database is updated throughout the year, and Physcape customers can monitor coding on a quarterly basis.
The cost of subscribing to Physcape depends on the size of your practice. Groups with five or fewer doctors pay $375 per quarter for the Physcape service that analyzes coding. For more information, visit the company's Web site at www.physcape.com .
Robert Lowes. Is your coding on target? Medical Economics Dec. 5, 2003;80:84.