February 2, 2009

Benchmarking Cautions
By Judy Sturgeon, CCS
For The Record
Vol. 21 No. 3 P. 6

To use benchmarking effectively (translation: to your advantage), it is necessary to understand that you are dealing with statistics. Remember that when reporting statistics, careful manipulation can make them imply just about anything, regardless of whether the implication is correct.

For example, if you were to ask 500 random people who are drinking in bars on a Wednesday afternoon whether they have a drinking problem, the response is likely to be quite different than if the same query was posed to their families. As this example illustrates, the population queried and the query itself can be critical. Keep that concept in mind when selecting your own benchmarks to ensure their validity for your application.

Auditing services, documentation improvement consultants, and financial reviewers all rely on benchmarking to a certain extent and for good reason. Simply explained, it’s a method to compare one entity with other entities. Even moms rely on benchmarking. Think of a frustrated mother telling her problem child that “all of your brothers are getting A’s and B’s, but you’re getting C’s and D’s!”  However, there are often multiple factors that must be considered in order to validate the criteria to the situation. What if the child scolded for poor grades is not as smart as his siblings or excels at playing football or playing the piano?

There are also myriad criteria against which to compare yourself. Complication and comorbidity (CC) and major CC (MCC) (secondary diagnosis) capture rates, overall case mix, medical/surgical case ratio, payer mix, productivity—there seems to be no end to the commonalities that are compared among our peers and competitors.

That in itself is a significant issue. To whom is the comparison being made? It wouldn’t be appropriate to benchmark a community hospital against the big medical center in the same county—or would it? If the end result is to identify the smaller hospital’s strengths and challenges compared with its geographical competitors, perhaps this is exactly what needs to be done. To identify best practices in coder recruitment and retention, it makes more sense to consider all of the area competitors. The little guys may not be able to offer salaries in line with the big-city coffers, but they may be able trump the competitors by identifying alternate enticements such as quality of life, opportunities for promotion, or the area’s school quality.

Consider benchmarks for CC and MCC capture rates. The first caution is to remember that if MedPAR (Medicare reporting) data are being used, by the time of publication, it is generally already two years old. The population in this data is generally adult and over the age of 65, making it necessary to also match the available data with the population of the provider being reviewed. If a facility is comprised mainly of deliveries and pediatrics with low severity, it isn’t likely to ever match the capture ratio of a major teaching facility with seasoned diagnosis-related group (DRG) coders.

An excellent plan is to aim for improvement in this area. Define a reasonable target with consideration for the actual patient mix and the resulting DRG mix. And make sure your benchmark facilities do not include the highly successful private hospital that was recently placed under a corporate compliance agreement with the Centers for Medicare & Medicaid Services (CMS) for noncompliant coding practices. Be certain that the expectation is realistic, correct, and compliant before you take aim.

Be wary of the vendor’s expectation to meet any specific number or percentage of improvement—for instance, a required target of a 90% CC capture rate or a 20% improvement in case mix. There are simply too many variables in the coding industry to lock in a single data point as a defining criterion for acceptable quality or reimbursement.

Examining this example further, if your coders are expected to achieve a 90% capture rate, they will be pressured to code incorrectly and pick up questionable secondary diagnoses in order to meet the expectation. Whether from lack of skill or lack of compliance, if they aren’t completely adherent to CMS coding rules, the risk is significant: recoupment of incorrect payment, a $10,000 fine per line item, and a triple indemnity penalty if fraud is identified. Remember that fraud is presumed if one knows, or should have known, that an action would cause incorrect payment. Professional coders and their managers should know the rules for correct coding and are expected to follow them at all times.

Also, a dictated percentage of case mix improvement is unrealistic. There are many issues that impact case mix, and the coding department has little or no impact on any of them except the coding. These issues include the following:

• Documentation quality among multiple care providers needs to be considered.

• Size does matter. One emergency tracheostomy and one ventriculostomy will cause a major bump in a small hospital’s case mix but will barely make a blip on the radar for a major medical center’s reporting.

• Holiday and summer changes to surgery schedules impact the available surgical DRGs that affect case mix.

• Patient mix needs to be a consideration. Medical or surgical? Transplants? Lots of moms and babies? All of these affect how much you can—or can’t—do with case mix.

Even trying to benchmark against other facilities for coder productivity isn’t as simple as it sounds. Some departments expect a minimum number of charts to be coded per day, while some calculate charts per hour. Some have extensive lost time and additional chores done by coders daily; others are strictly a production line with even basic abstracting being performed by someone other than the credentialed coders. Patient population affects coder productivity just as it affects case mix. A high percentage of simple delivery moms and babies will be coded much faster than those with a medical/surgical mix, extended lengths of stay, and medical complexity.

Benchmarking is vital, make no mistake about that. Provider economy is just as critical as anyone else’s, and it must be managed and optimized in the same way if financial success is to be achieved. Unless we can outline and justify reasonable expectations from our processes, we cannot begin to manage them. Whether you’re purchasing the product or developing the benchmark on your own, you will need to understand the qualities and variations within the population queried and the query itself to maximize results.

— Judy Sturgeon, CCS, is the hospital coding senior manager at The University of Texas Medical Branch in Galveston and a contributing editor at For The Record. While her initial education was in medical technology, she has been in hospital coding and appeal management for 20 years.