Hospitals not prepared
By our proprietary metrics and benchmarking, we have found that the average hospital has 19 percent of its inpatient coding cases at risk for a RAC audit and 9 percent of its cases underpaid by Medicare. From our consultative work with hospitals, our data mining experience and our past and current compliance initiatives, we believe that there would be significant opportunities for RACs in most hospitals. In actuality, once CBIZ has run a hospital’s real data through our RAC criteria, the results have reinforced the need for hospitals to make significant coding improvements. The RACs are real, and from where we stand too many hospitals are not prepared to deal with the consequences that follow.
Outpatient versus inpatient
One of the major RAC target areas is outpatient surgical procedures performed as inpatient procedures. For instance, cardiac device procedures like pacemakers can be done as an outpatient or inpatient procedure. However, there has to be a sufficient medical necessity determined by Medicare using Medicare InterQual guidelines for the procedure to warrant an inpatient stay, which is more costly and has a significantly higher reimbursement.
If the inpatient stay does not have the proper physician documentation and accompanying procedures, Medicare can reduce, or even in some cases, withhold payment. The RACs are looking at these inpatient cases to determine if the patient stay was necessary.
Data mining, not random audits
CBIZ is working with a major teaching hospital that had 1,299 cases over a 12-month period that would be subject to this RAC target area. In order to examine these cases, this hospital would have to spend significant sums of money and resources to pull each chart and examine all 1,299 cases individually – a cost of $100,000 minimum. A random audit may look at 100 or 200 of these cases; however, our results have shown that both of these approaches are a waste of time and resources.
Through our data mining, we found out that this hospital only had 118 cases at risk out of the 1,299, or 9 percent. Thus, it would be unnecessary and costly for a hospital to commit resources to examine all 1,299 cases. The random audit approach to this problem would be even more dangerous. Is there any certainty that the small sample would even correctly estimate the proper number of cases at risk? If only nine percent of the cases are at risk and a hospital is examining only 10 to 15 percent of the cases, the likelihood of proper mitigation in this target area is highly unlikely using a random sample methodology.
With data mining, a hospital can obtain an accurate snapshot of their risk exposure without committing significant internal and external resources. This hospital would have committed hundreds of hours or a comparable consultative price in trying to assess their risk.
A good data mining system will also shed light on the types of cases that continually subject the hospital to risk. Those cases, especially if they are ‘high-dollar’ cases, should be the hospital’s priority. CBIZ worked with one hospital that had 696 cases that would be subject to this list. However, only 6 percent of those cases were subject to any risk. The hospital really wanted to devote more of its resources to inpatient coding – in this instance, where 21 percent of their cases were at risk. By using data mining the hospital was able to prioritize its RAC efforts. Instead of examining all 696 cases in a RAC-targeted area that was not a comparatively high-risk area, the hospital devoted its resources to a more pressing RAC target area.
In terms of mitigation, the hospital examined those inpatient coding cases that put the hospital most at risk and made the necessary changes, such as coder education, hardening the patient record, if appropriate, through physician queries and setting up hospital-wide best practices which lead to appropriate coding. Without data mining, a hospital’s mitigation and overall RAC preparedness would be arbitrary and incomplete.
If a provider does not use data mining to evaluate its RAC exposure it will not be adequately prepared for the RAC. From a financial planning perspective, any RAC assessment without data mining is incomplete and therefore, not sufficiently serving the needs of the hospital.
From a medical records perspective, there cannot be an adequate understanding of the risks, or in some cases, the lack of risk the hospitals face from inappropriate coding. For case management, the knowledge that the hospital InterQual and Medicare guidelines are being consistently followed cannot be ascertained without data mining. Finally, from a compliance standpoint, the hospital cannot determine if they face substantive and repeated problems with national fiscal intermediary guidelines without using data mining.
Essentially, hospitals will not know vulnerabilities for the RAC, not to mention other compliance or integrity initiatives without data mining. For example, a hospital we are working with had no prior knowledge that approximately 70 percent of its short-stay inpatient cases – cases with an inpatient stay of less than three days – were at risk for RAC recoupment.
Third day transfers
Another hospital that we are working with was convinced they had a problem RAC-targeted area that involved third-day transfers to a skilled nursing facility. After examining all of their cases through data mining, we found that only 2 percent of their 933 total cases were at risk. This is not to state that those cases that make up the 2 percent are unimportant; however, the hospital can now place the proper emphasis – resources, time and money on their real RAC exposure.
Through data mining, hospital decision-makers can effectively prioritize their RAC response.
Sam Donio, the President of CBIZ KA Consulting Services, LLC. An authority on sound and compliant healthcare financial management, Mr. Donio has developed numerous products and services to meet the ever-changing needs of healthcare providers.
Samuel A. Donio, Jr. is president of CBIZ KA Consulting Svcs, LLC