The 2014 Inpatient Prospective Payment System (IPPS) “two-midnight rule” has forced hospitals to change many processes in order to identify the correct status for Medicare patients.
The change has many implications on the patient, hospital, and requirements of the physicians. Errors in status determination could cost hospitals millions of dollars, lead to increased audits with money rescinded, and foment greater patient dissatisfaction. In the past, hospitals could use claims metrics and established benchmarks to evaluate performance of utilization review (UR) processes to identify areas for improvement. The 2014 IPPS has changed all of this.
The IPPS Process Effect on Data Analytics
Overall, the benchmarks in place prior to new IPPS final rule are no longer applicable for Medicare or Medicare fee-for-service cases. One of the most frequent questions asked by hospitals of late is this: What is an acceptable observation rate? Unfortunately, there is no right answer. As with any new process, data is often slow to arrive, making establishing new benchmarks more difficult.
Adding to the complexity of the issue of new requirements is that hospitals vary in their interpretation of IPPS and how they implement these new processes. While we are in the early stages, it is difficult to provide an acceptable, comprehensive range on observation rates. Instead, we could examine the metrics that each facility should be tracking to identify trends (compared to national data) and potential areas of exposure that are at risk for audit.
Prior to Oct. 1, 2013, the Program for Evaluating Payment Patterns Electronic Report (PEPPER) has been a key evaluation tool – from an aggregate perspective – for Medicare fee-for=service cases. The PEPPER is basically divided into three areas: medical necessity, DRG validation and coding, and miscellaneous categories that cross over both categories.
For DRG validation and coding, the PEPPER is still a good source of analytics to use, but for medical necessity and DRG length of stay, this data may not be as applicable for comparison for two reasons:
- The variance in the implemented UR processes among hospitals is very broad at this point. A wide variance will make the “normal” or acceptable range larger and could give hospitals a false sense of security.
- Second and more importantly, the LOS focus areas will likely change to zero and/or one midnight. Currently, there are one- and two-day LOS categories. “Midnights” are now the key metric, so these categories will likely change.
Medicare and the Texas Medical Foundation (TMF) likely will evolve the PEPPER into some form of zero/one-midnight-stay categories. However the fundamentals of the PEPPER will remain the same (i.e., “this is where you stand compared to other hospitals in your state, across the country, and within your billing jurisdiction.”)
Note: Although I suspect that the LOS focus will transition to zero- and one-midnight stays, this doesn’t mean that the PEPPER will not still look at stays of two midnights or greater. If two-plus midnight cases are included, it likely will be targeting gaming or delays with certain diagnoses.
Which Areas Should Be Monitored Now?
There is no reason a hospital should wait for the PEPPER to change before evolving and developing internal evaluation metrics. A common misconception is that tracking metrics only can be used if there are benchmarks. The two starter metrics that hospitals can use now center on the current number of one-midnight inpatient and observation cases.
First, monitoring zero- and one-midnight inpatient cases is critical. These are the cases with the highest risk of audit. When tracking zero- and one-midnight inpatient cases, ensure removal or consideration for the one-midnight exception cases (inpatient-only list, AMA, death, transfer, and medical mechanical ventilation). If you can track the zero- and one- midnight inpatient cases, excluding the exceptions, this should help pinpoint some actionable trends and metrics.
Another potential area to monitor is the zero- or one-midnight stays with observation hours that total greater than two midnights.
While tracking inpatient cases helps identify audit risk, observation case metrics also could help identify cases that were under-billed and perhaps help determine that a DRG payment was more appropriate. However, the traditional observation rate (total observation cases/observation plus inpatient cases) provides less information than it did in the past.
Consider dividing observation cases into two groups: zero- or one-midnight stays in one group (likely correct status) and stays of greater than two midnights (likely incorrect status) in the other. Then calculate the percentage of cases in each category. Zero- and one-midnight observation cases should represent a high percentage of the total (perhaps around 90 percent). If you discover a significant percentage of greater than two-midnight observation cases in this group, this could suggest inefficient care or overuse of observation.
While benchmarks are being established, track these numbers for change from month to month or quarter to quarter. This trending will identify patterns (good and bad) and allow for more timely correction, if needed. You can also use guesstimations in the short term for expected value as noted above.
Data and metrics continue to be powerful tools in establishing a compliant UR program. Never let a lack of benchmarks stand in your way from following trends and using data to fine-tune your UR process.
About the Author
Dr. Ralph Wuebker serves as chief medical officer of Executive Health Resources (EHR). In this role, Dr. Wuebker provides clinical leadership within EHR and works closely with hospital leaders to ensure strong utilization review and compliance programs. Additionally, Dr. Wuebker oversees EHR's Audit, Compliance and Education (ACE) physician team, which is focused on providing on-site education for physicians, case managers, and hospital administrative personnel and on helping hospitals identify potential compliance vulnerabilities through ongoing internal audit.
Contact the Author
To comment on this article go to firstname.lastname@example.org