For insurance carriers, the first step to address and prevent the problem is to identify the offending providers. Querying the data within a defined period of time will reveal the transgressors. Instating a rules engine and monitoring system designed to “red flag” these providers and questionable procedures will alert carriers' special investigative units (SIUs) of problems. Establishing specific rules will also help ensure that the workflow moves in the proper direction, as seemingly innocuous procedures may not be flagged by all claim examiners as fraudulent. Additionally, using a rules engine will enable a carrier to apply consistent rules in the claim work environment.

From that point, individual carriers can decide the best way to proceed, perhaps with an independent medical examination (IME), which will serve to put fraud perpetrators on notice. The goal is to change provider behavior before they submit their automotive medical bills. Beyond individual carriers, enlisting the help of the National Insurance Crime Bureau (NICB) will help stop fraud by providing enough aggregate data for authorities to start surveillance on bad providers and pursue prosecution of repeat offenders.

Taking action to prevent automotive medical claim fraud should be part of all carriers' bill review processes. The more it becomes standard practice, the more carriers will be able to profile fraud from the moment the claim arrives. Thus, carriers can catch fraud before it becomes as costly a problem as CT scans are in PIP states.

The following review of data experiences has elicited some surprising and some not-so-surprising results.

Identify and Quantify Opportunities

We must first evaluate the claim workflow and document current processes to identify and quantify the opportunities for review. Then we can methodically analyze a series of standard metrics to spot potential areas for improvement. In the case of suspected issues in radiology, we focused on all radiology current procedural terminology (CPT) codes and billing data. The next step is using the data to identify a set of more unique, deep-dive analyses in specific areas that have high potential for fraud. In the case of radiology, the data drove us toward the “high-ticket” items such as CT scans and MRIs. To further segregate, we divided the data by U.S. region.

Once we identified the procedures for review and the geographic areas, we drew upon several years of data for comparison and baseline analysis. For the radiology review, we used closed-claim data and segregated the information by region: Midwest, Northeast, South, and West. We further divided the data into coverages and discerned between first and third parties. Here, we will focus solely on first-party billing data.

Once the data have been isolated and multiple years of experience are available for review, we can benchmark the data and compare regions. Because there is variation in PIP benefits from state to state, policy limits have been considered.

Delving into the data elicits several unique regional observations. The information used for this review encompasses three years of data — from 2005 to 2007 — totaling more than 1.5 million bills and first-party claims only.

Average Charge per Claim

Figure 1 demonstrates the average charge per first-party claim nationally, divided into Northeast, South, Midwest, and West.

As evident in the observed charge data, claim severity is on the rise. Because claim severity is the ultimate measurement and indicator in the claim environment for why dollars are spent, it is important to investigate the drivers. Judging from this trending information, we anticipate that claim severity will continue to increase through 2008.

Increases in provider charges have also demonstrated higher indemnity payouts (severity). Figure 2 provides an analysis and comparison of nationally and regionally based average paid claim results.

Overall, the national average is on the rise through the 2007 year of loss claims. This is true despite the dramatic decrease experienced in the Northeast region. Data is based on year of loss, so one would expect a decrease in 2007 claim averages, as these claims are less mature than 2005 and 2006 claims. The dramatic decrease in the Northeast is likely exaggerated by the large policy limit PIP states of New York and New Jersey.

Average Charge per Unit

Providers bill by “units” on medical bills. The units tell the insurer how many times a procedure was performed, usually per day. The data depicted in Figure 3 demonstrates again that provider charges are increasing substantially when reviewed according to units. These changes may stem from increased provider fees or the more frequent use of high-dollar medical services. Units are potential drivers and clues when reviewing the data to discern why these observations are occurring.

The amounts displayed for the average unit is noteworthy because provider charges — broken down to the unit values — have an impact on the average paid claim. The only way to identify these claim issues is to identify the driver and use a rules engine as a tool in recognizing the procedures for which unit values are on the rise. Claim representatives can then make decisions as to whether the number of units are appropriate to pay.

Thus far, the data has demonstrated that severity, payments, charges, and unit costs are all on an upswing. The next step is to determine what set of procedures or procedure may be affecting the severity results. To accomplish this, we need to aggregate similar procedures into groups. This also enables us to focus solely on the areas of billing that affect severity, as not all areas will.

Reviewing the code groups as depicted in Figure 4 and associated billing data indicates that the services keeping claim representatives busy may not be driving the severity. As noted in Figure 4, the “units” are touch points for claim representatives. Sixty-five percent of bills are related to physical therapy; 27 percent are classified by the Healthcare Common Procedure Coding System (HCPCS) as durable medical equipment, dental, ambulance, and so forth; six percent are related to radiology billing; and two percent are related to surgery.

The differences in the observations are related to the provider charges and allowances. In particular, although radiology represents only six percent of the units (touch points) for payment, this category comprises 30 percent of the charges and 29 percent of the allowances. A similar, yet smaller portion is related to surgery. For this review, we will focus only on the radiology portion, as it is the most significant.

Procedure Code Analysis, Trended Nationally

It is important to consider how the procedure code groups have contributed to the allowances and compare them nationally. Figure 5 illustrates that the allowances for radiology are now narrowing the gap with physical medicine services. Allowances for radiology services have grown 25 to 30 percent of total allowances in the claim data.

As stated before, creating families or groups of like procedures will provide comparisons that will simplify the identification of issues. We use Figure 6 to demonstrate the comparisons in the groups for allowance of payment.

In Figure 6, it is apparent physical medicine and DME have had little change and impact on the allowance on a yearly basis. Radiology demonstrates the largest variance, followed by the surgical codes. Now that the group of codes has been identified, we can isolate the contibutors (drivers) of the codes that are major contributors to the severity changes.

Increased Radiology Allowances

CT scans are the procedures that most significantly contribute to increased radiology allowances. On average, CT scan charges have risen 28 percent since 2005 and continue to swell. These CT scans in particular are driving the increase:

• Cervical CT without contrast
  (CPT code 72125)
• Head/Brain CT without contrast
  (CPT code 70450)
• Lumbar CT without contrast
  (CPT code 72131)
• Thorax CT with contrast
  (CPT code 71260)
• Abdominal CT with contrast
  (CPT code 74160)
• Pelvic CT with contrast (CPT code 72193)

Figure 7 depicts the yearly increase in CT scan allowances on a national basis.

Other than high charges, we need to consider various reasons and billing behavior changes. In addition to the hike in charges/allowances, the data show that CT scans are performed more frequently and closer to the date of loss than in prior years. Although this is yet to be determined as a driver, it should be considered. More and more states are enacting legislation to protect the benefits paid to emergency responders and/or emergency care. This means that emergency bills must be presented faster and ahead of other care that may occur on the case. Reviewing the types of bills and the order in which they are paid is a vital part of claim workflow.

Mitigation

There are a number of ways to mitigate the issue. The steps we would advise to review this and similar issues would be:

• Use an analytical approach to establish if these observations are occurring in claim operations.
• Once a data review has occurred, identify the drivers or CPT codes. These codes can be input into a rules engine and flagged for review. Rules engines have the capability to flag CPT codes within a specific time frame and frequency.
• Medical review should also be used in conjunction with this flagging mechanism to discern whether the procedure was medically necessary and performed in accordance with industry standards.

Performing CT and other diagnostic testing before they are deemed medically necessary constitutes fraud. Using the analytic approach to trending will assist in the efficient identification of issues and also create solutions for reviewing medical bills appropriately.