In July, 2013 the Insurance Research Council (IRC) published acompelling study entitled “Interstate Differences in MedicalUtilization in Auto Injury Claims”. The study was based on anational sample of more than 20,000 BI liability claims and nearly6,000 PIP claims closed with payment in 2012 from private passengerauto insurance companies that account for approximately 50 percentof the U.S. market¹.

The study examined interstate variances in the use of commondiagnostic procedures and treatments for claims where the mostserious injury reported was a sprain or strain of the back or neck. These injury outcomes represented 59 percent of the BIoutcomes in the study². More specifically, thestudy looked at the frequency of use for each of the following: Magnetic resonance imaging, computed tomography,chiropractors and physical therapists (PT). The differences amongtort states for chiropractor and PT treatments, which are the focusof this paper, are shown in Figure 1³ below:

On the surface, as one examines variances in the frequency ofthe utilization and the average number of visits for thesetreatments by state, the expanse between the differences isdisturbing. For those providers in the states with the lowestaverages, are they simply more conservative in their use oftreatments? Or, could they be that much more effective indiscerning the need for treatment? And, when treatment isdeemed necessary, are their treatments significantly moreeffective?

As for the states with the highest averages, what are thedrivers of these behaviors? Do conditions in these statescreate an opportunity or need to cost shift? From aninjury causation perspective, one would expect that states withmore energetic collisions would tend to see more severe injuriesand injury outcomes that require more treatment. Theprinciple being that the higher the impact energy, the greater thepotential for injury, and consequently, the greater the potentialneed for treatment. Could an explanation of the variancesobserved in the previously noted IRC study be that accidents incertain states are, in general, more or less severe in terms of thecollision energy?

Let's examine potential surrogate measures for collision energyin the tort states studied. Since airbags generally deploy inmoderate to more severe auto collisions, a higher frequency ofairbag deployments documented in total loss evaluations and vehiclerepair estimates for the same year should indicate states whereimpact energy is, on average, higher. Additionally, a higherrate of traffic-related fatalities would also suggest higheraverage collision energy. Conversely, one might argue thatthe absence of fatal accidents or airbag deployments, as documentedin total loss evaluations and vehicle repair estimates for the sameyear, would suggest a higher frequency of lower energy accidents,which would be consistent with injury outcomes where strains andsprains are the most severe injury reported.

Regression analyses among these variables showed some surprisingresults. On an individual basis, both fatalityrates⁴ and the frequency of airbag deployments inliability auto physical damage claims⁵ in the statesstudied could explain a meaningful percentage of variance observedin the frequency of PT utilization. More specifically, asfatality rates and the rate of airbag deployments decrease, theaverage frequency of PT utilization increases. Thisobservation is consistent with the hypothesis that the absence offatal accidents or airbag deployments suggests that there is ahigher frequency of lower energy accidents that would be consistentwith injury outcomes where strains and sprains are the most severeinjury reported.

Surprisingly, neither of these variables showed any significantability to predict variance for the average frequency ofchiropractic utilization or variances in the average number ofvisits made for PT or chiropractic treatment. In other words,variances in the average frequency of chiropractic utilization orthe average number of visits made for PT or chiropractic treatmentcannot be meaningfully explained by these surrogate measures ofcollision energy. This is troubling since it is wellunderstood that collision energy is a proximate cause of autoinjury and injury severity.

So, what can explain the significant discrepancies seen in theIRC findings? One possible explanation is that physicians andchiropractors are trained to take a history from the patient,conduct a physical and then relate the findings of the previous twosteps into his/her diagnosis and treatment recommendations. These diagnoses and treatments rarely, if ever, include anobjective and measured understanding of the collision energyexperienced by the patient. The ultimate diagnosis andtreatment is based on the healthcare provider's perceived severityof the collision as described to him/her by the patient. Additionally, standards of care are also not linked to collisionenergy experienced by a patient. As a result of thesecircumstances, the opportunity for extreme variances observed inthe IRC study exists.

Injuries are actually caused by a very specificphysical stress or strain, or specific combinations of stresses andstrains. In auto accidents, these requisite stresses andstrains can be found to exist or not exist based on the amount ofcollision energy, position of the occupant in the vehicle useof restraint systems, and an analysis of the physics of theaccident, among other factors. Additionally, whenrequisite stresses and strains exist, they also must exceed anindividual's tolerance, before the injury can becaused. And, the greater the requisite stresses andstrains, the greater the potential for a more severe injury. As a result, the amount of energy in the accident should beconsidered to evaluate the severity of an injury, if it was indeedcaused by the accident.

Scientific research regarding injuries produced by lower energyauto collisions (i.e., tests with impact related changes invelocity ≤ 10 mph) provides further insight into the variancesdocumented in the IRC study. A meta-analysis of this researchshows that well over 75 percent of the approximately 4,000 humansubject test exposures produced no injury⁶. For tests that actually produced symptoms, over 60 percent of thetest subjects noted symptoms for less than one day. Thelongest reported period of symptoms from a subject participating ina test with an actual impact was 14 days. Assuming threetreatment days per week, a maximum number of expected visits wouldbe six. It should be noted that this calculation is anestimate of the maximum visits, not the average number of visitspreviously noted in the IRC study.

Crash severity analytics show that an auto insurer can expectthat roughly 50 percent or substantially more of the third partyinjury claims evaluated are associated with a low energy collisionfor at least one vehicle in the accident⁷. As aresult, there is clear overlap in the frequency of collision eventsassociated with outcomes documented in the referenced IRC study andthe scientific research regarding injury outcomes in lower energyauto collisions. From a chiropractic care perspective, onlythe states found in the IRC study with very low frequency ofutilization and very low average number of visits appear to havecare outcomes that somewhat match with what is found in thepreviously noted scientific research. With respect to PTcare, frequency of utilization appears to somewhat match with thesame scientific research. However, the average number ofvisits matches only in those states with a very low average numberof visits. These observations are consistent with thefindings from the regression of the data in the IRC study withsurrogate collision energy metrics.

When adjusting soft tissue injury claims resulting from autoaccidents, how can the chaos and excesses highlighted by the IRCstudy be consistently avoided? What can be done to accuratelyassess the need for treatment and when needed, itsduration?

Claim technologies exists today which calculate and characterizecollision energy experienced by claimants from information inrepair estimates and other metrics of vehicle damage. More precise measurements of collision energy can alsobe extracted from a vehicle's event data recorder, commonlyreferred to as the “black box”. With collision energy data,claim technology can also evaluate the physics of an accident andits causal relationship to a claimed injury. Additionally, ifthe accident is capable of causing the claimed injury, the expectedduration of symptoms can be assessed by comparing the injury toresults observed at comparable energy levels in thepreviously-referenced human subject testing and scientificresearch.

As the data from the referenced IRC study suggests, periods ofPT or chiropractic treatment may often exceed what is expected froma low energy accident capable of producing astrain. In these cases, a review of medical billscan complement a scientific evaluation of injury causation andsymptom duration. Medical bill reviews, which can also betechnology based, highlight irregularities in invoices for medicalcare, including inappropriate treatments for the diagnoses, unusualpricing for procedures, and delays or gaps in treatment. Whenirregularities identified from a medical bill review aresignificant and treatment duration exceeds the expected duration oftreatment from a scientific evaluation, questionable treatment andrelated costs can be more comprehensively and confidentlyaddressed.

Integrated technology platforms that offer both scientificcausation analytics from vehicle repair/damage data andcomprehensive medical bill review analytics will soon be availableto facilitate a more seamless and comprehensive evaluation ofinjury claims. The result promises a more objective,efficient and effective way to consistently and accurately evaluateinjuries. Or, in other words, provide valuable assistance toclaims organizations working to appropriately diminish thesignificant expanse in outcomes for both the utilization andduration of certain treatments.

*******

1 – Insurance Research Council, “Interstate Differences inMedical Utilization in Auto Injury Claims”, 2013, page 9.

2 – Ibid, page 5.

3 – Ibid, pages 31, 35.

4 – U.S. Department of Transportation Fatal Accident ReportingSystem, 2012.

5 – CCC Information Services Inc.

6 – CCC Information Services Inc.

7 – CCC Information Services Inc.