The potential for fraud in healthcare by providers is readily recognized in all aspects of society. The high fees charged by doctors for their services regardless of outcome is frequently the source of humor, derision, and anger. The reason for this potential for fraud is what is known in the business of healthcare finance as the “agent relationship;” the individual who profits from the sale of the service is also the arbiter of necessity of the service. The patient is informed that they need a lumbar fusion, hysterectomy, root canal, or 35 chiropractic adjustments by the neurosurgeon, gynecologist, endodontist, or chiropractor, who will be paid for the service. For this reason, the doctor-patient relationship amounts to a leap of faith on the part of the patient that they are being sold services that are truly necessary for their condition. This leads to a fiduciary as well as ethical obligation on the part of the healthcare provider to the patient; the doctor is obliged to provide services that are appropriate for the patient’s condition in both quality and quantity. This responsibility is enforced by peer review groups and licensing boards, which curb potential abuse.
How then, does the independent medical (or chiropractic) examiner fit into the scheme of appropriate and necessary healthcare? Their purpose is to review the propriety of a specific treatment for a particular condition of an individual patient, usually for the third party providing reimbursement (usually an insurer), who in turn has a fiduciary responsibility to the patient to pay for treatment that is considered reasonable and necessary. The agent relationship for the provider of the IME is entirely different than for the healthcare provider. The IME provider has no ethical or fiduciary responsibility to the patient, and has only to answer to the insurer who has requested the IME. The reimbursement rate for the IME is usually set in advance, so that there is no opportunity for the IME provider to enhance recovery from an individual IME. Future work for the IME provider, however, is more likely if the insurer client is satisfied with the results of the IME. Satisfaction for the insurer with the results of the IME is more likely if it is a cost effective alternative to reimbursing for treatment without question. Therefore, there is a financial incentive for the IME provider to determine that the treatment in question is unrelated to a condition for which the insurer is responsible. Thus the agent relationship for the IME provider to the insurer requires a leap of faith on the insurer’s part that the report of the IME is an accurate characterization of the needs of the patient and the propriety of the treatment. There is no apparent fiduciary or ethical responsibility on the part of the IME provider to provide an accurate picture of the patient’s condition or treatment, other than the personal ethics of the individual IME provider. There are generally no peer review or licensing board disciplinary consequences if the IME provider chooses to serve only their own financial needs in the performance of the IME. Such an arrangement invites abuse.
The IME situation that encourages the greatest amount of abuse is the defense medical evaluation. The DME occurs in liability litigation where the defending insurer or attorney representing the insurer is not attempting to determine the necessity of treatment, but rather sets out to prove that either or both the treatment and condition allegedly related to the litigation is not the responsibility of the defending insurer. This situation most frequently arises in motor vehicle crash-related injuries. The DME provider is asked to give the defending attorney evidence that will help with a legal defense of the allegation that the treatment and injuries in question are related to the crash in question. There are no consequences for the DME provider for giving a less than truthful assessment of the situation; one that benefits the defending insurer and the DME provider. This arrangement not only invites abuse, it encourages it. It is important to note that not all DME providers abuse their position, however, it is equally important to note that there is no disincentive for such abuse.
The purpose of this discussion is to present the scientific weaknesses inherent in the self-serving DME or IME opinion, one that embraces junk science as its core (I define junk science as the use of scientific terms and formulae applied inappropriately for the express purpose of lending credence to an opinion that is clearly lacking in validity). The following are the primary scientific and logical transgressions of such opinions that invalidate them:
The use of risk retrospectively
Example: “Ms. Jones presents with clear evidence of a herniated disc. The risk of such an injury following the crash in question is minimal, therefore I find it highly unlikely that Ms. Jones sustained any injury beyond a mild muscle strain in the subject crash.” Discussion: Risk is a statistically-derived tool that is used prospectively to determine to probable proportion of a population that will experience an outcome. A correct use of risk is as follows: “one out of three people who sustain a whiplash injury and seek treatment will have some degree of residual neck pain 33 months post crash.” An incorrect use of risk would be “three people sustained a whiplash injury 33 months prior, therefore at least one but not two of them now have neck pain.”
Why is this wrong? Why doesn’t 10 coin tosses result in heads every other toss? Because of the effects of random variation. Since it is unknown how random variation will affect an individual outcome until after an event, the only valid measure of the outcome is the measure of the outcome, and not the probability of the outcome. In the case of an injury following a crash, the measure of the outcome is the evaluation of the injury by a qualified and competent practitioner. Injury risk is inconsequential, and not to be considered after the fact. The most absurd, yet appropriate example of this particular type of junk science is to use the statistic that risk of death in a plane crash in 1 in 1,000,000, and therefore so unlikely that it could not have occurred in the case in question.
The use of an average as a range
Example: “The average recovery time for whiplash is 8 weeks, therefore the first 8 weeks of Ms. Jones’ treatment was reasonable and necessary, but all subsequent treatment was not.”
Discussion: The average of a data set is a measure of the central tendency of that data set. It does not imply the range of the data set in any way. For example, the statement that the weight of an average US citizen is 165 lbs. does not rule out a 100, 250, or 800 lb. person. Likewise, regardless of the average recovery time, it is the range of recovery time that is of importance. As an average is a measure of the middle of the bell curve, using it as a cutoff point only correctly defines about half of the population. As discussed above, random variation dictates that an individual outcome can land anywhere on a distribution curve, from 50th percentile to 99.9th percentile. The actual outcome is determined by real determinants of the patient’s condition, such as history and evaluation. The use of an average obviously implies advanced scientific knowledge of the epidemiology of the condition, yet invariably, it is fabricated from the examiner’s experience. This is a classic example of junk science.
Misuse and misinterpretation of the biomedical literature
Example: “It is clearly indicated in the literature that chronic pain following whiplash is non-pathologic, and thus no treatment is indicated for such conditions.”
Discussion: The literature is frequently cited as a source of information by which the examiner can state that the patient in question does not have an injury associated with a particular crash, however, extrapolating the literature to an individual not specifically described in the literature is flawed on the same bases stated under #2. Thus, even if the above statement was true (it is not) it does not rule out that fact that an individual patient can have a pathologic response to a whiplash injury. The only way to account for random variation is to examine the facts of the case on their own merits. Additionally, most frequently the DME and IME-cited literature is selectively read, ignoring the majority of papers that contradict opinions expressed by the authors of the particular paper. Just as frequently, the papers are misread, misquoted, or not read at all, and the opinion is unrelated to the cited publicatio. Another popular ruse is to use the expression “the literature shows that…” without giving any specific cites, usually because none exist. Many cites are actually based on abstracts of papers that have not been read in their entirety.
Setting a standard of care based on an individual opinion
Example: “Such conditions do not typically require care for more than a few weeks”
Discussion: The appropriate duration of type of treatment, regardless of guidelines, is a case-by-case determination that is made by practitioners on a daily basis. How long a particular individual will need treatment cannot be determined until the individual has recovered or the treatment ceases to be effective. Average values are not appropriate in determining the outcome of a real event, as discussed under #2. Most particularly, the personal opinion of a reviewing practitioner versus the treating practitioner regarding the quantity or type of treatment required for a particular condition is an open invitation for abuse of the position by the reviewer. As there are no ethical constraints preventing the reviewer from disallowing treatment (it is simply personal clinical opinion), and there are considerable financial incentives for doing so, I do not believe that this is a workable method for determining the reasonableness of treatment.
IME and DME examiners should be held to the same ethical standards as practitioners. Peer review and disciplinary action from examining boards should be available to monitor the practices of IME/DME providers.
The financial incentive combined with lack of ethical constraints inherent in the IME/DME system must be recognized and freely discussed by licensing boards and other regulatory agencies.
The use of junk science methodology by the IME and DME to deny treatment and remuneration for injury, particularly given the lack of ethical constraints on such practices, should be viewed as an act of unprofessional conduct, in precisely the same manner that overcharging and overtreating a patient would be viewed, since both are financially motivated.
Reasonableness of treatment remains an important problem for all parties. An unbiased, well-trained group of providers who do not provide IME/DME services should be formed by the regulatory boards to provide treatment review. Their work product should be standardized and open to statistical review.
Dr. Michael Freeman holds a Ph.D. in trauma epidemiology from Oregon State University, as well as a degree in biostatistics from that school. His thesis was on chronic pain following motor vehicle crash injuries. Dr. Freeman serves as clinical assistant professor at Oregon Health Sciences University School of Medicine, Department of Public Health and Preventive Medicine, as a trauma epidemiologist, where he teaches a course in injury and trauma epidemiology (with a focus on motor vehicle crash injuries). Dr. Freeman is originally trained as a chiropractic physician, and in that capacity has evaluated, treated, and followed well over 3,000 cases of crash-related injuries.
Dr. Freeman is a fully accredited crash reconstructionist with the Accreditation Commission for Traffic Accident Reconstruction (ACTAR accreditation #1581), originally certified in basic and advanced crash reconstruction with Northwestern University in Evanston, Illinois, and has reconstructed more than 1000 motor vehicle crashes of all severity, He also serves as a Consultant Forensic Trauma Epidemiologist to the Medical Examiner Division of the Oregon State Police, analyzing fatal collisions and the resulting injuries. For the past 6 years Dr. Freeman has been a part of a team that has designed and conducted more than 80 human volunteer and crash test dummy full-scale experimental crashes, focusing on minimal and no visible damage crashes. The results of these experiments are published in the Journal of Accident Analysis and Prevention, a peer-reviewed journal.
Dr. Freeman is the co-editor in chief of the Journal of Whiplash and Related Disorders, a peer-reviewed journal from Haworth Press, and the scientific co-chair the International Whiplash Trauma Congress, a scientific forum that emphasize research on injuries that result from minimal and no damage collisions. Dr. Freeman has written and published widely in the field of motor vehicle crash injuries and crash reconstruction, and his CV can be accessed on the Web at http://www.ohsu.edu/public-health/employees/faculty/freeman.shtml.