Keenan Trial Blog

By Don Keenan

Many of you know Michael Freeman but many of you don't so the following is an interview conducted by Alan Galbraith of my office with Michael overviewing the variety of expertise that he possesses for our cases.  What follows is Alan and Michael's interview:

  Allan Galbraith                 Michael Freeman

I recently had the opportunity to spend a couple of hours chatting with Dr. Michael Freeman, a forensic epidemiologist with the Oregon Health & Science University School of Medicine.  Dr. Freeman is also an outstanding expert witness for those of you of have cases which can make use of his expertise, which is essentially any case in which causation contested by the defense.  We spoke about several of recent cases as well as some current projects.  As you read through segments of our conversation below, notice how Dr. Freeman effectively utilizes concepts that are implicit in the Reptile in terms of how he analyzes the facts of a case and exposes the fallacies of the defense argument. 

At the risk of belaboring the obvious allow me to cite several examples-in the case of the car wreck plaintiff with pre-existing degenerative disc disease, he swift boats the defense by turning their perceived strength of the pre-existing condition against them to demonstrate that the statistical odds that the injury in question was caused by the defendant’s negligence are 7,300 to 1.  Concepts of community safety are present throughout his work-the manufacture and marketing of propofol in 50 ml vials encouraged and enabled the spread of infectious disease, needlessly endangering any patient of the healthcare provider who was using the product. The same is true of the corporate defendant placing a known dangerous driver into a much larger and heavier vehicle-the increased risk of death to an individual who is involved in a wreck with the dangerous driver can be statistically shown.   That no doubt appears obvious, but quantitating the danger is powerful evidence both in the sense that it makes concrete the enhancement of the danger by the defendant, but it also demonstrates that the defendant’s conduct needlessly placed the members of that community in danger. Lastly, we all know through personal experience that a favorite defense strategy is to defend on causation either through claims that the injury would have happened anyway even had the defendant not been negligent or that there are so many potential causes of the injury that it isn’t possible to identify the defendant’s conduct as being the causal factor.  Notice how you can make use of the Reptile through the simplicity of the statistical evidence that Dr. Freeman brings to the table-it exposes the hypocrisy/falsity of the defense position and the attempt to deceive the jurors (and the community) as to what has truly taken place.       

 

AG:  

Tell us who you are and what you do.

 

Dr. Freeman:

I am a forensic epidemiologist, and an affiliate professor of epidemiology and psychiatry at Oregon Health & Science University School of Medicine. I spend about 30 percent of my time doing academic work and research and about 70 percent of my time doing forensic consultations. My forensic consultations are divided around 20 percent criminal and 80 percent civil. The civil cases range from injury litigation (car crashes, falls, product cases) to medical negligence and birth trauma to toxic tort cases. Any kind of case in which causation is disputed and in which a quantification of relative risk, and thus probability of causation, is helpful to a fact finder is a case that I might get involved with as an expert. I have 3 scientists who work with me in my forensic practice, including a biomechanist, a statistician, and an epidemiologist.

 

AG:  

Tell us what you mean when you say quantification of relative risk.

 

Dr. Freeman:

In virtually all cases the plaintiff must prove, to some standard defined by a probability (typically >50%), that the negligent action of the defendant was the cause of the plaintiff’s injury or disease. The risk of the injury associated with the hazard resulting from the act of negligence is called the “absolute risk” of the hazard-related injury. Let’s say, for example, that a rear impact crash at 10 mph carries with it an absolute risk of spinal disk herniation for a belted driver of 1 in 85 (from epidemiologic study). The defense will use an engineer or biomechanist to provide similar testimony based on an analysis of crash forces, with the implication or overt claim that because disk injury after such crashes is rare that the attending physician attributions of injury cause to the crash are wrong. The plaintiff will often run out and get an engineer or biomechanical expert to counter such testimony, thinking that the standard they are trying to meet is to prove that the probability that the injury could have resulted from the crash is more than 50%. This is a result of a failure to recognize that the most important question that the plaintiff has to answer for a jury isn’t “how did this injury happen?” but rather “If the crash hadn’t happened would the plaintiff still have the same injury and same need for treatment?” This is known as the “but-for” question, and it gets at the heart of causation evaluation, which is answered by relative rather than absolute risk. Relative risk compares the risk of injury from the hazard to the risk that the injury would have occurred if the hazard had never happened (also known as the base-line risk). If the risk of the hazard is double or more than the baseline risk then the 50% threshold has been exceeded. Relative risk is applicable to virtually all causation questions: the likelihood that an airbag non-deployment caused an observed injury is the relative risk of injury when an airbag doesn’t deploy versus the baseline risk of the injury in the same crash when the airbag does deploy. The likelihood that an act of negligence during an operation caused an amputation is the relative risk of the injury associated with the act of negligence versus the baseline risk of amputation when there is no negligence. So part of the need for this approach comes from understanding that one of the most common defenses that is raised in civil litigation cases is the assertion that the condition or injury was going to happen regardless of the negligent action of the defendant. 

 

AG:  

Probably most people are familiar with what you’ve done in the motor vehicle wreck case and the defense is well this guy’s got an aging back.  He has degenerative disc disease.  He was a goner anyway in terms of his back.  We didn’t really hurt him.  How do you address that?

 

Dr. Freeman:

That’s a perfect example of the application of the “but-for” standard of causation, because in that kind of a case you’re talking about a person who has a condition that only may ultimately lead to pain and the need for treatment, which is degenerative arthritic changes in the spine, a condition that is no less common in the aging population than are wrinkles on the face. Is it reasonable for the defense to claim that something that existed in the plaintiff without causing symptoms for years only coincidentally became symptomatic on the day of the crash? When the injury is associated with a low speed crash the defense is forced to take a contradictory position, in which they assert through their DME that the plaintiff was so fragile he was going to have pain at any minute due to his arthritic spine, but at the same time through their biomechanist they assert there’s no way the crash could have caused the symptomatic activation of previously fragile but asymptomatic condition of the spine. I call it the “reverse kryptonite” defense, in that the crash is the only thing that can’t hurt the plaintiff.

If we apply the relative risk approach to a simple crash like you asked about, you have the disk injury risk of the hazard (the crash)  – which we can say is 1 in 100 (derived from an epidemiologic database). This 1 in 100 risk is compared to the chance that the previously asymptomatic disk degeneration was going to coincidentally become symptomatic and ultimately require surgery, starting on the same day as the crash. This figure can be estimated using an epidemiologic database of hospital discharges, which can tell me how often people of the same age and gender undergo surgery for a disk herniation that was not traumatically induced. Let’s say that the chance of spontaneous herniation requiring surgery in a given year for men the same age as the plaintiff is 1in 2,000. Since we are only interested in the same day as the crash we have to multiply the 1 in 2,000 by all 365 days in the year to get the risk of spontaneous development of a disk herniation on any given day (this is 1 in 730,000). If the risk of injury in the crash is 1 in 100, then the relative risk that the crash caused the injury is 730,000 divided by 100, or 7,300 to 1. This is equal to >99.99% probability of causation. The numbers I used for my example are pretty close to accurate for a 40 year old man exposed to a rear impact collision of 10 mph or less, but I could be off in my estimates by 1000 times and still my conclusion of causation would be correct. These sorts of results play perfectly into the Reptile model of case presentation.

 

AG:  

Can you give us an example of a case that you are currently working on?

 

Dr. Freeman:

This morning I did an analysis in a case that was a bit out of the ordinary. It involved a head-on collision between two pickup trucks, in which one pickup, operated by an intoxicated driver with marijuana and amphetamines in his system, passed another vehicle on a double yellow, and struck the other pickup with a husband and wife in it. Their pickup rolled a half turn, and the wife was extricated with serious injuries. The husband was more severely injured, but still alive and conscious when the vehicle caught on fire, and he was burnt to death, essentially in front of the wife. Truly horrific circumstances, with a very unsympathetic defendant. 

It turned out that the man had been recently given the pickup truck by his father, who owned a company that owned the vehicle. Ordinarily he would have been driving a smaller sedan. For a variety of reasons that there is no need to go into here, the fact that the defendant was driving the pickup was potentially an additional act of negligence that may invoke punitive damages if it could be proven that it contributed to the probability of the death of the plaintiff. I approached this as I would any other relative risk question, but I considered the baseline risk to be the probability of death when a pickup is struck by a sedan, and the risk of the hazard the increased risk of death over the baseline when a pickup is struck by another pickup. This is the kind of thing I enjoy most, which is the construction of an ad hoc epidemiologic analysis for a specific case, in which I may be the first person ever to describe such data in this way. I went to a federal database of police reports on crashes, which include a sample of around 50,000 reports annually. I was able to pull up the body type of all vehicles in all crashes, sort by those involved in head on collisions at highway speeds, and look for the survival outcome. 

I was able to find a national estimate of around 71,000 collisions that occurred between 2000 and 2008 that fit the criteria, and then I threw out all of the crashes except for pickup versus pickup and pickup versus a car. Ultimately, I arrived at a relative risk estimate of 5.6 for the hazard (pickup vs. pickup) relative to the baseline (pickup vs. car), which meant that had the defendant been driving his sedan that the risk of death to the plaintiff would have been reduced by 82%. Additionally, it also would have meant had the intoxicated driver been in a sedan rather than a pickup that he would have been 9 times more likely to have been the one who died that night.

 

AG:  

Can you tell us about a case that you have testified in recently?

 

Dr. Freeman:

I testified yesterday in a case in which a beautiful 10-year-old girl named Mercedes Mears died of an acute anaphylactic reaction that accompanied an asthma attack while at her school. She died just feet away from an EpiPen that her mother had brought to school for just such an emergency, and because the untrained health aide had no idea what to do to save this girl’s life.

The defense claimed that it wasn’t really anaphylaxis, but that if it were that she would have died regardless of prompt treatment. I went to the Nationwide Inpatient Sample hospital discharge database and pulled up around 500,000 hospitalizations of school-age children with asthma over a 4-year period, and found around 500 cases of anaphylaxis among them. The death rate among these children with asthma and anaphylaxis was zero. There really isn’t a come back to this kind of testimony; it’s just factual. 

 

AG:  

I know last week when we were talking with you, you talked about your involvement in Rick Friedman’s recent case in Las Vegas where he had a large verdict against Baxter in a case that involved a drug called Propofol.  Can you tell us a little bit about what was going on in that case and, and secondly, what your involvement in the case was?

Dr. Freeman:

A large endoscopy clinic in Las Vegas was running through 70 or 80 patients every day for upper and lower GI exams. They had three or four GI docs, and four or five nurse anesthetists (CRNAs) there to administer Propofol as a sedative during the procedures. The doctor who owned the clinic was very concerned with profits, and instructed the CRNAs to re-use Propofol left over from prior procedures. He was purchasing 50 ml vials of Propofol, but only needed 10-20 ml per procedure. Unfortunately, the CRNAs would re-use the syringes to get extra Propofol when it was needed during the endoscopy, and in doing so potentially contaminate the Propofol remaining in the vial with the blood of the patient. Obviously, the practice became a hazard to subsequent patients if the remaining Propofol contained a fairly prevalent blood borne like Hepatitis C. Ultimately, the practice spread infection to at least 115, and perhaps as many as 300 patients.

 

AG:  

And the claim was against the manufacturers of propofol, Baxter and Teva?

 

Dr. Freeman:

That is correct. Despite the fact that it was the actions of the CRNAs that spread the infection, the outbreak could not have occurred if the manufacturers did not sell 50 ml vials of Propofol. With 10 ml or even 20 ml Propofol vials the infections really couldn’t have happened. The heightened risk of Hepatitis C transmission because of anesthetic misuse, including the “multi-dosing” of Propofol made possible by the 50 ml vials, had been discussed for years in the medical literature, and thus was well known by the manufacturers.

 

AG:  

What were you asked to do in the case?

 

Dr. Freeman:

I was asked to evaluate the most likely transmission route of the infections. The defense brought a well qualified epidemiologist out of Berkeley who testified that there were many possible sources of infection transmission behind the outbreak, including improper hand washing, improper sterilization of the endoscopes, and reuse of some of the equipment that shouldn’t have been reused, like the biopsy forceps. In fact, the only cause that he claimed was not possible was the re-use of Propofol. It was frankly preposterous testimony. I testified that if what the defense expert said was true, then we would have expected a bunch of other bacterial infections along with Hepatitis C to have been spread among the other patients. You see, Hepatitis C is really only spread by blood-to-blood contact, and that’s why most of the new infections seen in the U.S. are as a result of sharing needles among IV drug users. Hepatitis C is virtually never transmitted by casual contact.

In my testimony I noted that 97% of infections that occur in endoscopy facilities when there is inadequate sterilization of equipment are bacterial infections like Pseudomonas and Salmonella. If it was truly inadequate “hand washing” that had caused the at least 115 Hepatitis C infections, then I would have expected more than 6,000 bacterial infections to have hit the hospitals at the same time as well. The fact that there were no bacterial infections associated with this Hepatitis C outbreak was a powerful piece of evidence that it indeed was the Propofol multidosing that was the cause of the infections. 

 

AG:  

And for anybody who's been living in a cave and hasn't heard, how did the case turn out?

 

Dr. Freeman:

There was a $104 million verdict, consisting of $7 million each to the plaintiffs Mr. and Mrs. Washington, and then an additional $90 million in punitive damages. Rick did a tremendous job for these very nice people.

 

AG:  

Let me cover one last area with you, if I can, and that's the work that you've been doing with shoulder dystocia.  What can you tell us about that?

 

Dr. Freeman:

A large part of my daily work as a forensic epidemiologist is to perform ad hoc epidemiologic analysis for cases, like the one I just told you about with the differing fatality risk based on different impacting vehicle types.  I've been involved in half a dozen or so shoulder dystocia cases in which there was an allegation that improper actions by personnel attending the birth increased the traction on the baby’s shoulder as it was stuck in the birth canal, resulting in a permanent stretch and sometimes tear injury to the brachial plexus.

In one of these cases there was an allegation that fundal pressure had been applied when the dystocia became apparent in a difficult birth with a large baby, something that is absolutely contraindicated by current obstetrical practices. The baby was born with an Erb’s palsy, an injury to the upper nerves of the brachial plexus. The defense asserted that the injury could just as easily have occurred without the fundal pressure, and that it couldn’t be proven that this maneuver was causal.

For me the question came down to a relative risk assessment, which was a comparison of the risk of a brachial plexus injury when fundal pressure or other traction is applied (the hazard) to the risk of the same injury when there are no traction procedures, or, as a best case scenario, when a Zavanelli procedure is performed, in which the baby’s head is pushed back into the birth canal and an emergency C-section is performed.

We pulled data, again from a national hospital discharge database, called the KID database, which has information on around 4 million births per year. We looked at a total of 5 years of data, so there were around 20 millions births from which we could look first for cases of dystocia (around 20,000 per year) and then cases of brachial plexus injury. We then sorted by baby size and other relevant risk factors, and looked at the rate of brachial plexus injury with and without traction/fundal pressure, and with a C-section. Ultimately we arrived at an increased injury risk of 3.3 when the traction/fundal pressure procedures were applied versus no procedures, and a 48 to 1 risk of injury when the procedures were compared with a C-section.

These cases have sparked an interest in proceeding with more in depth research on the topic, and we are currently in the process of getting data from individual states, which will allow us to link the mother’s file with the baby’s, and thus we can refine a model for predicting dystocia-related brachial plexus injury. We already know it’s not a good thing to increase traction on the baby, but ultimately, the hope is to eliminate these injuries by early identification of the most potent risk factors. So, like a lot of what we do, this research is an ongoing effort.

 

AG:  

I appreciate you taking the time to speak with us.

_________________________________________________________________

 

NEXT BLOG:  THE REPTILE IS NOT A DESTINATION BUT A JOURNEY OF UNDERSTANDING
 
UPCOMING SEMINARS:  

November 4-5

The New Deposition Template seminar

Houston, TX

 

November 18-19

Intro To The Reptile

Orlando, FL

 

December 2-4

Advanced Reptile Seminar

Costa Rica

 

December 16-17

Intro To The Reptile

Phoenix, AZ

 

January 20-21

Reptile In Trial seminar

Los Angeles, CA