Monday, February 6, 2017

Jury Duty or I’ll Never Take My Ankle for Granted Again

Nasrudin the Fool became a magistrate and presided over a case. The first party presented his argument, and when he was done, Nasrudin said, “I believe you are right.” The other party presented his argument, and when he was done, Nasrudin said, “I believe you are right.” The first party then said, “Your honor, we cannot both be right,” to which Nasrudin replied, “I believe you are right.”
—Sufi parable

In the above story, as well as in the many Sufi parables featuring the character Nasrudin, who is usually identified as a fool, it is easy to take the wrong lesson, for although the things he says and does might sound foolish, it is the insight of the Sufis that Nasrudin’s foolishness is the flip side of wisdom. There is usually some truth that we can extract from his seeming idiocy that vindicates Nasrudin when he stumbles upon a profound insight.

I performed jury duty recently, and the experience reminded me of Nasrudin’s dilemma. Our case involved a woman who accused a surgeon of negligence in the treatment of her ankle fracture. Each side presented expert witnesses who, among them, presented confusingly contradictory accounts of what would be expected to happen in an injury such as the one suffered by the plaintiff and what a surgeon treating such a condition should be expected to do. (What is the standard of care to which a medical practitioner ought to adhere?) Although, after five days, I and my fellow jurors became so used to some of the relevant medical terms that we could almost fool someone into thinking that we knew what we were talking about, nevertheless, even during our final deliberations, I found myself contemplating what we did not understand about the relevant medical and bio-mechanical processes involved in the plaintiff’s condition.

The plaintiff is a woman in her forties who had slipped off of a five-foot-high wall and broken her ankle in three places. This is called a trimalleolar fracture, because, among the bones that make up the ankle joint, hers had fractured in three (tri-) places. The first two of these places are the lower (or distal) ends of two related bones, the tibia (the larger bone that supports the calf of the leg) and the fibula (the thinner bone that runs more or less parallel to the tibia). The lower end of the tibia is called the medial malleolus while the lower end of the fibula is called the lateral malleolus. You can feel these two malleoli easily. They are the bony bumps on each side of your ankle. The fibula, and therefore the lateral malleolus, is always on the outside of the ankle. The third part of a trimalleolar fracture is the rearward (posterior) part of the tibia, often referred to as the posterior malleolus. The malleoli fit over the inner ankle bone (talus), similar to the way that the cup part of a ball bearing fits over the ball (only, the talus is almost squarish rather than round). Together these bones form the ankle joint that gives the foot its characteristically versatile range of motion.
cutaway of bones of ankle


An anatomical feature of the ankle that turned out to be crucial to the case is called the syndesmosis, or syndesmotic ligaments. This structure connects the lateral malleolus to the tibia. If this connection is torn loose, then the tibia and fibula will drift apart, making the ankle unstable, essentially destroying the ankle as soon as the patient tries to walk on it. Determining whether the syndesmosis is intact or in need of repair is a critical step in any surgery on a trimalleolar fracture. Our case would turn partly on whether or not the surgeon actually made a determination of whether or not the syndesmosis needed repair.

 Complicating matters, we were given to understand that there are cases where a patient comes almost straight down on their foot very hard, driving the ankle bone (talus) upward into the tibia with a wedge-like force that shatters the lower end of the tibia (called a pylon fracture). In such injuries, the cartilage between these bones will be severely damaged, and in such cases, arthritis will eventually set in because of the loss of cartilage, the only question being, how soon? On the other hand, in cases where the ankle is violently twisted, the ankle will usually be broken in the three places mentioned above: the lower end of the fibula and, also, the lower end of the tibia in two places (medial and posterior). In our case, the injury to the woman’s ankle was primarily the second kind of injury (trimalleolar fracture), but it also had some of the characteristics of the pylon injury. After all, she had fallen from a height and landed on her foot with great force. The defendant argued that this factor made the plaintiff’s admittedly rapid and severe arthritis more likely to be explained by the accident (post-traumatic arthritis) rather than by the surgery or post-operative treatment (negligence).

The crucial question that we as jurors were charged with resolving was whether by applying an objective Standard of Care (capitalized here to emphasize its importance), the surgeon had treated the patient reasonably and competently (the legal term we heard more than once is reasonably prudent or reasonable prudence). If not, had his negligence contributed to the harm suffered by the patient, or did his negligence have no effect? Was the patient harmed only as the result of her original injury? (I noted that the plaintiff’s case was, at times, argued so extremely that one might almost be forgiven for imagining that her attorney wanted jurors to imagine that the surgeon had pushed her off of the wall, thereby causing the original injury.)

The patient’s ankle quickly became swollen on the day of the accident. After first seeing the patient, the surgeon put off operating for ten days because the swelling made too-soon-an-operation inadvisable. As one expert witness explained, if extremely swollen skin is cut open during an operation, it can become next to impossible to close incisions properly. One of the tedious aspects of the case was that everyone kept reciting and belaboring this and other aspects of treatment that both sides stipulated had been well-handled. I had to wonder whether this was done in order to allow the defendant to show that he had done something right, and the plaintiff to show a sense of fair-play by acknowledging it. In any case, it seemed unnecessary to mention it more than once.

There is more than one right way to repair a trimalleolar fracture. Different ankle surgeons perform the procedure in different ways, depending on their training and experience. Although each case is unique and the surgeon must be guided by the actual conditions he finds during the procedure, still, two different surgeons could perform this surgery using different techniques, and both would be judged to be competent. Depending on the specifics of the injury, the surgeon can apply various devices in various ways to the areas that need repair.

The surgeon in this case had performed similar surgeries before, and he developed a tried-and-true plan after viewing x-rays of the injured ankle. After manually arranging the shattered pieces of the fibula so that they were more or less properly aligned, he fastened them together with a long narrow plate that he affixed to the fibula with nine screws. (I have since seen x-rays of similar fibula repairs where only eight screws were used.) If the surgeon believes that there is a possibility that the syndesmosis, which holds the fibula to the tibia, might have been torn during the fracture, he leaves one of the screw holes in the plate open, in case he needs to do an additional repair. So the surgeon’s plan has built-in contingencies: he plans to perform the minimum amount of surgery that should adequately fix the problem without subjecting the patient to unnecessary procedures; however, he anticipates that he might need to do more than the minimum once he has eyeballed the injury during the procedure. In this case, the surgeon not only could see the medial and lateral malleoli of the ankle with his eyes by looking through the separate incisions made for each, but he also had the benefit of a fluoroscope, which allowed him to take x-ray images in real time and gave the surgeon the advantage of being able to move the camera to see the ankle from different angles. Especially important in this case is that the fluoroscopic image gave the surgeon views of those parts of the ankle that he could not eyeball directly through the incisions.



 The surgeon usually performs a test to determine whether the syndesmosis is intact. This test is sometimes called the hook test. The surgeon simply uses a dental pick (or dental pick-like device?) to yank on the fibula to see whether it pulls away from the tibia. If it pulls away, that is called a positive result, meaning that the syndesmotic ligaments are positively not holding the fibula to the tibia. Unless something is done to repair the connection between the two bones, then, as soon as the patient tries to walk on the foot, the bones will begin spreading apart. If this is there is a rift, the surgeon should use the hole he left in the middle of the plate to insert what is called a syndesmotic screw. One cannot miss this screw on an x-ray, because it is much longer than the other screws that hold the plate to the fibula. The syndesmotic screw, unlike the other, shorter screws in the plate, is screwed not only through the fibula but into the tibia itself, mechanically holding the two bones together in lieu of their compromised syndesmotic connection. Next, on the inside of the ankle, the medial malleolus is fixed (or fixated, as foot and ankle surgeons prefer to say) with another long (but bigger) screw. In this way, the ankle joint is stabilized from two sides.

Unlike the surgeon in this case, some surgeons fixate the medial malleolus before they deal with the fibula, often because they believe in performing two different tests of the ligaments, one of which requires both sides of the ankle to be fixed before the test. Regardless of the order in which each of the malleoli is dealt with, both sides of the ankle need to be cut open, but one incision must be sewn up before making the second. The surgeon cannot go back and forth from one side to the other working through open incisions. In this case, the salient point is that the surgeon had to perform the hook test before he sewed up the incision over the fibula and made the next incision on the opposite side of the ankle. Once he closed the incision on the lateral (fibula) side, the window for performing the hook test closed, too.

The case seemed to turn on whether a couple of screws that the surgeon did not fixate should have been. In his operative report, he did not say whether he performed the hook test. He testified that it is his habit to perform this test routinely, but he had no specific memory of having done it. In his defense, I might compare this lapse to my habit of locking the door to my home every time I go out but not always specifically remembering that I locked it. When I come home and find the door locked, only this tends to confirm that I did lock it. Since a surgeon’s actions are considerably more critical to the wellbeing of his patient, this kind of forgetfulness seems less forgivable than someone forgetting whether they locked their door. (Though, the possibility that a criminal might get into my house while I am away is certainly a serious potential consequence of my seemingly less serious forgetfulness.)

The surgeon in this case did not use the syndesmotic screw that he would have been expected to use if the hook test showed excessive play of the fibula in relation to the tibia. Either the surgeon did not perform the test and incompetently assumed that he did not need to use this additional screw, or he did perform the test and perhaps legitimately determined that the additional screw was not needed. In either case, the extra hole he had left in the plate for the syndesmotic screw remained conspicuously empty and obvious on post-operative x-rays. Did this omission have a significant impact on the patient’s wellbeing? Perhaps the added screw actually was not needed, and although it became a bone of contention whether or not it was, I for one remained doubtful.

The post-operative x-ray films are all that we have to go by, and these have received mixed reviews. There is an apparent fissure between the medial and rear portions of the tibia, and each side’s expert witnesses offered conflicting opinions as to whether this fissure was too wide and misaligned or close to the best that could be expected. The plaintiff’s experts said that the “step-off” between the two pieces of the tibia was unacceptably off, while the defendant’s reminded us that we were looking at an image that is a bit magnified as well as distorted and that the step off between one piece of the tibia and the other is acceptable and the best that any surgeon could have done given the damage from the original fracture.

A similar problem arises when interpreting the x-ray images of the fibula. Is it too far away from the tibia or is it actually rather close. It looks to the layman’s eye as if there might be a gap, but is it more or less than two millimeters? Evidently, even physicians who are use to looking at x-rays can disagree on this question. The defense argued that if there had genuinely been a loss of syndesmosis then the gap would have gotten wider and wider until it would have been obvious to everyone. This seems not to have happened, leading me, at least, to conclude that there was no evidence that the syndesmosis screw was necessary.

Another screw he did not use would have been fixated through the back of the tibia to secure the posterior malleolus or rear fragment. This was the third break in the trimalleolar fracture. Why he did not fixate this part of the fracture? Should he have done so? This seemed to me to be the most important question in the case. Expert witnesses who testified for the plaintiff argued that the surgeon should have fixated the posterior fragment, while those who testified for the defendant said that he did not need to do so.

There is a rule of thumb that if the posterior fragment is 25 to 30 percent of the malleolar surface of the tibia (the upper part of the socket that covers the talus), it should be fixated or secured with a screw. There were differences of opinion as to how large the fragment was in this case. The radiologist, who reported on the initial x-rays shortly after the ankle fracture, thought that the fragment was considerably more than thirty percent, possibly as much as half of the tibia’s surface, and although the surgeon did not think that it was that large, he admitted that it was at least 25 to 30 percent. Whatever the actual size, it seemed to be large enough that the rule of thumb should have kicked in, and the surgeon might have been expected to fixate the fragment.

In his operative report, the surgeon claims that after he fixated the lateral and medial malleoli, he used fluoroscopy to examine the posterior fragment and concluded that the ligaments (the ones holding the posterior fragment, not to be confused with the syndesmotic ligaments) had led the rear fragment back into its proper alignment. (The technical name this is the “vassal principle.”) He judged this “reduction” (literally, “leading back” rather than the common meaning of the word: “lessening in size or volume”) to be good enough. Perhaps his assumption was that the joint had been so badly damaged in the accident that it was never going to be restored to its pre-accident state. Even so, it might have helped the defendant’s case if, at this point during the operation, he had snapped some still pictures with the fluoroscopic x-ray machine so that there would be a permanent record of what he saw that led him to the conclusion that further repair of the posterior fragment was not needed, but there are no fluoroscopic images from the operation, and another problem is that the posterior fragment does not show up on most of the x-ray films because of the way the foot is turned.

One of the arguments for fixating the rear fragment is that reinforcing it should secure the talus underneath the “malleolar articular surface of the tibia.” This under-surface of the tibia is like the cup of a socket that fits over the blockish-shaped talus. It is the sliding of these two bones against each other (lubricated by cartilage) that makes them a joint that allows movement of the ankle as a whole. If you can imagine the cup-like part of the socket broken at the rear, you can see how someone might suggest that not securing this broken posterior fragment might allow the talus part of the socket to slip out from under the weakened rear edge. In this case, it was the plaintiff’s contention that the weakened rear fragment gave way and the talus slipped (or did it grind its way) backward toward the heel, causing enormous pain and making it difficult for the plaintiff to regain the ability to walk, even after a year and a half of healing and physical therapy. A year and seven months after the accident, an x-ray demonstrated that the talus had, indeed, slid backward under the tibia.
Slide showing talus of ankle joint with tibia lifted away

What else could explain the plaintiff’s condition if it was not due to the surgeon’s failure to fixate the posterior fragment of the tibia? The defendant’s argument seemed to be that the original injury had begun a process of wearing away the cartilage until there was enough room for the talus was able to work its way backward underneath the tibia’s malleolar surface. Is this plausible? One would need to be an expert to determine this, and the defendant had expert witnesses who made such a claim, but the plaintiff’s experts disagreed. The jurors were left to sort this out with a combination of applied common sense and a good deal of head-scratching.

What seemed eminently clear, however, was that there were many deficiencies in the treatment of the patient. Aside from there being no evidence from his records that the surgeon had performed the hook test, he did not have a tangible reason for not fixing the troublingly large posterior fragment. The plaintiff’s attorney might have been unfair in his attempt to plant the narrative that the surgeon had failed to fixate the rear fragment because of fear or stubbornness or even laziness (because he did not want to reposition the patient and make yet a third incision, which was necessary to access the fragment and fixate it), but this attribution of motive seemed to ring truer when we saw how the doctor continued to treat this patient in the months following surgery. There were very few x-rays made after the surgery and two of the x-rays that were made went mysteriously missing. The patient’s repeated complaints of excruciating pain were treated with apparent skepticism by the surgeon who also mentioned in a couple of later reports, half a year after the accident, that it might be a good idea to take a new x-ray, but he never actually ordered one.

Add to this the fact that virtually all of the surgeon’s reports seemed to varying degrees to be plagued with errors, some of which—but perhaps not all of which—could be attributed to abominable “auto-population’ software that imports information from one report to the next, requiring a diligent doctor to spend enormous amounts of his time correcting the resulting software errors. Ironically, this kind of software is intended to save time by filling in repetitive information from previous reports, but if it imports inapplicable information, it is obviously doing more harm than good. In one case the surgeon said that he did notice that the abbreviation “estab” had printed as “Nestab,” and it persisted despite his attempt to correct it.

As a professional proofreader, I could not help but notice that the surgeon made plenty of his own mistakes that he did not seem to notice. Instead of writing that the patient “fell off” a wall, he wrote that she “fell on” a wall. In another instance, he again used the word “on,” this time where he obviously meant “to.” Frequently, words were misspelled. When the jurors considered these kinds of errors, it was a challenge to pull back and remember that the test for negligence might not include spelling, but it might include accurate medical information. Still, what we had to look at was whether the surgeon did what he was supposed to do in his treatment of the patient and whether any negligence on his part “proximally caused” the damage suffered by the plaintiff.

It was difficult to determine whether the surgeon had caused harm to the patient in the way that he repaired the lateral and medial malleoli, but it seemed plausible that he might have caused harm by not fixating the posterior fragment. What the jury was left with was an accumulation of evidence that the surgeon had been careless with his operative and other reports, and that he did not stay on top of the patient’s treatment with, for example, timely x-rays so that he might have caught the slippage of the talus long before he did (which was eighteen months later). How did the talus slip backward? Was this not evidence that the posterior fragment should have been fixated? Even if that is debatable, the fact remains that, by his own admission, the surgeon believed that the posterior fragment was large enough that he should have fixated it. One of his defenses was that there are studies that show that fixation of somewhat large rear fragments is not always necessary, but the surgeon did not know this at the time. Indeed, this, combined with the fact that the talus did slip under the rear fragment, does seem to suggest that this was one of those cases where fixation should have been done but was not.
Every member of the jury thought that the surgeon’s treatment of the patient was at least careless, and carelessness is really either a lower form of negligence or identical to it, depending on how one wants to judge. (Extreme carelessness certainly seems to be another way of describing negligence.) The only question was whether the surgeon’s level of negligence rose to the status of a violation of the standard of care. Essentially this standard says that the surgeon must conform to what a “reasonably prudent” surgeon would do. (Some of the defendant’s expert witnesses defended the notion of minimal competence.)

Deliberations

While two jurors, out of the total of seven, worried that we needed to follow the rule of innocent until proven guilty in judging the defendant, we all knew that the bar in this case was set at “the preponderance of the evidence.” If it looked as if the defendant was negligent to a 51 percent certainty, then we had to find in favor of the plaintiff. I have since learned, through research, that in malpractice cases, courts have ruled that the standard of care is determined by the well-being of the patient rather than what is minimally required of the average practitioner. In a 1974 malpractice case, Helling v. Carey, the Supreme Court sided with the plaintiff, favorably citing an earlier (1903) opinion of Justice Oliver Wendell Holmes that “what usually is done may be evidence of what ought to be done, but what ought to be done is fixed by a standard of reasonable prudence, whether it usually is complied with or not.” 


Justice Oliver Wendell Holmes Jr. (1841-1935)
The Court in that 1974 case thus came down on the side of saying that “reasonable prudence” means going beyond what you would do if you merely looked at the odds and decided that a conservative course of treatment was safe, for example, by not fixating a posterior fragment. In the Helling case, a forty year-old woman developed glaucoma and argued that her ophthalmologist should have tested her for it even though the incidence of glaucoma in forty-year-olds is very low. The Court ruled that he should have tested her for it even though it was unlikely that the test would be positive. I can understand medical practitioners finding such a ruling crazy-making. Should a lawyer’s client sue his lawyer for not anticipating that opposing counsel was going to come up with some whacko legal argument and that an eccentric judge, against all reason, would rule in favor of the client’s adversary? (I maintain that, as a rule, all liability theories ought to be tested by assuming that the defendant might be a lawyer, if for no other reason than that this drives lawyers nuts, but also because lawyers tend to assume that regulations that apply to other professionals should not apply to them.)


In the end, the majority of jurors insisted that there was a sufficient level of negligence to find in favor the plaintiff and to award at least some damages. The plaintiff’s attorney had asked for $900,000, but only one on the jury went as high as $350,000. The lowest number ever suggested was zero dollars, but this was not taken seriously once we agreed to find the defendant negligent and his actions the cause of at least some the plaintiff’s suffering. The amount of damages had to go up and finally stood at $225,000. One might imagine that the higher amount of $350,000 had no effect, but I suspect that it acted as a kind of magnet, preventing the amount of damages from falling below $200,000 and may even have helped to pull the amount up above $200,000, facilitated by the fact that several jurors wanted to award $250,000. I was actually surprised that the final amount was $225,000 because I thought that the number of jurors who advocated $250,000 would prevent a compromise from going lower, but there was a very friendly atmosphere in the jury room, and strong opinions did not lead to intransigence.

Neither party in the suit asked for the jury to be polled, seeming rather resigned to what the jury had decided after more than three hours of deliberation.

Some final thoughts: A jury trial in a medical malpractice case seems characteristic of American jurisprudence and has been for a very long time; however, if this case had taken place in Canada, it would not have been heard by a jury at all. There such cases are always heard by a judge in a bench trial. The attorney for the plaintiff in our case was given to emotional displays, personal comments, and other histrionic interjections that, in front of a lone judge—or even a panel of judges—would have been lost. Such a judge would also have known the law so that a three hour -plus deliberation would have been unnecessary. For this and related reasons, medical malpractice insurance is much cheaper in Canada than it is in the United States.

Another thought I had was that I could not help wondering whether this case would give surgeons the message that they had better be sure to document routine things that they often did without documentation, and that even if they thought that they might do more harm than good by fixating a screw, they had better do it anyway—even if doing so causes harm in a particular case and leads to a different law suit, potentially putting medical professionals in the position of being damned whether they do or do not.

Other issues that were raised but almost as if only to muddy the waters:

1. X-ray films are notoriously difficult to learn how to read. They are two-dimensional representations of three-dimensional objects, after all, and it is easy to miss the actual relationships of object that appear on a “radiographic’ image. Is an object that appears flat actually tilted? Where two or three bones come together, are they touching and if so how?

2. After her last visit to the defendant’s office, the plaintiff almost immediately went to another surgeon for a second opinion. This surgeon recommended and performed an ankle fusion in which the tibia and talus were mechanically resurfaced and fixated together with several screws. The result, according to the plaintiff, has been a reduction if not an elimination of pain and some success in walking without help, although she evidently had to learn over again how to walk. A primary complaint was that her adjusted gait forced her to use her knee and hip unnaturally so as to cause friction and new areas of pain and dysfunction. She maintained in her suit that, not only had the surgeon’s negligence led to a botched surgery, but every contingency since then was his fault, including past and future surgical remedies to make her as whole as possible. She therefore expected her damage award to include north of $100,000 for a total replacement of her ankle with an artificial joint.

Another problem with the ankle fusion was that it essentially destroyed evidence and any chance of looking at the original surgery for purposes of the trial. (It was mentioned a few times in the course of the trial that a CAT scan or other three-dimensional image might have provided helpful evidence, but no such image was ever ordered.) An expert witness for the defendant was asked under cross-examination why he had not directly examined the plaintiff, and he pointed out that the defendant’s surgical competence could no longer be determined by examining the patient but only through reports and x-ray evidence of a surgery that no longer existed.

3. Ankle fractures can be classified by more than one system, only one of which was explained. (Another was mentioned by name but never explained.) The one that was explained is the Weber classification of fractures of the fibula. Three categories in the Weber system are A, B, and C, which represent how high the break is on the fibula. A is below the level of the ankle, B is at the level of the ankle and C is above the ankle.

The surgeon thought that he was dealing with a level C fracture—above the ankle. The trouble is that a long segment of the fibula from the ankle to a point above it had been shattered. If the break had started at the top, then it would have been a level C, but one of the expert witnesses argued that the break was actually a level B, because the break should be classified according to where it started—in this case, he believed, at the ankle—not by its highest point. The issue that this raises is whether the surgeon should have assumed that there was damage to the syndesmosis. If he thought it was a level C, then he should have assumed—and he admitted that he did assume—that there was at least an 85 percent chance that the syndesmosis was damaged. Since it probably was a B fracture, the likelihood was actually much less, but the surgeon did not know that. This spoils the argument that even if the surgeon forgot to do the hook test (not that he conceded that he forgot to do it), it did not matter. The surgeon would have been negligent not to do what he thought he had to do.


4. This is similar to the issue of the fixation of the rear fragment. There are studies that show that fixating it is not that big of a deal, but the surgeon did not seem to know this at the time of the operation, believing instead that it was a big deal, and that the size of the fragment should have persuaded him to fixate it.

No comments:

Post a Comment