Aortic dissection: Be suspicious or the autopsy will make diagnosis
Aortic dissection: Be suspicious or the autopsy will make diagnosis
It was Friday in the emergency department (ED), approaching midnight, and the midnight shift emergency physician had just come on duty. One of the first patients he saw was a woman in her 50s who had come to the ED with a complaint of chest pain. She complained of severe pain in her anterior chest and the interscapular area. The pain radiated into her upper left arm and into the neck. The pain was dull in nature, and she described it as "exactly like when I had my heart attack." She denied any "ripping" or "tearing" character to the pain.
Her electrocardiogram (ECG) was highly suspicious for acute myocardial infarction (AMI) with ST-segment elevations in the inferior leads. While there definitely appeared to be ST-segment elevations, they were relatively minor and not clearly diagnostic of AMI. Her blood pressure was high, not unlike many patients with AMI. She was given sublingual nitroglycerin and intravenous morphine sulfate. Her blood pressure came down slightly, but her pain continued. Laboratory studies had been sent, and a portable chest X-ray had been ordered. Both the laboratory results and the chest X-ray were delayed because it was a typically busy midnight shift on a weekend, and the department was literally overflowing with patients. Radiology, in particular, was particularly slow because of several multiple traumas that required large numbers of X-rays.
The patient’s pain continued, and she was given additional morphine sulfate and was started on a nitroglycerine drip. Still, her pain did not improve. Everything was ready for the thrombolytics — laboratory studies drawn, intravenous lines started, medication mixed, and consent obtained — and the physician and nurses were anxious to proceed, there having already been a significant delay while waiting for the chest X-ray. Because the ECG did not clearly fall within the hospital’s protocol criteria for thrombolytics, the physician consulted with the cardiologist on call. The emergency physician faxed the ECG to the cardiologist. He agreed that, while the ECG was not clearly diagnostic of AMI, in the face of the patient’s known previous AMI, clinical presentation (pain "just like when I had my heart attack"), and abnormal ECG suggestive of AMI, thrombolytics were indicated. The only thing holding up the thrombolytics was the chest X-ray. The patient’s pain was simply not responding to intravenous morphine and nitroglycerine.
Finally, the chest X-ray was done. The thrombolytics were mixed and at the bedside ready to go when the X-ray was brought in. It should come as no surprise after this buildup that the chest X-ray was not normal — it showed a markedly widened mediastinum. There was nothing subtle about the chest X-ray, which was strongly suggestive of aortic dissection. The emergency physician called the cardiologist back, he came in and took the patient to the catheterization laboratory where the diagnosis of aortic dissection was made. The patient was then taken to the operating room for surgical repair of the dissection.
This is a true story, and one that many emergency physicians could tell. It highlights the fact that the clinical diagnosis of aortic dissection is not always obvious, and aortic dissection can easily be confused with AMI. While sudden "tearing" chest pain that radiates to the back may be the classic presentation of aortic dissection, many (perhaps most) patients with aortic dissection do not have such a presentation. This is further complicated by the fact that a patient with aortic dissection might also, as a consequence of the dissection, have an AMI.
It is only natural to relax when you have confirmed the diagnosis that you had suspected clinically (e.g., AMI in a patient with chest pain). After all, you only expect a patient to have one problem at a time. However, in a patient with aortic dissection, an intimal flap may obstruct a coronary artery and lead to AMI. In such cases, missing the primary cause of the AMI and administering thrombolytics will be disastrous.
The consequence of giving thrombolytics to a patient you thought only had an AMI, but actually had aortic dissection alone or aortic dissection associated with AMI, is quite likely to be death. Fortunately, the physician in this case was able to resist the strong urge to give thrombolytics, despite the pressure to give thrombolytics to all AMIs as quickly as possible ("time is muscle"), and he waited for the patient’s chest X-ray to be done.
In this case, the physician was able to make the diagnosis of aortic dissection because the screening chest X-ray was markedly abnormal and, because of the X-ray findings, the physician was able to make a presumptive diagnosis of aortic dissection, confirmed by a definitive diagnostic test (angiography). In some cases, the X-ray findings will be highly suggestive of aortic dissection. Chest X-ray findings are, however, often much more subtle, and aortic dissection is a diagnosis that is missed with alarming regularity. In a recent study of 236 patients with documented aortic dissection, only 72% of clinicians initially suspected aortic dissection.1 An alarming 28% of patients were not correctly diagnosed until autopsy. A diagnostic rate of 72% represents some improvement in the past decade, but, given the consequences of missed aortic dissection, clearly remains much too low. The solution is to increase our index of suspicion when dealing with patients with chest pain and to learn to spot the sometimes very subtle clues of aortic dissection. The low rate of correct diagnosis is all the more worrisome given that: "1) thoracic aortic dissection is the most common lethal disease affecting the aorta, and is two to three times as common as abdominal aortic aneurysm (AAA) rupture; 2) the absolute incidence by autopsy studies has risen two- to fourfold in the last 30 years; and 3) the mortality is as high as 1% to 2% per hour in untreated patients."2 I suspect that most emergency physicians would think that AAA rupture is more common that aortic dissection.
Aortic dissection occurs when blood enters the media of the aorta (the middle layer), usually because of a tear in the intima (the inner lining of the aorta), and splits the wall of the aorta longitudinally. This creates a false lumen in the wall of the aorta that may extend proximally, distally, or in both directions. It is this extension of the false lumen that may lead to decreased flow to, or obstruction of, the various branches of the aorta. This, in turn, leads to many of the secondary symptoms of aortic dissection (e.g., neurological symptoms due to decreased flow through the carotids).
Most patients (> 90%) with aortic dissection experience sudden, severe chest pain.3 While the classic description of aortic dissection pain is "ripping" or "tearing," at least one study has shown this type of pain to be relatively rare.4 It was, for example, not present in the patient whose case history was presented above.
The cases presented below all involve aortic dissections that were allegedly "missed." These cases provide lessons that will be useful in avoiding the risk of "missed" aortic dissection (e.g., handling of the "rule out AMI" patient). They also provide reinforcing examples of some of the legal and risk management principles that have been discussed previously in ED Legal Letter (e.g., X-ray follow-up procedures). Finally, some of those cases introduce some interesting legal issues we have not previously addressed (e.g., assumption of duty and liability of consultants contacted by telephone). While the cases below all involve aortic dissection, the legal principles discussed are of general applicability.
Liability Requires Causation
Case No. 1: Carmen C. Gomez v. Tri City Community Hospital, Ltd. d/b/a Tri-City Community Hospital.5
In this case, the Court of Appeals of Texas dealt with the issue of causation in a case involving a patient who died of aortic dissection. Recall that, in order to prevail in a medical malpractice action, a plaintiff must prove all the following elements:
1) duty: i.e., a physician-patient relationship;
2) negligence: i.e., breach of the applicable standard of care;
3) proximate causation: i.e., the breach of the standard of care must be a cause of the patient’s damages, sufficiently close in the sequence of events leading up to the injury that the physician should be held responsible;
4) damages: i.e., the plaintiff must have suffered a compensable injury.
On April 20, 1995, Jeronimo Carrasco was brought by ambulance to the ED at Tri City Community Hospital in Jourdanton, TX, with a complaint of back pain. He was admitted for "observation," and released the following day, April 21, 1995. At the time of his release, he was still complaining of back pain.
On April 22, 1995, Mr. Carrasco returned to the hospital with a complaint of continuing back pain and an inability to stand up. In addition, he now complained of not having had a bowel movement for the past four days. He was admitted to the hospital. A chest X-ray was done in the course of his evaluation on April 22 that revealed a "significantly widened mediastinum," and "an increase in the size of the cardiac silhouette."6
On April 24, two days after admission, a radiologist reviewed the chest X-ray and dictated his report. In addition to the report of his findings, as indicated above, the radiologist stated that: "In the setting of back pain, consideration should be given for aortic dissection."7 The report also contained a note that: "Ward [was] notified 4-24-95."8
Sometime on April 24, Mr. Carrasco’s condition deteriorated, and he was transferred by air to the Methodist Hospital in San Antonio. A CT scan performed there revealed that Mr. Carrasco was suffering from a Type 1 dissection of the thoracic aorta. He was taken to the operating room emergently where he was found to have a ruptured thoracic aorta dissection. The dissection was repaired with a synthetic graft, and Mr. Carrasco was "fairly stable following the surgery."9
The following day, Mr. Carrasco suffered another pericardial tamponade, and emergency surgery revealed a "new bleeding site into the pericardium" and "abundant blood" in the left chest.10 Mr. Carrasco arrested and resuscitative efforts were unsuccessful.
The plaintiffs sued Tri City Community Hospital and the emergency physician who had treated Mr. Carrasco. The plaintiffs settled with the emergency physician, and the lawsuit proceeded against the hospital. Prior to trial, the hospital moved for summary judgment, and the trial court granted a no-evidence summary judgment on a finding that the plaintiffs had failed to introduce any evidence of causation. That is, while the hospital may have been negligent in its treatment of Mr. Carrasco, and Mr. Carrasco had obviously suffered damages (death), the plaintiffs had failed to introduce any evidence that sufficiently linked the negligent acts of the hospital to Mr. Carrasco’s death. The judge, therefore, refused to allow the case to proceed to trial. The plaintiffs appealed this ruling to the Court of Appeals of Texas.
It is important to understand the standard of review in a case such as this and, as a result, how difficult it can be for a defendant to prevail on a motion for summary judgment. Short of getting a plaintiff to drop a lawsuit before, summary judgment is the next best alternative. In a motion for summary judgment, the moving party (usually the defendant) argues that there is no issue of fact for a jury to consider and that the court should, as a matter of law, dismiss the case. In considering such a motion, the judge is obligated to look at all the evidence in the light most favorable to the opposing party (in this case, the plaintiffs) and to grant the motion only if a reasonable jury could not find for the nonmoving party.
On appeal, the appeals court also will "look at the evidence in the light most favorable to the [party] against whom the summary judgment was rendered, disregarding all contrary evidence and inferences."11 This obviously gives the appellant (the plaintiffs in our case) a huge advantage in the court’s weighing of the evidence. With the entire benefit of the doubt in the plaintiff’s favor, he should prevail if he can introduce any evidence that a juror might construe as establishing his case.
Exactly how much evidence is it necessary for a plaintiff to bring forward in order to survive a summary judgment motion? Not much in this case: "A no-evidence summary judgment is improperly granted [i.e., the trial court should be reversed] if the respondent [plaintiff in our case] brings forth more that a scintilla of probative evidence to raise a genuine issue of material fact."12 The court went on to define "more than a scintilla" as any evidence that "rises to a level that would enable reasonable and fair-minded people [i.e., the jurors] to differ in their conclusions."13 That is, if the plaintiffs had introduced any evidence that, if viewed in a light most favorable to the plaintiffs, would be sufficient for reasonable people on a jury to differ in their conclusion, the case should have been allowed to proceed to trial. Having restated the accepted rule for consideration of the appeal, the court proceeded to search for the requisite "more than a scintilla" of evidence.
The plaintiffs had introduced evidence, in the form of their expert’s affidavit, that, had the diagnosis been made on April 22, the day of admission, and Mr. Carrasco been taken to surgery at that time, he would have had a greater chance of survival. It was clear from the medical record that the X-ray was not read until April 24, two days after Mr. Carrasco’s admission. According to the court, there was sufficient evidence for reasonable people to conclude that, had the report of the chest X-ray been provided to Mr. Carrasco’s doctors on April 22, the correct diagnosis would have been suspected at that time. Based upon the medical record and the plaintiffs’ expert witness’ affidavit, there was evidence that there was a breach of the standard of care — the failure to read and communicate the results of the chest X-ray on the day of admission. The court subsequently was able to link the negligent act and injury (i.e., establish proximate causation) from the evidence by making the following inferences: 1) had Mr. Carrasco’s physicians received the report on April 22, they would have considered the correct diagnosis; and 2) had the diagnosis been made on April 22, the surgery would have been elective (the rupture likely did not occur until April 24 when the patient’s condition suddenly deteriorated). In addition, there was evidence that Mr. Carrasco would have had a greater chance of surviving had he been taken to the operating room on April 22. The grant of summary judgment for the trial court was reversed, and the case was remanded for trial.
Commentary: In defending a malpractice case, all elements of the plaintiff’s case that can be attacked, are attacked. If the facts of the case make it difficult to defend the alleged breach of the standard of care, the case is not lost. The plaintiff also must prove that the breach of the standard of care proximately caused the patient’s injury.
The "Rule Out AMI" Patient
Case No. 2: Mindy Sommers v. Dr. Lisa Friedman and Wisconsin Patients Compensation Fund.14
Jay Sommers, the husband of the plaintiff, Mindy Sommers, was admitted to St. Mary’s Hospital in Madison, WI, on Oct. 7, 1986, after complaining of the sudden onset of chest pain. Dr. Lisa Friedman, a member of the St. Mary’s medical staff, became his primary treating physician. After conducting initial tests (editor’s note: the hospital’s "rule out AMI" protocol?), Dr. Friedman concluded that Mr. Sommers had not suffered an AMI. She was, however, unable to diagnose the cause of his pain. She consulted other physicians, including a cardiologist, and further tests were performed, including a treadmill/stress test and an upper gastrointestinal series, both of which were normal.
Following those consultations and the completion of the ancillary tests, Dr. Friedman informed Mr. Sommers and his wife that "heart attack, angina, and gastrointestinal problems had been ruled out as causes of his pain."15 On Oct. 12, Dr. Friedman discharged Mr. Sommers, providing him with copies of all his medical records and urging him to see a physician upon return to his home in Arizona. Unfortunately, Mr. Sommers died a few hours later as a result of aortic dissection.
Mindy Sommers sued Dr. Friedman and claimed that she was negligent in failing to diagnose and treat her husband’s aortic dissection. After six days of testimony, the jury found Dr. Friedman not liable. The plaintiff appealed. Among the issues raised on appeal was the trial court’s admission of the testimony of three physicians, two residents, and a cardiologist, who testified that they, too, had examined Mr. Sommers and had taken his history and reached the same conclusions as Dr. Friedman. That is, they had not suspected aortic dissection either. Plaintiff’s counsel objected to the admission of this testimony on the grounds that it was not relevant and was highly prejudicial in that it "did not bear upon whether Dr. Friedman exercised the applicable standard of care and skill, but merely provides [her] an escape from responsibility based upon the understandable reaction by the jury that if [other] doctors failed to diagnose the ailment, then Dr. Friedman, by default, must have met the standard of care.’"16
The appeals court discussed the basic rules regarding the admission of evidence. The basic rule is that evidence will generally be admitted if it is "relevant," that is, any "evidence having any tendency to make the existence of any fact that is of consequence to the determination of the action more probable or less probable than it would be without the evidence."17 However, evidence will be excluded, even if relevant, if it is "prejudicial": "if it has a tendency to influence the outcome [of the case] by improper means’ or if it appeals to the jury’s sympathies, arouses its sense of horror, provokes its instinct to punish,’ or otherwise causes a jury to base its decision on something other than the established propositions in the case.’"18
The trial court had ruled that the testimony of the other doctors "was relevant to the degree to which Dr. Friedman considered all of the medical information available to her when she diagnosed and treated Jay Sommers."19 In addition, the testimony of the doctors was relevant to the defense’s assertion that "an aortic dissection is a relatively obscure illness.’"20 According to the trial court, the testimony, therefore, was relevant, and the appeals court agreed.
The trial court agreed with plaintiffs’ counsel that there was "potential for misuse of this evidence by the jury to conclude that [because] these other doctors did not make the appropriate diagnosis either that that [fact] might be taken as a basis to relieve Dr. Friedman of her responsibility to exercise the requisite standard of care."21 The court then undertook a balancing of the probative value of the evidence, relative to its relevant purposes, vs. its improper possible prejudicial effect. The trial court had concluded that a limiting instruction would sufficiently reduce the possibility of prejudice and had admitted the evidence. The trial court’s limiting instruction was as follows: "With regard to diagnoses and opinions received from other treating physicians, you may consider that testimony as it relates to the quality of medical care actually given to Jay Sommers, but you are instructed that Dr. Lisa Friedman is obligated to provide medical care consistent with the standard of care just described irrespective of the opinions of other treating doctors."22 The appeals court agreed that this limiting instruction had been sufficient to avoid the possible misuse of the testimony by the jury and affirmed the decision below.
Commentary: We admit thousands of patients to "rule out AMI," often to dedicated chest pain units where relatively rigid protocols are followed to rule out AMI (serial cardiac enzymes, monitoring, and stress testing). Often, it is emergency physicians who staff those units and make the ultimate decision whether to admit or discharge the patient. When making this decision, it is crucial to not focus narrowly on whether AMI has been ruled in or out. A narrow focus on AMI (and unstable angina) as the only possible diagnosis (i.e., patients are essentially automatically discharged from the chest pain unit if they fulfill the rule out protocol’s criteria), creates a risk of sending home patients who did not have AMI, but had other serious diagnoses (e.g., aortic dissection and pulmonary embolism).
Assumption of Duty
Case No. 3: Schmitz v. Blanchard Valley OB-GYN Inc., et al.23
While this was not an ED case, this 1989 case involved a discussion of the "assumed duty" doctrine, a doctrine that has particular relevance to emergency physicians, especially when treating complex disorders. In this case, the Court of Appeals of Ohio considered the plaintiff’s appeal, which was based on the trial court’s refusal to give a requested jury instruction on assumption of duty.
The defendant physicians cared for Madonna Schmitz during her pregnancy and subsequent delivery at Blanchard Valley Hospital in Findlay, OH. Ms. Schmitz, who had a history of coarctation of the aorta and other cardiovascular disorders, developed hypertension during her pregnancy. It was the defendant physicians’ failure to consult a cardiologist that was the basis of the plaintiff’s claim.
On Aug. 10, 1982, defendant Dr. Allan Tong determined that Ms. Schmitz was suffering from an elevated blood pressure that he diagnosed as pregnancy-induced hypertension. He instructed her to "get bed rest."24 Subsequently, Ms. Schmitz’s blood pressure returned to normal. On Sept. 10, 1982, Dr. Tong again found Ms. Schmitz’s blood pressure to be elevated at the time of an office visit and again prescribed bed rest. On Sept. 13, 1982, defendant Dr. Emil Zeigler found that Ms. Schmitz’s blood pressure was still elevated and that she had developed edema. He admitted her to the hospital. On the evening of admission, Ms. Schmitz went into labor and delivered a 4-pound, 8-ounce baby girl. Unfortunately, on Sept. 15, 1982, two days following the birth of her daughter, Ms. Schmitz died, the result of an aortic dissection.
At trial, the plaintiff requested the court to instruct the jury on the "assumed duty" doctrine. According to the plaintiff’s requested jury instruction, the doctrine provides that, when a physician has a duty to obtain a consultation from a specialist and fails to do so (in this case a cardiologist), the physician then, as a result of failing to obtain the indicated consultation, assumes the duty of caring for the patient consistent with the standards of care applicable to the specialist (in this case, cardiology). The court refused this proposed jury instruction, and the jury subsequently returned a verdict in favor of the defendants. The refused jury instruction was the sole claim of error cited by the plaintiff on appeal.
The court of appeals, relying on an earlier Minnesota case, explained that liability based upon the assumed duty doctrine requires more than simply a breach of the duty to consult a specialist. There is a second element to the doctrine. The consequence of failing to obtain an indicated specialty consultation is that the physician is held to the standard of care of the specialist. It is the standard of care of the specialist that provides the basis for the second element of the doctrine. In order to prevail on a theory of assumed duty, the plaintiff must show that the physician not only failed to obtain the specialty consultation, but also failed to meet the standard of care of the specialist (in this case cardiology):
It is important to note, however, that the mere breach of duty to refer a patient to a specialist for treatment will not of itself make out a prima facie case of negligence against the general practitioner. *** It must appear that the breach of the duty to refer to a specialist in fact caused the plaintiff’s injury, and this can be shown only if the treatment the plaintiff received was in some way inferior to the treatment he would have received from a specialist. Thus, in order to make out a case of negligence based on a breach of duty to refer a patient to a specialist for treatment, the plaintiff must also present evidence from which the trier of fact may determine that in the treatment which he in fact administered, the defendant failed to exercise that degree of skill, care, knowledge, and attention ordinarily possessed and exercised by specialists in good standing under like circumstances.25
The court held that, even if it were assumed that the physician had a duty to refer Ms. Schmitz to a cardiologist and the defendants breached that duty, the plaintiff failed to show that the treatment Ms. Schmitz received was in some was inferior to the treatment she would have received from a cardiologist. In fact, plaintiff had fallen far short of the burden of proof in that he had failed to even introduce evidence at trial that would support a conclusion that a cardiologist would have been better qualified to manage the care of Ms. Schmitz under the circumstances.
Commentary: In a specialty such as emergency medicine, where we truly do see everything, it is crucial that we recognize our limits and make appropriate use of all specialists available to us.
The Chest X-ray in Aortic Dissection
Case No. 4: Alba Fernandez, et al. v. Corporacion Insular de Seguros, et al.26
On Nov. 4, 1991, Hiram Fernandez was taken to the emergency department of Federicao Trilla Hospital in Carolina, Puerto Rico, after awakening with complaints of chest pain, severe leg cramps, and vomiting. According to Mr. Fernandez, who was 64 years old, the chest pain had stopped after he vomited, but the severe leg pains had not abated. Mr. Fernandez failed to tell the emergency physician that he was taking medications for asthma and hypertension. Based upon the fact that the chest pain had resolved, all vital signs were normal or within a borderline range, Dr. Pedro Rivera Bermudez tentatively made a diagnosis of a "pinched nerve," but nonetheless ordered an ECG, a urinalysis, and an X-ray of the lumbosacral spine and the left leg.27 All of these tests were normal.
A few hours later, Dr. Rivera was relieved in the ED by Dr. Ricardo Martinez Cortinez, whose examination confirmed that Mr. Fernandez was stable and had no complaint of chest pain. Because Mr. Fernandez’ white blood count was slightly elevated, Dr. Martinez ordered a second urinalysis, as well as a chest X-ray, to rule out any urinary or pulmonary infection that might have been causing the vomiting. The chest X-ray revealed that Mr. Fernandez had a "dilated aorta, which Dr. Martinez attributed to normal borderline hypertension in a patient of Fernandez’s age."28 Like Dr. Rivera, Dr. Martinez diagnosed a pinched nerve. Dr. Martinez referred Mr. Fernandez to a neurologist and discharged him.
The next day, Mr. Fernandez was reportedly unable to recognize his family members. He was taken to his personal physician, Dr. Abelardo Vargas, who examined him, repeated the ECG, which was normal, and "ordered an upper gastrointestinal series to determine whether there was a hiatal hernia," which was to be performed the following day.29 Unfortunately, on the way to undergo the upper gastrointestinal series, Mr. Fernandez collapsed and died. An autopsy revealed that he had died as the result of an aortic dissection.
The family sued the two emergency physicians, Dr. Vargas, and the hospital. After trial, the jury returned a verdict for the defendants. The plaintiff’s appeal on various grounds was subsequently denied.
Commentary: While the defendants prevailed in this case, a different jury certainly might have found the defendants liable on those facts. There are a number of important risk management lessons to be learned from this case. First, this is yet another case where a second emergency physician may have overly relied upon the conclusion of the physician who had previously seen the patient. This problem is seen with both return visits and, as in this case, at shift change, if patients are transferred from one physician to new physician coming on duty. When a physician begins his or her shift, if patients are accepted from the previous physician, it is critically important to discuss the patient and agree upon exactly what role the second physician is to assume. As the second physician, unless your role is limited to, for example, making sure a final laboratory test is negative before the patient is discharged, it is critical to evaluate the patient for yourself. Do not rely exclusively on the evaluation by the previous physician. You should always "start fresh," and try to avoid being unduly influenced by the judgment of the previous physician.
The fact that you know that the physician who saw the patient earlier is an excellent physician is irrelevant. Even brilliant physicians make mistakes. In addition, the physician might have made a very reasonable, although ultimately incorrect, judgement based upon the presentation of the patient the time he or she was seen earlier and now, with the passage of some time, the patient’s problem might have progressed to a point where the correct diagnosis can be made.
Shift change is a dangerous time for patients in the emergency department, as well as in other areas of the hospital. The transfer of care of patients from the outgoing physician to the oncoming physician compounds the risks. Some departments control this risk by not transferring patients from one emergency physician to another. That is, if a physician starts a patient, that physician finishes the patient. If your department allows the transfer of patients, the safest procedure is for the oncoming physician to re-evaluate all patients "from scratch."
Second, always be wary of a widened mediastinum on chest X-ray. This is the most common chest X-ray finding in aortic dissection. In this case, the patient was 64 (certainly within the age range in which aortic dissection should be considered), had chest pain (apparently with radiation to his legs), and hypertension. Although the patient apparently failed to volunteer that he had hypertension, this information might have been available to the physicians. In short, this was a man at high risk of aortic dissection, and a definitive imaging study was certainly in order.
A normal, upright chest X-ray does not rule out aortic dissection. It is, however, the appropriate screening test for the diagnosis. Even in those cases in which you have a very high index of suspicion that the patient has aortic dissection and intend from the onset to obtain a definitive diagnostic test, a chest X-ray undoubtedly also will be part of the patient’s initial evaluation. The screening upright chest X-ray is abnormal in 80%-90% of patients with aortic dissection. In most cases, a normal upright chest X-ray will sufficiently screen for aortic dissection in the work-up of a patient who will receive thrombolytics for AMI.
Chest X-ray findings suggestive of aortic dissection include:
• mediastinal findings;
• widened mediastinum;
• extension of aortic shadow beyond calcified wall;
• blurred aortic knob or localized bulge;
• aortic enlargement;
• double density of the aorta (false lumen less radiopaque);
• loss of space between aorta and pulmonary artery;
• right side of film;
• deviation of trachea/nasogastric tube to the right;
• shift and elevation of right mainstem bronchus;
• deviation of right paraspinal line;
• left side of film;
• depressed left mainstem bronchus;
• new pleural effusion;
• apical cap (localized apical hemothorax).30
Of the chest X-ray findings that are suggestive of aortic dissection, a widened mediastinum (greater than 8 cm) is the most sensitive — approximately 75% of chest X-rays in patients with aortic dissection will exhibit this finding.31
Liability of On-Call Physicians
Case No. 5: Linda J. McKinney, et al. v. Frances C. Schlatter, MD, et al.32
Mr. Lanny McKinney arrived at the ED of Middletown (OH) Regional Hospital on Feb. 11, 1994, at approximately 4 a.m. He complained that he began experiencing chest pain at approximately 12:30 a.m., which was followed by the onset of abdominal pain. He was seen by the emergency physician, Dr. Fran Schlatter, who gave him two doses of Maalox, which relieved his chest pain. Dr. Schlatter consulted by telephone with Dr. Joseph Solomito at approximately 5 a.m., and again briefly at approximately 6 a.m. regarding Mr. McKinney’s diagnosis. During both of those telephone consultations, Dr. Solomito told Dr. Schlatter that he "did not think the cause of McKinney’s pain was cardiac in nature."33
Dr. Schlatter contacted Mr. McKinney’s family physician at approximately 7:30 a.m., and he said that he would see Mr. McKinney later that morning. Mr. McKinney was discharged from the ED at approximately 8:30 a.m. Mr. McKinney died later that morning, prior to seeing his family physician, due to aortic dissection.
A medical malpractice lawsuit was brought against Dr. Schlatter and Dr. Solomito. The initial trial resulted in a directed verdict for Dr. Solomito (i.e., the court entered a verdict for Dr. Solomito without allowing the jury to consider the issue) and a jury verdict for Dr. Schlatter. The basis for the court’s directed verdict was its ruling that, because all Dr. Solomito did was consult by telephone and never actually saw the patient, there was no physician-patient relationship. Therefore, because Dr. Solomito owed no duty to Mr. McKinney, he could not be held liable for his death.
The Court of Appeals of Ohio reversed that decision holding that:
A physician-patient relationship can exist by implication between an emergency room patient and an on-call physician who is consulted by the patient’s physician but who has never met, spoken with, or consulted the patient when the on-call physician 1) participates in the diagnosis of the patient’s condition; 2) participates in or prescribes a course of treatment for the patient; and 3) owes a duty to the hospital, staff, or patient for whose benefit he is on call.34
The court of appeals then remanded the case for further proceedings as to Dr. Solomito to determine whether a physician-patient relationship existed between him and Mr. McKinney under the facts of the case.
A second jury trial, commencing on March 30, 1998, had Dr. Solomito as the sole defendant. At issue in determining whether Dr. Solomito owed a duty to Mr. McKinney was the third element in the appeals court’s holding: "Owes a duty to the hospital, staff, or patient for whose benefit he is on call." That is, was Dr. Solomito on call for the hospital or Mr. McKinney when he was called and participated in his treatment. Testimony on this point was conflicting. Dr. Solomito acknowledged that he was on call for his practice on Feb. 11, 1994. However, he did not state that he was on call for the hospital or its ED. Dr. Schlatter, on the other hand, testified that she called Dr. Solomito because he was the cardiologist on call.
After the conclusion of testimony, Dr. Solomito’s counsel moved that certain call records from the hospital be admitted into evidence. The plaintiff’s counsel objected to the late admission of the call records, but the court overruled the objection and admitted them into evidence. These records indicated that Dr. Solomito’s practice had been on call for cardiology on Feb. 10, 1994, but that another cardiology group had been on call on Feb. 11, 1994. Dr. Schlatter’s calls to Dr. Solomito were made at approximately 5 a.m. and 6 a.m. on Feb. 11. Following trial, the jury returned a defense verdict. The plaintiff appealed the judgment claiming that admission of the call records after the conclusion of testimony was prejudicial error. The Court of Appeals of Ohio rejected this argument and affirmed the judgment.
Commentary: This case was interesting in that it turned on an interpretation of the hospital’s call records. Because the call records were ambiguous on their face as to when call started and ended and, because the records were admitted after the close of testimony, there was no testimony as to how the call schedule was actually implemented, the court had to interpret the call records. In so doing, the court of appeals almost certainly reached an incorrect conclusion: "The one-page calendar does not indicate any beginning and ending time for the days, therefore the 24-hour period applies."35 That is, the court ruled that call began and ended at midnight, a call schedule that I have never experienced. The result was that Dr. Solomito could not be liable because, when the calls were made to him (5 a.m.-6 a.m. on Feb. 11), he was, according to the court, not on call for the emergency department. He, therefore, owed no duty to the plaintiff. If the jury considered the fact that the emergency physician had specifically called Dr. Solomito to be evidence that he was on call, they apparently found it unconvincing. The appeals court apparently found it irrelevant.
X-ray Follow-up
Case No. 6: Gordon v. James Liquori, MD, et al.36
On Jan. 2, 1988, Richard Gordon experienced severe chest pain and was admitted to Mesa Lutheran Hospital. He was under the care of Dr. James Liquori, a cardiologist. Dr. Liquori initially believed that Mr. Gordon’s chest pain was caused by an AMI, esophageal spasms, pneumonia, pericarditis, or a dissecting aortic aneurysm. By the following day, Dr. Liquori had determined, based upon the laboratory tests and X-rays, that the chest pain probably was associated with pneumonia or pericarditis.
On Jan. 5, an echocardiogram was performed. Although the echocardiogram showed an abnormally large aortic root, this was not noted by Dr. Liquori when he first reviewed the echocardiogram on Jan. 6. On Jan. 6, Dr. Liquori again reviewed the echocardiogram and concluded that it showed a "markedly dilated ascending thoracic aorta, which indicated aortic abnormalities."37 Because of the abnormal echocardiogram, he ordered a CT scan of Mr. Gordon’s chest, which was performed on Jan. 7 at 3:42 p.m. That same day, Dr. Neal Junck, a radiologist employed by the hospital, "interpreted the CT scan to reflect an abnormality in the aorta without concluding that it was caused by a dissecting aortic aneurysm."38 Dr. Junck dictated his report, but did not immediately contact Dr. Liquori. The parties disputed whether Dr. Liquori had told Dr. Junck to contact him immediately if the CT scan suggested an aortic abnormality.
The next morning, Jan. 8, Dr. Liquori read Dr. Junck’s report and ordered an aortic angiogram, which showed that Mr. Gordon had a "dissecting aortic aneurysm."39 Mr. Gordon was taken to surgery at 2 p.m. that day but, during anesthetic induction, his aorta ruptured and he suffered "profound circulatory impairment."40 Although the surgeon was able to repair the aorta, Mr. Gordon was left in a vegetative state and died on Jan. 20.
Mr. Gordon’s survivors filed a medical malpractice compliant against Dr. Liquori and the hospital. They alleged that an earlier diagnosis of the aortic dissection would have improved Mr. Gordon’s chance of survival and that Dr. Liquori had been negligent in not making the diagnosis in a more timely fashion. As to the hospital, the plaintiffs alleged that Dr. Junck, a hospital employee, had failed to promptly communicate the abnormal results of the CT scan to Dr. Liquori, and that the nurses had failed to provide the radiology department with pertinent clinical data that would have alerted Dr. Junck to the urgency of the situation.
In response to interrogatories, Dr. Liquori indicated answered that: "It is [the] Defendant’s position that any alleged delay in diagnosing Mr. Gordon’s dissection probably did not play a causative role in his death. However, in the event there is evidence presented that it did, Defendant contends that Dr. Junck contributed to the delay by not advising Dr. Liquori of the abnormal CT scan, in accordance with Dr. Liquori’s specific request, which failure resulted in an approximate 12-hours [sic] delay."41 The hospital, on the other hand, indicated that its cardiology expert would testify that Dr. Liquori was negligent in not making a timely diagnosis. That is, it appeared during discovery that the two defendants would blame each other.
During the week before trial, Dr. Liquori filed a motion in limine to preclude plaintiffs from introducing any evidence concerning the defendants’ contingent expert witnesses, each of whom would criticize the other defendant. A motion in limine is a pretrial motion requesting the court to prohibit the opposing counsel from referring to or offering evidence on a matter that is, arguably, so prejudicial that curative instructions would not prevent the prejudice. The parties had both decided that, if the case went to trial with both defendants still in the case, neither defendant would call its standard of care experts to testify against the other. The court granted Dr. Liquori’s motion and ruled that the plaintiffs could not comment on the defendants’ experts’ failure to testify. It was this ruling that formed the basis of the plaintiffs’ appeal after the jury returned a verdict in favor of both defendants.
On appeal, the plaintiffs’ argued that the trial court erred in prohibiting them from:
1) offering evidence that defendants had reached an undisclosed agreement to not present evidence of each other’s negligence, and that the failure to disclose their agreement until three weeks before trial "worked a fraud upon the court"
2) commenting on the defendants’ failure to call these experts at trial.
The court of appeals held that:
1) had the defendants entered into the alleged agreement, and there was no evidence introduced at trial of such an agreement, it would not have affected the integrity of the trial (i.e., there was no "fraud upon the court");
2) the trial court did not err in not allowing the plaintiffs to comment on the defendants’ failure to call their expert witnesses at trial. The judgment below in favor of the defendants was affirmed.
The court noted that a trier of fact (the jury in this case) may draw an adverse inference from a party’s failure to present testimony only under limited circumstances. Factors to be considered by a trial court when deciding whether to allow comment regarding an uncalled witness are: "1) whether the witness was under the control of the party who failed to call him or her . . . 2) whether the party failed to call a seemingly available witness whose testimony it would naturally be expected to produce if it were favorable . . . and 3) whether the existence or nonexistence of a certain fact is uniquely within the knowledge of the witness . . ."42 These factors are to be applied in the conjunctive ("and"), rather than the dysjunctive ("or"). That is, the party wishing to comment on the failure to present the testimony must satisfy all of the factors. According to the court, in this case, the defendants’ uncalled expert witnesses "arguably meet the first two factors outlined above, they clearly do not meet the third."43 The trial court, therefore, was correct in not allowing plaintiffs to comment on the failure of the defendants to present their own opposing expert witnesses.
Risk Management Strategies
1. Think about vascular possibilities. Many of the missed diagnoses that result in malpractice cases for emergency physicians are vascular in nature. It is, therefore, only prudent to liberally consider those diagnoses (aortic dissection, abdominal aortic aneurysm, pulmonary embolism, subarachnoid hemorrhage, and coronary artery disease).
2. Maintain a high index of suspicion for aortic dissection. If you do not think about aortic dissection, you will not diagnose it with acceptable sensitivity. We know that as many as 65% of aortic dissections initially are missed. If you do not have a significant number of negative results when ordering CT scans (or other definitive studies) for the purpose of ruling out aortic dissection, you are not looking hard enough for the diagnosis and will miss it.
3. Be especially careful when a patient has chest pain and has one or more risk factors for aortic dissection. There are many conditions that result in an increased risk of aortic dissection:
• increased age — peak incidence > 50 and <70 years;
• male sex (3:1);
• history of hypertension (60%-90% of patients);
• connective tissue disorders (Marfan’s, Ehlers-Danlos);
• Turner’s syndrome;
• aortic coarctation;
• third trimester of pregnancy;
• congenital bicuspid or unicuspid aortic valve;
• Ebstein’s anomaly;
• aortic valve stenosis;
• familial incidence;
• illicit drugs (cocaine and methamphetamine);
• iatrogenic (surgery or cardiac catheterization);
• trauma (though blunt trauma usually produces aortic rupture).44
4. Never give thrombolytics for AMI without obtaining at least an upright chest X-ray to screen for aortic dissection. If the chest X-ray is suggestive of aortic dissection, obtain a quick definitive diagnostic test (computed tomography [CT] or transesophageal echocardiograph [TEE]). Even if the screening chest X-ray appears normal, a quick definitive test also is indicated when the clinical suspicion of aortic dissection is high.
5. Remember that most patients with aortic dissection have an abnormal chest X-ray, but a normal chest X-ray does not exclude the diagnosis. As many as 90% of patients with aortic dissection will have an abnormal chest X-ray.
6. When aortic dissection is suspected, mobilize appropriate consultants on a STAT basis. Consult cardiovascular surgery as soon as the diagnosis is made or, when you have any significant clinical suspicion of aortic diagnosis, even before the diagnosis is definitively made.
7. When transfer for definitive care is appropriate, arrange the transfer expeditiously. Transfer protocols should be in effect.
8. If aortic dissection is suspected, immediately obtain a definitive diagnostic test (TEE, angiography, CT, or magnetic resonance imaging) and report the results of that test, if positive, on a STAT basis.
9. When evaluating patients with chest pain, never forget that the definitive diagnosis of AMI does not rule out the concurrent presence of an aortic dissection.
10. When the diagnosis of aortic dissection is made, begin appropriate medical treatment immediately (blood pressure control). Remember, if using a vasodilator such as nitroprusside, administer an intravenous beta-blocker before or simultaneously with the vasodilator so as to avoid reflex tachycardia.
11. Do not restrict your consideration of aortic dissection to patients who have "ripping" or "tearing" pain. While this description of pain may be considered "classic" for aortic dissection, it is often not present.
12. Have a protocol in place ahead of time so that there is no confusion as to what definitive diagnostic tests are available, and which test will be used under the various possible circumstances. If obtaining a definitive diagnostic test requires the patient to be transferred, have appropriate transfer protocols in place.
13. Be especially certain to consider aortic dissection when chest pain is associated with neurologic or vascular symptoms and/or signs.
14. Aortic regurgitation in a patient with chest pain should trigger heightened concern of the possibility of aortic dissection.
15. If a patient has chest pain radiating to the back, or from the back to the front, be especially concerned about the possibility of aortic dissection.
16. A physician should expect to have a significant negative workup rate for aortic dissection. Just as a physician who never sends a patient to the operating room with a normal appendix will miss appendicitis, a physician whose work-up rate for possible aortic dissection approaches 100% will miss the diagnosis.
Endnotes
1. Spittal PC, Spittal JA Jr, Joyce JW, et al. Clinical features and differential diagnosis of aortic dissection: experience with 236 cases (1980 through 1990). May Clin Proc 1993; 68:642-651.
2. Hals G. Acute Thoracic aortic dissection: Current evaluation and management. Emerg Med Rep 2000; 21:1.
3. Bourland MD. Aortic dissection. In: Rosen P, et al., eds. Emergency Medicine: Concepts and Clinical Practice, 4th edition. Mosby-Year Book; 1998:819-1,825.
4. Armstrong F, Bach DS, Carey LM, et al. Clinical and echocardiographic findings in patients with suspected acute aortic dissection. Am Heart J 1998; 136:1,051-1,060.
5. 4 S.W.3d 281 (Texas App. 1999).
6. Id. at 282.
7. Id.
8. Id.
9. Id. at 283.
10. Id.
11. Id.
12 Id.
13. Id.
14. 493 N.W.2d 393 (Wisc. App. 1992).
15. Id. at 462.
16. Id. at 463-464.
17. Id. at 466, fn 4.
18. Id. at 464, fn 3.
19. Id. at 466
20. Id.
21. Id. at 465.
22. Id. at 467.
23. 580 N.E.2d 55 (Ohio App. 1989).
24. Id. at 56.
25. Id. at 57-58.
26. 79 F.3d 207 (3rd Cir. 1996).
27. Id. at 208.
28. Id.
29. Id.
30. Hals G. Acute thoracic aortic dissection: The clinical challenges of acute thoracic aortic dissection. Emerg Med Rep 2000; 21:2, 13.
31. Id. at 12.
32. 1999 Ohio App. LEXIS 81.
33. Id. at *1.
34. Id. at *2-3.
35. Id. at *4, fn 2.
36. 895 P.2d 523 (Ariz. App. 1995).
37. Id. at 524.
38. Id.
39. Id.
40. Id. at 525.
41. Id.
42. Id. at 527.
43. Id.
44. Hals, supra note 2, at 5.
Subscribe Now for Access
You have reached your article limit for the month. We hope you found our articles both enjoyable and insightful. For information on new subscriptions, product trials, alternative billing arrangements or group and site discounts please call 800-688-2421. We look forward to having you as a long-term member of the Relias Media community.