When a 50-year-old nurse was admitted to her service comatose in a state of rhabdomyolysis, neurohospitalist Christina M. Burch, MD, was stumped. There were no external signs of trauma, and computerized tomography (CT) of the head, laboratory studies, cerebrospinal fluid analysis, and toxicology screen were unrevealing. Although brain magnetic resonance imaging (MRI) later showed diffuse white matter changes and bilateral watershed infarcts, there was no clear medical explanation for her patient's condition.
Dr. Burch, who lives in Goodlettsville, TN, had seen dozens of fishy cases before, so she decided to take snapshots of her patient's body to discern if she had missed a pattern of bruising. And when a lawyer showed up the next day revealing that her patient (his client), had failed to show up in court on the morning of a custody hearing, Dr. Burch enlisted the help of her nurses to count the skin puncture wounds revealed in the photographs. After comparing them with the medical record and discovering one intramuscular injection site on the right anterior thigh, which remained unaccounted for, she called the police. It wasn't long before homicide detectives discovered that a paralytic agent was missing from pharmacy stores where her patient's ex-husband, a nurse-anesthetist, worked.
“If nothing is making sense, look outside the body,” advised Dr. Burch, who has learned that, at times when medical knowledge fails to explain a neurologic deterioration, it's useful to get a thorough social history. Because she has become highly sensitized to suspect behavior over the years, she has adopted a “three strikes” rule: if three people — hospital personnel, family members, or other visitors — come to her saying “Something's not right,” she calls the authorities.
This is exactly what she did when she was called to consult on a 38-year-old, otherwise healthy woman, who had become comatose after admission for a cholecystectomy. The history revealed that the patient had been awake and alert at 6 AM when she was taken to the bathroom by the nurse, but when the nurse returned at 6:30 AM, she found the patient unresponsive, with her husband asleep at the foot of the bed. After a code was called and the patient had been transferred to the intensive care unit, nurses approached Dr. Burch with reports of blasé behavior by the husband, who had, incidentally, also asked two of them out on dates. Later, after the woman's death, when the district attorney arranged for the body to be exhumed, the coroner determined that she had not died of natural causes. Dr. Burch, who had suspected that her patient had been smothered with a pillow when a brain MRI revealed findings consistent with anoxia, pointed out that clinicians often obliterate all traces of evidence during a code.
“Someone comes in coding, or at least in extremis, and everyone jumps into action, working desperately to save a life and erasing all proof while we do it,” said Dr. Burch. Absent confirmation of fibers in the nose and mouth and other concrete substantiation of homicide, the case never made it to trial.
“I'm often asked if people get away with murder, and I think sometimes they might,” said Lisa K. Mannix, MD, a coroner in Butler County, OH, and a neurologist with a headache specialty. “Suffocations and asphyxias are particularly challenging to prove forensically, even without a code, whereas strangulations leave more obvious signs.”
Dr. Mannix pointed out that “angel of death” murders are likely more common than other homicides in the hospital setting, but even those go undetected. “When we hear rumors of a history of domestic violence, or of more than one acquaintance or spouse who has previously died, or we're faced with a situation of multiple neurologic admissions or symptoms for which we don't have an answer, those are red flags,” she said.
Those “red flags” are what finally aroused the suspicions of Ronald B. Ziman, MD, associate clinical professor of neurology at the David Geffen School of Medicine at University of California Los Angeles, when a 40-something woman presented to his hospital with an electrodiagnostically-confirmed case of motor neuropathy from which she initially improved, but later suffered a mysterious relapse. When he found Mees' lines in her fingernails, arsenic poisoning was suspected and soon confirmed by blood and urine tests. Based on nail and hair analysis, it was apparent that she had had two distinct periods of exposure, so the police were notified.
The patient's husband had been seen regularly bringing her lunch while on the rehab floor, and consequently, the food was analyzed and confirmed to be laced with arsenic. Unlike cases with more ambiguous evidence, the man was convicted and sentenced to jail. Dr. Ziman, who has seen a few other cases of heavy metal toxicity (lead, mercury, and manganese) over the years, only occasionally identifying the source, said: “The real question is how many cases of willful poisoning I have missed. I fear there may have been others.”
Indeed, according to an article published earlier this year in the American Journal of Forensic Medicine and Pathology, homicidal poisoning in a medical setting is often difficult to discover; the symptoms of poisoning may simulate symptoms of disease, and even when poisoning is suspected, the symptoms are often nonspecific, offering few clues about potential lethal agents. Additionally, many poisoning victims exhibit a lag in the presentation of symptoms, giving the perpetrator enough time to destroy evidence and to conceal his or her role. By the time the victim starts showing signs of compromised health or is found dead, the incident is difficult to connect to the specific individual, and the evidence may be irretrievably destroyed.
“Many cases in which a patient is neurologically impaired are not brought to the attention of the coroner,” Dr. Mannix said. She suggests that if a family member makes accusations of foul play the physician should encourage them to call the police. Another option in the case of unexplained fractures or head injuries, for example, is to call in the hospital social worker who can alert adult protective services.
“Pediatricians are taught from day one how to look for signs of child abuse, but we get little or no education about elder abuse, nor are we taught when to suspect violence as a cause of affliction,” said Dr. Burch, explaining that this applies not just to elders, but to trauma victims whose attacks did not involve guns or knives, but rather blows that caused carotid or vertebral dissection and other sometimes-not-so-obvious aggressions. “If we don't look for it, we'll miss it and send our patients back into a dangerous environment,” she said.
WHEN TO GET A TOX SCREEN
The criteria for toxicology screening are evolving as stroke specialists have come to recognize that they are valuable in more than one demographic. Brian Silver, MD, associate professor of neurology at the Warren Alpert Medical School of Brown University, recently served as lead author for a paper in the April 30 edition of Neurology, which described the findings of a retrospective chart review examining which stroke patients at Henry Ford Hospital in Detroit underwent urinalysis drug testing. [See Neurology Today's coverage of the report, “Drug Testing Disparities Found Among Stroke Patients — Is it Racial Profiling?”: http://bit.ly/18CnUs3.]
“We've recently changed our protocol so that all stroke patients are supposed to get toxicology screens regardless of age, gender, or race,” said Dr. Silver, the director of the Stroke Center at Rhode Island Hospital. Several years ago, while working in Detroit, Dr. Silver admitted an 89-year-old woman with a basilar occlusion resulting in widespread brainstem damage, and her urine toxicology turned up positive for cocaine. When he asked the family about exposures, he discovered that the patient lived with her 61-year-old son, who was addicted to crack. “We learned that he had injected her with cocaine because she was ill and he thought it would provide her medical relief,” Dr. Silver said.
“If you have a clinical situation without a good explanation, or one in which the symptoms are strange, you're always on firm ground ordering a toxicology screen,” advised neuropathologist and forensic pathologist Bruce H. Wainer, MD, PhD, former chief forensic pathologist at East Baton Rouge Parish in Louisiana, adding that it's best to send the specimen to an outside reference lab that is able to screen for a large number of chemical agents.
Several years ago, he had been asked to perform an autopsy on a pregnant woman who had been in a chronic vegetative state due to a traumatic brain injury, when she suddenly died for no apparent reason a few hours after her C-section. He found no anatomic explanation for her death, but three weeks later the toxicology results revealed she had ten times the therapeutic level of meperidine (Demerol) in her blood. “If someone undergoes a sudden neurologic deterioration for unknown reasons and dies, you should report it to the medical examiner or coroner,” he advised.
Because there are neurologists now working full-time in institutional settings, they are privy to more conversations with visitors to the hospital room, and often develop the kind of trusting relationships with staff that fosters disclosure of unusual family behavior.
“Although these cases are infrequent, we should always be on the alert,” Dr. Mannix advised. “Unfortunately, the proliferation of crime shows has helped bad guys get smarter and come up with more creative means of violence.”
Dr. Avitzur, a neurologist in private practice in Tarrytown, NY, holds academic appointments at Yale University School of Medicine and New York Medical College. She is an associate editor of Neurology Today and chair of the AAN Medical Economics and Management Committee.
WHEN TO REPORT DEATHS TO THE CORONER
Lisa K. Mannix, MD, a coroner in Butler County, OH, and a neurologist with a headache specialty, advises that deaths reportable to the coroner include any hospital (institutional) cases of:
- Dead on Arrival: Any person pronounced dead on arrival at any hospital/ER or within 24 hours of admission to a hospital unless the patient has been under the care of a physician for natural disease that is reasonable for death.
- “Delayed death, where the immediate cause of death may actually be from natural disease. However, injury may have occurred days, weeks, months, or even years before death and is responsible for initiating the sequence of medical conditions or events leading to death. The most common examples include past traffic accidents with debilitating injury and long-term care in a nursing home, and hip fractures of the elderly where there is a downward course of condition after the injury.”
For more on when to report deaths to the coroner, see http://bit.ly/12qaZpL.
—Orly Avitzur, MD
If you have come across a case in which you suspected foul play while in the hospital setting, please e-mail Dr. Orly Avitzur at firstname.lastname@example.org