Falling Into the Fire: A Psychiatrist's Encounters with the Mind in Crisis

by Montross, Christine

  Though Blaffer Hrdy is talking about nonhuman primates, it is not hard to imagine the “direst circumstances” that might face human mothers, leading them to abandon or harm their children. A study published in the American Journal of Psychiatry that analyzed infanticide in seventeen countries emphasizes the contribution of stressful psychosocial factors. The study found an unambiguous “pattern of powerlessness, poverty, and alienation in the lives of the women” who had killed their children.

  “Because killing one’s own infant is so abhorrent to us,” Blaffer Hrdy writes, “there is a tendency to compartmentalize the mother’s actions . . . to consider her behavior in isolation from her circumstances, even though they are functionally related.”

  The laws in America regarding infanticide reflect this isolated abhorrence. England and at least twenty-one countries worldwide grant leniency to mothers who can demonstrate that they killed their children during a postpartum mental disturbance. These laws, based upon Britain’s Infanticide Act of 1922, adopt the premise that a woman who has committed infanticide may have done so because “the balance of her mind [was] disturbed by reason of her not having fully recovered from the effect of giving birth.” The result of this interpretation is that the maximum charge these women can face is not murder but rather manslaughter. The definition of “infanticide” varies from one country to the next, but in New Zealand the law applies to the murder of children who are as old as ten. America, in contrast, has no federal or state laws that specifically apply to infanticide. Not only may these women be tried for murder, but, as was the case for Andrea Yates, the prosecution may seek the death penalty.

  Psychiatrists already know that intense stress increases a new mother’s risk for postpartum mood disorders. And social scientists have repeatedly demonstrated that increased parental stress is a risk factor for child abuse and neglect. Stress has also, in various forms, been correlated with a mother’s risk of killing her children.

  Twenty-five-year-old Lashanda Armstrong had her own share of maternal stress when she drove her van into the frigid Hudson River in 2011, killing herself and three of her four children, ages five, two, and eleven months. The fourth child, her ten-year-old son, who was born when Lashanda was only fifteen, managed to slip out the van door and swim to safety. Though Armstrong was universally described by family and friends as a concerned and highly devoted mother, news reports revealed she was struggling amid difficult life circumstances. A supervisor at her children’s day care reported that Armstrong had recently described feeling “so alone.”

  “She’s a single parent. She takes great care of her kids, goes to school and works,” said the supervisor. “She really needed a helping hand.” Armstrong’s son had revealed to a teacher that his mother and stepfather, Jean Pierre, were fighting frequently due to Pierre’s alleged infidelities. Armstrong was apparently trying to obtain a court order so that he could not have contact with the children. The last time she had left their two-year-old in his care overnight, the child was found by police wandering a city street, barefoot, in a wet sweat suit. The night that Armstrong drove her children into the river, her family had contacted police, fearing that Armstrong and Pierre were “tussling.” The police reported that the couple had previously had episodes of domestic problems and that an order of protection had been issued—and subsequently violated by Pierre—in the hours immediately prior to the tragedy. Direst circumstances.

  The forensic psychologist Geoffrey R. McKee created a “Maternal Filicide Risk Matrix” in an attempt to help clinicians further assess a mother’s risk of killing her child. The matrix identifies sources of maternal stress that might combine to make a mother vulnerable to thoughts of child harm. Notably, the presence of a psychiatric disorder is only one of many risk factors. Other potential risk factors that exacerbate the mother’s risk of filicide include the following:

  teenage motherhood

  below-average IQ

  less than a twelfth-grade education

  no prenatal care

  history of trauma (including physical and sexual abuse as well as childhood loss of her own mother)

  denial of pregnancy

  negative attitude toward pregnancy

  unassisted birth

  nonhospital delivery

  difficult birth

  absent, abusive, mentally ill, addicted parents during her own childhood

  violence in partnership

  substance abuse in partner

  divorce

  single parenthood

  financial instability

  unemployment

  relocations

  low socioeconomic status

  having two or more children if under age seventeen

  many children in her care

  child difficult to care for

  lack of sleep

  There is no shortage of anecdotes that highlight the danger in this perfect storm of stressors. Andrea Yates is surely the best-known example. In addition to her documented history of postpartum psychosis, she had for a time lived with her husband, Rusty, and their four children in a 360-square-foot bus. The family had moved into a house by the time Yates delivered her fifth child, but she was still expected to homeschool the older four in the bus while simultaneously caring for her newborn daughter. She had had a string of psychiatric hospitalizations from 1999 to 2001 and an overdose attempt in 1999 after she disclosed a thought of stabbing one of her children. During a period of inpatient hospitalization, she had been so profoundly impaired and unable to care for herself that she had to be spoon-fed.

  I used McKee’s matrix as a guide in thinking about Anna and the potential risk she posed to her son. Anna was not a teenager, and she had graduated from high school. She was married, though she painted her husband in a less-than-supportive light. Still, there was no evidence of domestic violence or substance abuse in the home. Anna had referenced help from her mother-in-law and sister, though she felt guilty about turning to them for assistance too frequently. She and her husband had stable finances and employment. She had looked forward to the pregnancy and continued to express sincere desire to be a mother—a good mother—to her son.

  I understood the immense importance of diagnosing Anna correctly and remained aware that a diagnostic misstep could have disastrous consequences. Still, I was feeling increasingly confident that Anna’s symptoms had certain traits that designated them as more obsessive than psychotic. First and foremost, as distressed and fearful as Anna was, she was not out of touch with reality. Over the course of our discussions, she was consistently able to realize that these thoughts were fears and not plans. In addition, Anna’s responses to the thoughts were characteristic of someone with an anxiety disorder like obsessive-compulsive disorder, rather than a primary psychotic disorder. She tried to avoid the thoughts, and she tried to avoid the situations that brought about the thoughts. Therefore, when Dawn suggested that Anna tell her mother-in-law not to bring the baby in for visits anymore, Anna was relieved and in utter agreement. Finally, and more important, Anna’s “visions” were ego-dystonic—that is, they were repugnant to her and to how she perceived herself. She was deeply and constantly troubled by the thoughts of harming her son, as opposed to someone like Andrea Yates who might have found comfort in a delusional plan that would grant her children eternal salvation. Anna’s visions of bending over her son’s lifeless body “so sad at what I’ve done” and of herself in jail had the feel of compensatory rituals. In the same way that a germophobe might prevent the feared infection by repeatedly washing his hands or following a ritualized pattern of cleaning, Anna seemed to be preventing herself from acting on her intrusive thoughts of harming her son by adding her own punitive epilogue to the horrifying film.

  After consultation with colleagues and supervisors and multiple discussions with Anna, I began a gradually escalating course of exposure therapy with her in which she would visit w
ith (or “be exposed to”) her son and then we would meet to discuss the thoughts and feelings she had experienced during her time with him.

  I wrote orders for Anna to have structured visits with her son, first on the unit, then leaving the unit with her husband and her son for a few hours at a time, and eventually spending time at home with her son, by herself. At each stage Anna reported increased anxiety, a response that was consistent with an obsessive anxiety disorder. “The patient reports that intrusive thoughts were exacerbated yesterday during visit with son and husband,” my treatment notes read. “This was very distressing to her.” I make an additional note about her appearance during our meeting: “The patient is slightly disheveled and nervous-appearing, but cooperative. She taps her fingernails together and on the table. Her hands are tremulous.” As her visits with her son increase in duration and in independence, I note that Anna “is having an appropriate increase in anxiety level with continued exposure to her son and approaching discharge.” Meanwhile, Dawn’s nursing notes indicate that she remains dubious.

  “The patient returned from a four-hour pass,” she writes. “Reports that it went fair. Stated ‘I was with the baby for a half hour and had thoughts of stabbing him.’ M.D. still plans to have patient spend time with baby alone despite these ongoing thoughts.”

  There was a part of me that felt exactly the way Dawn did. I, too, feared that encouraging Anna to be alone with her son was too dangerous, that the stakes were too high and that I should err on the side of caution. And yet I also felt that in these moments I needed to internalize the very message I was trying to help Anna believe in and hold: that far more women have thoughts of killing their children than actually do; that fearing something does not make it happen; that we have, since Phil Resnick’s landmark study in 1969, begun to understand more and more about the women who do kill their children so that we can use clinical evidence to help us know those families who are more at risk and those who are less so.

  On the twentieth day of her hospitalization, Anna was home alone with her son on a four-hour pass when he had an asthma attack.

  “I went right into mom mode,” she said to me on her return, recounting the incident. “He couldn’t breathe and was gasping, and I just grabbed the nebulizer like it was second nature, hooked up the albuterol, and gave him a treatment.” She smiled. “It felt so great to help him like that.”

  One day later Anna was discharged from the hospital with plans to continue treatment on an outpatient basis for obsessive-compulsive disorder. Even after Anna’s three inpatient weeks of daily observation and treatment, and her noticeable improvement, I wasn’t 100 percent sure of my diagnosis.

  Defining the maladies that plague psychiatric patients is an interpretive science. Visions and voices and fear and despair cannot be captured by CT scan or measured in the amplitude of EKG waves. Try as we might, we simply cannot predict which of our patients will kill themselves, which will murder their children, and which will leave the hospital healed, never to return. The reliable portraits and profiles we do have of patients who commit horrific acts are too often, like that of Andrea Yates, available to us only in retrospect, after terrible and irreversible damage has already been done.

  With that hindsight, however, we are able to begin to build a framework of understanding as to the symptoms and circumstances that lead women to kill their children. As the field of research begun by Dr. Resnick continues to deepen and expand, the act of filicide may remain unthinkable, but it can be less incomprehensible to those of us who see or hear about it. If we continue to respond to the idea of child murder by mothers with disgust and scorn for the woman who commits the crime, as we so consistently do, we discourage all mothers—even those who would never harm their children—from feeling safe enough to seek help from the terrifying thoughts that plague them. We cannot prevent all instances of filicide, but if women felt that their disclosure of filicidal thoughts might be met with sympathy and support rather than repulsion and shame, we might have an opportunity to help certain mothers to think more clearly, or to imagine another, better way out.

  I never saw Anna again after her discharge. I took solace in the fact that I also never saw her in the headlines. Maybe that meant that my diagnosis of her was right. In any case, the discomfort I felt watching her leave the hospital with her suitcase, her husband, and her son has stayed with me. And yet I had to trust that the child would be safe. It was an awful, uncertain feeling. It somehow seemed right, though, given the fact that it was the exact uncertainty I had asked Anna to trust in and bear.

  (CHAPTER FIVE)

  Dancing Plagues and Double Impostors

  The mind has great influence over the body, and maladies often have their origin there.

  —Molière

  In my second year of residency training, I spent a month working the overnight shift in the freestanding psychiatric hospital where I am now an attending physician. In theory, my main responsibility during this time was to evaluate people who came to the ER overnight. If they needed inpatient treatment, I would admit them. If not, I would send them off with a list of resources I hoped would be helpful: names of outpatient therapists or psychiatrists, instructions for how to become wait-listed for a day hospital program, addresses and times for local AA or NA meetings.

  Some of these decisions were obvious. I admitted a man so paranoid that he had not eaten for days, afraid that he was being poisoned. An alcoholic woman who had no desire to stop drinking had been dragged in by her desperate daughter. With no legal right to hold her against her will, I let her go. Many decisions were not so clear-cut. A woman who had made a suicidal comment to a friend now swore that it was hyperbole. Was she telling me the truth, and would she be safe to leave? Or was she genuinely suicidal but denying it because she didn’t want to be hospitalized? A man who routinely claimed he was hearing command hallucinations to inject rubbing alcohol into his veins requested admission at the end of every month, when his assistance checks had run out. At the first of the next month, like clockwork, he would sign himself out of the hospital, stating that his voices had miraculously abated.

  In reality, my responsibilities extended far beyond being the gatekeeper of psychiatric admissions. Because this was a freestanding psychiatric hospital and therefore most of the patients were otherwise medically well, I was the only doctor in the hospital overnight. This meant that I was also responsible for any medical issue that might arise on the hospital wards. Frequently I was paged for minor requests: a patient with a headache wanted some Tylenol, or a smoker was in desperate need of a nicotine patch. Sometimes I was called to evaluate a patient with chest pain or to see someone who had taken a fall. Occasionally there were true medical emergencies. When the patients’ medical needs were beyond the basic level of care that our psychiatric hospital could provide, they had to be sent out to a medical hospital’s emergency room to be treated.

  When assessing a patient who needs medical care, different doctors have different thresholds of discomfort, different hierarchies of decision making. I think of it as something like a pain threshold. My own ability to tolerate physical pain is high, a lesson I learned after enduring hours upon hours of labor without medication before our daughter was born. Yet I am also risk-averse, and I err on the side of caution when it comes to patient care. I have colleagues who pride themselves on sending only the very sickest patients out for medical treatment or on admitting only those psychiatric patients who are clearly at the most severe and imminent risk. They joke of being impenetrable “walls” in the emergency room. They hold it as a point of honor that they do not waste ER doctors’ time with psychiatric patients who will surely sleep off their elevated blood-alcohol levels or whose acute chest pain is almost certainly a ploy for narcotics.

  I don’t want to waste the time of my colleagues in emergency rooms either, but my threshold for sending patients from the psychiatric hospital to the medical hospital is low. This doesn’t bother
me. I see it as recognizing my own limitations. And probably it’s also partially driven by the CYA, as in “cover your ass,” school of medicine. CYA, as a philosophy, is passed down from doctors to medical students in the earliest days of medical training as a kind of inoculation against medical malpractice. It is, of course, an overtly crass and overly simplistic approach, and there are those who would disparage acting to CYA as practicing defensive—rather than clinically indicated—medicine. Nonetheless, I think the gist turns out to be a good gut check: If this ends up being something serious, could people reviewing the chart determine that I should have sent this patient out for medical evaluation? Could I reasonably be expected to have acted differently in my practice by other doctors—or by a court of law?

  Sometimes when I send a patient out because I suspect he needs medical attention, I am right, and sometimes I am wrong. Once I was working on the most acute unit of the psychiatric hospital, a ward reserved for patients who were floridly psychotic, or violent, or actively trying to harm themselves. A man who was being treated for opiate dependence was sent to my unit from one of the hospital’s general-treatment wards because he had become increasingly psychotic and difficult to manage.

  When he arrived on the unit, the patient spoke mumbled nonsense and required constant intervention to keep him from mistakenly wandering into other patients’ rooms. When I could understand what he was saying, he was describing women in bikinis looking in his second-story window and men with guns after him about a card game. I suspected he was delirious and sent him to a medical hospital. Delirium can mimic psychosis, with its visions and voices and false beliefs, but it arises from states of medical disequilibrium, like infections or electrolyte abnormalities. In the medical hospital, my patient’s blood was found to have precipitously low sodium levels, which had led him into a hallucination-plagued stupor. Had I not sent him out for his sodium to be repleted, he could easily have died.