The Psychopathic Psychiatrist and the Murder of his Wife

Colin Bouwer lectured his medical students on the perfect way to kill someone with insulin. Months later, he started poisoning his wife.

Sometime in 1999, Colin Bouwer stood in front of a room full of medical students at the University of Otago in Dunedin, New Zealand, and told them that injecting insulin between the toes was the perfect way to commit a murder.

South African psychiatrist Colin Bouwer convicted of poisoning his wife

He was the Head of the Department of Psychological Medicine. A senior academic. A psychiatrist whose specialty, in a detail that would be rejected as too heavy-handed in fiction, was anxiety: what happens to people when they are terrified. His colleague Sarah Romans later noted his fascination with suffocation, water torture, and drowning. His students would have listened respectfully. Some probably took notes.

Within months, he began killing his wife.

Not with a needle between the toes, as it turned out. Colin Bouwer’s actual method was more ambitious than his lecture suggested, and crueler in a way that only a physician could manage. He didn’t just poison Annette Bouwer. He manufactured a fake disease inside her body, then watched as an entire hospital system chased a diagnosis that didn’t exist while he kept feeding her the drugs that were killing her.

He sat at her bedside while she cried. He held her hand while she was wheeled into unnecessary surgery. He controlled every piece of information that flowed between Annette and the medical team trying to save her life.

Annette Bouwer was forty-seven and the mother of two children when she died.

And he did all of this at the hospital where he was one of the most senior clinicians on staff, surrounded by colleagues who had been trained to recognize psychopaths.

Carl Elliott, writing about this case in The New Yorker, noted that what was striking about Colin Bouwer was his ability to fool his colleagues, many of whom would have studied psychopaths in their medical training. Elliott wondered whether Bouwer’s success revealed more about the nature of psychopathy or about the character of New Zealanders.

I think it reveals something about the gap between how we teach psychopathy in textbooks and how it actually operates in the real world. We teach it as a checklist. It operates as a performance.

“She Wanted to Live for the Sake of Her Children”

Annette Bouwer was a 47-year-old physiotherapist, a devout Christian, a mother of two teenagers, and a woman who lived a clean, healthy lifestyle. She loved Dunedin. She and Colin lived in a gracious home on the slopes overlooking St Clair Beach, decorated in fresh lemon and mint-green tones, where she walked her English bull terrier, Cinnamon, along the golden sand. She did jigsaw puzzles and crosswords and knitting. She read murder mysteries.

She had met Colin at a Mensa meeting in South Africa in 1981. He was charming. She married him that same year, his third wife, though he told some people she was his first. They had two children, Greg and Anthea. In 1997, the family emigrated to New Zealand. Annette said she was glad to leave behind the crime that plagued South Africa.

Friends described her as vibrant, kind, and strong-minded, a woman who put others first. She was also known to hate taking medication of any kind, a detail that matters because it tells us something about how her husband had to administer the drugs that killed her: hidden in food or drink, without her knowledge or consent.

In late 1999, the dizziness started. Then blurred vision. Coordination problems. She saw an optician. Got new glasses. Nothing helped.

On the morning of November 20th, Colin called an ambulance. He’d found Annette comatose in bed. Her blood sugar was 1.3 mmol/L, a level low enough to kill. Paramedics injected glucagon and rushed her to Dunedin Hospital. She woke up. Nobody could explain why a healthy, non-diabetic woman had collapsed into hypoglycemic shock.

Four days after discharge, it happened again.

The doctors suspected an insulinoma, a rare tumor of the pancreas that overproduces insulin. It was the only diagnosis that fit. Annette underwent an arterial calcium stimulation test, an invasive and painful procedure. The results were misinterpreted as confirmation. Surgery was scheduled.

Before the operation, Annette’s friend Magda Brits visited her in the hospital. Annette cried in Magda’s arms, terrified of the twelve-hour surgery ahead, agonizing over the burden her illness was placing on her children during their exams. Magda asked if she wanted to pray together. Annette said yes. She asked God to bless the surgery and make it a success. She said she wanted to live for her children’s sake.

While in the hospital, Annette told staff she thought she was being poisoned.

No one thought that made any sense. Her husband was the Head of Psychiatry. Annette’s allegation was filed under the mystery illness, another symptom of something nobody could explain.

On December 13th, surgeons removed two-thirds of her pancreas. They found no tumor. Just a small nodule that explained nothing.

Annette was discharged on Christmas Eve. She phoned her mother in South Africa. Her mother remembered how exhausted she sounded. Annette told her, “I’ve never smoked, drunk, or taken drugs. So why is God doing this to me?”

She told her children she wondered why she was even there, since they seemed to be managing fine without her. That sentence has stayed with me since I first read it. A woman who six weeks earlier had begged God to let her live for her children was now questioning whether they needed her at all. That’s not depression talking. That’s the voice of someone being systematically erased from her own life.

Eleven days after discharge, she was dead.

January 5, 2000

The day before Annette died, a laboratory technician at the hospital received a blood sample from Colin. He had drawn it himself from Annette’s arm at around 4:30 in the afternoon. The blood glucose was 1.7 mmol/L, so dangerously low that the technician called the phone number on the request form to report the result. The technician also noted the sample was grossly hemolyzed: broken down, unsuitable for the additional insulin and C-peptide tests Colin had requested. He asked for a fresh sample. None came.

What the technician didn’t know was that Colin had picked up his final forged prescription just thirty minutes before drawing that blood. It was the only prescription that included insulin. A thousand-unit vial of Humalog. Enough to kill.

The children were not home. Greg and Anthea had spent the day at Anne Walsh’s house, watching a cricket match. Walsh was Colin’s lover, a fellow psychiatrist at the university. The timing was not coincidental.

Sometime during the night, Colin administered the insulin. He may have also given Annette a cocktail of sedatives. Clonazepam was later found in her blood at several times the therapeutic level, along with massive amounts of metformin, glibenclamide, glipizide, and citalopram. He did not call an ambulance. He did not take her to the hospital. Death from hypoglycemia is typically preventable; intravenous glucose can reverse it in minutes. Colin Bouwer, who had twice before watched his wife revived by exactly this treatment, let her die at home.

On the morning of January 5th, he called Dr. Andrew Bowers.

Bowers arrived at the house and found the bedroom in disarray. Annette’s body was on the bed, the bedclothes soiled with vomit. There were traces of Instagel, a glucose paste, around her lips, suggesting someone had made a token effort at resuscitation. But Instagel cannot reverse the kind of overdose Annette had been given, and Colin knew that. He knew the difference between a gesture and a treatment.

Bowers confirmed she was dead. At the autopsy two days later, the pathologist would note five needle marks at the crook of Annette’s left elbow, from where Colin had repeatedly drawn her blood the previous day, and a single additional puncture mark in her right arm. Her right lung weighed nearly twice as much as her left, swollen with the pneumonia caused by aspirating her own vomit. Her brain, which should have been preserved whole for a neuropathologist to examine for the telltale damage of prolonged hypoglycemia, was described as “normal” and not kept. That failure would later become a point of contention at trial.

Colin asked Bowers to sign the death certificate. Cause of death: hypoglycemia from an insulinoma missed at surgery. Just sign the paperwork. She had been sick for weeks. Everyone knew it. The whole hospital had watched her decline. A signature, and it would be over.

Bowers had only been out of training for three years. Colin was the Head of Psychological Medicine, several notches higher in the academic hierarchy. Everything about the institutional dynamics of that moment pushed Bowers toward compliance: the power differential, the deference a junior physician owes a department chair, the sheer awkwardness of challenging a bereaved husband in his own home.

He said no.

Before signing, Bowers called the coroner, who indicated a hospital postmortem would suffice. Colin immediately objected. He told the doctors Annette was Jewish and therefore needed to be buried within forty-eight hours. He insisted she be cremated the following morning. But Orthodox Jews don’t cremate. And Annette’s funeral, when it was held, took place in an Anglican church.

The vicar who conducted the service later told investigators Colin’s behavior had been strange. The music played was neither Annette’s favorites nor treasured Christian hymns. Instead, it was a song about going off a cliff. He also noted that the house seemed to have been disinfected.

After the funeral, Colin gave away Cinnamon, Annette’s dog. He changed their telephone number. For a time, Annette’s family in South Africa had no way to reach the children. On the afternoon before her death, he had called his mother-in-law in South Africa to tell her Annette was dying, but not to bother coming to New Zealand, because Annette would be dead before she could arrive.

She was.

How to Turn a Hospital Into a Murder Weapon

Most spousal poisoners use what they have. Access to household chemicals, medications from the family medicine cabinet, and substances purchased online. Colin Bouwer had something far more dangerous: a medical degree, prescribing authority, and a position at the top of a hospital hierarchy that discouraged anyone below him from asking uncomfortable questions.

He exploited every one of these advantages.

Between November 16, 1999, and January 4, 2000, Bouwer wrote eleven forged prescriptions under fictitious patient names for a carefully selected combination of hypoglycemic drugs (glibenclamide, glipizide, metformin) and, on the final prescription, insulin. He ground the tablets with a mortar and pestle and mixed the powder into Annette’s food or drink.

The drug combination was not random. Glibenclamide and glipizide are sulphonylureas. They stimulate the pancreas’s own beta cells to secrete insulin, producing elevated insulin and C-peptide levels that look identical to an insulinoma. Metformin suppresses hepatic glucose production, further deepening hypoglycemia from a different angle. Together, they created a metabolic profile indistinguishable from that of an insulin-producing tumor. Any endocrinologist who saw the lab results would reach the same conclusion: insulinoma. Which is exactly what happened.

Why didn’t the hospital’s lab tests catch the drugs? Because the sulphonylurea assay used in New Zealand at the time had a lower limit of sensitivity of 176 μg/L, capable of detecting only a suicidally massive overdose. The immunoassay available in specialized labs elsewhere could detect sulphonylureas at concentrations as low as 0.5 μg/L, roughly 300 times more sensitive. Annette’s blood was screened once, sixteen hours after admission, while intravenous glucose was still running. The test came back negative, and that negative result closed the door on the poisoning hypothesis for the duration of her illness. Today, liquid chromatography-tandem mass spectrometry can detect sulphonylureas at nanogram-per-milliliter concentrations from a single blood draw. In 1999, Dunedin didn’t have it.

When Dr. Andrew Bowers, the internist managing Annette’s care, asked Colin to check the house for any drugs that might explain her condition, Colin did so and reported the house was clean. He was asked to investigate his own crime and cleared himself.

But Bouwer didn’t just hide evidence. He actively shaped the diagnostic process. He introduced himself to Bowers as “a physician, a psychologist, and a pharmacologist.” The latter two credentials were false, and he didn’t seem very knowledgeable about being a doctor. Bowers was stunned that Bouwer didn’t seem to understand the diagnostic tests associated with hypoglycemia.

Bowers also noticed Bouwer’s manipulative interpersonal tactics. “It was like hot and cold taps,” Bowers later said. “He tried to overbear me at first, but that didn’t work, so then he tried being friendly.”

What Bowers was describing, though he likely didn’t have the framework for it at the time, maps precisely onto what Christopher Patrick calls the “boldness” dimension of psychopathy: the capacity to remain socially dominant under pressure while rapidly adjusting tactics in response to feedback (Patrick, 2022). When dominance doesn’t produce compliance, switch to charm. When charm fails, try something else. The goal is control. The approach is calibrated in real time based on what works.

Bowers felt insecure. He questioned himself rather than questioning Bouwer. That self-doubt was not a personal failing. It was the predictable result of a massive power differential between a junior physician and the Head of Psychological Medicine. It is precisely the dynamic that institutional psychopaths count on.

And when doctors discussed referring Annette to a psychiatrist for the depression her mystery illness was causing? Colin requested that it be Dr. Anne Walsh, his lover.

The Mythical Autobiography

I’ve heard a lot of lies over the past thirty years. Murderers who pretended to be either victims or heroes. Men who insisted they were former CIA operatives. An imposter who swore he was once a missing child. These lies are often breathtaking in their ambition, and what they have in common is their target audience: people who either can’t check the facts or who love the liar too much to try.

Sitting across the table from a forensic evaluator, with records in front of me and no emotional stake in the story, these fabrications tend to collapse fast. Colin Bouwer wouldn’t have survived thirty minutes in an interrogation room. But he never thought he would have to. His lies and deceptions were built for colleagues, lovers, and family, people who cared about him and wanted to believe him.

His lies weren’t reactive. He didn’t bend the truth when caught. He constructed an entire fictional identity from the ground up and maintained it for decades.

He told people he had joined the African National Congress as a teenager in the 1960s. He described being detained without trial for six months and tortured by the South African secret police: standing naked with a brick hung from his testicles, sleep deprivation, near-drowning, electric shocks to his genitals, sodomy, solitary confinement. He claimed the abuse cost him a testicle. He said he’d survived by communicating with other prisoners in Morse code at night while practicing meditation and self-hypnosis.

None of it happened. The investigation that followed Annette’s death found no record of ANC involvement, no detention, no imprisonment. His actual military record was two months of service in the South African Defense Force before resigning.

He claimed to have counseled Nelson Mandela after his release from Robben Island. He claimed advanced degrees in pharmacology and specialized training in internal medicine. He told a friend that a previous wife had committed suicide after killing their children. He told Annette’s South African family she had cancer. He told former patients that his wife was dying and he hadn’t had sex in months, all while seducing them during therapy sessions. He told some people Annette was his first wife. She was his third. He would later pose as a Holocaust survivor in prison.

His sister-in-law revealed that he had once said New Zealand was the ideal place to commit the perfect murder.

Pathological lying on this scale isn’t about avoiding consequences. It’s about identity construction: building a self from scratch that is more heroic, more sympathetic, more self-serving than whoever actually exists underneath. The torture stories positioned him as a survivor, a man of moral courage and deep suffering, which made people trust him and lower their guard.

But I don’t think the lies were just theater. They were intelligence operations. When you tell someone you were brutalized by a fascist regime, you watch their face. You see what moves them. You see how far their empathy extends and in what direction. Every fabricated story was a probe, and every empathic response was a data point about the listener’s vulnerabilities. That is what made Bouwer dangerous. Not the lies themselves, but the information they generated about other people.

A Succession of Women

Megan Power, a journalist with South Africa’s Sunday Tribune, reported that during his time at Tygerberg and Stikland hospitals in Cape Town, Bouwer’s name was associated with multiple complaints. One colleague said only that Bouwer left behind “a lot of unhappiness” when he moved to New Zealand.

Former women patients in South Africa described being seduced by him. He would tell them that monogamy and fidelity were outdated concepts. He told his lovers that his perfectly healthy wife was dying of cancer and he hadn’t had sex in a long time. In 1996, two formal complaints of sexual misconduct were filed by former patients. Both were withdrawn; the women reportedly feared publicity. At least two additional complaints were never formally pursued for the same reason.

By the time the New Zealand police completed their surveillance of Bouwer after Annette’s death, investigators believed he was sexually involved with at least four hospital staff members across Invercargill and Dunedin.

Research shows that between seven and twelve percent of mental health practitioners in the United States acknowledge having had erotic contact with a client. The most dangerous subtype, the one most likely to have multiple victims and to use predatory rather than opportunistic tactics, is the narcissistic/antisocial personality type.

These are not therapists who cross a boundary in a moment of weakness or isolation. They groom. They isolate. They tell patients whatever story they need to create the opening.

Bouwer fits this profile. Telling patients his wife was dying of cancer reframed the sexual boundary violation as compassion, as rescue, as two struggling people finding comfort in extraordinary circumstances. The patient believes she is helping a man in pain. The therapist knows he is running a script.

What’s striking is how consistent the script was across years, countries, and relationships. The wife is always sick or dying. The marriage is always loveless. The patient or colleague is always the one who finally understands him. Over and over, because it worked, I confess that the consistency of the playbook bothers me more than the individual seductions. It suggests he had tested and refined the approach until it was reliable, the way a salesman tests a pitch.

And then there was Anne Walsh.

Walsh was a fellow psychiatrist at the University of Otago. Their affair intensified at a conference in Copenhagen in September 1999, though Walsh later maintained they didn’t sleep together until after Annette’s death. What is undisputed is that the timeline of the affair and the murder plan are virtually identical.

Within weeks of Copenhagen, the forged prescriptions began. The sicker Annette got, the more Walsh entered the family’s orbit. She met the children at the emergency department during Annette’s first hospitalization. On the day Colin stayed home to administer the final dose, Greg and Anthea were at Walsh’s house.

After the conviction, Walsh became the children’s guardian. She told a television interviewer that she knew Bouwer and that he was not a cold, calculating murderer. “He’s a very gentle man,” she said.

I have worked enough of these cases to know that this reaction is not stupidity. It is the result of cognitive investment so deep that accepting the truth would require dismantling everything you believed about your own judgment, your relationship, and yourself. Some people can’t afford that demolition.

“Stunningly Reckless”

Here’s the paradox of Colin Bouwer: for a man who lectured on the perfect murder, he was shockingly bad at committing one.

He emailed international hypoglycemia experts from his own computer, posing as a forensic psychiatrist, asking how likely an insulin injection was to be detected at postmortem. He forged eleven prescriptions under fictitious names, a paper trail a first-year detective could follow. He kept the mortar and pestle and the leftover drugs in his own house. He claimed his wife had prostate cancer, which is impossible for a woman. He fabricated a suicide note so unconvincing it became prosecution evidence. He shaved his head and told people he was undergoing chemotherapy; the South African doctors he claimed had treated him testified at trial, via satellite link from South Africa, that they had never met him. They also identified a letter on their hospital’s headed notepaper, purporting to describe his cancer treatment, as a forgery.

Carl Elliott called this behavior “stunningly reckless.” It is. But it tells us something important about how psychopathy actually works, something that the popular culture image of the cold, calculating mastermind consistently gets wrong.

What psychopaths reliably display is not superior planning ability but inflated self-assessment (Patrick, 2022). They believe they are smarter, more capable, and more in control than they actually are. The gap between perceived competence and actual competence is one of the most consistent features of the condition, and it’s the reason so many psychopathic offenders get caught doing things that, in retrospect, look almost comically stupid.

Bouwer didn’t email those experts because he was careful. He emailed them because he believed no one would connect the dots. The lecture to medical students about insulin murder wasn’t a slip. It was showing off. Narcissistic grandiosity told him he could say it out loud and nothing would happen, because the rules that governed ordinary people didn’t apply to him.

The six-week trial, which began in Christchurch in October 2001, put all this recklessness before a jury. The prosecution called 155 witnesses. At one point, the court was linked by satellite television to South Africa, where the doctors Colin claimed had operated on his cancer denied all knowledge of him. The defense called five witnesses, including Vincent Marks, one of the world’s leading authorities on criminal insulin use, who had been retained to review the medical evidence. Even Marks, whose job was to raise reasonable doubt, acknowledged that the circumstantial evidence was strong.

The jury was out for three hours and twenty-five minutes. On the same day as the guilty verdict, the Royal College of Psychiatrists of Australia and New Zealand revoked Bouwer’s fellowship. The judge imposed a life sentence with a minimum non-parole period of thirteen years, later increased to fifteen on appeal, citing the heinousness of the crime.

The Family Business

While Colin Bouwer sat in a Dunedin jail awaiting trial, his son, Colin Bouwer Jr., was arrested in South Africa for the murder of his wife, Ria.

Ria was twenty-three. She was found dead in the guest bathroom of their Kempton Park home. Her underwear was slashed. Toiletries were scattered around the room, staged to look like a break-in and sexual assault. Colin Jr. told police he had been out of the house for several hours with their seven-month-old daughter, Melissa.

It was a lie. After strangling Ria, Colin Jr. had called his mother, Mariette Kruger (Colin Sr.’s first wife), who came to the house and helped him fabricate an alibi, create the appearance of sexual assault, and rearrange the crime scene. Father and son, both convicted of murdering their wives. Both staged the aftermath to look like something it wasn’t.

Mariette was convicted of being an accessory after the fact and sentenced to two years. The judge noted that she harbored intense animosity toward Ria, whom she considered unsuitable for her son. Colin Jr. was convicted of murder in 2003. He had been described as a jealous, domineering husband who threw Ria out of the house on multiple occasions.

Colin Jr. reportedly saw his father only about 5 times over 20 years. That limited contact makes it difficult to attribute the family murder legacy to role modeling. You can’t learn to kill your wife by watching your father if you rarely saw your father.

But that’s not how heritability works. Twin studies of callous-unemotional traits in children, the developmental precursors of adult psychopathy, estimate heritability at roughly 0.7, meaning approximately seventy percent of the variation in these traits is attributable to genetic factors (De Brito et al., 2021). The estimate varies depending on which facets of psychopathy are being measured, and it should not be mistaken for a claim that psychopathy is inherited like eye color. But the core mechanism appears to involve reduced amygdala reactivity, producing diminished emotional response to other people’s distress. That’s not something you pick up from bad parenting. That’s wiring.

What you learn from your environment is how to handle the wiring. Colin Jr. didn’t grow up with a present father. He grew up with an absent one, a man who abandoned his family, lied about their fate, and moved on to a new wife and a new country. He grew up with a mother angry enough at her daughter-in-law to help stage a murder scene. Whatever genetic predisposition Colin Jr. carried, it developed in a context of abandonment and rage and the implicit understanding that other people’s lives could be disposed of when they became inconvenient.

The Psychiatrist Who Diagnosed Himself

In prison, Bouwer’s narrative about Annette’s death went through three distinct versions. First: I didn’t do it; her illness was real. Second: she killed herself after discovering my affair. Third, and this was the version he brought to his parole hearings starting in 2015, it was assisted suicide, based on a prior agreement between them that if either ever developed a terminal illness, the other would help end their life.

The Parole Board noted that no terminal illness existed and that the murder was clearly premeditated. They documented a pattern of ongoing “untruths” about his crime, his motives, and his health. In 2015 and 2016, they denied parole.

During this period, Bouwer converted to Messianic Judaism. He wrote about suffering, plagiarizing Dostoyevsky without attribution. He became tearful at hearings when discussing what he had done. The Board noted the tears. They also noted that his answers, when challenged, “did not seem compelling.”

I see this pattern in forensic evaluations so often that it has its own rhythm. The offender has been in treatment long enough to learn the vocabulary. He can say “empathy,” “accountability,” and “restorative justice.” He can produce tears at the right moments. But when you press on specifics, when you ask him to describe exactly what his victim experienced, to inhabit her perspective for even thirty seconds, the language goes flat. He returns to his own suffering. The Parole Board saw what any experienced forensic evaluator would see: a man who had memorized the script of rehabilitation without ever performing the role for real.

But the most revealing moment came in a letter Bouwer wrote from prison, in which he diagnosed himself:

“I have learnt that the pain of people like me with personality disorders is intense and not easily verbalised. I do not believe the medical profession, nor the general public, will ever understand the pain that ‘psychopaths’ endure.”

I want to set that next to something Annette said.

In a hospital bed, before surgery to remove two-thirds of her pancreas for a tumor that did not exist, Annette Bouwer prayed with her friend Magda. She asked God to bless the surgery. She said she wanted to live for the sake of her children.

A woman who wanted to live, and a man who wanted pity for what he did to her.

Bouwer knew the word for what he was. He could name it, discuss it, write about it with the kind of clinical detachment that passes for sophistication in a prison setting. What was absent was the part that matters: the capacity to hold his own pain and Annette’s in the same frame and recognize that they are not the same kind of pain. That hers was something he imposed on her. That was the consequence of the choices he made. He couldn’t do that. Not because he lacked the intelligence, but because the machinery required for that kind of moral reckoning was never installed.

The psychiatrist who diagnosed himself got the label right. What he couldn’t access was the meaning underneath it.

Bouwer was paroled in 2017 and deported to South Africa, banned from ever returning to New Zealand. He died of natural causes on August 15, 2018. He was sixty-seven.

What Annette’s Case Reveals

It would be satisfying to end this article with a list of failures. The hospital didn’t screen for sulphonylureas with adequate sensitivity—the clinicians who accepted a spouse’s report that the house was clean of medications. The ward staff who heard Annette say she was being poisoned and didn’t push it. But I’ve sat with this case long enough to know that satisfaction would be dishonest.

The sulphonylurea assay Dunedin had was the assay Dunedin had. The clinicians who asked Colin to search the house were following standard practice, not cutting corners. And the ward staff who heard a patient with unexplained neurological symptoms allege poisoning by the Head of Psychiatry were not negligent in failing to act on it. They were operating inside a hierarchy that made the allegation sound delusional. The system didn’t fail through incompetence. It functioned exactly as designed, and a psychopath exploited it.

That is the uncomfortable takeaway from this case. Medical hierarchy exists for good reasons. Seniority confers authority because experience matters, because hospitals need clear chains of command, and because patients benefit when someone is in charge. But the same structure that makes a hospital function also makes it vulnerable to a predator who holds rank within it.

Colin Bouwer’s prescribing authority let him obtain lethal drugs without suspicion. His department chair title made subordinates afraid to challenge him. His psychiatric training taught him how people think, what they fear, and how to use both. The institution didn’t just fail to catch him; it failed to catch him. It provided the camouflage.

Andrew Bowers refused to sign that death certificate, and he deserves credit for it. But notice what it took: a junior physician overriding every institutional signal telling him to defer to a senior colleague, on a hunch he couldn’t yet articulate, in the home of a man who outranked him. That is an extraordinary act of professional courage, and it is not a reasonable basis for building a safety system around. Safety systems that depend on individual heroism are not systems. They are luck.

What would actually help is structural: protocols that remove discretion from the people closest to the suspect. In equivocal hypoglycemia cases, LC-MS/MS screening should be standard rather than dependent on clinical suspicion. Home medication searches in suspected poisoning cases should be conducted by someone other than a family member. When symptoms resolve in a supervised setting and recur at home, the home environment itself should be treated as a variable in the differential, not an afterthought. These are not radical proposals. They are design changes that would make the hierarchy harder to exploit.

Annette Bouwer read murder mysteries. I have wondered, since I first encountered this case, whether she recognized the plot she was living inside. Whether the patterns from all those novels whispered to her that this was not a disease but a design. She told the hospital staff she was being poisoned. She told them directly. And there was no mechanism in the system to take that seriously, because the man she was accusing was the system.

That is not a story about people who should have tried harder. It is a story about an institution that was never built to catch the person at the top.

References

Cleckley, H. (1941). The mask of sanity: An attempt to clarify some issues about the so-called psychopathic personality. C.V. Mosby.

Corbett, J. (2001, November 25). The Colin Bouwer case: A fatal betrayal. New Zealand Herald. https://www.nzherald.co.nz/nz/the-colin-bouwer-case-a-fatal-betrayal/DHYB35ER7IIHNUORPIC5PI5QQQ/

Curtis, D. A., & Hart, C. L. (2023). Pathological lying: Theory, research, and practice. American Psychological Association.

De Brito, S. A., Forth, A. E., Baskin-Sommers, A. R., Brazil, I. A., Kimonis, E. R., Pardini, D., Frick, P. J., Blair, R. J. R., & Viding, E. (2021). Psychopathy. Nature Reviews Disease Primers, 7(1), Article 49. https://doi.org/10.1038/s41572-021-00282-1

Defiant Bouwer to be deported. (2017, September 4). Otago Daily Times. https://www.odt.co.nz/news/dunedin/defiant-bouwer-be-deported

Elliott, C. (2010, September 6). Mind game. The New Yorker. https://www.newyorker.com/magazine/2010/09/06/mind-game

Ferguson, C., & McLachlan, F. (2023). Continuing coercive control after intimate partner femicide: The role of detection avoidance and concealment. Feminist Criminology. Advance online publication. https://doi.org/10.1177/15570851231189531

Halty, L., Horcajo-Gil, P. J., Mesa, G. P., López-Ossorio, J. J., & González-Álvarez, J. L. (2023). Prevalence of psychopathy, intimate partner homicide, and suicide risk in Spain. Anuario de Psicología Jurídica, 33, 83–89. https://doi.org/10.5093/apj2023a1

Manning, P. J., Espiner, E. A., Yoon, K., Drury, P. L., Holdaway, M., & Bowers, A. (2003). An unusual cause of hyperinsulinaemic hypoglycaemia. Diabetic Medicine, 20(9), 772–776. https://doi.org/10.1046/j.1464-5491.2003.01010.x

Marks, V., & Richmond, C. (2007). Insulin murders. RSM Press.

Marks, V., & Richmond, C. (2008). Colin Bouwer: Professor of Psychiatry and murderer. Journal of the Royal Society of Medicine, 101(8), 400–408. https://doi.org/10.1258/jrsm.2007.071003

Patrick, C. J. (2022). Psychopathy: Current knowledge and future directions. Annual Review of Clinical Psychology, 18, 387–415. https://doi.org/10.1146/annurev-clinpsy-072720-012851

Psychiatrist who poisoned wife fights deportation. (2016, September 1). Radio New Zealand. https://www.rnz.co.nz/news/national/312311/dunedin-psychiatrist-who-poisoned-wife-fights-deportation

Sohn, J., & Hong, Y.-O. (2025). Are specific traits of psychopathy more predictive of intimate partner homicide? Feminist Criminology. Advance online publication. https://doi.org/10.1177/15570851251340075

Thomson, A., & Gifford, G. (2003, June 7). Bouwer’s ex-wife jailed over killing. New Zealand Herald. https://www.nzherald.co.nz/nz/bouwer-ex-wife-jailed-over-killing/5O72MMTB5C3DIIWVUK4RQO4MIM/

We believe he’s innocent. (2001, November 25). News24. https://www.news24.com/news24/we-believe-hes-innocent-20011125