Henry Cotton’s Surgical Bacteriology — Yanking Teeth and Colons, Killing a Third of Patients

Summary

Beginning around 1916 at the New Jersey State Hospital in Trenton, its medical director Henry Andrews Cotton announced that insanity was not a disease of the mind but a hidden bacterial infection — "focal sepsis" — seeded in the teeth, tonsils, sinuses, stomach, spleen, cervix, and above all the colon, and curable by cutting those organs out; the gap between his claimed 85 percent cure rate and the documented reality, in which his colon resections killed on the order of 30 percent of patients and cured essentially none, is the entire case. Cotton extracted teeth by the tens of thousands and performed serial abdominal surgeries on institutionalized, frequently non-consenting psychiatric patients, on the theory that a "detoxified" body would yield a sound mind.

The practice was not a fringe horror tolerated in obscurity. It was celebrated. Cotton, a protégé of the most powerful American psychiatrist of the era, Adolf Meyer of Johns Hopkins, was lauded in the press as a humane reformer who had abolished mechanical restraint and replaced it with "scientific" surgical cure. His 1921 book, The Defective, Delinquent and Insane, packaged focal-sepsis theory as the frontier of psychiatry, and visiting clinicians from Britain and the United States toured Trenton to watch the operations.

The reckoning was assembled but then buried. In 1924–25 Meyer commissioned a former student, the psychiatrist Phyllis Greenacre, to audit Cotton's results; she tracked his surgical patients and found the records chaotic, the cure claims unsupported, the recovery rate no better — in fact worse — than for unoperated patients, and the mortality catastrophic. Meyer suppressed her report rather than publish it, protecting Cotton and his own reputation, and the surgeries continued for years. Only Cotton's retirement in 1930 and his death in May 1933 finally ended them. The focal-infection theory of insanity collapsed entirely thereafter, and "surgical bacteriology" survives now as a textbook case of how a charismatic clinician, a surrogate endpoint, and a senior authority's cowardice can sustain a lethal, evidence-free operation on the powerless for the better part of two decades.

Timeline

1907

Cotton appointed director at Trenton

At roughly 30, Henry Cotton becomes medical superintendent of the New Jersey State Hospital, arriving as a modernizing reformer who abolishes mechanical restraints and straitjackets.

c. 1916

The focal-sepsis program begins

Cotton concludes that insanity is caused by chronic, low-grade infection ("focal sepsis") and begins systematically removing teeth as the first line of treatment, escalating to tonsils, sinuses, and internal organs.

1919

A prominent death

Margaret Fisher, daughter of the Yale economist Irving Fisher, dies of streptococcal infection at Trenton after a series of operations under Cotton's regime — one of many fatalities, but one connected to a powerful family.

1920–1922

Colectomies escalate

Cotton's own statistics for this window show colon-resection mortality on the order of 25–30 percent; he nonetheless reports cure rates as high as 85 percent.

1921

Cotton publishes his manifesto

The Defective, Delinquent and Insane presents focal-sepsis surgery as proven therapy; Cotton lectures in the U.S. and Britain to wide acclaim.

1922–23

Kopeloff and Kirby test the theory

At a New York hospital, Nicholas Kopeloff and George Kirby run a controlled comparison of focal-infection removal versus no surgery — one of psychiatry's first controlled trials — and find no therapeutic benefit.

Fall 1924

Meyer commissions Greenacre

Adolf Meyer of Johns Hopkins, Cotton's mentor, sends psychiatrist Phyllis Greenacre to audit Cotton's outcomes after mounting doubts.

1925

Greenacre delivers a damning follow-up

Tracking Cotton's surgical patients, Greenacre finds the recovery rate no better — and by some measures worse — than for unoperated patients, with very high mortality and unsupported cure claims.

1925

New Jersey legislative scrutiny

Testimony from former patients and staff reaches a state senate inquiry; hospital trustees and allied physicians defend Cotton, and the New York Times reports Trenton praised as "the most progressive institution in the world."

1925–26

Meyer suppresses the report

Rather than publish Greenacre's findings, Meyer shelves them, reassigns her, and shields Cotton; the surgeries continue.

Oct 1930

Cotton retires

Cotton steps down as active director; the most aggressive abdominal surgeries wind down, though some procedures persist under his lingering influence.

May 8, 1933

Cotton dies

Cotton dies of a heart attack; the focal-sepsis program ends with him, and the theory collapses into discredit over the following years.

A Reformer Builds a Theory of Everything

Cotton arrived at Trenton as a genuine modernizer. He banned restraints, improved sanitation, and brought a laboratory ethos to a custodial asylum, and that credibility is exactly what made the catastrophe possible. Trained in the new bacteriology and mentored by Adolf Meyer, Cotton fused two respectable ideas — that infection causes disease, and that hidden ("focal") sepsis in teeth and tonsils could cause distant illness — into one unfalsifiable claim: that all insanity was, at bottom, chronic infection poisoning the brain. The theory was seductive because it was mechanistic, optimistic, and surgical at a time when asylums offered no cures at all. But Cotton's evidence was the same trap that recurs across this catalog: a surrogate endpoint and the operator's own uncontrolled testimony. "Cure" meant a patient who had calmed enough to be discharged, attributed to the surgery by the surgeon, with no control group, no blinding, and no honest accounting of the patients who left in a coffin rather than through the front gate. On that foundation he declared cure rates up to 85 percent and began pulling teeth, then escalating inward.

From Extractions to the Colon: Industrialized Mutilation

The logic of focal sepsis had no natural stopping point, and Cotton followed it organ by organ. When tooth extraction did not cure a patient, the inference was never that the theory was wrong; it was that the infection lurked deeper — in the tonsils, the sinuses, the stomach, the spleen, the cervix, the testicles, and finally the colon. Patients who failed to recover were operated on again, and again, in a ratchet that could only escalate because the framework forbade the possibility that the treatment itself was the harm. The abdominal surgeries were the deadliest. By Cotton's own published figures, colon resection carried a mortality on the order of 25–30 percent; independent observers put some series higher still. These were not consenting volunteers but committed psychiatric patients, frequently operated on over their objection, in an institution where the director's authority over inmates was near-total. Cotton proposed colectomies even on children to "prevent" insanity and curb habits like masturbation. The surrogate endpoint — a discharged, "detoxified" patient — was tallied as success; the true endpoint, a recovered and living person, was rarely reached and routinely buried.

The Audit That Was Written and Then Hidden

The disconfirmation existed, on paper, years before the practice ended — and that is the darkest fact in the file. Two independent checks contradicted Cotton. Kopeloff and Kirby's controlled comparison, among the earliest controlled trials in psychiatry, found that removing "focal infections" produced no therapeutic benefit. More devastatingly, Phyllis Greenacre's 1924–25 follow-up, commissioned by Meyer himself, tracked Cotton's surgical patients and found the cure claims illusory: recovery was no better than among unoperated patients, the data were internally contradictory, and the mortality was appalling. Greenacre's was a controlled, longitudinal audit — exactly the evidence the field needed. Meyer's response defines the institutional failure. Rather than publish, he suppressed the report, declined to confront his protégé publicly, reassigned Greenacre, and let the surgeries continue. The state senate inquiry of 1925 fizzled against the testimony of loyal trustees and physicians who called Trenton the most progressive hospital in the world. No tribunal, no regulator, and no journal stopped Cotton. He retired in 1930 and died in 1933, eulogized by Meyer; the theory of focal sepsis dissolved only afterward, when no one was left to defend it.

Contributing Factors

01

Surrogate endpoint mistaken for cure

"Cure" was defined as a calmer, dischargeable patient credited to the surgeon, not as a recovered, surviving human being. Cotton optimized the proxy — apparent improvement, attributed to his operations — and declared an 85 percent success rate while the real endpoint, a living and well patient, was rarely reached. When the metric rewards the operator and ignores the dead, scale multiplies harm.

02

An unfalsifiable theory that escalated on its own failures

Focal sepsis could not be disproved from inside: a patient who did not recover simply harbored deeper infection, justifying the next, more dangerous operation. A framework in which treatment failure becomes evidence for more treatment has no internal brake and ratchets inevitably toward the colon — and the morgue.

03

Prestige and mentorship substituted for evidence

Cotton's credibility flowed from Adolf Meyer, American psychiatry's leading figure, and from his own reputation as a humane reformer who had abolished restraints. That borrowed authority functioned as the approval no controlled trial had granted, manufacturing consensus and deflecting scrutiny precisely when the death toll demanded it.

04

Harm concentrated on a captive, non-consenting population

The patients were committed psychiatric inmates — institutionalized, frequently operated on against their will, including children — with no power to refuse and no advocate. An intervention that operates on those who cannot say no has escaped the consent and evidence checks that constrain medicine, and that is where unconsented, lethal harm concentrates.

05

Disconfirming evidence suppressed by the authority that should have acted

The decisive audit was commissioned, completed, and then buried. Greenacre proved the surgeries did not work and killed at high rates, and Meyer hid the report to protect his protégé and himself. When the person empowered to stop a practice chooses reputation over disclosure, the harm continues for exactly as long as the cover-up holds.

Aftermath

The material consequence is counted in bodies and mutilated survivors: hundreds of major excisions, with abdominal-surgery mortality on the order of 30 percent and higher, and thousands more patients left toothless, organ-deprived, and no saner — a death toll that, on Greenacre's accounting, bought no measurable recovery at all. The durable ripple is twofold. Methodologically, the episode became a standing argument for controlled trials and honest follow-up in psychiatry: Kopeloff and Kirby's comparison and Greenacre's audit are remembered as early demonstrations that uncontrolled cure claims are worthless and that surgery for mental illness must be tested, not asserted. Ethically, it became a case study in the cost of suppressed evidence and protected reputations — Meyer's decision to bury the audit is now read as a graver failure than Cotton's zeal, because it was the failure of the safeguard. The focal-infection theory of insanity is wholly discredited, though its descendant — the popular intuition that hidden "toxins" or gut infection drive mental illness — never fully died. "Surgical bacteriology" survives as a byword for a confident, mechanistic, lethal therapy validated only by its inventor, abandoned not by a verdict but by an obituary.

Lessons

Measure the patient's survival and recovery, not the operator's claimed cures. If your success rate is self-reported by the person performing the procedure and your denominator quietly omits the dead, you are counting a surrogate that flatters the operator while it kills the subject. Demand independent, longitudinal outcomes before you believe any cure rate.
Distrust a theory that turns its own failures into a mandate for more. When non-response is read as "the problem is deeper, so escalate" rather than "the treatment may be wrong," you have an unfalsifiable engine that ratchets toward greater harm. Build in the result that would make you stop, and honor it.
Treat mentorship and reputation as flags, not proofs. A revered backer or a reformer's halo is a reason to demand controlled evidence, not a reason to waive it. Borrowed prestige is how an untested, dangerous practice acquires the consensus that data never earned it.
Watch hardest over those who cannot refuse. When an intervention is performed on the committed, the incapacitated, or children, the ordinary checks of consent and refusal are absent — so the external scrutiny must be greatest exactly where it is usually weakest.
Publish the damning audit; suppression is the deeper crime. If you commission an investigation, you are bound by its findings. Burying a report that shows a practice does not work and kills people does not protect anyone but the perpetrator — and it makes the silencer the author of every harm that follows.