Focal-sepsis surgery (Cotton)

1. Introduction

Version followed: the historical record of Henry A. Cotton's focal-sepsis programme at the New Jersey State Hospital at Trenton, mainly as the historian Andrew Scull reconstructs it in Madhouse: A Tragic Tale of Megalomania and Modern Medicine (Yale University Press, 2005) and as reviewers and medical historians summarise that work. The operations are not recommended and are not a treatment for anything.

What it was. "Focal sepsis" (also called focal infection) was the theory that hidden, chronic infections somewhere in the body (in the teeth, tonsils or internal organs) caused mental illness. It followed that removing the infected part would cure the madness. Cotton, a psychiatrist trained at Johns Hopkins under Adolf Meyer and in Munich under Emil Kraepelin and Alois Alzheimer, led the state hospital in Trenton. From the 1910s he had patients' teeth extracted and then moved on to removing tonsils and abdominal and pelvic organs. He reported cure rates of about 80–85 percent. Those claims rested on a broad definition of cure and on a high, unreported death rate after abdominal surgery. [source: Nuland 2005, NYRB; CMAJ review of Scull 2005; Hektoen International 2023]

Its core idea of change, as its proponents stated it. Mental disorder was treated as the downstream effect of a bodily infection, a "bacteriological" model of madness. Cure was supposed to come from finding the infected focus and removing it surgically. Cotton was influenced by the discovery that syphilis can cause brain disease, and he extended that model to other disorders. [source: Hektoen International 2023; Tyrer 2020, Psychiatric Times] Scull argues that Cotton was not a lone renegade. His work was enthusiastically received in the profession, it spread to Britain (notably through Thomas Chivers Graves), and sceptics were reluctant to challenge it. [source: Publishers Weekly review of Scull 2005; Tyrer 2020]

Why it matters to psychotherapists today. The episode is a standard case study of three things: a plausible-sounding biological theory used to justify invasive treatment without controlled evidence; patients who could not effectively refuse; and a professional network that suppressed an internal investigation (Phyllis Greenacre's 1925 audit). Clients still meet invasive or unproven "root-cause" treatments sold for mental symptoms, and some families have relatives who were treated in state hospitals of this era. [source: Nuland 2005; Publishers Weekly; Tyrer 2020] [synthesis for the present-day relevance]

2. Framework

2a. Historical phases of the practice (outline only)

Appointment and early reform (c. 1906–1907). Cotton became superintendent at Trenton. Nuland gives 1907, when Cotton was 31, and says he first removed the restraining devices used on inmates. The Hektoen vignette gives 1906. A CMAJ review dates his role as chief psychiatrist to 1916–1933. [uncertain: sources give different start dates] [source: Nuland 2005; Hektoen 2023; CMAJ review]
Start of the operations (c. 1913–1916). Dental extractions came first, then tonsils, then removal of abdominal and pelvic organs, often in search of an infection the previous operation had not found. Hektoen gives "around 1913". Carlat, summarising Scull, gives 1916. [uncertain: start date varies] [source: Hektoen 2023; Carlat 2005]
Publicity and spread (early 1920s). The New York Times praised the work in June 1922. Cotton and Graves presented together at the Medico-Psychological Association meeting in London in July 1923. Cotton claimed that first-year recoveries had risen from 37 percent to 85 percent. Similar operations were carried out at Rubery Hill Hospital in Birmingham, England, in the late 1920s and 1930s. Cotton also opened a private hospital in Trenton. [source: Tyrer 2020; Carlat 2005; Hektoen 2023]
Investigation and suppression (1925 onwards). Meyer sent his former trainee Phyllis Greenacre to audit the results. Her report is summarised below. Meyer suppressed the study, apparently to protect Cotton. Hektoen reports that a New Jersey Senate inquiry, prompted by complaints, found no wrongdoing and praised Cotton for saving the state money. [source: Nuland 2005; Publishers Weekly; Hektoen 2023] [uncertain: the Senate inquiry rests on a single secondary source]
Persistence and decline (to 1933 and after). Cotton kept to his methods until he died of a heart attack in 1933. Nuland states that over the following two decades radiological, bacteriological, biochemical, clinical and microscopic studies undermined focal sepsis and the related theory of autointoxication. [source: Nuland 2005; Carlat 2005; Hektoen 2023] Nuland and the CMAJ review disagree on whether the theory was already debunked by 1933 or took longer to fall. [uncertain]

2b. Phases of the psychotherapist's work

Session number and length follow the client's needs and the setting [synthesis]. The phases below are an organising frame [synthesis]; move between them by the criteria given, and loop back as needed.

Recognition and historical literacy. Know the history well enough to recognise its pattern today: a hidden bodily cause, a removal or purge, cure claims without controls, and dismissal of critics. Move on when a client's situation shows that pattern, or a client or family raises the history.
Assessment. Map what the client believes is causing their symptoms, what they have been offered or have undergone, who is selling it, and the risks (medical, financial, suicide, safeguarding). Move on when you have a working picture and know whether medical review, safeguarding or crisis action is needed.
Safety, safeguarding and medical routing. Act on any immediate risk, involve the client's own physician for anything somatic, and follow child-protection or adult-safeguarding routes where someone cannot protect themselves. Move on when the immediate risk is managed.
Processing and repair. Work with medical trauma, bodily loss, betrayal, grief, family history and shame, and with mistrust of mental-health services. Move on when the client can hold the experience with less distress and has re-engaged with supported care.
Advocacy and follow-up. Support complaints, records requests and commemoration, and help the client keep up with supported care. Bring what you learn into your own professional reflection. End or taper when the client's goals are met.

3. Guiding principles

No treatment by removal without evidence. Any claim that removing a body part, organ or "focus" will cure mental symptoms is treated as unproven unless independent evidence shows otherwise. Practically: never endorse, suggest or refer to such a procedure. Route all somatic questions to the client's own physician. [source: Nuland 2005; Tyrer 2020] [synthesis for application]
Plausible mechanism is not proof. Cotton's theory sounded scientific and built on real findings about syphilis. Practically: ask what controlled outcomes and independent replication exist, not just whether the story makes sense. [source: Hektoen 2023; Tyrer 2020]
Watch how "cure" is defined and who is counted. Cotton's high cure rates came from a broad definition of cure and left out the deaths. Practically: when a client cites success rates, ask together how cure was measured and what happened to people it did not help. [source: CMAJ review; Nuland 2005]
Consent is the patient's to give. Hektoen reports the operations were done without informed consent. Scull describes some patients being taken to surgery against their resistance. Practically: support clients' right to refuse, to seek second opinions and to have time to decide. [source: Hektoen 2023; Publishers Weekly]
Critics are not uncaring. Tyrer notes that critics were cast as uncaring, because no one can argue against treating infection. Practically: in teams and with clients, keep room for doubt about popular treatments. [source: Tyrer 2020]
Institutions can protect their own. Greenacre's findings were suppressed. Practically: when a colleague or service harms clients, use formal reporting routes rather than relying on informal reassurance. [source: Nuland 2005; Publishers Weekly] [synthesis]
Separate legitimate medical questions from unproven ones. Some links between infection and psychiatric symptoms are recognised and still studied (for example PANS/PANDAS in children, per NIMH). Practically: do not dismiss a client's medical concern. Do not diagnose either. Route it to qualified medical assessment. [source: NIMH 2025]
Never advise on medicines or procedures. Practically: do not suggest starting, stopping or changing any medicine, antibiotic, supplement or procedure. Say it is a question for the prescriber or physician. [synthesis]
Respect the explanatory model and stay honest about evidence. Practically: explore the client's beliefs with curiosity, and say plainly what the evidence shows. [synthesis]
Protect children and adults at risk. Practically: follow local child-protection and adult-safeguarding procedures when someone who cannot consent faces unnecessary invasive treatment. [synthesis]
Protect patients' identities. Practically: when discussing the history with clients or teams, describe patients by role and keep relatives' records private. [synthesis]
Hold the profession's history honestly. Practically: acknowledge that psychiatry's own leaders backed this programme. Clients' mistrust can be well founded. [source: Publishers Weekly; Nuland 2005]

4. Intervention inventory

The interventions below are the psychotherapist's own work; steps not drawn from a cited source are marked [synthesis]. None is a step of the practice. The psychotherapist never provides, supervises or refers to focal-sepsis surgery or to any invasive procedure sold for mental symptoms.

4.1 Building your own historical literacy

Purpose / when: Before working with clients or families affected by early-20th-century somatic treatments, or when teaching the history. [synthesis]
Steps: (1) Read a reliable secondary history (Scull 2005) and at least two independent reviews of it, noting where they disagree (dates, death rates). (2) Learn the pattern: hidden bodily cause, removal, cure claims without controls, suppressed audit. (3) Note the British strand (Graves, Rubery Hill) as well as Trenton. (4) Keep a short reference note with sources, so you can answer questions accurately. [source: Nuland 2005; Tyrer 2020; CMAJ review] [synthesis for steps]
Example language (illustrative): "What we know about Trenton comes largely from a historian's reconstruction of the hospital records, and the reported figures vary between accounts."
Watch for: Overstating certainty (for example, quoting one exact death rate as fact). Sensationalising. Naming patients. You are on track when you can describe the history accurately and briefly, with its uncertainties.

4.2 Screening for invasive or unproven somatic "cures" for mental symptoms

Purpose / when: At intake and whenever a client mentions a physical explanation, a "root cause" practitioner or a planned procedure for anxiety, depression, psychosis, obsessions or similar. [synthesis]
Steps: (1) Ask an open question about every treatment, test or practitioner the client has used or plans to use for their mental symptoms, including dental, surgical, "detox", gut or infection-focused approaches. (2) Ask who recommended it, what it costs and what they were told it would do. (3) Note anything invasive, irreversible or expensive. (4) Ask whether their own physician knows. (5) Record it factually. [synthesis]
Example language (illustrative): "Some people try physical treatments for how they feel emotionally. Have you tried, or been offered, anything like that, such as tests, procedures or programmes aimed at a hidden cause?"
Watch for: Shame or defensiveness. Reassure the client that you are asking to keep them safe, not to judge. The screen is working when clients tell you about non-conventional treatments without being prompted.

4.3 Exploring the client's explanatory model

Purpose / when: When a client strongly believes their mental symptoms have a hidden bodily cause. [synthesis]
Steps: (1) Use a structured cultural and explanatory-model interview (such as the DSM-5-TR Cultural Formulation Interview) to ask what they think causes the problem, why now, and what would help. (2) Reflect the model back without arguing. (3) Ask what it would mean to them if the cause were not (only) physical. (4) Identify common ground, for example that wanting a clear, fixable cause makes sense when suffering is severe. [synthesis]
Example language (illustrative): "It sounds as though finding one physical cause would make this feel fixable. Can you tell me what it would mean if the answer turned out to be more mixed?"
Watch for: Debating too early, which tends to entrench the belief. Collusion, meaning agreeing in order to avoid conflict. It is working when the client becomes curious about other contributing factors.

4.4 Collaborative medical routing

Purpose / when: Whenever a client raises a possible physical cause. Some infection–psychiatric links are recognised and still studied; NIMH describes PANS and PANDAS as a sudden onset of OCD or restrictive eating in children before puberty, with no confirming lab test, diagnosed after other conditions are ruled out. [source: NIMH 2025]
Steps: (1) Take the concern seriously. (2) With consent, encourage review by the client's own physician or paediatrician. (3) For children with sudden-onset OCD or food restriction, suggest the family raise it with the paediatrician. NIMH points families to the International OCD Foundation or the PANDAS Physicians Network if their local provider is unfamiliar with these conditions. (4) Do not diagnose, and do not suggest tests, antibiotics or other treatments yourself. (5) Keep psychological care going alongside; NIMH lists CBT among symptom management for PANS/PANDAS. [source: NIMH 2025] [synthesis for steps 1, 2 and 4]
Example language (illustrative): "That's a real medical question, and it belongs with your doctor. I can help you think about what to ask, and we can keep working on the anxiety here in the meantime."
Watch for: Two opposite errors: dismissing a legitimate medical concern, or appearing to endorse an unproven procedure. Do not conflate recognised but debated conditions such as PANS/PANDAS with focal-sepsis surgery. It is working when the client has an appointment with an accountable physician and stays engaged with you.

4.5 Psychoeducation using the history

Purpose / when: When a client is weighing an invasive "root-cause" treatment, or asks whether bodily theories of mental illness have been tried. Use it only if the client is interested; it is not a lecture. [synthesis]
Steps: (1) Ask permission to share a piece of history. (2) Summarise neutrally: a 1920s theory that hidden infections caused madness, high claimed cure rates, high unreported deaths, a suppressed audit, a theory later abandoned. (3) Draw out the general lesson, that cure claims need controlled evidence and informed consent. (4) Ask what the client makes of it. Do not equate their current practitioner with Cotton. [source: CMAJ review; Nuland 2005] [synthesis for steps]
Example language (illustrative): "In the 1920s some doctors were sure hidden infections caused mental illness, and they reported high cure rates. Later checks told a very different story. What strikes you about that?"
Watch for: The client feeling mocked or lectured. Keep it brief and collaborative. It is working when the client starts asking evaluative questions of their own.

4.6 Working with a client drawn to an invasive procedure for mental symptoms

Purpose / when: A client intends to undergo, or is paying for, an invasive or irreversible intervention promoted as curing mental symptoms. [synthesis]
Steps: (1) Explore their hopes and fears using a motivational, non-confrontational stance. (2) Weigh expected benefits against risks, costs and irreversibility together. (3) Encourage an independent medical second opinion. (4) Help them list questions for the practitioner (see 4.7). (5) Respect their autonomy. If they decide to go ahead, stay available and keep up safety planning. (6) Never refer or endorse. [synthesis]
Example language (illustrative): "You've been suffering for a long time, and this offers hope. Can we look together at what you know about it, and what you'd want to know before deciding?"
Watch for: Desperation, financial pressure, a practitioner discouraging second opinions, and promises of certain cure. All of these raise the risk. It is working when the client takes time, gets independent advice, or makes a decision with full information.

4.7 Informed-consent literacy

Purpose / when: To strengthen a client's ability to evaluate any proposed procedure. This is a direct counter to the lack of consent recorded at Trenton. [source: Hektoen 2023] [synthesis for steps]
Steps: (1) Explain that the client has the right to clear information, time, a second opinion and refusal. (2) Help them draft questions in their own words: what evidence shows this helps mental symptoms, how success is measured, what the risks and alternatives are, what happens if it does not work, and who profits. (3) Role-play asking those questions if the client is anxious about doing so. [synthesis]
Example language (illustrative): "You're allowed to ask how they know it works, and to say 'I need time to think.'"
Watch for: Practitioners who are evasive or pressure for quick decisions. It is working when the client feels able to ask and to refuse.

4.8 Working with parents seeking invasive treatment for a child's mental symptoms

Purpose / when: A parent is pursuing unproven invasive or irreversible procedures for a child's behaviour, anxiety or obsessions. [synthesis]
Steps: (1) Validate the parent's worry and wish to act. (2) Explore what has been proposed and by whom. (3) Encourage paediatric review (see 4.4). (4) Keep the child's own voice central: seek the child's views and assent at an appropriate level. (5) If the child is at risk of significant harm, follow 4.9. [source: NIMH 2025 for paediatric routing] [synthesis]
Example language (illustrative): "You clearly want to help. Let's make sure your child's doctor is part of this, and talk about what your child thinks too."
Watch for: Coercion of the child, repeated procedures, doctor-shopping, a child's distress about medical visits. It is working when decisions run through an accountable paediatrician and the child's wellbeing is monitored.

4.9 Safeguarding and reporting

Purpose / when: A child, or an adult who cannot protect themselves, is being subjected to, or is at imminent risk of, an unnecessary invasive procedure, or is being harmed by a programme promising a physical cure for mental symptoms. [synthesis]
Steps: (1) Consult your supervisor and your service's safeguarding lead. (2) Follow your jurisdiction's mandatory-reporting and child-protection or adult-safeguarding procedures. (3) Record facts, not opinions. (4) Tell the family about the referral where this is safe and permitted. (5) Keep supporting the client. [synthesis]
Example language (illustrative): "I'm worried about your child's safety, and I have a duty to share that concern with the people whose job is to check it. I'd like to talk you through what happens next."
Watch for: Delay while you seek certainty, which is not required for a referral. Keep within your role: you are raising a concern, not investigating. It is working when the protective agencies are involved and the child is safe.

4.10 When a colleague or institution promotes such a practice

Purpose / when: A colleague, service or employer recommends, provides or refers clients to invasive procedures for mental symptoms without evidence, or suppresses concerns about outcomes. This is the Greenacre situation in modern form. [source: Nuland 2005; Publishers Weekly] [synthesis for steps]
Steps: (1) Document what you observed. (2) Raise it first through supervision or with the colleague, if that is safe. (3) If unresolved or serious, use your organisation's clinical-governance or whistleblowing route and your licensing board's or regulator's complaint process. (4) Avoid any personal involvement. If you have a financial or other role in a clinic that sells such procedures, that is a multiple relationship that can impair objectivity (APA Ethics Code, Standard 3.05). (5) Protect client confidentiality throughout. [source: APA Ethics Code for Standard 3.05] [synthesis for other steps and ethics points]
Example language (illustrative): "I have concerns about the evidence behind referring clients for this, and I'd like to raise it formally so it can be reviewed."
Watch for: Informal reassurance standing in for review, retaliation, and "everyone does it" reasoning. It is working when there is a documented, independent review.

4.11 Working with clients harmed by unnecessary invasive procedures

Purpose / when: A client underwent invasive or irreversible procedures, such as dental, surgical or other, promoted as treating their mental symptoms, and lives with the consequences. [synthesis]
Steps: (1) Let the client tell the story at their own pace, without graphic detail being required. (2) Assess for medical-trauma responses, grief for lost function or body parts, body-image distress, anger and self-blame. (3) Make sure ongoing physical consequences are being managed by their physician. (4) Offer an evidence-based trauma-focused or grief-focused therapy where indicated, with consent. (5) Address self-blame directly: the client acted on expert advice or under pressure. [synthesis]
Example language (illustrative): "You trusted people who told you this would help. What happened to your body was not your fault."
Watch for: Re-traumatisation when medical settings come up, avoidance of necessary medical care, and suicidal thinking (see 4.17). It is working when distress reduces and the client can use the medical care they need.

4.12 Addressing institutional betrayal and mistrust of services

Purpose / when: The client, or their family, distrusts psychiatry or psychotherapy because of past harm, or because they know histories such as Trenton. [synthesis]
Steps: (1) Acknowledge that the mistrust has grounds: leading figures in the profession supported Cotton, and an investigation was suppressed. (2) Do not get defensive. (3) Make your own practice transparent: what you do, its evidence, and the client's right to question, complain and stop. (4) Agree how the client will tell you if they feel pressured. [source: Publishers Weekly; Nuland 2005] [synthesis]
Example language (illustrative): "Psychiatry has got things badly wrong before, and people were hurt. You have every right to ask me how I know what I do here helps, and to stop if it doesn't."
Watch for: Mistrust spilling into rejection of all care. Collusion with blanket anti-treatment views. It is working when the client can test trust step by step.

4.13 Family-history work with descendants of state-hospital patients

Purpose / when: A client has learned that a relative was a patient at Trenton or a similar hospital of the period, possibly one who underwent such operations or died after them. [synthesis]
Steps: (1) Explore what the client knows and how they learned it. (2) Make space for grief, anger, shame or curiosity. (3) Discuss family narratives, such as secrecy, "madness in the family" or stigma, and how those narratives shape the client. (4) If the client wants to find records, help them plan for emotionally difficult discoveries. Access to historical patient records varies by state and archive. [uncertain: not verified] (5) Consider ways to commemorate or reframe the relative's story. [synthesis]
Example language (illustrative): "What has it been like to discover this about your great-grandmother? What do you want to do with what you've learned?"
Watch for: Fear of hereditary mental illness. Offer accurate information and medical or genetic counselling referral if the client asks. Watch also for distress from graphic records. It is working when the story becomes part of the family history rather than a secret.

4.14 Supporting records requests and complaints

Purpose / when: A client wants to request their own medical records, complain about a practitioner, or report a procedure to a regulator or medical board. [synthesis]
Steps: (1) Clarify what the client wants to achieve. (2) Help them identify the right body, such as the practitioner's licensing board, a health ombudsman or a consumer-protection agency, and, for legal questions, a lawyer. (3) Help them prepare emotionally for the process and its possible outcomes. (4) Do not give legal advice. [synthesis]
Example language (illustrative): "Would it help to work out together who the right place to complain to is, and what you want to come out of it?"
Watch for: Retraumatisation during proceedings and unrealistic expectations. It is working when the client feels they have a voice, whatever the outcome.

4.15 Hope and meaning after a failed "cure"

Purpose / when: A client invested heavily in a procedure, emotionally, financially or physically, and it did not help. [synthesis]
Steps: (1) Name the disappointment and the loss. (2) Explore sunk-cost feelings and self-blame. (3) Separate "this did not work" from "nothing will work". (4) Re-open supported options: for psychosis, for example, NICE guideline CG178 covers treatment and management in adults. [source: NICE CG178 page] [uncertain: specific recommendations] (5) Rebuild short-term goals. [synthesis]
Example language (illustrative): "You put so much into this, and it didn't give you what was promised. That's a real loss. It doesn't mean there's nothing that can help."
Watch for: Hopelessness and suicidal thinking (4.17), and searching straight away for the next unproven procedure. It is working when the client re-engages with supported care.

4.16 Distinguishing supported from unsupported treatments with a client

Purpose / when: A client is confused by competing claims, for example online "root-cause" content alongside mainstream care. [synthesis]
Steps: (1) Introduce a few plain criteria: independent controlled studies; endorsement by guideline bodies such as NICE; transparency about risks and failures; no promise of certain cure; no discouragement of second opinions. (2) Apply them together to one claim the client cares about. (3) Show the client where to find guideline-body pages. [source: NICE CG178 page for an example of a guideline body] [synthesis]
Example language (illustrative): "Let's check this together. Who tested it, against what, and what happened to the people it didn't help?"
Watch for: Overwhelming the client with information. It is working when the client can apply the questions independently.

4.17 Suicide-risk assessment and crisis planning

Purpose / when: After harm from a procedure, a failed cure, a family discovery, or whenever hopelessness appears. [synthesis]
Steps: (1) Ask directly about suicidal thoughts, intent and plans, using your service's standard risk-assessment approach. (2) Build a collaborative safety plan: warning signs, coping steps, people to contact, and professional and crisis contacts. (3) Discuss safe storage or removal of means without discussing methods. (4) Give crisis resources: 988 Suicide & Crisis Lifeline (US, call or text 988); Samaritans 116 123 (UK and Ireland). (5) Follow up at the next contact. [synthesis]
Example language (illustrative): "When people feel let down like this, sometimes they think about not wanting to be alive. Has that been on your mind?"
Watch for: Sudden calm after despair, withdrawal and giving things away. Act according to your service's protocol. It is working when the client uses the plan and stays connected.

4.18 Screening for food restriction or "cleansing" rationales

Purpose / when: Some present-day hidden-cause narratives involve restricting food or purging the body in order to remove a supposed source of mental symptoms. These can overlap with eating difficulties. NIMH also notes that restrictive eating can be a presenting feature in PANS. [source: NIMH 2025] [synthesis for the rest]
Steps: (1) Ask non-judgementally about eating, restriction and any cleansing regime. (2) Screen for disordered eating with your service's standard tool. (3) Route medical risk to the client's physician, and refer to a registered dietitian or an eating-disorder service. (4) Give support information: the National Alliance for Eating Disorders. (5) Give no diet, fasting or supplement advice. [synthesis]
Example language (illustrative): "Have you changed how you eat, or cut things out, to try to fix how you're feeling?"
Watch for: Weight loss, dizziness, secrecy and rigid rules. All of these need medical review. It is working when the client is under medical and dietetic care.

4.19 Teaching the history in training and supervision

Purpose / when: In supervision, team teaching or continuing education, to build the habits that would have caught Trenton earlier. [source: Tyrer 2020] [synthesis for steps]
Steps: (1) Present the case in outline, with sources and stated uncertainties. (2) Draw out Tyrer's lessons: reliance on case reports rather than comparisons; practice changed on untested theory; critics dismissed as uncaring. (3) Discuss Greenacre's audit and its suppression as a lesson about independent review. (4) Ask participants to apply the lessons to a current practice in their own field. [source: Tyrer 2020; Nuland 2005]
Example language (illustrative): "What would an independent audit of our own outcomes look like, and who would we let see it?"
Watch for: Treating the episode as a distant curiosity. It is working when teams adopt outcome monitoring and open review.

4.20 Monitoring your own outcomes and inviting review

Purpose / when: Routine practice, as the positive counterpart to the lessons of Trenton, where cure claims went unchecked for years. [source: CMAJ review] [synthesis for steps]
Steps: (1) Use routine outcome measures and record deterioration as well as improvement. (2) Bring cases that are not improving to supervision. (3) Invite client feedback on the alliance and on whether the work is helping. (4) Report adverse events through your service's procedures. [synthesis]
Example language (illustrative): "I'd like to check regularly whether this is actually helping you, and to hear it if it's not."
Watch for: Counting only the successes, which mirrors the broad definition of cure in the history. It is working when deterioration is noticed early and acted on.

5. Handling problems

The problems below are the psychotherapist's own synthesis [synthesis], except where a source is cited.

A client is set on an invasive procedure and won't discuss it. Do not argue. Ask permission to come back to it later, keep safety planning in place, and encourage an independent medical opinion. Stay in the relationship so the client can return to you if it goes badly (see 4.6). [synthesis]
The client experiences questions as dismissal ("you think it's all in my head"). Repair the rupture. Say you take the physical concern seriously, which is why it belongs with a physician, and that you are not doubting their suffering. Check how they experienced the conversation. [synthesis]
Overwhelm when discussing harm or family records. Slow down, ground the client, set the material aside, and return to stabilisation before any trauma processing. Do not ask for graphic details. [synthesis]
Non-adherence to supported care after mistrust. Explore the reasons (see 4.12). Negotiate small, reversible steps, and respect the client's right to decline. Never pressure; that would repeat the history. Do not advise on medicines; send those questions to the prescriber. [synthesis]
A recognised medical question gets mixed up with an unproven one. For example, a parent says their child has PANDAS and is seeking an unproven invasive treatment. Keep the two apart: support paediatric assessment, which NIMH describes as ruling out other conditions with no single confirming test, and raise concerns about anything invasive or unproven with the paediatrician and, where the child is at risk, through safeguarding (4.9). [source: NIMH 2025] [synthesis]
Your own anger at the history, or at a current practitioner. Take it to supervision. Moralising can alienate clients who chose the procedure in good faith. [synthesis]
A colleague dismisses concerns as "just being cautious" or "anti-innovation". This repeats the dynamic Tyrer describes, in which critics were cast as uncaring. Use formal review routes (4.10). [source: Tyrer 2020] [synthesis]
The client wants you to confirm or deny a specific history. For example: "Was my great-grandfather killed by Cotton?" Be honest about what is known and what is uncertain. Point the client to archives and historians. Avoid speculating. [synthesis]
Suicidal crisis. Follow your service's protocol and give crisis resources: 988 (US), Samaritans 116 123 (UK and Ireland) (see 4.17). [synthesis]
Risk to a child or an adult at risk. Safeguarding takes priority over the therapeutic frame. Explain your duty, refer, and stay involved where you can (4.9). [synthesis]

Harm account

What was actually done

From the 1910s until his death in 1933, Henry A. Cotton, superintendent of the New Jersey State Hospital at Trenton, subjected large numbers of psychiatric patients to the removal of supposedly infected body parts in order to cure their mental illness. A Publishers Weekly review of Scull's history says thousands of patients. [source: Nuland 2005; Publishers Weekly; Hektoen 2023; Carlat 2005] The sources list, at the level of categories, removal of teeth and tonsils, then of internal organs: parts of the bowel, the stomach, gall bladder, spleen, appendix, uterus and cervix, ovaries and testicles. Tyrer also mentions injections of so-called vaccines. [source: Nuland 2005; Tyrer 2020; Hektoen 2023] Hektoen notes Cotton had no surgical training but often operated himself. [source: Hektoen 2023] He extended the approach to his own family and himself, having their teeth removed. [source: Hektoen 2023; Publishers Weekly] He ran a private clinic in Trenton alongside his state post, and his ideas were promoted to the press, administrators and politicians. [source: Carlat 2005; CMAJ review] In Britain, Thomas Chivers Graves championed related surgery at Rubery Hill Hospital in the late 1920s and 1930s. [source: Tyrer 2020, citing Scull 1996]

What it looked like to patients and families. To patients and families it looked like ordinary hospital medicine: tooth extractions and then operations on the throat or abdomen, presented as treatment for the mental illness and carried out at the state hospital without informed consent [source: Hektoen 2023; Nuland 2005]. Some patients were operated on against their will and resisted being taken to surgery [source: Nuland 2005; Publishers Weekly]. After losing their teeth, some could not speak or eat properly, and Greenacre found patients kept in overcrowded, unwholesome conditions [source: Hektoen 2023]. Patients' complaints led to a New Jersey Senate inquiry, which found no problems and praised Cotton for saving the state money [source: Hektoen 2023] [uncertain: reported only by a secondary source]. Cotton kept believing in the method until his death [source: Hektoen 2023]. Warning signs in a family history: teeth or organs removed from a relative during a psychiatric admission, and a death recorded after "surgery for infection" in a state hospital [synthesis].

How it caused harm

Deaths. Nuland reports that about a third of surgical patients died. The CMAJ review gives an unreported death rate of 30 percent. Carlat, summarising Scull, gives 30 percent for abdominal surgery, later revised to nearer 45 percent. Hektoen gives 30–40 percent for bowel operations. [uncertain: figures differ between accounts] [source: Nuland 2005; CMAJ review; Carlat 2005; Hektoen 2023]
Maiming and disability. Greenacre's audit found maiming and disfigurement and found the treatment useless. Many patients could not speak or eat properly after losing their teeth. [source: Nuland 2005; Hektoen 2023]
Absence of consent. The operations were done without informed consent. Some patients were operated on against their will, and some resisted being taken to surgery. [source: Hektoen 2023; Nuland 2005; Publishers Weekly]
False claims of cure. Cotton's 80–85 percent cure claims rested on a broad definition of cure and on omitting the deaths. Other centres could not replicate them. [source: CMAJ review; Hektoen 2023]
Suppression and delay. Greenacre found flawed protocols, chaotic records and falsified results. Meyer suppressed her study. Hektoen reports that a state Senate inquiry found no problems. The evidence stayed out of view for more than a decade. [source: Nuland 2005; Hektoen 2023; CMAJ review] [uncertain: the Senate inquiry is reported only by a secondary source]
Professional enabling. Scull argues the profession's enthusiasm and reluctance to challenge Cotton allowed the harm to continue and spread. [source: Publishers Weekly]

Recognising it today

Focal-sepsis surgery itself ended with the theory's decline. [source: Nuland 2005] Its pattern can still be recognised [synthesis]:

a single hidden bodily cause offered for a wide range of mental symptoms;
invasive, irreversible or costly removal, "cleansing" or extraction offered as the cure;
high success rates without independent controlled studies, and silence about people it did not help or harmed;
pressure to decide quickly, discouragement of second opinions, and dismissal of critics as uncaring or closed-minded;
the practitioner or clinic profiting directly.

Do not conflate this pattern with recognised medical questions. NIMH describes PANS and PANDAS as studied conditions with defined diagnostic criteria, no confirming test and ongoing research, managed by physicians. The concern is unproven invasive treatment, not the medical investigation of symptoms. [source: NIMH 2025]

What to do instead and how to help those harmed

Supported care. Mental-health conditions should be assessed and treated according to guideline bodies. For psychosis and schizophrenia in adults, see NICE guideline CG178. [source: NICE CG178 page] [uncertain: specific recommendations] Physical concerns go to accountable physicians. For children with sudden-onset OCD or restrictive eating, NIMH describes paediatric assessment with CBT among symptom management. [source: NIMH 2025] The psychotherapist never refers to an unproven invasive procedure.
Survivors' support. Trauma-informed, consent-centred therapy, attention to bodily loss and medical trauma, and help using the medical care the person needs (4.11, 4.15). [synthesis]
Family support. Descendants of state-hospital patients may need help with grief, stigma and records searches (4.13). [synthesis]
Reporting duties and legal routes. Safeguarding referrals for children and adults at risk (4.9). Complaints to licensing boards, regulators or health ombudsmen. Clinical-governance and whistleblowing routes for practitioners (4.10, 4.14). A lawyer for legal questions. [synthesis]
Crisis resources. 988 (US); Samaritans 116 123 (UK and Ireland).
Survivor organisations. No survivor organisation specific to focal-sepsis surgery is known [uncertain].

Sources

Web sources:

Books and chapters cited:

Scull, A. (2005). Madhouse: A Tragic Tale of Megalomania and Modern Medicine. Yale University Press. (Details as given in the reviews above.)
Scull, A. (1996). "Focal sepsis and psychosis: the career of Thomas Chivers Graves (1883–1964)." In 150 Years of British Psychiatry, Volume II: The Aftermath (as cited by Tyrer 2020).

← Previous · XVI-051Deep sleep therapy (Chelmsford)Next · XVI-053 →Psychic driving (Cameron, MKUltra)