Estimated reading time at 200 wpm: 42 minutes

Prescribing and a prescription are not the same thing. The prescription is the visible endpoint. A signed line. A ledger entry. A medication authorised.

Prescribing is everything that must happen before that endpoint can properly be reached. A diagnosis must be formulated against recognised criteria. Authoritative clinical guidelines must be considered. Capacity must be assessed. A consent dialogue must take place, meeting the Montgomery standard of material disclosure. If the medication is unlicensed for the condition being treated, that departure must be justified. Monitoring must be arranged. A contemporaneous record of the whole must be made.

These are not optional extras. They are the prescribing. Many will be surprised to learn that prescription is not the endpoint. It is a signpost. The authorisation is a visible marker — a moment when the duties of assessment, diagnosis, capacity evaluation, consent dialogue, and risk-benefit analysis crystallise into a written instruction. But the broader duty of prescribing does not end there. It continues. The GMC’s own standards make this clear, and they extend the prescriber’s responsibility well beyond the signature. [Click this to scroll to that part]

The system rarely sees it this way. It treats repeat prescribing as an administrative function — re-authorising an existing instruction, clearing a ledger. Productivity is measured by prescriptions processed, not by duties discharged. A signature is demanded. But the time and the information that would make that signature defensible are not supplied.

The result is that a prescription which appears to represent care may represent something else entirely. It may be an unchecked inheritance, passed from one clinician to the next. Each assumes someone else did the work. Often, no one did.

This article is about what happens when that inheritance reaches a consultant psychiatrist who stops to look at what they are being asked to sign.

It opens with a fictional scenario. A consultant psychiatrist — one of three appearing before the GMC for the same case — faces cross-examination at a Fitness-to-Practise Tribunal. The patient did not die from a side effect of the medication. The Coroner was critical of the duties of care. The questions the GMC’s counsel asks are the questions the regulator asks. The fictional scenario is not a record of actual proceedings. But it is drawn from the standards the GMC applies. It is not too far from reality. Those who have attended such hearings will know. The author has been present to witness and given expert evidence at the GMC.

After the cross-examination, the article steps back. It examines the legal and regulatory framework that encases every prescribing decision. It explains what that framework is, how it is composed, and why a generation of psychiatrists may not have noticed its construction.

Many doctors were trained in the era of Bolam. They learned that clinical justification, validated by peer consensus, was enough. It is not in the current framework. It is necessary, but it is not sufficient. The world has moved on. The profession’s understanding has not always moved with it.

The aim of this article is not to alarm. It is to illuminate. It argues three things. First, that prescribing is not a clerical act. Second, that clinical justification alone is not enough. Third, that the duties wrapped around every prescribing decision are not optional.

They are the framework against which every doctor – every psychiatrist – will be judged. They apply whether the system makes them easy to discharge or not.

Fictional scenario at the GMC’s MPTS

Context: The patient, a 34-year-old man with a recorded diagnosis of emotionally unstable personality disorder (borderline type), died by suicide while under the care of the community mental health team. He had been prescribed risperidone 2 mg twice daily by Dr X, a consultant psychiatrist, approximately eighteen months before the patient died. The Coroner’s inquest returned a narrative verdict and was critical of the standard of clinical care. The Coroner’s report was forwarded to the GMC. Dr X is appearing before a Fitness-to-Practise Tribunal (MPTS). The GMC’s counsel, Ms Talbot, is examining him. The scenario is fictional but not too far from reality.

Ms Talbot: Dr X, you prescribed risperidone to this patient for approximately eighteen months. What condition were you treating?

Dr X: Emotionally unstable personality disorder, borderline type. He had a long history of emotional dysregulation, recurrent self-harm, and difficulty maintaining relationships. He was in considerable distress.

Ms Talbot: And where in your clinical records have you set out the diagnostic criteria you applied to reach that diagnosis?

Dr X: The diagnosis was already in his notes when I took over his care. It had been made by a previous consultant.

Ms Talbot: I am not asking who made it, Dr X. I am asking where — in your own clinical records — you documented the basis on which you accepted it.

I have before me the ICD-10 Diagnostic Criteria for Research which is the same as your Blue Book but operationalised. For disorders of adult personality, the general criteria at F60 require G1 to G6, to be satisfied. You can refer to a copy of those criteria before you in Doc115.

Those criteria must be satisfied before any specific personality disorder can be diagnosed. You may refer to Doc116 for the sub-criteria on Borderline PD. None of those matters were recorded in the records you made – were they?

Dr X: I accept that the diagnosis that was in the record. I did not personally reapply the ICD-10 criteria.

Ms Talbot: Would you agree that no psychiatrist — yourself, the previous consultants, anyone whose notes you inherited — ever generated evidence in the clinical records that this diagnosis was properly made against the WHO criteria?

Dr X: The diagnosis was well established in the notes. It had been repeated by several clinicians.

Ms Talbot: It may have been repeated several times, Dr X, but none of the psychiatrists had ever generated evidence in the records that the condition was properly diagnosed — would you agree? It was a label repeated by several people who each assumed someone else had done the work. In other words it was a fictional diagnosis, repeated several times. I hope that is not what you mean by “established.”

Dr X: I believed the diagnosis was correct.

Ms Talbot: Let me ask you directly: do you believe that a diagnosis of personality disorder should be made by the proper application of the WHO criteria set out in ICD-10?

Dr X: In practice, no psychiatrist uses the criteria in that way. We form a working clinical impression based on the patient’s history and presentation. We do not sit with a checklist.

Ms Talbot: That is an extraordinary statement, Dr X. You are telling this Tribunal that a diagnosis carrying significant consequences — prolonged prescribing of an antipsychotic medication with some potentially serious side effects — was made on the basis of a “working clinical impression,” without any documented application of the internationally recognised diagnostic criteria for the condition. Is that your evidence?

Dr X: That is how psychiatric diagnosis works in practice.

Ms Talbot: I see. Perhaps you would suggest the same approach to abdominal surgery. A surgeon forms a working impression of which organ to remove. No need to document how the diagnosis was arrived at. No need to justify the reasoning. Just a label, passed from one surgeon to the next, each assuming the previous one did the work. Would that be acceptable, Dr X?

Dr X: I th.. think — that is not a fair comparison.

Chairman of the Tribunal: Ms Talbot, that’s enough. You have made your point. Please confine questions to matters of evidence.

Ms Talbot: I am obliged, Mr Chairman. I withdraw the comparison and apologise for my deviation. May I continue.

Chairman: Nods to Ms Talbot.

Ms Talbot: Dr X, let me direct you to Good medical practice, which states that in providing clinical care you must “adequately assess a patient’s condition, taking account of their history, including symptoms and relevant psychological factors” [GMP, para. 7(a)]. The word is must, Dr X. Not should. Not where practicable. It is a duty of care. Where in the records is the evidence that you discharged that duty in relation to this patient’s diagnosis?

Dr X: It is not recorded.

Ms Talbot: Yes. It is not recorded at all. And where there is no record, there is no evidence that the duty was discharged — is that fair observation?

Dr X: I accept that the documentation does not reflect my clinical thinking.

Ms Talbot: The documentation is the only evidence this Tribunal could have of what might have been your clinical thinking. If it is not in the record, it does not exist for our purposes.

Let us move to the prescription itself. You prescribed risperidone. Are you aware that risperidone is not licensed for the treatment of emotionally unstable personality disorder?

Dr X: I am aware, yes. But it is widely used in practice by many consultant psychiatrists to treat personality disorders. It is common practice in the profession. There is evidence it can help with affective dysregulation and other symptoms.

Ms Talbot: Evidence which you did not cite in the clinical record.

Dr X: I did not cite it, no.

Ms Talbot: You have a copy of CG78 before you in Doc110. Are you familiar with NICE clinical guideline CG78, which addresses the treatment of borderline personality disorder?

Dr X: I am.

Ms Talbot: Let me read into evidence from CG78. Paragraph 1.3.5.1: “Drug treatment should not be used specifically for borderline personality disorder or for the individual symptoms or behaviour associated with the disorder (for example, repeated self-harm, marked emotional instability, risk-taking behaviour and transient psychotic symptoms).“

Paragraph 1.3.5.2: “Antipsychotic drugs should not be used for the medium- and long-term treatment of borderline personality disorder.“

These are not ambiguous statements, are they, Dr X? They seem to cover a whole class of medications.

Dr X: No, they are clear.

Ms Talbot: And yet you prescribed an antipsychotic — risperidone — for the medium- to long-term treatment of a patient with a recorded diagnosis of borderline personality disorder. The GMC’s prescribing guidance states that you should take account of clinical guidelines published by national bodies such as NICE [GMCRx, para. 14]. It also states that you must only prescribe if it is safe to do so, and that it is not safe to prescribe without sufficient information about the patient’s health [GMCRx, para. 26]. I put it to you that there no evidence that you considered CG78, or how you came to depart from it?

Dr X: It is not recorded.

Ms Talbot: Where is the balance? A clinician departing from a national guideline that explicitly states antipsychotics should not be used for the condition being treated has a professional duty to document the reasons for that departure. The duty arises from the obligation to take account of NICE guidance [GMCRx, para. 14], and from the standards on record-keeping which require you to include the information shared with patients and the decisions made [GMP, para. 69–70]. There is no documentation of any of this.

Dr X: It was not documented.

Ms Talbot: So we have no proper diagnosis, an unlicensed prescription, contrary to national guidance, with no documented consideration of the guideline you were departing from, and no recorded justification. Let us turn to the matter of consent. Before you prescribed risperidone to this patient, did you assess his capacity to consent?

Dr X: I did. He was an intelligent man. He understood what I was proposing.

Ms Talbot: This patient had a diagnosis of emotionally unstable personality disorder. He had a history of recurrent self-harm and emotional dysregulation. By definition, those features indicate some functional impairment that could affect his decision-making. You had evidence, from the very diagnosis you accepted, that his capacity might require assessment in relation to a decision about unlicenced prescribing of antipsychotic medication. The GMC’s guidance requires you to start from the presumption of capacity but also requires you to be alert to signs that patients may lack capacity and to assess it where there is reason to doubt it [DMC, para. 81–82]. Given the functional impairment inherent in the diagnosis you accepted, capacity needed to be tested — did it not, Dr X?

Dr X: I formed the view that he had capacity. I did not conduct a formal assessment.

Ms Talbot: Yes – there is no record of that — formal or otherwise. The GMC’s guidance states that assessing capacity is a core clinical skill, that you should be able to draw reasonable conclusions during your dialogue with the patient[DMC, para. 82 and 83]. This is absent from the records you made – isn’t it?

Dr X: I did not make a specific entry about capacity testing.

Ms Talbot: So we have your word, today, that capacity was checked but not assessed — a man with emotional dysregulation and recurrent self-harm, about to be prescribed an antipsychotic off-label — and a clinical record that is entirely silent on the matter. Would you agree that a doctor who had actually assessed capacity in compliance with the GMC’s standards ought to have generated documentary evidence of that assessment?

Dr X: In an ideal world, yes.

Ms Talbot: Yes – it’s not an ideal world. Let me ask you about the consent dialogue. Did you explain to this patient that risperidone was not licenced for his condition?

Dr X: I do not recall discussing the licencing issue specifically.

Ms Talbot: Did you explain to the patient that the national clinical guideline for his condition — CG78, which I have just read to you — does not recommend the use of antipsychotic medication, either for the condition or for its individual symptoms?

Dr X: I do not believe I framed it in those terms.

Ms Talbot: Did you explain the material risks the patient would need to weigh; perhaps metabolic changes, hyperprolactinaemia, sexual dysfunction, and a range of common dystonias associated with this medication? Did you tell him that there were alternatives — psychological therapies, other medications, no medication? Did you tell him he could refuse?

Dr X: I discussed the medication in general terms. I do not recall specific details.

Ms Talbot: Yes – there is no record of that discussion. Decision making and consent states that you must give patients the information they want or need to make a decision, including diagnosis, prognosis, options — including the option to take no action — and the potential benefits and risks of harm [DMC, para. 10]. That is a must, Dr X. Not a should. A mandatory duty. There is no evidence that you discharged that duty – is there?

Dr X: My note from that consultation is brief.

Ms Talbot: Yes, it is. I have it here. Following one line of eight in the record: “MSE – emotional dysregulation, poor sleep, urges to self harm, attention-seeking and manipulative behaviours,” it reads: “Impression – EUPD. Deemed to have capacity. Risperidone 1 mg od commenced. Increase to 2mg bd over 3 weeks. Review in 6 weeks.” That is the entirety of the entry of your rationale. Would you accept that this falls short of the standard required by Good medical practice, which states that records must be clear, accurate, and that they should include the information shared with patients and the decisions made? [GMP, para. 69–70] The word is must again. A mandatory duty.

Dr X: In retrospect, it is a brief entry. Many psychiatrists write notes like that.

Ms Talbot: We are dealing with your entry – not many psychiatrists. Your entry is basically empty – isn’t it Dr X? It provides no evidence that any of the mandatory duties were discharged. It provides no evidence of what information was given, what the patient was told about the unlicenced nature of the prescription, what the patient was told about the NICE guideline, what risks were disclosed, what alternatives were discussed, or whether consent was freely given. It tells this Tribunal nothing. You seem to think that what ‘many psychiatrists’ do is the standard. It isn’t.

Let us continue. You prescribed an unlicensed medicine. The GMC’s guidance on unlicensed medicines requires you to be satisfied that there is sufficient evidence or experience of using the medicine to demonstrate its safety and efficacy, to take responsibility for monitoring, and to make a clear, accurate and legible record of all medicines prescribed and your reasons for prescribing an unlicensed medicine [GMCRx, para. 105(a)–(c)]. Each of those is a mandatory duty. There is no evidence that you delivered the required standards and duties.

Dr X: The reasons were not recorded. The monitoring was planned but not consistently followed through.

Ms Talbot: Not consistently followed through. Let me put it more directly: there is no record of monitoring at all for eleven of the eighteen months the deceased was on risperidone. The GMC’s guidance states that you must make sure that suitable arrangements are in place for monitoring, follow-up and review [GMCRx, para. 92]. That is a must. Again no evidence you discharged the duty.

Dr X: I had intended to. The service was under considerable pressure. I was under pressure. One of the balls dropped.

Ms Talbot: Nonetheless we are dealing with duties. The duty is to ensure monitoring is in place. The evidence of compliance would be records of monitoring being arranged and results being acted upon. Would you agree that the record contains no evidence that these duties were discharged?

Dr X: Y…y..yes.

Ms Talbot: Dr X, I want to return to something you said earlier. The issue was about capacity is delineated in the Mental Capacity Act 2005, and the addition from Montgomery that shifts care to towards a patient-centric position. You are familiar with these matters. How do you approach them? [DMC, para 12, 23]

Dr X: Yes. But we are not lawyers. I trained to diagnose and treat, not to parse statutes and regulatory guidance. The standards you are putting to me — capacity assessments, consent dialogues, documented consideration of NICE etc — these are legal constructs. They have been imposed on a profession that is already drowning. I did not go to medical school to become a legal clerk.

Ms Talbot: Dr X, I am going to put it to you that you have misunderstood something fundamental. I can agree with you that they legal constructs imposed on the profession from outside; from a body of case law, statute and inquiries. However you and your profession are where you are. They are your own regulator’s injection of legal accountabilities into what good clinical practice looks like. They are the things a competent psychiatrist must do or should be doing.

You do not need a law degree to know that you must verify a patient’s diagnosis before prescribing an antipsychotic. That is not a legal requirement. It is a clinical one. You do not need to be a lawyer to know that you must assess and document whether a patient with emotional dysregulation and recurrent self-harm has the capacity to consent to a medication. That is not a legal technicality. That is a clinical duty owed to a vulnerable person – and yes the law encases it.

Dr X: We will have to agree to disagree.

Chairman: Ms Talbot, can we move on.

Ms Talbot: Yes Mr Chairman.

Let me turn to the patient’s death. He died by suicide. I am not suggesting, and the GMC does not suggest, that risperidone caused his death. But the Coroner was critical of the standard of care he received, and particularly of the multiple duties of care for which there was no documentary evidence of compliance. The Coroner noted that there was no evidence that the diagnosis had been properly formulated against recognised criteria, no evidence of capacity assessment, no evidence of a consent dialogue, no evidence that NICE guidance had been considered, no evidence of justification for unlicensed prescribing, and no evidence of meaningful monitoring. The Coroner’s words were “a systemic failure of clinical duties of care.” Do you accept that description?

Dr X: I accept that the records were not satisfactory.

Ms Talbot: Dr X, let me be precise. The records were not just unsatisfactory. And the absence of those relevant documentation means there is no evidence before this Tribunal that you discharged the duties the GMC imposes on every doctor who prescribes medicines. Those are the must duties. The ones from which there is no exemption for a busy clinic or a pressured service. I put it to you that you cannot demonstrate — from your own notes — that you made an adequate assessment of this patient’s condition, that you considered the relevant clinical guidelines, that you obtained informed consent, that you disclosed the unlicensed status of the medication, that you arranged for monitoring, or that you reviewed the prescription appropriately. Is that a fair summary of the facts?

Dr X: …Yes.

Ms Talbot: One final matter. Did you, at any point during the eighteen months you prescribed risperidone to this man, raise a concern about the workload or the systems in your service that were making it difficult for you to meet the duties?

Dr X: Not in writing. We all knew the service was ridiculously over-stretched. We were all drowning in overwork. It’s a national problem. Everyone knows about it. If we take the time to document in the way you are suggesting, then more patients will suffer – and probably die – they just won’t be seen. We simply do not have the time.

Ms Talbot: The GMC’s guidance requires you to raise concerns and to act promptly if you think patient safety is being seriously compromised by the service or system you are working in [GMP, para. 75(b)] — a must. Did you consider that a service in which a consultant psychiatrist cannot document the evidence of duties being discharged is a service in which patient safety may be seriously compromised?

Dr X: I thought I was doing the best I could in very difficult circumstances. It’s the same across nearly all mental health services – the same among my consultant colleagues.

Ms Talbot: Dr X, I am sure you believe that. Your colleagues in their many numbers are not here defending their licence to practise. But the question this Tribunal must answer is whether your best met the professional standards expected of a consultant psychiatrist — standards expressed, in many cases, as duties bearing the word must. The record you have left behind provides no evidence that those duties were discharged. And that, Dr X, is the case against you. The standards are not subject to the pressures of service or lack of resources, else there might as well be no standards at all. I have no further questions, Mr Chairman.

The Legal Wrapper: How the Framework Encases Every Clinical Decision

A prescribing decision is not a single act. It is a sequence of duties. Each layer of duty surrounds the clinical core, and each layer must be discharged — and evidenced — independently. The clinical justification sits at the centre. But it cannot be reached without passing through every layer of the wrapper. What follows describes those layers. They are drawn from the GMC’s professional standards and they apply to every prescription, repeat or new, in every setting.

The GMC create duties of care. It is not there to find ‘negligence’ in tort. Those duties of care would have been driven by failures arising from inquiries, Coroner’s investigations and negligence case law. To be blunt the GMC has been sharpening their guillotine! GMC guidance stands in the uncomfortable place of quasi-legislation. Not statute. Not secondary legislation – but in a particular aspect of true legality. Quasi-legislation is not half-baked law. It is about regulating individuals in a profession. That is its legal force.

Layer 1: Diagnosis

The first layer is the diagnosis. Before any medication is prescribed, the condition being treated must be properly identified and recorded.

The GMC requires that a doctor must adequately assess a patient’s condition, taking account of their history, symptoms, and relevant psychological and social factors [GMP, para. 7(a)]. This means the diagnosis must be formulated, not merely stated. The criteria must be applied. The reasoning must be documented.

A label without a formulation is not a diagnosis. It is a placeholder. And a prescription built on a placeholder rests on sand.

Layer 2: The clinical guideline

Once a diagnosis is established, authoritative guidance must be considered. The GMC states that a doctor should take account of clinical guidelines published by NICE and other authoritative sources [GMCRx, para. 14].

This does not mean a doctor must follow every guideline. It means the guideline must be considered. If the doctor departs from it, the departure must be reasoned and recorded.

Where the record is silent, there is no evidence that the guideline was considered at all. Silence reads as ignorance, not as reasoned departure.

Layer 3: Licensed status

Every medicine has a licence. The licence specifies the conditions for which the medicine is approved, the doses, and the populations.

When a doctor prescribes within the licence, this layer is thin. When they prescribe outside it — off-label — the layer thickens. The GMC requires that the doctor must be satisfied there is sufficient evidence or experience to demonstrate the medicine’s safety and efficacy for the proposed use [GMCRx, para. 105(a)]. The doctor must take responsibility for monitoring and follow-up [GMCRx, para. 105(b)]. And the doctor must make a clear, accurate record of the reasons for prescribing an unlicensed medicine [GMCRx, para. 105(c)].

These are mandatory duties. They cannot be delegated to a guideline or assumed from common practice.

Layer 4: Capacity

Every adult patient is presumed to have capacity. But the presumption is not a conclusion. It is a starting point.

The GMC requires that a doctor must start from the presumption that every adult patient has capacity [DMC, para. 81]. Where there is reason to question capacity — and functional impairment from a mental health condition may provide such reason — it must be assessed. The assessment must be recorded.

A patient with a diagnosis of emotionally unstable personality disorder, experiencing emotional dysregulation and recurrent self-harm, may have fluctuating capacity. To assume capacity without assessment is not an application of the presumption. It is a failure to engage with it.

Layer 5: Consent and information

Consent is not a signature on a form or a tick-box exercise. It is the outcome of a dialogue.

The GMC requires that a doctor must give patients the information they want or need to make a decision [DMC, para. 10]. This includes the diagnosis, the prognosis, the options — including the option to take no action — and the potential benefits and risks of harm of each option.

The standard of disclosure is not what a responsible body of doctors would say. It is what a reasonable person in the patient’s position would want to know. That is the Montgomery test. It requires the doctor to understand what matters to this particular patient before deciding what to disclose.

For an off-label prescription, the patient must be told that the medicine is not licensed for their condition. They must be told what is known and what is uncertain. They must be told that there are alternatives — even if those alternatives are not immediately available. And they must be told they can say no.

A note that reads “Risperidone 1 mg commenced” records none of this. It records an action. It does not record a decision.

Layer 6: Monitoring and review

A prescription is not a one-time event. It is an ongoing responsibility.

The GMC requires that a doctor must make sure that suitable arrangements are in place for monitoring, follow-up and review [GMCRx, para. 92]. Review is particularly important where patients are frail, have multiple illnesses, are on medicines with serious side effects, or where the BNF recommends blood tests or other monitoring [GMCRx, para. 94].

Monitoring that is planned but not executed is not monitoring. Results that are available but not checked are not review. A prescription that continues for months without anyone asking whether it is still needed, still effective, or still safe has left the framework entirely.

Layer 7: The record

The record is not a bureaucratic afterthought. It is the evidence that every other layer of the wrapper was discharged.

The GMC requires that records must be clear, accurate, contemporaneous, and legible [GMP, para. 69]. They should include the clinical findings, the drugs proposed or prescribed, the information shared with the patient, the concerns expressed by the patient, the decisions made, and who made them [GMP, para. 70].

Where the record is silent, there is no evidence. And where there is no evidence, the regulator will conclude — on the balance of probabilities — that the duty was not discharged. The clinical justification does not fill the gap. The record is the only proof.

Layer 8: Escalation

The final layer applies when the system itself makes compliance with the other layers difficult or impossible.

The GMC requires that if patients are at risk because of inadequate resources, policies, or systems, a doctor must first protect patients and then raise the concern [GMP, para. 75(b)]. If a colleague may not be fit to practise and may be putting patients at risk, the doctor must ask for advice and, if still concerned, must report it [GMP, para. 75(c)].

These are not optional. They are mandatory. A doctor who continues to prescribe in a system that makes safe prescribing impossible, without raising the concern, is not being pragmatic. They are breaching the final layer of the wrapper — the duty to speak.

How the layers work together

Each layer is independent. A breach of one is not cured by compliance with another.

A doctor may have a clear diagnosis, consider the guidelines, and arrange monitoring — but if there is no record of the consent dialogue, the wrapper is broken at layer 5.

A doctor may have an excellent clinical rationale, documented in detail — but if capacity was never assessed, the wrapper is broken at layer 4.

A doctor may do everything else right — but if the medicine is unlicensed and the patient was never told, the wrapper is broken at layer 3.

The GMC does not weigh the layers against each other. It does not ask whether the clinical outcome was good. It asks whether each duty was discharged. And it asks where the evidence is.

Why clinical justification is not enough

A doctor who says “I can justify this clinically” has addressed one question. It is the question at the centre of the wrapper. But the GMC asks seven other questions, and each one requires its own answer, supported by its own evidence.

The doctor who cannot produce that evidence — for diagnosis, for capacity, for consent, for guideline consideration, for monitoring, for review — is not safe, however sound the clinical reasoning. The wrapper is there to ensure that clinical reasoning is not the only thing standing between a patient and harm.

When the wrapper is intact, the clinical decision is supported by a structure that makes it defensible. When it is broken, the clinical decision stands alone. And under scrutiny, a decision standing alone is a decision waiting to fall.

Discussion

The world in which many psychiatrists trained

For much of the second half of the twentieth century, the legal framework governing clinical negligence in the United Kingdom was dominated by a single case. Bolam v Friern Hospital Management Committee [1957] established that a doctor was not negligent if they acted in accordance with a practice accepted as proper by a responsible body of medical opinion. That standard — clinical justification, validated by peer consensus — became the lens through which a generation of doctors understood their accountability. If you could explain your clinical reasoning, and if a body of your peers would support it, you were safe.

Bolam was always a civil standard. It was born in a negligence claim and never left the civil domain. But its influence extended far beyond the law of tort. It became a professional pedagogy. It taught doctors that the question to ask was: can I defend this clinically?

For decades, that question appeared to be sufficient. And there was a reason.

The regulatory backdrop: a criminal standard of proof

Until reforms initiated in the early 2000s — following the Shipman Inquiry and the Foster review of the GMC’s fitness to practise procedures, and culminating in the establishment of the Medical Practitioners Tribunal Service (MPTS) — the GMC was required to prove its case to the criminal standard: beyond reasonable doubt.

That high evidential bar meant that a doctor facing a GMC allegation could deploy Bolam-style reasoning as a powerful evidential shield. If a responsible body of medical opinion would support the clinical decision, how could the GMC prove beyond reasonable doubt that the doctor’s fitness to practise was impaired? The question was not whether Bolam was the GMC’s test — it never formally was — but whether the criminal standard of proof gave Bolam-style peer consensus a regulatory significance it never possessed in law.

It did. The high bar made it difficult for the GMC to establish impairment in the face of plausible clinical justification. And that difficulty shaped a generation’s understanding of what accountability meant. The question “can I defend this clinically?” appeared to be the whole answer, because for much of those doctors’ careers, it had been.

The shift to the civil standard — and what it changed

The GMC’s move from the criminal to the civil standard of proof — the balance of probabilities — was not merely procedural. It changed the nature of the regulatory inquiry.

Under the civil standard, the GMC no longer needs to eliminate reasonable doubt. It needs only to find it more likely than not that the doctor’s conduct fell below the standards expected. The absence of a record of capacity assessment is no longer neutral. It is evidence from which an inference may be drawn. The silence in the clinical notes — the missing consent dialogue, the undocumented consideration of NICE guidance, the absent monitoring schedule — is no longer a gap the doctor can fill with oral testimony and expect the benefit of the doubt. On the balance of probabilities, what is not recorded probably did not happen.

The GMC’s own framework for assessing fitness to practise reflects this shift. The standards describe good practice, and not every departure from them is considered serious [GMP]. But in assessing seriousness, the GMC considers the extent of any departure from those standards, any relevant context — including systems factors — and the doctor’s insight and remediation [GMP]. A doctor who cannot demonstrate, through contemporaneous records, that the mandatory duties were discharged faces an evidential difficulty that did not exist under the criminal standard.

The GMC’s mindset has changed — but the profession’s has not

The regulator has moved from a quasi-criminal, Bolam-influenced model to a civil, compliance-based one. The standards now impose procedural duties that exist independently of clinical defensibility: adequate assessment of the patient’s condition [GMP, para. 7(a)], consideration of clinical guidelines [GMCRx, para. 14], capacity assessment [DMC, para. 81–82], consent dialogue on the basis of material information [DMC, para. 10], arrangements for monitoring and review [GMCRx, para. 92], and clear, accurate, contemporaneous records [GMP, para. 69–70]. These are not clinical duties. They are legal and regulatory duties wrapped around the clinical act.

But many doctors — particularly those trained in the Bolam era, when the GMC had to meet a criminal standard — have not internalised this shift. They continue to believe that if they can explain why a medication was a reasonable choice, they have answered the question. They have not. They have answered a different question — the question the courts asked in 1957, and the question the GMC appeared to ask when the burden of proof was higher. The question the GMC asks now is broader, more demanding, and less forgiving of procedural gaps.

This is not a failure of individual competence. It is a structural failure of professional education and clinical governance. The GMC publishes the standards. It updates them. It applies them in fitness to practise proceedings. But the profession has not systematically ensured that clinicians understand how those standards have evolved, and what the evolution means for their daily practice. The result is that the very clinicians who supervise, appraise, and lead services may be operating with an outdated model of what accountability means — and may not know it.

What this means for the individual psychiatrist

The practical consequence is that a doctor who believes clinical justification is the whole answer is practising in a regulatory environment that demands more. The “wrapper” described earlier in this article — diagnosis formulated and recorded, guidelines considered, capacity assessed, consent obtained on the basis of material information, monitoring arranged, records kept — is not optional. It is the framework against which the GMC will assess conduct if something goes wrong.

Psychiatry is vulnerable because it operates more like an art and not a science. A psychiatrist can have a brilliant clinical rationale for a prescribing decision. They can cite studies, invoke peer practice, and construct a Bolam-compliant defence. But if the record contains no capacity assessment, no consent dialogue, no documented consideration of NICE guidance, and no monitoring schedule, the clinical justification does not rescue them. The GMC is not asking whether the decision was clinically defensible. It is asking whether the mandatory duties were discharged — and where the evidence is.

This is not an argument against clinical judgement. It is an argument that clinical judgement, on its own, is no longer enough.

A final word on prescribing and prescription

There is a persistent macro-cultural misunderstanding, encouraged by the design of the services in which many psychiatrists work, that prescribing and a prescription are the same thing. They are not.

The prescription is a waypoint. It is the visible authorisation of a treatment plan — a signed line, a ledger entry, a medication dispensed. Prescribing is the clinical work that precedes that waypoint and, critically, the clinical work that follows it. The prescription is not a boundary. The duties of prescribing flow through it and extend well beyond it.

This is not always how the profession has understood the matter. A culture that treats repeat prescribing as an administrative function — the re-authorisation of an existing instruction, the clearing of a backlog — has encouraged the belief that once the prescription is signed, the prescriber’s job is done. The GMC’s standards tell a different story.

GMCRx, paragraph 92 states that whether prescribing on a repeat or one-off basis, the doctor must make sure suitable arrangements are in place for monitoring, follow-up and review. That is a duty of care. It is not discharged by signing the prescription. It is discharged by ensuring that someone will check whether the medication is working, whether it is causing harm, and whether it is still needed.

GMCRx, paragraph 98 says that the doctor should agree with the patient how their condition will be managed, including a date for review, and must make clear records of these discussions. The review date is not an administrative aspiration. It is part of the prescribing process, and it must be planned at the point the prescription is written.

GMCRx, paragraph 100 says that at each review, the doctor should confirm that the patient is taking the medicines as directed and check that they are still needed, effective and tolerated. This is an ongoing duty. A prescription signed six months ago, with no review in between, is a prescription that has drifted out of the prescriber’s sight but not out of their responsibility.

DMC, paragraphs 56 to 58 embed the same principle within the consent framework. Decisions should be reviewed. If the patient’s condition has changed, if new information has become available, if significant time has passed, the decision to continue treatment must be revisited. Consent is not a one-time event. It is a continuing process, and the prescription is merely the first formal record of it.

GMP, paragraph 65 requires the doctor to contribute to continuity of care — sharing relevant information with others involved in the patient’s care, checking that a named clinician has taken over responsibility when the doctor’s own role ends. GMCRx, paragraph 52 requires the doctor to share all relevant information when handing over care, including information about the patient’s current and recent use of medicines and any previous adverse reactions. GMCRx, paragraph 55 requires the doctor to tell the patient’s GP about changes to medicines, the length of intended treatment, monitoring requirements, and any new allergies or adverse reactions identified.

Taken together, these duties describe a process that begins before the prescription is written and continues long after it. The prescription is a signpost within that process. It says, ‘Hello this patient has started treatment with medications’. That will have flowed from duties of assessment, diagnosis, capacity evaluation, consent dialogue, and clinical justification crystallise into a written instruction. But the prescriber’s responsibility does not end there. It ends when the medication is stopped, or when the patient is safely transferred to another responsible clinician. Until then, the duties persist.

A useful comparison may be drawn with surgery. The surgical treatment of a patient does not end when the incision is closed. The surgeon is expected to remain actively involved in post-operative care — reviewing the patient, managing complications, making decisions about mobilisation and discharge — until the patient is stable and properly handed over to another clinician. No surgeon would claim that their responsibility ended when they put the scalpel down. The surgical community would regard that as an abandonment of the patient.

Psychiatric prescribing should be no different. The prescription initiates or continues a course of treatment whose effects — therapeutic and adverse — unfold over time. The prescriber who signs and walks away, leaving no monitoring, no review date, and no arrangement for handover, has not completed an episode of care. They have launched a pharmacological intervention into a patient’s life and walked off while it is still in flight.

This is not an argument for defensive practice. It is an argument for understanding what prescribing actually is. It is not a clerical act. It is a clinical process that extends across time. The prescription is a waypoint within that process, not the end of it. The job of prescribing does not stop with the written prescription. It stops when the treatment is safely concluded or safely handed on. Until then, the prescriber remains responsible — not because the GMC says so, but because that is what it means to treat.

Conclusion

The psychiatrist – any doctor – who authorises a prescription stands alone, even if they have consulted with an MDT, with pharmacists or colleagues. Why? Every prescription or repeat prescription carries individual responsibility and is accountable to the Regulator.

That is the central truth this article has sought to illuminate. When a prescription causes harm, the system that made safe prescribing difficult — the clinic that crammed numerous scripts between appointments, the service that did not provide pharmaceutical support or adequate records — is nowhere in the dock. Only the doctor is there. And the questions the doctor will be asked are not about the system. They are about what the doctor did, what the psychiatrist checked, and what the psychiatrist documented.

This is not an accident. It is an arrangement. The liability is individual. The conditions that make compliance with mandatory duties difficult are systemic. And the gap between the two is where conscientious doctors will be lost.

The prescription is not the act

Much of the difficulty arises from a misunderstanding that the system encourages and that the profession has not sufficiently resisted. It is the belief that prescribing and a prescription are the same thing. They are not.

Prescribing is the sequence of clinical and procedural acts that must be performed — and evidenced — before a prescription can properly be signed: diagnostic formulation against recognised criteria [GMP, para. 7(a)], consideration of authoritative clinical guidelines [GMCRx, para. 14], assessment of capacity [DMC, para. 81–82], a consent dialogue that meets the Montgomery standard of material disclosure [DMC, para. 10], justification for any departure from licensed indications [GMCRx, para. 105], arrangements for monitoring and review [GMCRx, para. 92], and a contemporaneous record of the whole [GMP, para. 69–70]. The prescription ought to be the visible endpoint of work that must be evidenced, not assumed. Therefore it has to be ‘seen to be done’.

When the system treats the prescription as the act — allocating a signature slot crammed between busy clinics, measuring productivity by scripts processed, conflating administrative re-authorisation with clinical approval — it is demanding an endpoint without the process that makes the endpoint defensible. And when a doctor signs under those conditions, the signature represents an assurance the doctor may not have been equipped to give.

The mental model is outdated

The profession has not universally grasped the transformation that has occurred in the regulatory framework. A generation of psychiatrists trained in the era of Bolam internalised the belief that clinical justification — being able to explain why a treatment was reasonable, validated by peer consensus — was the answer. For much of their careers, under a GMC that had to meet a criminal standard of proof, that belief appeared to hold.

It no longer does. The GMC now assesses fitness to practise on the balance of probabilities, not beyond reasonable doubt. The standards impose procedural duties that exist independently of clinical defensibility — capacity must be assessed, consent must be obtained on the basis of material information, guidelines must be considered, monitoring must be arranged, and records must be clear, accurate, and contemporaneous. These are not clinical duties. They are legal and regulatory duties wrapped around the clinical act. And where the record is silent, the GMC will draw the inference — on the balance of probabilities — that the duty was not discharged.

The clinical justification does not fill the procedural void. A psychiatrist who has a brilliant clinical rationale for prescribing risperidone to a patient with emotionally unstable personality disorder, but has no record of the diagnostic criteria applied, no documented consideration of NICE CG78, no capacity assessment, no consent dialogue disclosing the unlicensed status of the medication, and no monitoring schedule, is not safe. Not because the clinical reasoning was wrong, but because the wrapper that makes a prescribing decision lawful and defensible is absent. And its absence is evidence of a departure from mandatory standards that the GMC cannot ignore.

The conscientious doctor carries the burden

What has emerged from the analysis in this article is not a comfortable picture. The escalation routes that are supposed to support doctors who try to comply. The conscientious psychiatrist who refuses to sign unsafe prescriptions, who documents the gaps in the record, who raises concerns about the system, risks being moved on, denied references, and penalised financially. The profession’s support infrastructure has not kept pace with the demands its regulatory standards impose.

And yet the standards remain. They are not going to bend. The GMC’s guidance is clear that the duties apply wherever a doctor practises, whatever the field of medicine or practice setting [GMP]. The system’s failure to provide the conditions in which compliance is possible does not extinguish the duty. It creates a second duty: to raise the concern, to protect the patient, and to refuse to be the conduit through which unsafe prescribing reaches the patient. The doctor becomes the last line of ‘defence’ for the patient in often overburdened systems.

This is a heavy burden. It is not distributed equally. It falls on the individual clinician, often the locum, often the doctor with the least institutional protection, precisely because they are the ones who see the records with fresh eyes and recognise what the system has normalised. The article does not offer false comfort. The burden is real. But it is the burden the profession carries, and the only safeguard the patient has.

A word to those who design the system

This article is addressed primarily to psychiatrists who are operating in a domain of quasi-science. But the arguments it has made apply with equal force to those who design and manage the services in which psychiatrists work. Non-medical managers who set the throughput expectations, who allocate the script slots, who measure productivity by prescriptions processed — they are not in the dock when a prescribing error occurs. But the systems they create determine whether a doctor can comply with the duties the GMC imposes.

A manager who has never read Good medical practice or the prescribing guidance does not know that cramming several repeat scripts between busy clinics is a recipe for corner-cutting and failure of duties of care. They see a backlog and a ledger to clear. They do not see the diagnostic, capacity, consent, monitoring, and record-keeping duties each script represents. The result is a service designed for efficiency that is structurally incompatible with the standards the doctors it employs are required to meet. That is not efficiency. It is a liability, and it should be understood as such by everyone who bears responsibility for service design.

The framework has changed

The legal and regulatory framework in which psychiatry is practised has changed, quietly but fundamentally, over the past two decades. Montgomery rewired the consent standard. The GMC lowered its evidential threshold and raised its procedural expectations. The duties that encase every prescribing decision are now more demanding, more specific, and more auditable than they were when many of the profession’s senior clinicians trained.

The profession’s understanding of that framework must change with it. Not through fear of the GMC, but through a clear-eyed recognition that prescribing is not an administrative act, that clinical justification is not enough, and that the wrapper of procedural duties is not optional. It is the framework that protects patients, and it is the framework against which every psychiatrist — however experienced, however well-intentioned, however pressured — will be judged.

The prescription is not the act. It is the record of the act. And if the act was never performed, or was performed but never recorded, the prescription is evidence of nothing except the signature of a doctor who did not pause. The profession must learn to pause. The system must learn to let it.