Restriction on Use of Electroconvulsive Therapy (Section 94)

Few treatments in psychiatry carry the symbolic weight of electro-convulsive therapy (ECT). For decades, in films and in the popular imagination, the image of a patient convulsing under electric current without anaesthesia became shorthand for institutional cruelty. The Mental Healthcare Act, 2017 confronted that image head-on. It did not abolish ECT, because the evidence base for modified ECT in severe depression, catatonia and suicidality remains strong. Instead, it drew three sharp legal lines: ECT may never be given as emergency treatment, it may never be given without muscle relaxants and anaesthesia (the so-called "unmodified" or "direct" ECT), and it may never be given to a minor except through a tightly guarded gateway of guardian consent and Board permission. This chapter unpacks the statutory architecture — primarily Section 94 (emergency treatment) and Section 95 (prohibited procedures) — and situates it within the constitutional jurisprudence of bodily autonomy, dignity and informed consent that gives these restrictions their deeper meaning.

Where the ECT restriction actually lives in the Act

A point of caution at the outset, because aspirants frequently misremember the section number. The Mental Healthcare Act, 2017 deals with electro-convulsive therapy across two adjacent provisions, and it is worth being precise about which clause does what. Section 94 is headed "Emergency treatment"; its sub-section (3) carries the categorical bar that ECT may not be used as a form of emergency treatment. The substantive restriction on how, and on whom, ECT may be administered in ordinary practice lives in Section 95, headed "Prohibited procedures". Section 95(1)(a) forbids ECT without muscle relaxants and anaesthesia; Section 95(1)(b) forbids ECT for minors; and Section 95(2) creates the narrow exception for minors. Section 96, the immediately following provision, governs psychosurgery. So when an exam paper refers loosely to "the restriction on use of electro-convulsive therapy", it is referring to this Section 94-95 cluster read together. Treat the two as a single scheme: 94(3) closes the emergency door, and 95 polices the front door.

This chapter therefore reads the provisions as a unit. For the broader rights framework into which they fit, see our introduction to the Mental Healthcare Act and the subject hub.

From custody to capacity: the paradigm the restriction serves

The 2017 Act repealed the Mental Health Act, 1987 and came into force on 29 May 2018. Its preamble announces a deliberate break from the custodial logic of earlier law: it is "An Act to provide for mental healthcare and services for persons with mental illness and to protect, promote and fulfil the rights of such persons during delivery of mental healthcare". The statute was drafted to align India's domestic law with the United Nations Convention on the Rights of Persons with Disabilities (CRPD), which India ratified in 2007 and which insists that disability — including psychosocial disability — does not extinguish legal agency.

This matters enormously for the ECT provisions. Under the old regime, treatment decisions flowed downward from the institution to the patient. Under the 2017 Act, the default is that a person retains capacity to make their own treatment decisions unless and until that capacity is shown to be absent, and even then the decision is mediated through their advance directive or nominated representative. The ECT restrictions are not free-floating prohibitions; they are concrete expressions of this rights-based paradigm. A treatment as invasive as ECT is precisely the kind of intervention the Act wants surrounded by procedural protection rather than clinical discretion alone.

Section 94(3): no ECT in an emergency

Section 94 permits a registered medical practitioner to give any medical treatment, including treatment for mental illness, to a person with mental illness — at a health establishment or in the community — subject to the informed consent of the nominated representative where one is available, but only where it is immediately necessary to prevent death or irreversible harm to the person, the person inflicting serious harm on himself or others, or serious damage to property flowing directly from the mental illness. The Explanation clarifies that emergency treatment includes transporting the person to the nearest mental health establishment for assessment.

Two limits hem this power in. Section 94(2) says emergency treatment cannot extend to medical treatment unrelated to the emergency. And Section 94(3) states flatly: "Nothing in this section shall allow any medical officer or psychiatrist to use electro-convulsive therapy as a form of treatment." The legislative logic is sound. Emergency treatment under Section 94 is, by definition, treatment given without the full consent architecture of the Act — it is a stop-gap to avert imminent catastrophe, authorised on the informed consent of the nominated representative where one is available, and otherwise on the practitioner's own assessment of immediate necessity. ECT is a planned procedure requiring anaesthetic work-up, fasting, cardiac assessment and a controlled recovery; it is never a true emergency intervention in the way a sedating injection might be. By carving ECT out of the emergency power, the Act ensures that this particular treatment can only ever be reached through the deliberative, consent-based route, never through the shortcut. Section 94(4) caps emergency treatment at seventy-two hours, or until assessment at an establishment, whichever is earlier, extendable to seven days only during a declared disaster or emergency.

Section 95(1)(a): the end of unmodified ECT

The single most consequential clause in the scheme is Section 95(1)(a), which prohibits "electro-convulsive therapy without the use of muscle relaxants and anaesthesia". This is the statutory abolition of what clinicians call unmodified or "direct" ECT. In unmodified ECT, the electrical stimulus induces a full motor seizure in a conscious, un-relaxed patient — historically associated with fractures, dislocations, dental injury, terror and the wider cultural stigma that attached to the procedure. Modified ECT administers the stimulus under general anaesthesia with a muscle relaxant, so that the therapeutic cerebral seizure occurs without the violent peripheral convulsion.

The significance of the ban can only be understood against India's clinical reality before 2018. Unmodified ECT remained in use in a number of public psychiatric institutions, often justified on grounds of cost, the shortage of anaesthetists, and the volume of patients. The 2017 Act made no allowance for these constraints. The prohibition is unconditional and admits of no exception: there is no proviso permitting unmodified ECT where an anaesthetist is unavailable. The legislative choice was to treat dignity and safety as non-negotiable rather than as resource-dependent. The practical consequence has been that establishments unable to provide anaesthesia and muscle relaxation must simply cease offering ECT, not offer it in its unmodified form — a hard but coherent line that places the burden of compliance squarely on the system, not the patient.

It is worth noticing what the clause does not say. It does not require the patient's consent to be in any particular form, nor does it set out the frequency or maximum number of ECT sessions; those matters are left to clinical regulation and to the general consent provisions of the Act. What 95(1)(a) does is fix an irreducible minimum standard for the manner of administration. The result is that consent and modification operate on different planes: a capacitated adult may consent to a course of ECT, but no quantum of consent can convert unmodified ECT into a lawful procedure. The modification requirement is a mandatory feature of the treatment itself, not a waivable safeguard — a structure that mirrors the way the law treats certain irreducible safety standards in other regulated activities, where consent to the activity never extends to consent to its prohibited form.

Section 95(1)(b) and 95(2): the near-ban on ECT for children

Section 95(1)(b) prohibits "electro-convulsive therapy for minors". A minor, under the general law, is a person below eighteen years of age. The default position of the Act is therefore that children may not be given ECT at all. But the Act stops short of an absolute ban. Section 95(2) opens a guarded gateway: "Notwithstanding anything contained in sub-section (1), if, in the opinion of psychiatrist in charge of a minor's treatment, electro-convulsive therapy is required, then, such treatment shall be done with the informed consent of the guardian and prior permission of the concerned Board."

Three cumulative conditions must therefore be satisfied before a minor receives ECT. First, the treating psychiatrist must form a clinical opinion that ECT is required — a clinical gate. Second, the guardian must give informed consent — a familial gate. Third, and crucially, the concerned Mental Health Review Board must grant prior permission — an independent, external gate. The third condition is what distinguishes ECT for minors from almost every other treatment decision under the Act; the Board's prior permission inserts a quasi-judicial check between the doctor and the child's brain. Note also that the relaxation in 95(2) applies only to the minors bar in 95(1)(b); it does not dilute 95(1)(a). A minor lawfully given ECT under the Board-approved route must still receive it modified, with muscle relaxants and anaesthesia, because the unmodified-ECT prohibition has no exception at all.

The role of the Mental Health Review Board

The "concerned Board" referenced in Section 95(2) is the Mental Health Review Board constituted under Chapter XI of the Act, a quasi-judicial body whose functions include registering and reviewing advance directives, adjudicating applications relating to supported admission, and protecting the rights of persons with mental illness. By routing minor-ECT through prior Board permission, the legislature transformed what would otherwise be a private clinical decision into a reviewable administrative act, susceptible to record-keeping, reasoning and challenge.

This design echoes a recurring constitutional instinct: that the most invasive interferences with the person should not be left to unilateral professional judgment but should attract an independent layer of scrutiny. The same instinct animates Section 96, which permits psychosurgery only with both the informed consent of the person and approval of the Board. ECT for minors and psychosurgery are thus the two procedures the Act singles out for prior Board clearance — a signal of how gravely the legislature viewed irreversible or deeply invasive interventions on those least able to protect themselves.

The requirement of prior permission is deliberate and exacting. The Board's clearance must precede the treatment; it is not a post-facto ratification or an after-the-event review. This sequencing is what gives the safeguard its bite, because once ECT has been administered its effects cannot be undone. The structure also implies a duty to document — the psychiatrist's opinion that ECT is required, the guardian's informed consent and the Board's reasoned permission must all exist on the record before the first session. In practical terms, the gateway converts a clinical recommendation into a small adjudicatory file, and that file is precisely what makes the decision auditable and, where necessary, challengeable. For a child, whose voice in the process is mediated through a guardian, that independent record is the principal protection against an over-ready resort to a powerful and frightening intervention.

The constitutional anchor: bodily autonomy and informed consent

The ECT restrictions do not float in a statutory vacuum; they sit atop a now-settled constitutional foundation that a competent adult's body is not the state's to treat without consent. In Common Cause (A Regd. Society) v. Union of India, (2018) 5 SCC 1, a five-judge Constitution Bench held that the right to refuse medical treatment and to die with dignity is an inextricable facet of the right to life under Article 21, and grounded that holding expressly in bodily autonomy and informed consent. If a competent person may lawfully refuse life-saving treatment, then a fortiori the state cannot impose an invasive, stigma-laden treatment such as ECT without satisfying the consent and procedural conditions the statute lays down.

That principle had earlier been articulated in the criminal-procedure context in Selvi v. State of Karnataka, (2010) 7 SCC 263, where the Supreme Court held that the involuntary administration of narco-analysis, polygraph and brain-mapping techniques violates mental privacy and bodily integrity under Article 21, and may amount to cruel, inhuman or degrading treatment. The reasoning translates directly: a forcibly administered medical-scientific procedure on an unconsenting person is a constitutional wrong. The unmodified-ECT ban and the consent requirements around minor-ECT are the statutory operationalisation of exactly this Selvi logic within the psychiatric setting.

Privacy, dignity and the decisional self

The autonomy thread was constitutionalised at the broadest level in Justice K.S. Puttaswamy v. Union of India, (2017) 10 SCC 1, where a nine-judge Bench recognised the right to privacy as a fundamental right flowing from Articles 14, 19 and 21, and identified bodily autonomy and decisional autonomy — the freedom to make intimate decisions about one's own body — as core components of privacy, with human dignity as the foundational value. Any state interference must satisfy the threefold test of legality, legitimate aim and proportionality.

The ECT restrictions read as a model of Puttaswamy-compliant regulation. They pursue a legitimate aim (protecting persons with mental illness from harmful and degrading treatment); they are sanctioned by clear law (Sections 94-95); and they are proportionate — rather than banning ECT outright, which would deny patients an effective therapy, the Act bans only its dangerous form (unmodified) and conditions its use on the most vulnerable (minors) on layered consent. Dignity is honoured at both ends: the dignity of being protected from a brutal procedure, and the dignity of not being denied a beneficial one. The same dignity-and-autonomy reasoning underlies National Legal Services Authority v. Union of India, (2014) 5 SCC 438, where the Court located self-determination over one's own identity and body within Article 21.

Applying the proportionality lens concretely sharpens the analysis. The legitimate aim is twofold and, importantly, points in both directions — protecting the vulnerable from a degrading intervention, while preserving access to an evidence-based therapy for severe, sometimes life-threatening conditions such as catatonia and treatment-resistant depression with active suicidality. A blanket ban on all ECT would fail the proportionality test from the opposite side, by denying patients a beneficial treatment and thereby itself imperilling life and dignity. The Act's chosen instrument — prohibit the dangerous form, condition use on the most vulnerable, and otherwise route the decision through ordinary consent — is the least restrictive means that still achieves the protective aim. This is why the ECT scheme is so frequently cited in academic writing as a comparatively well-drafted illustration of proportionate health regulation, even where the same commentators criticise other parts of the Act as over- or under-inclusive.

The companion prohibitions: sterilisation and chaining

Section 95(1) does not stop at ECT. The same sub-section also prohibits "sterilisation of men or women, when such sterilisation is intended as a treatment for mental illness" (clause (c)) and being "chained in any manner or form whatsoever" (clause (d)). These companion prohibitions illuminate the legislative purpose behind the whole provision: it is a charter against a catalogue of historical abuses inflicted on the mentally ill — coerced sterilisation, physical chaining, and brutal unmodified shock.

The sterilisation prohibition resonates with Suchita Srivastava v. Chandigarh Administration, (2009) 9 SCC 1, in which the Supreme Court held that reproductive autonomy is a dimension of personal liberty under Article 21, and that the personal autonomy of a woman with intellectual disability over reproductive decisions must be respected; consent could not be diluted merely because of her condition. Section 95(1)(c) codifies that principle by forbidding sterilisation as a treatment for mental illness. The chaining prohibition, for its part, is absolute and unqualified — there is no therapeutic justification clause, in contrast to the carefully bounded permission for physical restraint in Section 97. Reading clauses (a) to (d) together, the unifying idea is that certain interventions are so degrading, irreversible or unrelated to genuine therapy that the law removes them from the menu of permissible options altogether.

The placement of these four prohibitions in a single sub-section is itself instructive. The legislature could have scattered them through the Act, but it chose to assemble them as a coherent declaration of what may never be done to a person with mental illness. That assembly creates an interpretive presumption against reading any of them down: where three of the four bans are plainly absolute, a court will be slow to read the fourth — the ECT-for-minors bar — as anything other than a strong default that yields only to the precise conditions spelled out in Section 95(2). The drafting, in other words, supports a protective rather than a permissive reading of the whole provision.

ECT and psychosurgery: a deliberate distinction

It repays attention to compare Section 95 with Section 96. Psychosurgery is not listed among the "prohibited procedures" in Section 95; it has its own provision in Section 96, which permits it as a treatment for mental illness only where both the informed consent of the person undergoing it and the approval of the concerned Board have been obtained. Two structural differences stand out. First, psychosurgery requires the consent of the person themselves, not a guardian or nominated representative, reflecting its irreversible character — only the patient's own informed consent will do. Second, while minor-ECT requires guardian consent plus Board permission, psychosurgery requires the person's own consent plus Board approval, and the Act notably does not create a guardian-substitution route for it.

The drafting choice tells a story about gradations of invasiveness as the legislature perceived them. Unmodified ECT and chaining are banned outright. Modified ECT is permitted with ordinary treatment-consent for adults, and through the special minors gateway for children. Psychosurgery — irreversible by its nature — is permitted only on the person's own informed consent plus Board approval. The Act thus calibrates the strength of the procedural safeguard to the gravity and reversibility of the intervention, an approach squarely consistent with the proportionality demanded by Puttaswamy.

How consent flows for adult ECT under the Act

For an adult, ECT is not listed in Section 95(1)'s outright bans (except in its unmodified form). It is therefore a permissible treatment, but it must be reached through the Act's general consent machinery rather than imposed at will. Where the person has capacity to make treatment decisions, their informed consent governs; ECT cannot be administered over the refusal of a capacitated adult. Where capacity is impaired, the Act looks first to a valid advance directive in which the person may, in advance, have refused or consented to specific treatments, and then to the nominated representative, who exercises a supported-decision-making function rather than pure substitution.

This is why the ECT restrictions cannot be studied in isolation. They are the sharp end of a consent architecture that runs through the whole statute. The bar in Section 94(3) and the modifications mandated in Section 95(1)(a) operate as floors that even the consent machinery cannot lower: a person cannot validly consent to unmodified ECT, and no nominated representative or emergency power can authorise ECT as emergency treatment. Consent regulates the permissible; the prohibitions define the impermissible regardless of consent.

The interaction with supported admission is also worth flagging. A person admitted as a supported patient under the Act does not, by virtue of admission, lose the protection of Sections 94-95. Admission status governs where and under whose oversight a person is treated; it does not dissolve the substantive bars on emergency ECT, unmodified ECT or unsanctioned minor-ECT. This is a frequent point of confusion: candidates sometimes assume that an involuntary or supported admission carries with it a general authority to treat. It does not. The treatment-decision and prohibited-procedure provisions continue to apply with full force, so that even a lawfully admitted patient retains the floor of protection that Sections 94 and 95 guarantee. The admission machinery and the treatment-restriction machinery are parallel tracks, and the more protective rule on each track prevails.

Enforcement, ground reality and the limits of legislation

A statute is only as strong as its implementation. Since the Act's commencement in 2018, clinical and human-rights commentary has documented continuing pockets of unmodified ECT in under-resourced institutions, and persistent concern about whether the prohibition in Section 95(1)(a) is uniformly observed where anaesthetic services are thin. The National Human Rights Commission and professional bodies have at various points flagged the gap between the statutory standard and the ground reality, particularly in high-volume public psychiatric establishments.

This gap does not weaken the legal proposition; it sharpens it. The Act's text is unambiguous — unmodified ECT is prohibited, full stop — so any continuing practice of it is unlawful, not merely sub-optimal. The enforcement challenge is administrative and financial: ensuring every establishment that offers ECT has reliable access to anaesthesia and muscle relaxation, failing which it must stop offering ECT rather than revert to its unmodified form. For the exam, the safe statement is that the law admits no exception for resource constraints, and that compliance is a duty of the establishment and the State, monitored through the State Mental Health Authority and the Mental Health Review Boards.

There is also a deeper jurisprudential point in the enforcement gap. A right that exists on paper but is routinely breached in practice tends, over time, to generate litigation that crystallises the standard further. The constitutional anchors already in place — Common Cause, Selvi and Puttaswamy — mean that a person subjected to unmodified ECT in breach of Section 95(1)(a) would have not only a statutory complaint but a constitutional grievance under Article 21, framed as an unconsented and degrading interference with bodily integrity. The statutory prohibition and the constitutional guarantee reinforce one another: the Act supplies the bright-line rule, and the Constitution supplies the remedy and the interpretive direction that the rule must be read in favour of the patient's dignity. That mutual reinforcement is the answer to the cynic who says the restrictions are merely aspirational.

Exam focus: how this topic is tested

Judiciary and CLAT-PG papers test this topic in predictable ways. Expect direct objective questions on which section bans unmodified ECT (Section 95(1)(a)), whether ECT can be given in an emergency (no — Section 94(3)), and the conditions for ECT on a minor (psychiatrist's opinion + guardian's informed consent + prior Board permission, Section 95(2)). A common trap is to ask whether unmodified ECT can ever be given with consent or in an emergency — the answer is no on both counts, because 95(1)(a) is unconditional and 94(3) is absolute.

Descriptive answers should weave the statutory scheme together with the constitutional anchors: Common Cause (refusal of treatment, bodily autonomy), Selvi (no forcible invasive procedures), Puttaswamy (privacy, decisional autonomy, proportionality) and Suchita Srivastava (reproductive autonomy, relevant to the sterilisation prohibition). The strongest answers note the calibration across Sections 95 and 96 — outright bans, conditional permissions for minors, and the distinct psychosurgery regime — and conclude that the restrictions are a proportionate, CRPD-aligned protection of dignity rather than a blanket prohibition of an effective therapy. Always cite the correct section number; confusing Section 94 (emergency) with Section 95 (prohibited procedures) is the single most common error.