Mental Health Parity in Group Mediclaim: Where Indian Wordings Still Fail the Test

Section 21(4) of the Mental Healthcare Act, 2017 is one sentence long:

Every insurer shall make provision for medical insurance for treatment of mental illness on the same basis as is available for treatment of physical illness.

IRDAI directed insurers to comply through a circular of 16 August 2018, and the consolidated health insurance master circular of 2024 carried the position forward. As a result, a blanket mental illness exclusion in a group mediclaim wording is now rare. If a broker's parity review stops at confirming the cover clause exists, it will pass almost every wording on the market.

That is the wrong test in 2026. An analysis published by Express Healthcare on 16 August 2026, titled "Mental health insurance in India: From policy wording to patient access", argues that the compliance gap has shifted from headline inclusion into the mechanics: what the wording defines as mental illness, which treatment settings trigger a payable claim, what internal caps apply, and whether an employee can actually find an empanelled facility. Each of those is a clause, and each clause can quietly take back what the cover clause grants.

This post is a clause-level review for a group mediclaim wording before renewal: where parity fails in practice, what to read line by line, and what to negotiate rather than accept.

A standard GMC is an indemnity product built around a 24-hour inpatient admission, plus a list of day-care procedures and, where the employer has bought it, an OPD benefit. Most psychiatric care fits none of these. Depression, anxiety disorders and most mood disorders are treated through outpatient psychiatrist consultations, therapy sessions and medication, sometimes for years, with no admission ever occurring. A wording can include mental illness in full and still pay close to nothing, because the events it responds to rarely occur in psychiatric treatment.

So the first question is structural: does the programme have an OPD benefit at all, and if it does, does the OPD wording cover psychiatric consultations and psychotherapy, or only physician visits, diagnostics and pharmacy? Plenty of OPD riders were drafted before mental health was on anyone's renewal agenda and list eligible expenses in a way that leaves a clinical psychologist's session outside the definition. That is a parity failure that never shows up in the exclusions list.

Day-care definitions compound the problem. Day-care lists were built for procedures that anaesthesia and technology shortened below 24 hours: cataract surgery, chemotherapy, dialysis. Structured psychiatric day programmes, where a patient attends a facility through the day and returns home at night, do not fit the standard day-care definition even though they are a recognised and cheaper alternative to full admission. Unless the wording extends day care to psychiatric treatment explicitly, the cheaper setting is the uninsured one, which pushes patients toward full admission or out of treatment.

The parity language sits in the cover clause. The caps sit in the annexure. Wordings still turn up at review with some or all of the following, and each one treats a psychiatric admission worse than a physical one:

• A psychiatric sub-limit per member or per family per policy year, while physical illness pays up to the full sum insured. This is the most common and the most direct parity failure.
• Room-rent treatment that maps psychiatric wards to a lower category than the general entitlement, dragging down every linked expense through proportionate deduction.
• Restoration or refill benefits worded to reinstate the sum insured for "illness" or "accident" in a way the insurer reads as excluding psychiatric admissions.
• Co-payments applied to mental illness claims specifically, where the base programme carries none.
• Waiting periods applied to mental illness in a group programme that has waived waiting periods for everything else.

None of these caps announces itself. A sub-limit is a row in a benefits table, and a room-rent category is a definition. This is why the review has to be run on the specimen policy wording and the benefits annexure together, before renewal terms are confirmed, and never on the sales presentation.

The Act defines mental illness as a substantial disorder of thinking, mood, perception, orientation or memory that grossly impairs judgment, behaviour, capacity to recognise reality or ability to meet the ordinary demands of life, and the definition expressly includes mental conditions associated with the abuse of alcohol and drugs. Many group wordings simultaneously carry a standard exclusion for treatment of alcoholism, drug or substance abuse. Read together, the wording promises cover on the Act's terms in one clause and withdraws part of the Act's own definition in another. At minimum, the broker should know which reading the insurer will apply at claim time and get it in writing.

Self-harm exclusions deserve the same scrutiny. Wordings routinely exclude "intentional self-injury" or "attempted suicide". Section 115 of the same Act presumes that a person who attempts suicide is under severe stress, and the hospitalisation that follows such an event is precisely the psychiatric emergency the parity provision exists for. An insurer that repudiates that admission on the self-injury exclusion is applying pre-2017 drafting to a post-2017 obligation, and the clause should be challenged at placement rather than argued after a repudiation.

Two more definitional checks. First, the Act's definition of mental illness does not include intellectual disability, so an exclusion there is legitimate; but some wordings stretch it into a broad exclusion for "behavioural, developmental or neurodevelopmental disorders" that can sweep in conditions the Act does cover. Second, check whether the wording anchors mental illness to a recognised classification such as the ICD or leaves the term undefined. An undefined term is decided by the claims desk, and repudiation patterns in group health show how much discretion an undefined term hands the insurer.

The Express Healthcare analysis pairs wording with access for a reason: a payable claim still needs a facility that can admit the patient and a cashless pipe that works. Mental health establishments in India must register with state mental health authorities under the Act, and the pool of standalone psychiatric hospitals is small relative to general hospitals. Insurer and TPA cashless networks are thinner still in this segment. The practical consequence is that a psychiatric admission is more likely than a physical one to end up in the reimbursement queue, where documentation demands are heavier and settlement is slower.

Supply is improving. Medical Buyer reported on 20 August 2026 that mental healthcare is emerging as a sustainable sector in India, with organised providers expanding. Networks lag provider growth, though, and a network list that was thin last year does not update itself because the sector grew.

Two checks before renewal:

1. Pull the current network list and count empanelled psychiatric facilities in the cities where the client's employees actually live, not nationally. A national count hides the fact that a two-city workforce may have no usable facility in either city.
2. Ask the TPA how many psychiatric cashless authorisations it processed on this insurer's book in the last twelve months. A near-zero number against a large insured base means the cashless route is untested, whatever the network list claims.

Utilisation data on mental health claims understates need, because disclosure is still suppressed. A report covered by People Matters on 21 August 2026 found that only 41 per cent of Indian employees feel comfortable discussing mental health at work. Read that number the way an underwriter should: the majority of the insured population is not yet presenting, and every point of improvement in workplace comfort converts hidden need into claims.

The cohort driving that conversion is already inside the workforce. Business Today reported on 19 August 2026 on how Gen Z employees are dropping the "personal reasons" cover story and naming mental health directly at work. For this cohort, the mental health benefit is a visible, tested part of the employment offer. A wording that pays nothing for therapy and caps psychiatric admissions will be discovered, and it will be discovered by the employees most willing to talk about it.

The cost context sharpens the point. Asia Insurance Review reported on 17 August 2026 that rising healthcare costs are the top worry for Indian corporate employees. Employers are answering that worry with the GMC, and the programme is absorbing more of employees' financial anxiety at exactly the time group health premiums are hardening. Fixing the mental health wording now is cheap. Retro-fitting it in two renewals' time, after utilisation has risen and the insurer is pricing the book harder, will not be.

Run this against the specimen wording and benefits annexure, in order. Every item is a specific clause to locate and read, and the answer to each is yes, no, or a follow-up question to the insurer.

1. Cover clause. Does the wording state that mental illness is covered on the same basis as physical illness, referencing the Mental Healthcare Act, 2017?
2. Definition. Is mental illness defined at least as widely as the Act, ideally anchored to the ICD? An undefined or narrower term fails.
3. Sub-limits. Does the benefits annexure carry any psychiatric cap below the full sum insured? If yes, that is the primary negotiation item.
4. Room rent. Do psychiatric wards map to the same room category and entitlement as general admissions?
5. Waiting periods. Are any waiting periods applied to mental illness that the group programme waives elsewhere?
6. Day care. Does the day-care definition or a specific extension cover structured psychiatric day treatment?
7. OPD. If the programme has an OPD benefit, do psychiatric consultations and psychotherapy sessions qualify, and what are the per-session and annual caps?
8. Exclusions. How do the substance-abuse, self-injury and developmental-disorder exclusions interact with the Act's definition and Section 115?
9. Restoration and refill. Do reinstatement benefits apply to psychiatric admissions on the same terms as physical ones?
10. Network. How many registered mental health establishments are empanelled in the client's actual locations, and what is the TPA's psychiatric cashless track record?

A broker who can hand the client this list, filled in with clause references from their own wording, has done more for parity than any wellness-week email the HR team will send this year.

The review produces a defect list. The negotiation converts it. Four positions worth taking into the renewal discussion:

First, ask for psychiatric sub-limits to be deleted outright rather than raised. The legal basis is Section 21(4), the commercial basis is that current psychiatric utilisation on most group books is still low, and an insurer holding the rest of the placement has room to concede. A raised sub-limit concedes the principle and leaves the clause in place for the claims desk to apply.

Second, buy OPD mental health cover with defined numbers: sessions per member per year, a per-session limit, and named eligible providers including clinical psychologists. A defined benefit gets used and gets valued. An undefined "wellness benefit" gets neither. On the same point, an employee assistance programme is a counselling service contract, and it does not discharge the insurer's parity obligation or substitute for indemnity cover. Keep the two separate in the benefits communication so employees know which one pays a hospital bill.

Third, get the day-care and reimbursement positions as endorsements, in writing: psychiatric day treatment payable under the day-care benefit, and admissions at any establishment registered under the Act payable on reimbursement at network-equivalent terms. Verbal confirmations from the placement discussion do not survive to the claims desk.

Fourth, demand claims data at every renewal: psychiatric claims count, cashless share, and repudiation reasons. The 41 per cent comfort figure will move, utilisation will follow it, and the employer should see that trend in its own data before the insurer prices it in.

Every position above depends on knowing what the client's exact wording says and how other insurers draft the same clause. Sarvada makes insurer group health wordings searchable at clause level, so a broker can compare psychiatric sub-limits, OPD definitions, day-care extensions and exclusion language across insurers before the renewal meeting, and hold each insurer to its own language. If you place or service group mediclaim programmes and want that comparison at your desk, request access to Sarvada.