The Burnout Number Indian Residency Does Not Talk About

The conversation about Indian residency burnout usually starts with hours. 80–100 hours a week. 36-hour shifts. Six days a week with one day off, sometimes none. These numbers are real, and they are the visible part of the iceberg.

The published burnout rate among Indian medical residents, across cohorts and specialties, is 40–60% depending on the institution and the measurement instrument. Studies from AIIMS Delhi (Singh et al., Indian J Med Res, 2018), JIPMER (Saini et al., Industrial Psychiatry Journal, 2017), and PGIMER (Grover et al., Indian J Psychol Med, 2018) converge on roughly the same range.

What the instrument is actually measuring

It is worth knowing what "burnout" means in those papers, because the word has been worn smooth by overuse. Most of these studies use the Maslach Burnout Inventory, which measures three separate things: emotional exhaustion, depersonalisation, and a reduced sense of personal accomplishment.

They are not interchangeable, and the second one is the one that should worry a hospital. Emotional exhaustion is what the resident feels. Depersonalisation — treating patients as objects, as tasks, as the appendix in Bed 12 — is what the patient feels. A ward can run for years on exhausted residents. It cannot run safely on depersonalised ones, and depersonalisation is the dimension most strongly associated with the kind of shortcut that ends up in an incident report.

So when a study reports "52% burnout," it is not reporting that half the residents are tired. It is reporting that half the residents are somewhere on a trajectory that ends in worse care.

The variable nobody counts

What the hours-focused conversation misses is the interruption-density contribution. A resident on a 36-hour shift in a low-load department burns out at a different rate than a resident on a 36-hour shift in a high-load department, even with identical hours. The mediating variable is the number of decisions made per hour and the cognitive cost of each.

The published work on this in non-medical contexts (Mark et al. on knowledge workers, various papers) suggests that the cognitive recovery cost of frequent task-switching may be as significant as the absolute hours. Applied to residency, this is the explanation for why two residents in the same programme, on the same rotation, on the same hours, have different burnout trajectories.

Anyone who has worked a ward knows the mechanism without needing the citation. You are three lines into a discharge summary when the nurse asks about Bed 9's potassium. You answer, return to the summary, and re-read the last paragraph to find out what you were about to say. The interruption cost thirty seconds. The resumption cost another ninety. Repeat that forty times in a shift and you have lost an hour to nothing, and — more importantly — you have spent the day never once holding a complete thought to its end.

That is the texture of a high-load shift. Not one big hard thing. Two hundred small ones, none of which is difficult, arriving in an order nobody chose.

Not all decisions cost the same

The useful move is to stop counting decisions and start sorting them.

Trivial and recurring: the paracetamol dose for a 14 kg child, the standard repeat interval for a post-insulin ABG, whether this creatinine needs the metformin held, what the handover line for a stable post-op day-2 looks like. These have a right answer that does not depend on judgement. Each costs maybe twenty seconds of working memory. There are dozens per shift.

Genuinely clinical: is this patient's breathlessness cardiac or septic, is this family ready for the conversation about ceilings of care, does this abdomen need a surgeon tonight. These are what a doctor is for. They are expensive and they should be.

The two categories compete for the same limited resource. Every trivial decision is drawn from the same pool as the clinical ones, and the pool empties. By hour twenty-six, the resident is not making bad clinical decisions because they have forgotten the medicine. They are making them because there is nothing left in the pool by the time the hard question arrives.

Why this is the tractable variable

The implication is uncomfortable. Reducing resident hours alone may not reduce burnout to the degree the public conversation expects. What might reduce burnout is reducing the number of decisions a resident has to make per hour, by automating the trivial ones — which dose of paracetamol, what's the standard ABG follow-up timing, what's the typical handover note for a stable post-op.

Hours require a policy change, a funding line, and more residents than currently exist. That fight is worth having and it will take a decade. Cognitive load per hour requires better tools, and the tools can ship this year. It is the rare case where the less important variable is also the only one an engineer can move.

There is a failure mode worth naming out loud, because it is the obvious way this argument gets abused: a tool that removes twenty minutes of trivial load per shift can be used as the justification for adding four more patients to the list. If that happens, the tool has not reduced burnout. It has increased throughput and left the resident exactly where they started, with a slightly better story told about them. Anyone deploying software on this argument should be measuring the resident's load, not the ward's.

The number we should be reporting

This is not a substitute for the longer conversation about hours, pay, and structural reform of residency in India. It is, however, the conversation that the existing tools can have now — by removing the trivial cognitive load from the resident's day. The hard cognitive load — the clinical reasoning, the patient conversation, the team coordination — is what the resident is there to do.

If I could add one field to the residency workload surveys, it would not be another hours question. It would be: how many times today were you interrupted mid-task? My guess is that it predicts the depersonalisation score better than the hours do, and it is the number a hospital could actually act on next month.

The burnout number Indian residency doesn't talk about is the cognitive load per hour. We should.