Overtime is a symptom, and the diagnosis matters
Overtime and compensatory off management in hospitals usually arrives on the agenda as a cost problem, which is the least useful way to frame it. Overtime hours are a signal about the staffing system, and the correct response depends entirely on what the signal means. Cutting overtime without diagnosing it simply moves the cost somewhere less visible — unpaid extended shifts, deferred leave, or a unit running below its own stated norm.
There are only a few underlying causes, and they need opposite treatments. Structural understaffing means the sanctioned strength cannot cover the roster, so overtime is permanent and predictable. Genuine demand spikes mean a mass casualty, an outbreak, or a seasonal surge created temporary need. Absence-driven overtime means the sanctioned strength is adequate but attendance is not. Process-driven overtime means shifts run long because handover, documentation, or discharge processes are inefficient.
The first analytical task is therefore to split the overtime pool by cause before touching policy. A hospital that applies an approval clampdown to structural overtime has capped nothing except its ability to staff the ward.
Telling structural understaffing apart from a genuine spike
The distinguishing feature is distribution over time, not volume. Structural overtime is flat and persistent: the same units, roughly the same hours, every month, largely on the same shifts. A genuine spike is exactly that — a departure from baseline with a discoverable cause and a return to baseline afterwards.
Plot overtime hours by unit by month for at least a year and the picture usually resolves itself immediately. A ward showing consistent overtime in every month of the year is not experiencing twelve consecutive emergencies. It is short-staffed, and the overtime is the informal mechanism by which the hospital has been funding the gap at a premium rate rather than sanctioning posts.
Compare the persistent figure against the cost of sanctioning the posts that would remove it. In many cases the persistent overtime bill approaches or exceeds the cost of the additional posts, without any of the benefits — no leave relief, no reduction in fatigue, no improvement in continuity.
Reading the overtime pattern
- Flat and persistent by unit: structural understaffing, sanction posts
- Sharp, dated, and self-resolving: genuine spike, no policy change needed
- Clustered on specific individuals: roster fairness or absence problem
- Concentrated at shift ends: process problem in handover or discharge
- Rising steadily month on month: attrition outpacing recruitment
Approval before the shift, not after
Most hospital overtime is approved retrospectively: the hours are worked, submitted at month end, and signed off because refusing to pay for work already done is neither fair nor defensible. This makes the approval control entirely fictional. The decision that mattered — whether the extra hours should be worked at all — was made by a nurse-in-charge under pressure with no authority and no alternative.
Moving approval before the shift changes the question from should we pay for this to should this happen. It requires a fast, low-friction request path available at 6 a.m. from a ward, and an approver reachable on every shift including nights. If the approval route is only available during office hours, retrospective approval will return immediately regardless of policy.
Retrospective approval should still exist for genuine emergencies, but as a flagged exception with a written reason that is reviewed. When the exception path carries the same friction as the planned path, it stops being the default. A platform such as HealUDoc can carry the request from the ward to an on-duty approver and record which route was used, so the proportion of retrospective approvals becomes a measurable control rather than an assumption.
Designing a compensatory off policy that people can actually use
Compensatory off is offered as the humane alternative to overtime payment, and it becomes the opposite when the roster cannot release the person to take it. Accrued comp-off that expires unused is worse than no comp-off at all: the hospital has taken the extra hours and given nothing in exchange, and the staff member knows it.
A workable policy needs four decisions made explicitly. What triggers accrual — full extra shifts only, or extensions above a threshold. What the accrual rate is, and whether night or holiday work accrues at a different rate. What the expiry window is, and whether it can be extended when the roster prevented use. And what happens on expiry — lapse, automatic conversion to payment, or encashment on exit.
The most important design choice is the release obligation. If a comp-off request within the validity window cannot be refused more than a defined number of times, the unit head has an incentive to plan for it. Without that, comp-off is simply deferred and eventually written off.
Comp-off policy decisions to settle in writing
- Accrual trigger: full extra shift, or extension beyond a set threshold
- Accrual rate, including any differential for night or holiday duty
- Validity window before expiry, and who may extend it
- Treatment on expiry: lapse, auto-conversion to payment, or encashment
- Limits on how often a unit head may decline a request within the window
- Whether comp-off and paid overtime can both apply to the same hours — they should not
The statutory and record-keeping backdrop
Hospitals in India operate under state Shops and Establishments legislation or the applicable factory-equivalent rules depending on the state and the nature of the establishment, alongside the wage and working-hours framework. These generally set expectations on maximum working hours, weekly rest, and the rate payable for overtime work, with the specifics varying by state.
The practical obligation is record-keeping. Attendance registers, overtime registers, and wage records need to be accurate and consistent with each other, and this is exactly where informally managed overtime creates exposure. Extra hours worked but recorded as regular attendance to keep an overtime figure low is a records problem, not a cost saving.
Because entitlements vary by state and by establishment category, the policy should be drafted with reference to the specific rules applicable to each location, particularly for hospital groups operating across states. A single group-wide policy that ignores state variation will be wrong somewhere.
Overtime as a leading indicator of roster failure
The most valuable use of overtime data is not cost control at all. It is early warning. Overtime rises before attrition does, because staff absorb additional load for months before they resign, and it rises before quality signals move, because tired teams compensate until they cannot.
Watch three derived measures rather than the total bill. Concentration — the share of overtime hours worked by the top few individuals in a unit, which reveals whether the load is distributed or falling on a handful of people who will eventually leave. Recurrence — how many consecutive months a unit has exceeded its own threshold. And the ratio of overtime hours to sanctioned hours per unit, which normalises for unit size and makes comparison meaningful.
HealUDoc dashboards can surface these alongside roster and attendance data so the conversation at the monthly review starts from a distribution rather than a total. A single hospital-wide overtime figure conceals precisely the pattern that would tell you which ward is about to lose three nurses.
“Our overtime bill was flat for a year, so nobody looked at it. When we broke it down, four nurses were carrying most of it in one ward. Three of them resigned within six months.”
A sequence for bringing overtime under control
Start with measurement, not policy. Twelve months of overtime by unit, by shift, and by individual will tell you which of the four causes you actually have, and the answer usually differs by department within the same hospital.
Then treat each cause with its own instrument. Structural gaps get post sanction, and the business case writes itself from the overtime data. Absence-driven overtime gets an attendance and leave-planning fix. Process-driven overtime gets a look at handover and discharge timing. Only genuine spike overtime needs no intervention beyond ensuring it is approved and paid properly.
Finally, set a threshold per unit above which the unit head is required to explain, and make the explanation part of the routine monthly review rather than a disciplinary event. The objective is a hospital where extra hours are a deliberate, approved, compensated response to something specific — not the invisible mechanism that keeps an under-sanctioned roster standing up.
