Costing a Hospital or Diagnostic Chain: Procedure, Bed-Day and Test
Your billed revenue is not your revenue, and an average cost per case is not your cost.
Summary
- Hospital costing needs three cost objects: cost per procedure, cost per bed-day by acuity, and both set against what each payer pays. Diagnostics adds cost per test.
- Cost per procedure has four layers: consumables, clinical time, theatre and equipment by the minute, and shared overhead, which is usually split by revenue and systematically wrong.
- Apportion each shared cost on what consumes it. Weighting bed-days by acuity is the single highest-value change most hospitals can make.
- Realised revenue minus the first three layers is payer contribution; subtract apportioned overhead too and the result is specialty profit.
Two cases can earn the same package rate and consume very different amounts of a hospital. One takes forty minutes in theatre, the other four hours. One patient recovers on a general ward; another spends days in ICU, where nursing effort per bed-day runs three to six times higher.
A cost system that gives every case a flat share of theatre and nursing books the two at the same cost. It is wrong about both, and the error does not average out.
The revenue side of the same case, from tariff billed to cash kept, is walked step by step in hospital pricing and payer mix. This is the cost side: what one procedure, one bed-day and one test actually consume.
Procedure, bed-day and payer: three cost objects
Hospitals need three, and most track none of them properly.
- Cost per procedure. What one case consumes.
- Cost per bed-day. What occupancy costs, by acuity level.
- Cost per payer, per procedure. The two above, set against what that payer actually pays.
Diagnostics has its own unit, cost per test. The hospital's three objects are built from the same four layers.
Four layers in the cost of one procedure
- Direct consumables. Drugs, implants, surgical consumables, disposables, blood products. Traceable to the case and usually recorded reasonably well, because they are issued against the patient.
- Direct clinical time. Surgeon, anaesthetist, theatre nursing, technicians, at loaded cost for the time the case actually consumed. That means theatre minutes, not a flat charge per case.
- Theatre and equipment. The OT itself, imaging, the equipment used. Costed per minute of use rather than per case, because a forty-minute case and a four-hour case are not the same.
- Apportioned indirect cost. Ward nursing, housekeeping, biomedical maintenance, administration, building, utilities. This layer decides most of the answer, and it is almost always apportioned by revenue.
Apportion each shared cost on what consumes it
Apportion ward nursing, housekeeping, biomedical maintenance and administration by revenue share, and your high-billing specialties carry the most overhead. That feels fair and it is systematically wrong. Cost apportionment covers choosing a basis in general; in a hospital, the drivers look like this.
What to apportion each shared hospital cost on
| Shared cost | Usual basis | Driver that reflects consumption |
|---|---|---|
| Ward nursing | Revenue | Bed-days weighted by acuity: an ICU day is not a general ward day |
| Theatre overhead | Revenue or case count | OT minutes used |
| Housekeeping | Revenue | Floor area, or bed-days |
| Biomedical maintenance | Revenue | Equipment usage hours by department |
| Imaging and lab support | Revenue | Tests or scans requested per case |
| Billing and claims | Revenue | Claims submitted, weighted by payer complexity |
| Administration | Revenue | Headcount, or patient episodes |
The claims line deserves attention. Processing a self-pay bill and processing a government scheme claim, with its documentation, pre-authorisation and likely resubmission, are not remotely the same effort. Apportion billing cost by payer complexity and the cost of serving each payer shifts noticeably.
Cost per bed-day, weighted by acuity
Acuity-weighted bed-days is the single highest-value change most hospitals can make. That is because nursing is usually the largest apportioned cost, and its effort per bed-day varies three to six times between ICU and general ward.
A worked month shows the size of the error. Take a month's nursing cost of ₹66 lakh across 2,400 general ward bed-days and 300 ICU bed-days. Weight an ICU day at three ward days, the low end of that range.
One month's ₹66 lakh nursing cost, apportioned two ways
| Bed-days | Weight | Weighted bed-days | ₹ per bed-day, spread evenly | ₹ per bed-day, by acuity | |
|---|---|---|---|---|---|
| General ward | 2,400 | 1 | 2,400 | 2,444 | 2,000 |
| ICU | 300 | 3 | 900 | 2,444 | 6,000 |
| Total | 2,700 | 3,300 |
Both methods recover the full ₹66 lakh (2,400 × ₹2,000 plus 300 × ₹6,000 is ₹48 lakh plus ₹18 lakh). But the even split undercosts every ICU day by ₹3,556 and overcosts every ward day by ₹444. A specialty whose patients spend days in ICU then looks far cheaper than it is.
Spreading nursing evenly makes intensive specialties look cheap and routine ones look expensive: exactly backwards.
Three ways of paying doctors, three costing treatments
Doctor compensation distorts the picture, and mixing up its three common models is a frequent error.
- Salaried consultants. A fixed cost spread across their cases by the theatre minutes each consumes, so their matters. A salaried consultant at low caseload has a very high cost per case.
- Fee-for-service or revenue share. Variable and directly attributable. Easy.
- Hybrid, with a minimum guarantee. Fixed up to a threshold and variable above it, so cost per case falls as volume rises until the threshold is crossed.
A hospital with a mix of all three cannot compare specialty profitability without normalising for this, and most do not. Every fixed cost in the building behaves this way; see why unit cost moves with volume and, for occupancy, the real economics of running a hospital.
Capacity consumed that no payer paid for
Every business has yield loss. In a hospital it looks like this:
- No-shows and cancellations. A cancelled theatre slot consumed scheduling, preparation and an empty OT.
- Denied and disallowed claims. Work delivered, cost incurred, revenue not received.
- Length of stay beyond the package. Days consumed that the package rate does not pay for.
- Readmissions within the warranty period. Delivered twice, paid once.
Each belongs in the cost of the cases that did get paid, because the capacity was consumed either way.
Payer contribution and specialty profit are different numbers
Set the cost layers against what each payer pays and two numbers come out, for two decisions.
Payer contribution = realised revenue per payer − cost layers 1 to 3 (consumables, clinical time, theatre and equipment), by payer, by procedure.
Specialty profit = realised revenue − all four layers, including apportioned overhead.
Payer contribution says whether a payer's rate covers what its cases consume: the question in a package negotiation. Specialty profit says whether a specialty carries its share of the building: the question in deciding what to grow. A payer that shows a loss once overhead is loaded can still be covering its direct cost and paying towards the building.
Realised means after package caps, disallowances, settlement deductions, write-offs and the financing cost of the collection period. In the waterfall on the hospital pricing page, a ₹2.4 lakh bill is approved at ₹1.9 lakh and realises ₹1.67 lakh. That page builds the revenue column; this one supplies the cost column.
Diagnostics: cost per test follows analyser utilisation
Diagnostics runs on different economics from the hospital core. Cost per test is built from reagent and consumables, the analyser's cost per run, technician time, and a share of collection logistics and reporting. The dominant variable is equipment utilisation. An analyser's cost is largely fixed. At 40% of capacity it is spread over half the tests it would carry at 80%, so the analyser cost per test roughly doubles. That is the analyser's share of cost per test; reagent is consumed test by test and moves with yield instead.
Yield loss here is repeat tests, QC and calibration runs, reagent wasted on small batches, and samples rejected on arrival. Small-batch running is the main driver, which is why consolidating runs is a costing decision as much as an operational one.
Channel then adds a revenue question: walk-in patients pay close to list, while B2B referral rates are negotiated one discount at a time and drift. Home collection is usually priced as a convenience and costs more than its fee once phlebotomist time, travel and failed visits are counted. Both are taken further in hospital pricing and payer mix.
Three procedures and ten tests, costed properly
For a hospital: take your three highest-volume procedures. Build cost per procedure for each with acuity-weighted nursing and OT-minute theatre cost. Then set it against realised revenue for each payer you serve, so payer contribution and specialty profit sit side by side.
For a diagnostics chain: take your ten highest-volume tests. Cost each one including analyser utilisation and yield loss, then split the realised rate by walk-in and by referral partner.
Either exercise takes a few days, and it gives every later pricing, package and empanelment decision a cost to stand on.
Our Pricing Maturity Assessment covers cost structure alongside the rest of your pricing.
Frequently asked questions
What is hospital costing?
Costing what each procedure, bed-day and test actually consumes: consumables, clinical time, theatre and equipment time, and apportioned overhead. Set against realised revenue by payer, it shows which cases, payers and specialties cover their cost.
How should a hospital cost a package rate?
Cost what the package contains at current prices: consumables and implants, clinical and theatre time for a typical case, and bed-days at the length of stay the package assumes. Compare that with the realised package rate, and count days beyond the package as yield loss.
How useful was this article?
One tap. It tells us what to write more of.
About the author
SRF Capital Studio
The next one
Get what we publish next, by email.
Working notes on raising, borrowing, protecting, growing and structuring capital in India. One email a week at most, and you can leave any time.
We use your address only to send this. See our privacy policy.
We store your address to send you these emails and nothing else. See our privacy policy.
Related reading
