The $18 Million Private Room Conversion and a Six-Week Deadline
A 220-bed community hospital is deciding whether to convert its 40 semi-private medical-surgical rooms to private rooms at an estimated capital cost of $18 million. The CFO will support the project only if projected payback is under seven years. A vendor white paper, cited by the facilities director, claims private rooms cut hospital-acquired infections by 30% and length of stay by 1.5 days. Baseline data for your hospital: 78% average occupancy, mean LOS 4.8 days, HAI rate 4.2 per 1,000 patient days. Sixty-two percent of reimbursement is fixed under DRG and capitated contracts. The board meets in six weeks and wants a recommendation with a defensible financial rationale, not a literature review.
1. Critique the vendor's two claims and state what additional evidence you would require before any of those effect sizes enter a financial projection.
Read the worked answer
Appraise before you multiply. Both claims are aggregate effect sizes detached from context.
- HAI reduction: Private rooms plausibly act through fewer roommates and easier cohorting, but published reductions are usually confounded by concurrent bundles (hand hygiene campaigns, CHG bathing, stewardship). Attribution of the whole effect to room type is a category error. Demand a case-mix-adjusted estimate with confidence intervals, from a setting comparable in size and acuity to yours, and a stated comparator (semi-private vs. private, not 'old unit vs. new unit').
- LOS reduction: 1.5 days is large relative to your 4.8-day baseline and improbable as a room-type-only effect. LOS is driven by case mix, discharge planning, and post-acute capacity. Ask whether the estimate survives adjustment and whether it reflects excess days or total days.
If no credible local or transferable estimate exists, the honest position is: use a range anchored at a small effect, with zero as the low case, and flag the assumption explicitly to the board.
2. Given that 62% of reimbursement is fixed, which financial levers actually matter here, and how would you build the business case around them?
Read the worked answer
Fixed reimbursement kills the 'freed bed-days = revenue' argument. Shortening LOS generates revenue only where there is unmet demand and a payer that pays per case or per diem. So the levers are:
- Cost avoidance, not revenue. Roughly 62,600 patient days per year at 78% occupancy; at 4.2 HAIs per 1,000 patient days that is on the order of 260 infections annually. Value them at variable cost of the excess stay, not fully loaded cost, and apply only an attributable fraction you can defend - never the vendor's 30%.
- Throughput, conditionally. At 78% average occupancy the capacity argument is weak unless peak-day or ED-boarding data show genuine constraint. Ask for peak occupancy and diversion hours before claiming admissions growth.
- Market and workforce effects. Physician recruitment, patient choice, and agency-nurse avoidance are real but soft; present them separately.
- Compliance/life-safety risk if any rooms fail current standards.
Frame two cases - a hard-benefit case (cost avoidance + defensible throughput) and a total case including soft benefits - and show what payback each yields.
3. The board insists on a single payback number for the meeting. How do you respond, and what risk communication would you put on one page?
Read the worked answer
Give a range with named assumptions, not a false point estimate.
- Present a sensitivity table: attributable HAI reduction {0%, 10%, 20%}, LOS effect {0, 0.3, 0.8 days}, and peak-occupancy scenarios {no constraint, 5% unmet demand}. Show payback under each.
- State plainly which cells fall outside seven years. If most do, say the project cannot be justified on hard financials alone and recommend a decision framed on mission, market, and risk.
- Separate what is known (your baseline rates, capital cost), estimated (cost per HAI), and assumed (attributable fraction of the design).
- Propose a decision with a learning component: phased conversion of one 10-room pod with pre/post measurement and a go/no-go gate, which converts an unresolvable argument into a cheap experiment.
Boards accept uncertainty; they punish surprise. The one-pager should let a skeptical CFO reconstruct your number from four visible inputs.