Demand forecasting in healthcare: capacity and supplies

Demand forecasting in healthcare compares expected service volumes with department actuals, staffed capacity and the supplies those volumes imply. Start with a frozen forecast and aggregate operational exports for the same period. Completed services show delivered activity; they do not reveal every request the organization could not accommodate.

Start with a defined service and period

A finance team at an illustrative US health system is reviewing September 2026 outpatient activity. Its forecast estimates completed services by department. This is an operational finance review, with department totals rather than individual records or clinical predictions.

Define what a service means before combining exports. An imaging exam, therapy visit and ambulatory procedure are separate counting units. Their counts can reconcile to the activity register, but their staffing requirements are not interchangeable. Keep new requests and completed services in separate columns if both are available.

Healthcare finance already works with utilization and cost-center information: CMS describes annual provider cost reports as containing utilization data, facility characteristics, and costs and charges by cost center. For a monthly review, use your own current department exports rather than treating an annual public report as a scheduling forecast.

The general demand forecasting guide covers the broader planning question. This workflow concentrates on comparing the supplied service forecast with the resources behind it.

Assemble the exports without changing the forecast

Bring the approved forecast snapshot, completed-activity summary, staffed-slot schedule and supply assumptions. Match department codes and month boundaries across them. Preserve the forecast approval date so the review does not compare actuals with a plan revised after the work occurred.

For an Excel workbook, let a Raw sheet contain aggregate activity: department in A, month in B, status in C and service count in D. Store the month as its first calendar date. On a Volume sheet, put department in A, frozen forecast in B, completed activity in C and staffed slots in D.

Populate C2 from the aggregate export with:

=SUMIFS(Raw!$D$2:$D$500,Raw!$A$2:$A$500,$A2,Raw!$B$2:$B$500,DATE(2026,9,1),Raw!$C$2:$C$500,"Completed")

Microsoft documents SUMIFS as adding values that meet multiple criteria; its sum and criteria ranges must have matching dimensions. Extend every range together when your export is longer. Covirage is not affiliated with Microsoft.

Reconcile the resulting completed counts to the source report before reading forecast error. A department missing from the capacity export needs an exception flag, not an assumed zero.

Compare forecast error and capacity separately

Place the following volume example in Volume, with headers in row 1 and departments in rows 2 through 4. Percentages are rounded to one decimal place; the workbook calculations use unrounded values.

Use these formulas in E2, F2 and G2, then fill down:

=C2-B2
=ABS(E2)
=IF(D2=0,"Check capacity",C2/D2)

The net gap is only 5 services: 1,655 minus 1,650. That hides the imaging increase and the shortfalls elsewhere. The absolute errors total 40 + 30 + 5 = 75 services. Weighted absolute percentage error against completed activity is:

=SUM(F2:F4)/SUM(C2:C4)

That is 75 / 1,655 = 4.5%, rounded. It measures this forecast's completed-volume error, not unmet demand or service quality. Retain department results because a small consolidated percentage can conceal a meaningful local difference.

Do not average the utilization percentages or pool the spare slots. A therapy slot does not become an imaging slot merely because it is unused. The forecast bias review explains why recurring directional differences deserve a separate check across periods.

Translate volume differences into supply dollars

Use supply ratios approved by the department and procurement. The following kit counts are illustrative accounting assumptions, not instructions about supplies required for a particular procedure. Each department uses its own kit type, so adding kit quantities across departments would not produce a useful purchasing total.

On a separate Supplies sheet, place department in A, forecast services in B, completed services in C, kits per service in D and cost per kit in E. Calculate forecast cost in F and actual-volume cost at the same standard in G:

=B2*D2*E2
=C2*D2*E2

The implied volume effect is $110: $14,310 minus $14,200. Its department components reconcile to $320 minus $90 minus $120 = $110. This holds kit costs and usage standards fixed, separating the effect of activity from purchasing prices and consumption differences.

Compare this calculation with actual inventory issues before calling it actual supply expense. Receipts, opening stock, waste and closing stock can explain why purchasing invoices differ from supplies consumed. Give procurement the department exceptions rather than an unexplained change in the total.

Where healthcare forecasting goes wrong

• Throughput becomes demand. A full schedule can cap completed activity while unserved requests rise. Review aggregate requests and waiting-list movements alongside completion counts.
• Unstaffed space becomes capacity. Use the approved operating schedule and usable service slots, with closures reflected in the period being reviewed.
• Different services share a denominator. Keep ratios within a defined department and service category; unlike activities need their own resource assumptions.
• Purchases become consumption. Standard supply needs, actual issues and vendor invoices answer different questions. Reconcile their differences before changing the forecast.
• Revised plans erase the comparison. Keep the approved snapshot and explain changes prospectively. Use the demand forecasting process to document that handoff.

Finance should close the review with a reconciled variance, an operations explanation and an owner for the next action. Repeat the comparison over consistent periods before treating a single month's miss as a persistent pattern.

Review your department files in Covirage

Upload a file or Google Sheet containing your aggregate department forecasts, completed volumes, capacity and supply assumptions; Covirage maps the columns, and its tools compute and check the comparisons. The external AI model explains the results with every figure traced to its rows. Create a free Covirage account, set up your business and bring a sample of your own data; creating the account takes about a minute.

See Covirage for healthcare teams for the industry context.