The Capital Request That Should Have Been a Rota Change.

The capital committee agenda lands on the desk a week before the meeting. Four requests, two of them substantial. A new MRI scanner because the department is “at capacity.” Additional endoscopy bed space because the wait list is “unmanageable.” Expanded ICU because bed pressure has become “structural.” A new pre-admission clinic because the current facility is “saturated.” Each request arrives with a business case. Each business case describes a real operational problem. And each of them, in my experience, is probably at least partly misdiagnosed.

This is not a criticism of the people who wrote the business cases. It is a pattern. Clinicians and operational managers describe the pain they see, and the default framing of that pain in most German hospital cultures is capital-shaped: we do not have enough of X, we need more of X, the solution is to acquire X. The framing is so embedded in how leadership discussions are structured that the alternative diagnosis — we have enough of X, but we are using it badly, and the solution is to change how we use it — is often not even considered as a formal option. The result is that meaningful proportion of the capital requested in any given year is actually purchasing operational inefficiency that could have been redesigned for zero marginal cost.

The FLOW layer of CuraOS exists to surface this pattern. Not as a defence against legitimate capital investment — hospitals genuinely do need new scanners, new endoscopy suites, new ICU beds, and new pre-admission facilities under the right conditions — but as a disciplined check, run before the business case is written, that filters out the capital requests whose real diagnosis is operational. The filter is simple. It takes about three hours per request. It saves, at hospital scale, seven-figure sums annually.

The observation: most hospital capital requests describe real operational pain accurately and diagnose the cause incorrectly. A four-question filter, applied before the business case is written, identifies which requests are genuinely capital and which are operational problems dressed in capital language.

Why capital framing is the default.

Three structural reasons make capital framing the default response to operational pain in German hospitals.

Capital investment is legible to governance. A capital request produces a specific document, a specific line item, a specific board decision, and a specific implementation timeline. Each of these artefacts is visible to supervisory structures and demonstrates management action. An operational redesign, by contrast, produces fewer artefacts: a changed rota, a different protocol, a new meeting cadence. The operational path is harder to report upward and harder to claim credit for.

Capital investment has a clear finance owner. A capital request is evaluated by the capital committee; the Finanzabteilung has the mandate to analyse it; the decision process is established. An operational redesign does not have a comparable decision structure. No single function in most German hospitals is explicitly mandated to evaluate “whether this capital ask is really an operational problem.” The absence of the mandate allows the capital framing to proceed unchallenged.

The clinical teams who write business cases are not rewarded for operational sophistication. A clinical team that proposes a capital solution appears to be advocating for its patients. A clinical team that proposes an operational redesign appears, to some eyes, to be accepting a status quo that the capital path would have transcended. The reputational economics favour capital requests, even when the operational path is clearly superior on cost-benefit grounds.

These structural forces are not malicious. They are the normal functioning of a governance system optimised for capital decisions. The FLOW-layer discipline is to install a compensating structural force — the four-question filter — that the existing governance does not naturally provide.

The four-question filter.

When a capital request is submitted — or more usefully, when the underlying pain is first identified, before the capital framing has solidified — four questions filter the request into capital-justified, operational-redesign, or hybrid categories. Each question is binary in answer form, but the aggregate pattern across the four determines the classification.

The first question — is the asset used at physical capacity? — is the most commonly failed. Most requests for additional capacity describe peak-hour saturation as though it were full-day saturation. When the utilisation is measured across the complete operating window, the asset typically has significant idle time outside peak hours. The operational path is to shift demand into the idle time rather than add capacity for peak. A CT scanner that is fully booked 09:00–13:00 and half-idle 13:00–18:00 is not a capacity problem; it is a scheduling problem.

The second question — is the demand profile actually uniform? — probes whether the peak can be softened. Many peaks in hospital operations are self-inflicted: elective surgery starts concentrated at 08:00 rather than distributed across the morning; ward round timing concentrates imaging requests in a narrow mid-morning window; discharge timing concentrates pharmacy demand in the same window. A non-uniform demand profile is a design choice that can be redesigned. The question is whether the redesign has been seriously attempted before the capital request was framed.

The third question — is the process serial where it could be parallel? — identifies hidden capacity. Many diagnostic and treatment processes run steps sequentially that could, with protocol modification, run in parallel. Preoperative assessment, imaging, laboratory testing, and anaesthetic consultation are commonly run in sequence when the clinical requirements would allow them to overlap. Each sequential pair that becomes parallel is a gain in effective capacity without any change to physical assets.

The fourth question — is the bottleneck the asset or the handoffs around it? — is often the decisive one. A scanner that is “over capacity” is often fine in terms of actual imaging throughput; the constraint is in the handoffs — transporter availability, porter scheduling, radiology reporting turnaround, clinical review of reports. A scanner that reports same-day has different effective capacity than one that reports the following day, even if the imaging volume is identical. Handoff-timing constraints masquerade as capacity constraints, and purchase of additional capacity does not resolve them.

What the filter changes when applied.

When the four-question filter is run consistently on capital requests, three operational effects emerge over time. The effects are cumulative and take eighteen to twenty-four months to stabilise.

The capital pipeline shrinks. The volume of capital requests that reach the capital committee declines, because many requests are rerouted to operational redesign during the diagnostic phase. Those that do reach the committee arrive with better documentation of why operational redesign has been considered and why the capital path is necessary.

Operational projects multiply. The projects that were formerly capital business cases become operational redesign projects, each with a named owner, a measurable target, and a 90-to-180-day delivery timeline. The proliferation of small operational projects changes the departmental rhythm — operational improvement becomes part of the ordinary management conversation rather than a specialised function.

Capital decisions become faster. Because the capital requests that do reach the committee are better filtered and better documented, the committee itself can deliberate faster. The backlog shrinks; the hit rate on approvals rises; the implementation timeline from decision to commissioning shortens.

The Schlüchtern FLOW work on capital discipline.

At Main-Kinzig-Kliniken Schlüchtern, the geriatric department installed the four-question discipline in 2021 as part of the broader operational programme under the research with Prof. Dr. Rainer Sibbel at Frankfurt School [1,2]. The initial application was to a specific ward-capacity request that had been framed as a case for physical expansion; the four questions surfaced a combination of peak-hour scheduling concentration (discharge-readiness reviews clustered at 14:00 rather than distributed across the day), serial-parallel inefficiency (physiotherapy assessments running sequentially rather than in parallel with medication reviews), and handoff-timing constraints (pharmacy turnaround for discharge medications) that together accounted for a material fraction of the apparent capacity shortfall.

The ward-capacity request was redirected into operational redesign. The redesign produced measurable throughput improvements across 2021–2022 without physical expansion. The capital request was not formally cancelled; it was deferred pending the operational effect. By 2023, the operational redesign had absorbed enough of the original pain that the physical expansion request was revised and resubmitted with a more narrowly defined scope, for which genuine capital investment was justified.

The pattern is representative. The four-question filter did not eliminate the capital investment; it sharpened it. The hospital ended up with a smaller capital spend, delivered faster, on a better-defined problem — and with the operational improvements in place before the capital asset arrived. This sequencing matters: an operational redesign installed after a capital purchase has to work around the purchased asset, whereas one installed before the purchase informs the purchase specification.

What to do on Monday.

Take the three most recent capital requests that have crossed your desk. For each, work through the four questions with the clinical team that wrote the business case. Do not frame this as scepticism of the request; frame it as a diagnostic that both sides run together, with a genuine expectation that some requests will survive the filter and some will be redirected.

For the requests that survive the filter, authorise the capital work as usual — but with the four-question diagnostic documented as part of the business case package. For the requests that fail one or more questions, open a parallel operational-redesign project with a 90-day and 180-day milestone. Do not cancel the capital request immediately; hold it as deferred pending the operational effect. If the operational redesign closes the pain, the capital request can be formally withdrawn; if not, it is resubmitted with a sharper scope.

Install the four-question filter as a standing item in the capital committee’s process. Every new request is evaluated against the four questions before the business case is finalised. The process adds approximately three hours of work to the upstream of each request; it removes approximately thirty percent of requests from the capital pipeline (my observational estimate across engagements, variable by hospital); and it changes the conversation about resources at the institutional level.

Do not apply the filter selectively to requests you are already sceptical of. The filter’s value comes from applying it consistently, including to requests you would otherwise approve without question. The purpose is to install the diagnostic discipline institutionally, not to provide a rhetorical weapon for rejecting individual requests.

The capital request that should have been a rota change is, in my experience, one in three — sometimes one in two. The hospital that installs the discipline to identify which third or half it is, routinely and without drama, unlocks both financial savings and a categorically improved relationship between operational management and capital decision-making.

Most of what feels like a capacity problem is a rhythm problem. Most of what feels like a capital need is an operational design flaw. The four questions are the filter. The filter is the whole FLOW discipline in miniature.