Three Cohorts, Named.

The argument in Post 6 was that a hospital department typically serves about ten operational cohorts, and that naming them — with volume, DRG family, margin, and clinical lead — is the foundational ORBIT move. In practice, the most common failure mode when teams try to do this is not that they never produce the list. It is that they produce a list of ten names without the underlying specificity, and declare the work done. The list becomes a slide. The slide gets filed. Nothing operational changes, because nothing operationally concrete was actually committed to.

This post drills into the harder move: naming three cohorts specifically, properly, with the four pieces of information that make each name operational. Three is a much more useful number than ten at the starting point, because three forces the discipline that ten allows you to avoid. Once you have done three cohorts with the full specificity, the next seven are mechanical. Doing ten without the specificity is theatre with a more elaborate stage.

The observation at the foundation of this post: naming three cohorts — with exact volume, DRG, named clinical lead, and margin — is the minimum threshold at which the ORBIT layer becomes operationally useful. Below that threshold, the department is still running on abstractions. At or above it, specific operational decisions become possible.

Why three is the right number to start with.

Three named cohorts is large enough to produce a pattern and small enough to do properly. Two is insufficient — the department can reasonably argue that its two largest cohorts are idiosyncratic and that lessons from naming them do not generalise. Three is the smallest number that produces pattern-level insights: three named cohorts demonstrate that the naming discipline works across different clinical pathways, different DRG families, and different clinical leads. Five is also a defensible starting point; beyond five, the effort required to do the naming properly grows faster than the organisational capacity to absorb the implications.

The practical argument for three is also about time. Naming three cohorts properly — pulling exact volume from coding records, confirming DRG distribution with the coder, assigning a specific clinical lead by name, calculating margin after allocated cost — takes about a day of analyst time and two hours of leadership time per cohort. Three cohorts is therefore roughly three days of analyst work and a half-day of leadership engagement. This is tractable. Ten cohorts at the same level of specificity is two weeks. Most hospitals abandon the effort halfway through ten; most can complete three.

Once three are done, the demonstration effect changes the internal conversation. Leadership teams that see what a properly-named cohort looks like operationally can authorise the next seven with far more clarity about what the output should be. The first three are the investment; the remaining seven are the return on that investment.

Which three to start with.

The choice of which three cohorts to name first is itself a strategic decision, not a data exercise. There are three defensible heuristics, each of which produces a different starting three and a different organisational signal.

The “largest volume” starting three. The top three cohorts by annual case volume produce the highest-leverage first three in operational terms. Improvements to pathways for these cohorts affect the largest number of patients, and naming them surfaces the largest proportion of the department’s actual work. This is typically the right starting three for a hospital that has not previously run any cohort-level operational analysis.

The “highest margin” starting three. The top three cohorts by margin per case (after allocated cost) produce the highest-leverage first three in financial terms. This is typically the right starting three when the Geschäftsführung has asked the medical directorate to articulate, for the first time, where the department actually earns its contribution. The exercise reveals which clinical work is financially sustaining the department and which is financially subsidised by other work.

The “most specialty-defining” starting three. The top three cohorts that represent what the department exists to do — the work that distinguishes it from competing services, the work around which its reputation is built. This heuristic produces the right starting three when the department is entering a strategic repositioning conversation, because naming the specialty-defining cohorts forces the leadership team to be specific about what the department is actually claiming to be good at.

For most hospitals, the largest-volume starting three is the right first move. The financial and strategic starting threes are better suited to hospitals that have already completed one full pass of the discipline and are entering deeper strategic conversations. The key is to pick one heuristic deliberately, name that starting three, and then use the demonstration to authorise a second pass if needed.

What naming each of the three actually requires.

The four questions below are the minimum specificity that distinguishes a named cohort from a labelled cohort. A cohort is labelled when it has a name in a slide; it is named when the named clinical lead can answer all four questions from memory, without reference to documentation. Most hospitals have cohorts that are labelled. Very few have cohorts that are named in this stronger sense.

What naming unlocks operationally.

Naming three cohorts properly changes the subsequent operational conversation in four specific ways that are invisible from the outside but transformative from the inside.

Capital allocation becomes evaluable. When the Chefarzt requests capital or staffing investment, she can now specify which of the named cohorts the request serves and what the expected impact on that cohort’s volume, margin, or pathway efficiency would be. The Geschäftsführung can evaluate the request against the cohort’s operational performance, rather than against an abstract clinical case. Decisions become faster and more defensible in both directions.

Pathway design becomes specific. Pathway improvement projects are usually framed as department-wide initiatives and end up producing diffuse benefits that are hard to attribute. When the project is framed as pathway improvement for a specifically named cohort — “we are redesigning the hip-fracture-with-anticoagulation pathway” — the project has a clinical lead (the named lead for that cohort), a defined population (the named cohort’s volume), a measurable outcome (change in length of stay or margin for that cohort specifically), and a bounded scope. Projects framed this way finish. Projects framed at department level do not.

External communication becomes precise. When the Chefarzt talks to referring physicians, to the regional health authority, to potential recruits, or to the press, she can describe what the department does in terms of specific named cohorts rather than in the language of “complex multimorbid elderly patients.” The specificity is itself a form of professionalism; referring physicians are more likely to send the kinds of cases the department names as its work, because they know the department will own those cases with operational clarity.

Workforce attention is calibrated. The Ward Round, the departmental meeting, the teaching case, the quality review — all of these can be structured around the named cohorts rather than around whatever happens to be in the ward at the moment. This sharpens clinical judgment over time, because the team develops specific expertise in the named pathways rather than diffuse expertise across everything.

The Schlüchtern starting three.

At Main-Kinzig-Kliniken Schlüchtern, the geriatric department began its ORBIT work in early 2020 with a starting three: hip fracture with co-managed geriatric complex treatment, acute-on-chronic heart failure, and acute delirium requiring formal workup. Each cohort was named with the four pieces of specificity — volume, DRG family, clinical lead, and margin after allocated cost. The cohort lead for hip fracture was the deputy chief, who had the largest existing clinical relationship with the hip-fracture caseload; the heart-failure lead was the consultant with the longest specific interest in the pathway; the delirium lead was a senior oberarzt with a research interest in geriatric psychiatry.

The choice of hip fracture as the anchor cohort was not arbitrary. According to PubMed-indexed research, Van Heghe and colleagues’ 2022 systematic review and meta-analysis of 37 studies covering 37,294 patients found that orthogeriatric co-management for hip fracture is associated with a mean length-of-stay reduction of 1.55 days, a 28% reduction in in-hospital mortality, a 14% reduction in one-year mortality, and a 19% reduction in delirium incidence [3]. The evidence base for hip fracture as a named, managed cohort is stronger than for almost any other geriatric presentation — which is exactly why the discipline of naming it explicitly, with a named clinical lead and a named margin, produces operational returns that the generic “we treat elderly patients” framing cannot. The cohort with the strongest published evidence was, consequentially, the cohort with the largest potential upside once it was named.

The three cohort leads were given quarterly reporting responsibility for their cohort’s volume, margin, pathway efficiency, and readmission rate. They attended the monthly departmental operations meeting to present their cohort’s status. They participated in pathway improvement projects specifically scoped to their cohort. Over the course of 2020–2022, the three named cohorts became the primary unit around which the department’s operational conversation was organised. The subsequent seven cohorts, named across 2022–2024, were added with far less friction than the original three, because the naming discipline was already internalised.

By 2025, under the research programme with Prof. Dr. Rainer Sibbel at Frankfurt School [1,2], the department operated with all ten cohorts named at the level of specificity described in this post. The operational outcomes across the wider transformation — 17.6% length-of-stay reduction, 36% throughput improvement, 42% case volume growth — are documented in Posts 16 and 21 of this series. The ORBIT-layer claim is specifically that the cohort naming discipline was a necessary prerequisite for those improvements, not a parallel achievement. Without naming, leadership attention could not be calibrated to where the operational leverage actually sat.

What to do on Monday.

Pick a starting three. Use the largest-volume heuristic unless there is a specific reason to pick a different heuristic. The three cohort names should be clinically specific enough that a coder could match them against ICD-10 groupings without ambiguity.

For each of the three, commission the four pieces of specificity. Volume from the last twelve months of coding records. DRG distribution from the coder (which DRG does this cohort most commonly generate, and what is the BWR). Margin per case after allocated cost, from controlling; a range is acceptable if precise allocation is not available. Clinical lead by name — a specific person in the department who accepts operational ownership of the cohort’s pathway, margin, and readmission rate.

Hold the first cohort-review meeting three weeks after the naming is complete. The three cohort leads present, each for five minutes: cohort volume in the most recent month, notable pathway issues observed, coding capture status, upcoming pathway improvement work. The leads take five more minutes to discuss one operational theme that cuts across the three cohorts — because the act of identifying such a theme is itself a test that the naming discipline is working.

Continue the cohort-review meeting monthly. After six months, the three named cohorts will have become the primary unit around which department operations are discussed. At that point, commission the next three cohorts, using the same four-question specificity. The expansion will be noticeably easier than the first three.

A department that can name three cohorts properly is a department that has crossed the threshold at which the ORBIT layer becomes operational. A department that lists ten without the specificity is still running on abstraction, regardless of how thorough the slide looks. The specificity is the work. The specificity is the whole ORBIT layer.