The 24-Hour Coding Loop.

Three admissions, one Monday. A 58-year-old in cardiology with recurrent NSTEMI on top of known heart failure. A post-operative patient from general surgery with bowel resection and an escalating infection pattern. An elderly patient admitted through the emergency department with a femoral neck fracture. Three different specialties, three different DRG families, one identical structural problem.

Each of these patients will spend somewhere between four and ten days in hospital. The clinical complexity of each case will be substantial — multiple secondary diagnoses, procedures, complications, comorbidities — and the documentation will be produced in the ordinary cadence of physician notes, nursing entries, therapy records, and discharge summaries. Then, two to five days after discharge, each chart will arrive on a coder’s desk. The coder will open a record that is now closed. The physicians who managed the patient are on different shifts. The nursing team has turned over half the beds. The therapists have moved on to dozens of other patients. The social workers are already processing the next Monday’s admissions.

The coder works from what is written. What was observed but not written will not be coded. What was written but not structured to support a specific OPS complex-treatment bundle will not trigger it. What might have warranted a higher-complexity DRG — a CCL-raising secondary diagnosis, an intensive-care threshold, a complex-treatment code — but was documented in the flat, clinical-handover register of the physician note, will code as the middle-tier DRG. Nobody will know what was lost. The case will close. The hospital will bill what the documentation technically supports.

This is the default in most German hospitals, across every specialty with meaningful DRG-driven revenue. It is also the single largest revenue leak in the CAPTURE layer, and the one most resistant to being fixed — because almost nobody inside the organisation experiences the loss directly. The coder cannot code what is not there. The physician has discharged the patient. The controller sees a downstream aggregate that has always been the aggregate. And the Geschäftsführer sees a margin line that, across hundreds of such cases monthly, has been quietly eroding for years.

Why procedure-driven reimbursement amplifies the loss.

The German G-DRG system has evolved, over successive reform cycles, toward heavy weighting of procedural and complex-treatment OPS codes [1]. This matters because procedures can be under-captured in ways that diagnoses typically cannot. A diagnosis either exists in the chart or it does not. A procedural or complex-treatment bundle — an intensive-care complex treatment, a stroke-unit complex treatment, a geriatric frührehabilitative Komplexbehandlung, a cardiac or oncological intervention with specific duration and team-composition thresholds — requires documentation that meets specific structural requirements across multiple disciplines [2,3].

All of which can be technically present in a clinical sense and yet fail to be coded, because documentation written days after the fact does not meet the structural requirements that the coding framework demands [4].

The pattern repeats across specialties. In cardiology, complex-treatment codes for heart failure management, arrhythmia interventions, and combined cardiac-renal decompensation depend on documentation quality that degrades rapidly after discharge. In intensive care, OPS 8-980 complex treatment depends on team-composition logs and assessment documentation that is genuinely difficult to reconstruct retrospectively. In surgery, the combination of primary procedure code, complication codes, and nursing-intensity secondary diagnoses produces DRG assignment variance that can span multiple thousands of euros per case. In geriatrics, OPS 8-550 complex treatment — combined with a hip fracture, the differential between a baseline I47C assignment (BWR 1.389, base reimbursement approximately €6,104) and an I34Z assignment (BWR 2.629, base reimbursement €11,552) is +€5,448 per case in base reimbursement [1], or approximately +€8,624 when including the Pflegeerlös component.

The scale of this leakage is rarely measured directly by hospitals. It appears instead as an absence — cases that should have coded into a higher-complexity DRG but did not, distributed across a case mix, visible only through a specific retrospective audit. Across operational data from Main-Kinzig-Kliniken Schlüchtern [5] and comparable work across regional hospital and long-term care contexts, per-case under-capture in the range of €2,000–€6,000 is a defensible working estimate for complex inpatient cases; specific DRG differentials like the I47C-to-I34Z gap routinely exceed that. At hospital scale, across hundreds or thousands of cases annually, this compounds into annual figures that most controllers have never seen quantified.

What fallbegleitende Kodierung actually is.

The German-language term for concurrent coding — fallbegleitende Kodierung — translates literally as “case-accompanying coding.” The translation matters. It is not coding about the case, performed after the case has completed. It is coding alongside the case, performed while the clinical evidence is still being generated and can still be shaped.

In practice, fallbegleitende Kodierung is a daily rhythm with three structural elements.

First — a coder is procedurally connected to the ward huddle within 24 hours of admission. “Procedurally connected” can mean physically present at the morning ward round, or remotely connected via a documentation review platform, or a scheduled standing 10-minute daily call between the coder and the ward registrar. The specific mechanism matters less than the consistency. What matters is that within 24 hours of a patient arriving, a coder has looked at the documentation and either confirmed it supports the anticipated coding or flagged specific gaps that need to be closed.

Second — documentation gaps are communicated to clinicians while the patient is still present and the evidence is still being generated. If the coder sees that the admission narrative for a post-operative case does not document the specific complication pattern that would trigger a CCL-raising secondary diagnosis, the prompt reaches the clinician within a day. If the coder sees that an ICU patient’s documentation will likely fall just short of the team-composition threshold for OPS 8-980 complex treatment, the structural fix is raised while there is still time to correct it. If the coder sees that a hip-fracture case with geriatric comorbidities is being documented as a simple surgical case rather than a geriatric co-managed case warranting OPS 8-550 — a differential that routinely exceeds €5,000 per case in base DRG reimbursement — the feedback reaches the clinician while the patient is still on the ward. The clinician corrects the documentation in real time. The complexity, which was always present clinically, becomes coded.

Third — final coding is completed on or near the day of discharge. Because the groundwork has been done daily, the discharge coding is largely ratification rather than reconstruction. The case closes. The bill generates. Liquidity improves because cash enters the system days or weeks earlier than the retrospective model allows.

None of this is conceptually complicated. It is structurally rare.

Why most hospitals have never built it.

The structural rarity of fallbegleitende Kodierung — despite its well-documented benefits in both the German Medizincontrolling literature and in published operational cases — is worth examining on its own terms. Three obstacles consistently appear.

The first is staffing model. Most German hospitals have inherited a coding department that was designed for retrospective work. Coders sit in a central office, process cases in batches after discharge, and have no structural touchpoint with the wards. To install concurrent coding, that department must be redesigned — coders need either geographic proximity to wards or reliable remote presence, they need clinical schedule alignment, and they need the professional standing to flag gaps to physicians in real time. The redesign is not expensive. It is politically hard, because it asks a department that has operated a certain way for fifteen years to operate differently.

The second is clinical acceptance. Physicians are rightly protective of their clinical time. A coder interrupting ward round to flag a documentation gap is, without preparation, perceived as bureaucratic friction rather than operational support. Installing the bridge requires that physicians understand — concretely, with numbers — what the gap is costing and why closing it is in the clinical team’s interest, not just controlling’s. This conversation is rarely held because it requires the Chefarzt and the Leitung Controlling to sit down and agree on the framing together. Many hospitals’ organisational culture has never built the routine for that conversation.

The third is measurement. Hospitals that have never quantified their retrospective-coding leak have no business case for the concurrent-coding investment. The return is invisible because the loss is invisible. This is the classic operational constraint of the CAPTURE layer — the problem cannot be seen until it is measured, and until it is seen, the investment to fix it cannot be justified. Breaking this loop requires a specific act of commissioning: a two-week audit of recent discharges against the documentation gap, which produces the euro number that makes everything else tractable.

What the 24-hour loop produces — the Schlüchtern evidence.

At Main-Kinzig-Kliniken Schlüchtern, a concurrent-coding rhythm was installed in the geriatric department between 2019 and 2021 as part of broader operational restructuring that produced measurable results across the full CuraOS stack. The rhythm itself involved a daily 15-minute coder touchpoint with the ward registrar, a weekly joint case review between the department lead and the Leitung Medizincontrolling, and a quarterly audit of top-10 DRG coding patterns against benchmark.

The operational effects that have been formally documented and used in the research programme with Prof. Dr. Rainer Sibbel at Frankfurt School [6] are summarised in the figures below. What is relevant for the CAPTURE layer specifically: complex-treatment coding capture rose substantially across the eligible geriatric case mix, case mix index rose in parallel, and time from discharge to billing — which is a liquidity metric rather than a revenue metric — compressed meaningfully without any change in clinical throughput or headcount.

None of these gains required additional staff. The coding department operated at the same headcount as before. What changed was rhythm — the same coders, working the same hours, connected differently to the clinical flow. The pattern is replicable across specialties. The cross-sector case in the CAPTURE layer is whether leadership is willing to commit to the rhythm change.

What to do on Monday.

If you are a Chefarzt, CEO, or Geschäftsführer reading this and you suspect your hospital is losing revenue to retrospective coding — but you have not quantified the loss — there is a specific first move that takes about two weeks and costs effectively nothing.

Commission an audit of 50 randomly selected discharge records from the most recent completed quarter. The audit has a specific question: does the clinical documentation, across physician notes, nursing documentation, therapy records, and social work assessments, support a higher-complexity DRG than the one that was coded? If yes, calculate the euro gap per case. Aggregate across the 50 and project across annual discharge volume.

You now have the number. The number justifies everything else.

The second move is organisational — not yet structural. Book a meeting between the relevant department lead (for the Chefarzt of whichever service has the largest complex-treatment volume — intensive care, cardiology, surgery, geriatrics), your Leitung Medizincontrolling, and your Geschäftsführer. Put the euro number in the room. Ask the specific question: what would it take to install a daily 15-minute touchpoint between the coder and the clinical team, within 24 hours of admission?

You will get three answers. One will be about staffing. One will be about physician acceptance. One will be about software. The staffing answer is usually false — the touchpoint doesn’t require additional headcount if the existing coding team is re-scheduled. The physician acceptance answer is genuine and requires a specific conversation about the euro impact and the clinical benefit of fresh documentation feedback. The software answer is a red herring — no existing coding software will solve a rhythm problem.

The third move, only after the first two, is structural. Redesign the coding department’s daily schedule to include the ward touchpoints. Install the documentation prompt loop. Audit after 90 days. Audit again at 6 months. Adjust.

This is not a 5-year transformation. It is a 6-month rhythm change. What makes it rare is not complexity but the willingness of leadership to spend political capital on a revenue leak that has been quietly acceptable for years.

Coding at discharge is already too late. The clinical evidence has gone cold. The documentation cannot be reshaped. The complexity that was clinically present will not be coded, and nobody inside the organisation will feel the loss. But the loss is real, it is cumulative, and at any hospital with meaningful volume of complex-treatment-eligible inpatient cases — ICU, cardiac, surgical, oncological, geriatric — it is almost certainly in the mid-to-high six figures per year, and for larger hospitals, well into seven.

The 24-hour loop is how you stop it.