Coding at Discharge Is Already Too Late.

A hospital generates revenue the way a farm generates harvest. The crop is already grown — in beds occupied, in operations performed, in complex-treatment hours delivered, in the clinical complexity that arrives through the emergency department every Monday morning. What happens between the clinical work and the invoice determines how much of that crop reaches the barn. Most German hospitals, measured honestly, leave somewhere between fifteen and forty percent of it in the field.

The entire CAPTURE layer of hospital operations exists to address this. Not as a billing function — billing is downstream. Not as an audit function — audit is after the fact. CAPTURE is the discipline that connects what was clinically done to what is documented, what is documented to what is coded, what is coded to what is invoiced, and what is invoiced to what is actually paid. At every interface, revenue leaks. At every interface, a specific operational rhythm can prevent the leak.

Most hospitals have never named this layer. Coding sits in one department, documentation in another, billing in a third, audit response in a fourth, and the medical directorate in a fifth. They meet quarterly. No one person owns the interfaces between them. Revenue disappears through the gaps and nobody experiences the loss directly.

This is the observation that opens every CAPTURE-layer engagement: your hospital’s biggest revenue problem is probably not pricing, payer mix, or volume. It is the rhythm between clinical work and the documentation that describes it.

The single biggest leak is timing, not accuracy.

When I have asked hospital controllers why their case mix index is lower than peer benchmarks, the most common answer is “our coders need more training” or “we’re waiting on the new DRG grouper update.” Both answers are usually wrong. The coders are usually well-trained; the grouper is usually current. What is broken is almost always the rhythm between the coder and the ward.

Here is the structural problem. In the default German hospital model, coders sit in a central office and receive discharge records two to five days after the patient leaves. By that point, the physicians who managed the case are on different shifts, the nursing team has rotated, the therapy records are filed, and the social worker has moved on to the next admission. The coder opens a file that is now closed [1,2]. What was observed but not written will not be coded. What was written but not structured to support a complex-treatment bundle will not trigger it. The documentation does not support what the clinical complexity justified.

This happens across specialties. In cardiology, heart-failure complex treatment codes depend on documentation quality that degrades rapidly after discharge. In intensive care, OPS 8-980 complex treatment requires team-composition logs and assessment records that are difficult to reconstruct retrospectively. In surgery, the combination of primary procedure, complication codes, and nursing-intensity secondary diagnoses produces DRG assignment variance that can span several thousand euros per case [3]. In geriatrics, the differential between a baseline hip-fracture DRG (I47C at BWR 1.389) and a geriatric-co-managed DRG (I34Z at BWR 2.629) is roughly +€5,450 per case in base reimbursement [3].

None of these cases would be mis-coded if the documentation had been complete. None of them would be complete if the coder had been in the conversation. The rhythm is the fix.

Why the conversation never happens.

A Geschäftsführer asked me last year why his hospital had never implemented concurrent coding despite documented cases across the German Medizincontrolling literature showing sustained CMI improvements on similar case mix. The answer was not technical. It was political, and it took three conversations to surface.

His coding department had been optimised for retrospective throughput. The team processed cases in weekly batches, flagged audit-risk cases for MD review, and closed cases on a four-to-six-day rolling cycle. Moving to concurrent coding would require the coders to change their daily rhythm, establish daily contact with ward teams, and work to a different cadence than the controlling team they reported into. None of that was expensive. All of it was politically difficult.

Physicians were the second obstacle. A coder appearing on ward round to flag a documentation gap was, without preparation, perceived as bureaucratic interference rather than operational support. Installing the bridge required that the Chefarzt and the Leitung Medizincontrolling sit down and agree on the framing together — what the gap is costing, why closing it is in the clinical team’s interest, who has the authority to ask for the specific documentation changes. That conversation had never been had, because the hospital’s organisational culture had never built the routine for it.

The third obstacle was measurement. The hospital had never quantified its retrospective-coding leak. Without the euro number, the business case for the rhythm change was invisible. Without the business case, no senior executive would spend political capital on it. And without someone spending political capital, the status quo persisted indefinitely.

What concurrent coding actually is — in one sentence.

A coder, connected to the ward within 24 hours of admission, who flags documentation gaps while the patient is still present and the clinical evidence is still being generated.

Everything else is implementation detail. The specific mechanism matters less than the consistency — physical presence on morning ward round works, a scheduled 10-minute standing daily call works, a remote documentation review platform with structured feedback loops works. What must not change: within 24 hours of arrival, a coder has looked at the documentation and either confirmed it supports the anticipated coding or flagged specific gaps that need to be closed before discharge.

Three structural elements distinguish hospitals that have installed concurrent coding from hospitals that have not.

First — the coder is procedurally connected to the ward huddle. Not as an observer. As a working member whose professional judgment about documentation adequacy is heard and acted on. The coder flags a gap; the physician acts on it the same day. If the feedback loop takes longer than 24 hours, the rhythm is broken.

Second — the conversation happens while the evidence is still being generated. If the admission narrative for a surgical 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 an intensive-care patient’s documentation will likely fall short of the team-composition threshold for OPS 8-980 complex treatment [4], the structural fix is raised while there is still time to correct it. If a hip-fracture case with geriatric comorbidities is being documented as a simple surgical case rather than a geriatric co-managed case, the feedback reaches the clinician while the patient is still on the ward.

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. The rhythm pays for itself many times over in recovered revenue, but installing it requires that leadership treat CAPTURE as a named operational layer rather than as the administrative function that generates invoices at the back end of clinical work.

What the Schlüchtern operational data shows.

At Main-Kinzig-Kliniken Schlüchtern, the geriatric department installed a concurrent-coding rhythm between 2019 and 2021 as part of broader operational restructuring. 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 the top ten DRG coding patterns against benchmark.

The operational results across the full CuraOS stack, sustained over three consecutive fiscal years under the formal research programme with Prof. Dr. Rainer Sibbel at Frankfurt School [5,6], are in the figures below. Relevant specifically to the CAPTURE layer: complex-treatment coding capture rose substantially across the eligible case mix, case mix index rose in parallel, and time from discharge to billing compressed meaningfully without any change in clinical throughput or coding department headcount.

The operational pattern is replicable across specialties. The research programme formally covers the geriatric case mix; application of the same concurrent-coding rhythm in cardiology, intensive care, surgical, or oncological contexts is a matter of organisational transfer rather than methodological discovery. What is required is the same: the 24-hour coder-clinician touchpoint, installed as rhythm rather than project.

What to do on Monday.

If you are a Geschäftsführer or Chefarzt 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 fifty randomly selected discharge records from the most recent completed quarter, spanning three or four specialties. The audit asks one 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 fifty 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 Chefarzt of whichever service has the largest complex-treatment volume in your hospital (intensive care, cardiology, surgery, geriatrics), your Leitung Medizincontrolling, and the Geschäftsführung. 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 does not 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 six months. Adjust.

This is not a five-year transformation. It is a six-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 compounds into the mid-to-high six figures per year for smaller hospitals and well into seven for larger ones.

The 24-hour loop is how you stop it.