Documentation Gaps in Court: the Presumption of §630h BGB
What is not in the record is presumed in court not to have happened. How §630h BGB, the German burden-of-proof statute for treatment errors, shifts the burden of proof — and what departments can do about it structurally.

Dr. Sven Jungmann
CEO

The brief comes from the regional court; the treatment lies six years back. The allegation: after the procedure, the required close monitoring did not take place, hence the injury. In the department, everyone is certain that monitoring happened — that is how it is always done on this ward. In the chart, on the subject: no entry.
The attending from back then works at another hospital, the resident abroad. At this point the case changes direction — not because of the medicine, but because of a rule of presumption.
The presumption rule of §630h (3)
If the treating party has failed to record a medically required material measure and its result in the patient record, contrary to §630f — the German Civil Code's provision on treatment documentation — then "it is presumed that he did not take this measure" (§630h (3) BGB) [1]. The BGH, Germany's Federal Court of Justice, applies this line in settled case law: the absence of documentation of a measure subject to recording indicates that the measure was omitted (among others, judgment of 11 November 2014 – VI ZR 76/13) [2].
The rule is more precise than its reputation makes it.
- It applies to material measures. Subject to recording is what is, from a professional point of view, material to the current and future treatment (§630f (2)) [3]. Documentation serves the treatment — what is medically irrelevant need not be written anywhere for the litigation either.
- It is rebuttable. Counter-evidence remains possible: witnesses, examination of the parties, the rest of the documentation [4]. Only, six years later the witnesses are scattered and the memories generic; "that's how we always do it" rarely convinces courts. The most reliable witness that remains is the record itself.
- It is not an independent ground of liability. The gap does not by itself establish liability; it shifts the burden of proof [4]. That, however, is often enough to lose a case that could have been won on the medicine.
The asymmetry that follows is the real point for department leadership: documentation happens in the present, under time pressure, for the treatment. The record is read years later, at leisure, by people looking for gaps. A documentation culture that keeps this second reader in mind comes about through standards and working conditions — hardly ever through appeals to individual diligence.
The electronic record itself takes the witness stand
In 2021, the BGH extended the rule into the digital world: electronic documentation that, contrary to §630f (1), does not make subsequent changes recognizable carries no positive evidentiary weight that the documented measure actually took place (judgment of 27 April 2021 – VI ZR 84/19) [5]. The evidentiary question thus has two levels. If the entry is missing, the presumption works against the hospital. If the entry exists, it helps only if the software keeps changes visible. What §630f demands of digital systems in detail is covered in the article "§630f BGB in practice" (in German).
Where it gets expensive: the bridge to causation
§630h contains a second, heavier burden-of-proof figure: in the case of a gross treatment error that is fundamentally capable of causing the injury that occurred, causation is presumed in the patient's favor — and the same applies where a required finding was not obtained (paragraph 5 sentence 2) [1]. The dangerous combination in litigation: a documentation gap leads to the presumption that a finding was never obtained at all; from there the path to the causation presumption is short. The BGH recently marked the boundary: the reversal of the burden of proof under paragraph 5 sentence 2 presupposes an established error in obtaining findings (judgment of 4 June 2024 – VI ZR 108/23) [6]. For the department the sober lesson remains: the unremarkable finding belongs in the record too — it, especially, proves that it was obtained.
A third front runs alongside: for consent and proper informed-consent discussion, the treating party bears the burden of proof from the outset (§630h (2)) [1] — here no documentation gap is needed for the hospital to carry the burden. Consent documentation is thus structurally just as litigation-relevant as treatment documentation, and in experience it is the one with the bigger gaps.
What a department can ensure structurally
The presumption rule punishes randomness. Four things can be organized:
- Define what counts as a material measure on each ward. Postoperative checks, consent discussions, critical handovers, findings obtained including unremarkable results — as a department-wide standard, rather than leaving it to the individual diligence of particular shifts.
- Make promptness possible. Whoever can only document at the end of a twelve-hour shift documents with gaps; the "direct temporal connection" of §630f begins with the duty roster and the working tools.
- Know the change-proofness of your own systems. VI ZR 84/19 supplies the concrete test question for every hospital information system configuration and every add-on tool: does the original content remain recognizable, along with the time of change?
- Spot checks with the right question: would this record tell a third party, six years from now, what we knew when and what we did when? That question tests something different from the completeness of form fields — and it is the same one an expert witness will ask later. How structured, referenced data works in expert assessment is shown in the article on medical expert assessment (in German).
Whoever goes through these four points once a quarter has addressed the largest part of the evidentiary risk with onboard resources — before any software decision. And whoever goes through them with their own attendings, rather than sending them out as a directive, has a good chance they will still be lived in six years.
If you want to play through the question "What did we know when?" against your own documentation, write to us. Ongoing analysis of documentation, law and clinical AI comes in our weekly briefing Visite (German; English edition Grand Rounds is in preparation).
Sources
- §630h BGB — Beweislast bei Haftung für Behandlungs- und Aufklärungsfehler. https://www.gesetze-im-internet.de/bgb/__630h.html
- BGH, Urteil vom 11.11.2014 – VI ZR 76/13. https://dejure.org/dienste/vernetzung/rechtsprechung?Gericht=BGH&Datum=11.11.2014&Aktenzeichen=VI+ZR+76/13
- §630f BGB — Dokumentation der Behandlung. https://www.gesetze-im-internet.de/bgb/__630f.html
- Die unvollständige Behandlungsdokumentation. Fachbeitrag, Rechtslupe. https://www.rechtslupe.de/zivilrecht/die-unvollstaendige-behandlungsdokumentation-389828
- BGH, Urteil vom 27.04.2021 – VI ZR 84/19. https://dejure.org/dienste/vernetzung/rechtsprechung?Gericht=BGH&Datum=27.04.2021&Aktenzeichen=VI+ZR+84/19
- BGH, Urteil vom 04.06.2024 – VI ZR 108/23. Nachweis über https://dejure.org/gesetze/BGB/630h.html
Sources accessed in July 2026.
Editorial analysis, not legal advice.


