The Department in Five Years: What Remains of the Administrative Work
Think a hospital department's administrative work five years ahead and it falls into three piles: one disappears, one moves, one stays — and should stay. A sorting exercise for chief physicians.

Dr. Sven Jungmann
CEO

The job interview has been running for twenty minutes when the candidate asks the question she evidently came with: "How much of my time will I spend on documentation here?" The chief physician has two answers to choose from, a promotional one and an honest one. She takes the honest one: "At the moment, about a third."
The number is no exaggeration. In the MB-Monitor 2024, a membership survey by the Marburger Bund — the German association of employed physicians — among 9,649 employed physicians, the share for administration and documentation stood at around three hours per working day; 28 percent of the respondents were thinking about changing careers or leaving the profession [1]. That is self-report from an association survey, but it fits what time measurements find elsewhere: a US observational study covering 430 hours of ambulatory practice arrived at 27 percent direct patient time against 49 percent screen and desk work [2]. The US numbers come from outpatient practices and transfer only to a limited degree; the order of magnitude still surprises nobody who runs a ward.
More interesting than the inventory is a sorting exercise. Take a department's administrative work and think five years ahead, and it falls into three piles: one that disappears, one that moves, and one that stays — and should stay.
The pile that disappears
The first pile is transfer work: typing up the prior report that arrived as a fax. Gathering the lab values from three systems. Entering the same diagnosis again into the fourth form. This work has one property in common: the information already exists; it merely sits in the wrong place in the wrong format. There is no medical reason why a licensed physician must move it.
Honesty includes the pace. The methodologically most robust study of AI documentation tools — a randomized comparison across 238 physicians from 14 specialties — found a saving of a good 40 seconds of writing time per note [3]. The honest unit for today's tools is the minute. So the pile disappears in layers, over years, to the degree that information arrives in structured form where it is needed. But it is the only pile that can disappear completely, because nothing about it is medical.
The pile that moves
The second pile changes time, place and responsibility. The history reconstruction on the day of discharge moves to admission, where the information first appears. The discharge letter, today a trailing task of the last day, begins on the first. Tasks that require no medical qualification move to professions that can be hired for them.
The infrastructure for this migration is currently becoming binding. Since October 1, 2025, hospitals have been obliged to fill the elektronische Patientenakte (ePA — Germany's national electronic patient record) [4]. The European Health Data Space regulation (EHDS) makes discharge reports a priority data category from March 2031, to be available in structured, interoperable form [5]. You can read both as one more obligation. You can also read them as a rebuild: documentation turns from a report trailing the care into a part of the care itself — captured at the moment it comes into being, usable by everyone who comes after.
The pile that stays
The third pile was never administration. It has merely lain under the other two for years and been mistaken for them: the epicrisis — the case summary — as an act of thinking that condenses, weighs and justifies a course. The consent conversation. The assessment that turns twenty findings into a recommendation. The sign-off — the signature with which someone takes responsibility for a content.
This pile will grow rather than shrink. When drafts become cheap, review becomes the actual medical work on the document — why verification is becoming more important than generation (in German) can be shown well from the error research on AI texts. That is not bad news. It is documentation returning to what it once was in training: an instrument of clinical thinking.
A day in the year 2031
The same department, five years later, without a technical revolution — just with cleanly sorted piles. The ward round takes as long as it does today, but the physician looks at the patient, because the record is already right when she enters the room. The M&M conference answers the question "What did we know when?" from the record, in minutes. The preliminary letter is ready on the day of discharge because it was begun at admission. And the candidate from the job interview is by now a specialist — she spent her residency mostly on medicine.
Nothing about this is utopian. Every single scene fails today not for lack of technology, but because the three piles lie unsorted on top of each other — which is why either everything is supposed to be automated, or nothing.
How to tell whether your department is on its way
- Do you know your own distribution? Before any tool decision comes the measurement — how a department surveys its own time distribution (in German) is a one-week project.
- Is information captured where it comes into being — or reconstructed at the end?
- Does work actually disappear, or does it shift into the evening? In a US survey, extensive documentation work at home was associated with nearly doubled odds of burnout (OR 1.9; an association, not causation) [6]. After-hours time is the most honest metric for real relief.
- Is the share of time with patients growing? That is the only metric for which the whole rebuild is worth it.
- Does sign-off remain medical — enforced technically, too?
If you find sorting exercises like this useful: in our weekly briefing Visite (German; English edition Grand Rounds is in preparation), we sort what is actually moving in clinical documentation — and what is merely being announced.
Sources
- Marburger Bund. MB-Monitor 2024: Hohe Belastung, unzureichende Entlastung. Mitgliederbefragung, n = 9.649, Erhebung 27.09.–27.10.2024. https://www.marburger-bund.de/bundesverband/themen/marburger-bund-umfragen/mb-monitor-2024-hohe-belastung-unzureichende
- Sinsky C, Colligan L, Li L, et al. Allocation of Physician Time in Ambulatory Practice: A Time and Motion Study in 4 Specialties. Ann Intern Med. 2016;165(11):753–760. doi:10.7326/M16-0961.
- Lukac PJ, Turner W, Vangala S, et al. Ambient AI Scribes in Clinical Practice: A Randomized Trial. NEJM AI. 2025;2(12):AIoa2501000. doi:10.1056/AIoa2501000.
- Ärzteblatt. Ab Anfang Oktober ePA-Nutzung für Praxen verpflichtend. 2025. https://www.aerzteblatt.de/news/ab-anfang-oktober-epa-nutzung-fur-praxen-verpflichtend-13f0e40a-c05e-4417-9c90-521f93730fb7
- Verordnung (EU) 2025/327 über den europäischen Raum für Gesundheitsdaten (EHDS). Amtsblatt der EU, 5. März 2025. https://eur-lex.europa.eu/legal-content/DE/ALL/?uri=CELEX:32025R0327
- Gardner RL, Cooper E, Haskell J, et al. Physician stress and burnout: the impact of health information technology. J Am Med Inform Assoc. 2019;26(2):106–114. https://academic.oup.com/jamia/article/26/2/106/5230918


