The 17-Minute Round: Why Departments Should Measure Physician Time Before Buying Tools

Tuesday, 8:50 a.m., an internal medicine ward. In the workflow plan someone wrote for the quality manual years ago, the chief physician's round is listed at 17 minutes per room. Nobody remembers who set the number or what it was based on. On good days it is roughly right. On the others, a second, silent shift runs behind the round: the coding specialist's query, the phone call with the family, the discharge letter that has been waiting for sign-off since Friday.

Every department carries numbers like these. The attending would say she spends an hour a day on letters. The chief physician estimates his residents' administrative share at a third. Management has seen a vendor slide that spoke of four hours. None of them has measured.

That would hardly matter as long as nothing depended on it. But precisely these numbers become the basis for decisions as soon as documentation tools, staffing plans or digitization budgets are negotiated. Which is why it is worth looking closely at where the familiar figures come from — and how a department arrives at its own.

What the surveys show — and what they are

The MB-Monitor 2024 — the Marburger Bund physician survey, in this edition covering 9,649 employed physicians in the autumn of 2024 — arrives at an average of three hours per day for administrative and documentation tasks; 49 percent of respondents frequently feel overburdened, and 65 percent are dissatisfied with their hospital's IT [1]. It is a membership survey based on self-reports, conducted by an interest group — that belongs in the framing. What is striking is the stability: the MB-Monitor 2022 also arrived at three hours.

The Deutsches Krankenhausinstitut (DKI), a German hospital research institute, surveyed the hospitals themselves in the summer of 2024 — 225 general hospitals and 98 psychiatric clinics [2]. The result: physicians and nurses spend just under three hours per day on documentation and reporting duties. Arithmetically, that ties up around 59,500 of Germany's 165,200 full-time physician positions — roughly 36 percent of medical capacity, converted, occupied with nothing but paperwork.

The two surveys support each other. Except that the DKI figure is also an estimate — this time by hospital administrations about their staff. What happens when estimate and measurement are placed side by side is shown by an older study.

Estimating and measuring are two different things

In 2015, HIMSS Europe published a study of German acute-care hospitals, commissioned by Nuance, a vendor of speech recognition — the conflict of interest belongs in the first sentence [3]. The roughly 120 physicians surveyed estimated their documentation share at a mean of 35 percent of the working day. The time actually recorded was 44 percent, about four hours daily; for chief physicians it was five and a half. The study is small, ten years old and vendor-funded — of little use for absolute claims. For one claim it serves well: nine percentage points lay between self-estimate and measurement, and the estimate was the lower of the two.

The methodological reference for honest time measurement is a US time-motion study from 2016 [4]: trained observers accompanied 57 ambulatory physicians for 430 hours. 27 percent of practice time went to direct patient contact, 49 percent to screen and desk work; on top came one to two hours of record work in the evening. The setting is the American outpatient clinic; the numbers do not transfer to a German ward. The method does: observation instead of recollection.

What time measurements capture — and what they don't

Anyone planning their own measurement should know the pitfalls that published studies also hang on:

• Parallel work. A physician who documents the progress note while on the phone with the family doctor is doing both at once. Every measurement needs a rule for which activity counts in that case — and this rule shifts the result by percentage points.
• The observation effect. People who are shadowed work with more focus and skip the detours. Short, repeated measurement windows dampen the effect; it never disappears entirely.
• Category boundaries. Does the coding query count as documentation? Placing orders in the hospital information system? Preparing the tumor board? Depending on where the line is drawn, the same working day produces very different charts.
• The dynamics. Documentation duties grow faster than category catalogs age. Patient classification under the PPBV — Germany's nurse staffing assessment regulation — alone ties up an average of 9.3 full-time equivalents per hospital and year according to a DKI survey, extrapolated to around 9,800 full-time equivalents nationwide. Mostly in nursing, but an object lesson in how quickly new tasks emerge that appear in no old category [5].
• The role question. Residents, attendings and chief physicians have different documentation profiles. An average across all of them hides exactly the places where something could be changed.

How a department measures its own distribution

1. Define categories before measuring. Six to eight are enough: direct patient contact, documentation, communication, walking and waiting, training and education, other. Agree a written rule for parallel work — for instance: the activity that could not be interrupted is the one that counts.
2. Choose an ordinary week. Not the week before the audit and not the holiday season. Several weekdays, all medical roles, on-call hours included.
3. Choose the method. External observation yields the most honest data but costs one observer per person observed. Structured self-logs at fixed intervals are the practicable middle ground. The retrospective estimate on Friday afternoon is not a measurement.
4. Involve the Betriebsrat — the works council — and the team before the first measurement day. Processes are being measured, not people; results are analyzed only in aggregate. Anyone who explains this only afterwards will not get data a second time.
5. Document the baseline — with method, measurement week and category rules — and re-measure with the same method after every change. Otherwise every improvement remains an anecdote.

Three questions before you buy

Vendors like to argue with the big numbers — the three hours, the 36 percent. The robust evidence on AI-assisted document generation, by contrast, speaks of minutes per encounter; we have compiled it in our article on the physician letter pipeline (in German). Both can be true: the total burden is large, the effect of a single tool more modest than the brochure. What mediates between the two is your own baseline.

1. Which study does the time saving in the offer rest on — and is it based on observation or on self-report?
2. Does the vendor support a before-and-after measurement with your method and your categories?
3. What applies if your baseline deviates from the assumption in the offer?

How a selection process can be structured beyond that is described in our comparison of digital tools for hospitals (in German).

An honest measurement has a side effect that outlasts any procurement: the team sees that the department counts in the currency that matters — time at the bedside. If you would like to follow the methods and results of such measurements: Visite, our weekly briefing on documentation and AI in German healthcare (German; English edition Grand Rounds is in preparation).

Sources

1. Marburger Bund. MB-Monitor 2024: Hohe Belastung, unzureichende Rahmenbedingungen (Befragung von 9.649 angestellten Ärzt:innen, 27.09.–27.10.2024). https://www.marburger-bund.de/bundesverband/themen/marburger-bund-umfragen/mb-monitor-2024-hohe-belastung-unzureichende
2. Deutsches Ärzteblatt (aerztestellen.aerzteblatt.de). DKI-Umfrage: Jeden Tag drei Stunden Bürokratie. 01.10.2024. https://aerztestellen.aerzteblatt.de/de/redaktion/dki-umfrage-jeden-tag-drei-stunden-buerokratie
3. Deutsches Ärzteblatt. Klinikärzte verbringen 44 Prozent ihrer Zeit mit Dokumentation (Bericht über die HIMSS-Europe-Erhebung im Auftrag von Nuance). 24.03.2015. https://www.aerzteblatt.de/news/klinikaerzte-verbringen-44-prozent-ihrer-zeit-mit-dokumentation-fae3bd56-39ce-4977-977b-9b288e69695f
4. Sinsky C, Colligan L, Li L, et al. Allocation of Physician Time in Ambulatory Practice: A Time and Motion Study in 4 Specialties. Annals of Internal Medicine. 2016;165(11):753–760. doi:10.7326/M16-0961
5. Deutsches Krankenhausinstitut. Krankenhaus Barometer 2025 (Kapitel 2, Aufwand der Patienteneinstufung nach PPBV). https://www.dkgev.de/fileadmin/default/Mediapool/1_DKG/1.7_Presse/1.7.1_Pressemitteilungen/2025/2025-12-29_Anlage_DKI-Krankenhaus-Barometer.pdf