Medical timeline vs table: when each one wins
Short version: a timeline wins when the question is about time itself: how long, how far apart, what overlapped, what came before what. A table wins when you need an exact date, an exact quote or a count. In a 2019 JAMIA pilot, physicians answered medication history questions correctly more often with a timeline on all 5 tasks. For a review file, use both, drawn from the same cited rows.
Limits: Tool facts come from each vendor's own website, read 2026-11-12; we have not used those products. The case on this site is synthetic, with no real patient.
The evidence: a medication timeline tested against a table
Belden and colleagues at the University of Missouri, with Catherine Plaisant of the University of Maryland, designed an open-source medication timeline for ambulatory care and published it in the Journal of the American Medical Informatics Association (Belden et al., JAMIA 2019;26(2):95-105). Their starting point: most EHRs show medication history "only in a data table or clinician notes."
They then ran a small remote pilot. Practicing physicians answered 5 medication questions about the same synthetic patient, shown either as a table or as the timeline. Here is the published result, task by task:
| Task | Correct with table | Correct with timeline | Mean time, table | Mean time, timeline |
|---|---|---|---|---|
| 1. Spot which drug is not current | 50% (n=8) | 64% (n=14) | 44 s | 69 s |
| 2. Check all past prescriptions | 17% (n=12) | 91% (n=11) | 68 s | 50 s |
| 3. How many years a drug was prescribed | 44% (n=9) | 93% (n=14) | 22 s | 16 s |
| 4. New drugs started in a given interval | 50% (n=10) | 62% (n=13) | 65 s | 30 s |
| 5. Dose changes in a given interval | 0% (n=10) | 18% (n=11) | 36 s | 30 s |
Source: Belden et al., JAMIA 2019, results table, read 2026-11-12. The authors note the speed differences did not reach statistical significance and that remote, unsupervised testing may not reflect real use.
3 things stand out. The biggest gains came on tasks about spans: past prescriptions and duration. Task 1, a lookup, was slower on the timeline. And task 5, dose changes, stayed hard in both views, which the authors read as a sign that some tasks need their own design. That's why the duration bar format on this site splits a bar where a dose changes.
The older finding: intervals vs exact dates
The same paper cites an earlier test of Lifelines, a timeline built at the University of Maryland, in which people answered 31 questions about personal histories using either a timeline or a table. The timeline did better for comparing intervals and connecting events across categories. The table did better for finding exact dates.
That split is the most useful rule we found, and it matches what you see on the interactive synthetic case: the 112-day gap is obvious on the chart and invisible in the table, while the date of the second MRI takes 1 glance in the table and a click on the chart.
Task by task: which view to use in a record review
| Question you're answering | Better view | Why |
|---|---|---|
| Was there a gap in care, and how long? | Timeline | Empty space reads faster than subtracting dates |
| Which drugs overlapped? | Timeline (duration bars) | Overlap is visible; in a table you compare start and end across rows |
| Did symptoms change after a procedure? | Timeline with a trend panel | Before and after sit side by side on 1 axis |
| Who treated the patient in a given month? | Timeline (swimlanes by provider) | Parallel providers show as parallel rows |
| What exactly did the note on 2024-11-25 say? | Table | Exact text and page cite in 1 row |
| On what date was the MRI done? | Table | The Lifelines result: tables win on exact dates |
| How many PT visits were there? | Table | Counting rows beats counting marks |
| Which records did I review? | Table | A list is the record; a picture isn't |
| Explaining the case to someone else | Timeline, with the table attached | The picture orients; the table proves |
What this means when you choose a tool
- Don't give up the table for the picture. Every claim in your report has to point to a page. A tool that offers only a graphic, with no cited rows behind it, can't support that.
- Do ask for a drawn view for time questions. Gaps, overlaps and durations are where a table slows you down. If the tool lacks a graphic view, export and draw one.
- Insist that the picture is drawn from the rows. If you fix a date in the table, the mark should move. If the 2 can be edited separately, they'll drift.
- Test on time questions. When you trial a tool, ask it the 5 kinds of question above and time yourself in each view. The checklist has the steps.
Limits of this evidence
The JAMIA pilot was small, remote and about medication histories in primary care, not about full medical-legal files. The Lifelines study wasn't about medicine at all. Neither tested AI-built timelines; both used clean data. In real files, the rows can be wrong before anything is drawn, which is why the checklist starts with the data, not the picture. We found no published study comparing AI-built medical timelines with tables for physician record review. If you know of one, email aj@medrecords.ai and we'll add it.
Questions about timelines and tables
Do physicians read a timeline faster than a table?
For some tasks. In the JAMIA pilot the timeline was faster on 4 of 5 medication tasks, but the speed difference wasn't statistically significant. The clearer gain was in correct answers.
When should I use a table instead of a timeline?
When you need an exact date, an exact quote, a count, or a list of records reviewed. Tables also carry page cites better than charts.
Is there research on AI-generated medical timelines?
We found none comparing AI-built timelines with tables for physician record review as of 2026-11-12. The studies above used clean, hand-built data.