An
interactive swimmer plot is a longitudinal visualization that displays the
journey of multiple subjects across a common time axis. Each horizontal lane is
one subject; its length shows duration within the observation period; markers
show events exactly where they occur.
This lets
users read duration, sequence, timing, and frequency of events across all
subjects at once, then drop into any lane for detail. Filtering
by site or cohort/arm narrows the view to a specific site
or treatment group, without losing the shared time axis.
A
closer look: tumor response over time
Each lane
is a subject. The grey bar shows study
duration. Triangles mark tumor assessments, colored by visit
response: dark green (CR), light green (PR), yellow (SD), red (PD)) — the color
sequence traces the subject's disease trajectory. The inward-pointing
triangle (◄) marks study discontinuation, showing when a subject exited the
study.
Tooltips carry
supporting details on hover — assessment day, sum of diameters, change from
baseline, target/non-target response, new lesion status, and visit response —
keeping the chart uncluttered.
The lane
can carry other events too. A baseline diagnosis marker at or before
day 0 could surface tumor type, stage, and date of diagnosis on hover — adding
disease context without extra columns.
Together,
these elements let a reviewer answer at a glance what would otherwise need
cross-referencing tables: who had durable responses, who progressed early,
whether discontinuations cluster, and how response relates to time on study.
Other
clinical research scenarios
The same
pattern applies across other study needs:
- Adverse events — lanes
show time on treatment; markers show AE onset, severity, and resolution
- Dosing and exposure —
lanes show treatment duration; markers show dose changes or interruptions
- Visit compliance — lanes
show enrollment duration; markers show completed, missed, or delayed
visits
Why it
works
The
swimmer plot combines subject-level detail, sequencing, event comparison,
duration, and pattern recognition in one view. Scanning many lanes together
makes it easy to spot long versus short durations, early versus late events,
event clusters, gaps in activity, and outliers — valuable whenever when something
happened matters as much as what happened. Interactivity
(filtering, highlighting, tooltips) lets reviewers zoom from all subjects to
one, without losing the timeline as context.