Everyday charting

Charting for a patient on a clinical trial

Your note is now the study's source record. Chart the time on the clock, or explain it to a monitor three weeks from now.

7 min read built on 5 full opinions updated 2026-10-15

Written by a med-surg RN, ten years, day shift. Why there is no name on it · How this guide is made

Denise reacting to this topic
All sections

Your patient is on a study. The research team will copy dose times, draw times and symptoms out of your note and into the study record. Later a monitor checks that record against your chart, line by line. That makes your note the source, so it has to show what happened and exactly when.

Key points

  • Chart the actual time you started the dose, finished it and drew each sample. Never the scheduled time.
  • Tie every timed draw to the dose. Chart when the infusion ended, then when each tube came out.
  • If anything went off schedule (late, missed, held, wrong arm, interrupted), chart what happened, when, why and who you told. Whether it counts as a protocol deviation is the research team's call.
  • New symptom? Chart the onset time, what you measured and who you notified. Whether the study drug caused it is for the investigator to decide.
  • Charting after the fact? Label it a late entry and give both times.

What to write

Dose times

How it often gets charted Better
Study drug given per protocol.Study drug [NAME] 200 mg IV infusion started 0915 L forearm PIV, ended 1022. Pump paused 0940-1002 for occlusion alarm, line flushed, restarted. Total volume infused 250 mL.
Oral study med given AM.Study capsule x2 taken by pt 0807 with water, observed by this RN. Pt fasting since 2200 per pt report.
Dose held.0800 study dose not given. Pt vomiting x2 since 0630. Dr. [NAME] and study coordinator [NAME] notified 0812. Instructed to hold, will reassess 1000.

Draw times

How it often gets charted Better
PK labs drawn.PK sample 1h post-infusion drawn 1105 from R antecubital (opposite arm from infusion). Infusion end time 1022. Sample handed to research lab tech [NAME] 1110.
Trough drawn before dose.Pre-dose trough drawn 0752. Study dose started 0801.
Labs late due to pt in CT.4h PK sample scheduled 1422, drawn 1451. Pt off unit for CT 1405-1445. Coordinator [NAME] notified 1455.

New symptoms

How it often gets charted Better
Pt having reaction to study drug.1035 pt reports itching on chest and arms, onset about 10 min ago per pt. Raised red wheals noted on anterior chest. BP 128/76, HR 96, RR 18, SpO2 98% RA. No SOB, no throat tightness per pt. Dr. [NAME] notified 1038.
Pt c/o HA, tolerated well otherwise.1300 pt reports frontal headache 5/10, onset 1230. Denies vision change. BP 142/88. No prior headache today per pt.
Pt fine, no complaints.1500 pt denies nausea, headache, rash, SOB when asked. Skin intact, no rash on chest or arms.

Look twice at that last pair. The study team often needs to know you looked for a symptom and it wasn't there. Charting that you asked is evidence. No complaints only tells the reader the patient didn't bring anything up.

Template

Timed study dose

[DATE] [TIME OF ENTRY]
Study: [STUDY NAME OR NUMBER]. Pt enrolled per research team.
Study drug [DRUG NAME] [DOSE] [ROUTE] started [START TIME] via [SITE/LINE].
[FOR INFUSION: Rate [RATE]. Ended [END TIME]. Interruptions: [NONE / PAUSED START-STOP TIME, REASON].]
Total given: [AMOUNT OR VOLUME]. Kit/lot number if on label: [NUMBER].
Pre-dose vitals [TIME]: BP [ ], HR [ ], RR [ ], Temp [ ], SpO2 [ ].
Post-dose vitals [TIME]: BP [ ], HR [ ], RR [ ], Temp [ ], SpO2 [ ].
Pt denies [SYMPTOMS ASKED ABOUT] when asked at [TIME].
Administered by [YOUR NAME, CREDENTIALS].

Timed draw, off schedule

[DATE] [TIME OF ENTRY]
Study: [STUDY NAME OR NUMBER].
[SAMPLE NAME, e.g. 2h post-dose PK] scheduled [SCHEDULED TIME], drawn [ACTUAL TIME].
Reference time: study dose ended [DOSE END TIME].
Reason for difference: [WHAT HAPPENED, e.g. pt off unit for imaging from [TIME] to [TIME]].
Drawn from [SITE], [ARM OPPOSITE INFUSION / SAME ARM, REASON].
Sample handed to [NAME, ROLE] at [TIME].
Study coordinator [NAME] notified at [TIME] by [PHONE / PAGE / IN PERSON].
[YOUR NAME, CREDENTIALS]

New symptom in a study patient

[DATE] [TIME OF ENTRY]
Study: [STUDY NAME OR NUMBER]. Last study dose [DATE] [TIME], ended [TIME].
At [TIME] pt reports [SYMPTOM IN PT'S OWN WORDS, NO QUOTATION MARKS].
Onset [TIME] per [PT / OBSERVED BY THIS RN].
Observed: [WHAT YOU SAW, LOCATION, SIZE, SEVERITY SCORE].
Vitals [TIME]: BP [ ], HR [ ], RR [ ], Temp [ ], SpO2 [ ].
Pt denies [RELATED SYMPTOMS ASKED ABOUT].
[DR. NAME] notified at [TIME]. Orders received: [ORDERS / NONE].
Study coordinator or investigator [NAME] notified at [TIME].
Actions taken: [WHAT WAS DONE AND WHEN].
Reassessed at [TIME]: [FINDINGS].
[YOUR NAME, CREDENTIALS]

Late entry

[DATE] [TIME OF THIS ENTRY] Late entry for [DATE] [TIME OF EVENT].
[WHAT HAPPENED, WITH ACTUAL TIMES].
Times taken from [PUMP HISTORY / LAB ACCESSION / OWN NOTES MADE AT THE TIME].
[YOUR NAME, CREDENTIALS]

Words to avoid

Word or phrase Why it hurts Use instead
per protocolThe monitor has nothing to check it againstThe actual times, amounts and site
on schedule, on timeHides the real clock timeThe clock time, e.g. drawn 1105
approximately, around, ~Study windows run in minutesThe exact time. If it's an estimate, say whose: onset about 1230 per pt
reaction, side effect, drug-relatedAssigns cause. Causality is the investigator's callWhat you saw and when
tolerated wellA judgment with no data behind itVitals, plus what the patient denied when asked
WNLNormal by what range? Which measure?The values
late, delayed (alone)No reason, no new timescheduled [TIME], done [TIME], reason [REASON]
done, completedDoesn't say what or whenThe task, the time, who did it

Charting the planned time instead of the real one

The usual mistake is charting the planned time instead of the real one.

Say the protocol wants a PK sample 60 minutes after the infusion ends, with a 10-minute window. The pump alarmed, so the infusion actually ended at 1022 instead of 1000. You drew at 1105. That's 43 minutes after the real end, so it's out of window. Your MAR, though, shows the dose completed at 1000, because the scheduled time auto-filled. And you charted the draw at 1100 because that's when you meant to do it. Now the study record shows a clean on-time draw. It never happened.

The monitor lines up the pump history, the lab accession time and your note, and gets three timestamps that disagree. That becomes a query. Weeks later the coordinator has to come find you, and nobody remembers what the pump did that morning. If nobody can sort the times out, the sample may be unusable. The patient got stuck for nothing.

Pre-charting and copy-forward do the same damage. A note written before the task, or carried over from yesterday, holds the times from the plan. What actually happened isn't in it. On a regular med-surg day, clicking boxes might tell the story well enough. On a study day every box has a clock on it.

Bottom line

On a study patient, chart every dose, draw and symptom at the clock time it really happened. The time on the schedule doesn't go in your note.

Official guidance

  • Protocol deviations cluster around timed assessments and dosing windows. A missed visit or procedure, or a treatment given or delayed outside the window the protocol allows, is something the site documents and reports. Your timed entry is how the team sees it. (University of Utah, CCRPS)
  • Results of protocol-required evaluations and an accurate case history for each enrolled participant go in that participant's source records. Clinic and nursing notes count as source too, along with the study forms. (GWU Clinical Research, HPTN)

In court, 2014–2023

What happened in 5 court cases

None of the cases in our set involves a clinical trial. What they do show is someone reading the chart months or years later to work out the order of events between an exposure and a symptom. That's exactly what the monitor does.

Sajbel v. Secretary of Health and Human Services: United States Court of Federal Claims, 2017. The decision builds its timeline out of timed entries. It records the dose to the minute: "At 32 minutes of age, B.B.T. received the pediatric hepatitis B vaccination, Recombivax." Then it uses the nurses' timed skin assessments (4:21 am, 6:30 am, 6:41 pm) to show when the skin findings changed. The special master found for the petitioner. For you, the point is what the record let the reader do. It put a dose and a new finding on the same clock.

Thompson v. Secretary of Health and Human Services: United States Court of Federal Claims, 2023. The opinion says plainly when a measurement was estimated and never actually taken: "Paramedics did not take J.T.’s temperature but noted that it was normal to the touch." Later an ER record shows the parent said the child had stopped a prescribed anticonvulsant. A study nurse runs into both: a vital sign someone estimated, and a dose the patient didn't take. Both got read closely, years later.

Schultz v. Secretary of Health and Human Services: United States Court of Federal Claims, 2020. This one shows how a wrong interval spreads. The decision points out that later records were "consistently, if erroneously, suggesting that her December 1st stroke onset had occurred one week after receipt of the flu vaccine". The real gap was five to six weeks. Once a wrong interval gets carried forward, every note after it repeats it.

Cottingham v. Secretary of Health and Human Services: United States Court of Federal Claims, 2022. The decision goes visit by visit and notes what each record leaves out: "This record, however, does not indicate that Ms. Cottingham was experiencing near black-outs or low-grade fevers." The special master denied the fee request for lack of reasonable basis. Our lesson is narrower. A symptom the patient mentions that never makes it into your note doesn't exist for whoever reads the chart later.

Tibbs v. Bunnell: Kentucky Supreme Court, 2014. The question here was whether an event report filed through a patient safety system had to be produced in a malpractice suit. Along the way the court quotes the federal statute's carve-out: protected patient safety material "does not include a patient’s medical record, billing and discharge information, or any other original patient or provider record." Put simply, the chart is the original record and it stays one. A deviation form or event report filed somewhere else doesn't replace what you write in the chart.

Court decisions

Published decisions, linked to CourtListener.

All decisions behind this guide

From nurses online

Quoted as written. Opinion, not a source.

Other guides

Advice, not law.