Chart Fabrication: 5 Signature Patterns That Prove the Nurse Didn't Turn Your Client

The core clinical fact of every stage III and stage IV pressure-ulcer case is the same. Unrelieved pressure causes tissue death. If a resident is repositioned on a legitimate q2h schedule, combined with a pressure-redistribution mattress and appropriate nutrition, a stage IV is very nearly impossible in a resident who came in with intact skin.

So when you open the CNA flow sheets and see meticulous repositioning — right side, back, left side, right side, back, left side — for 90 days running, and the resident is dead of sepsis from a sacral wound that reached bone, you don't have a care record. You have a fiction. The question is where the seams are.

Stage IV pressure ulcers do not develop in residents who were actually turned. The chart that says otherwise is the chart that will fall apart on cross.

Here are five patterns we see again and again. Each one is independently provable. Stack two or three together and the facility will not want a jury looking at their records.

Pattern 1: Same-Minute Batch Entries

When a CNA backfills at the end of her shift — because she did not actually turn the resident every two hours, because there was no way she could have given her assignment — she sits at the med cart with the EHR open and fills in the blanks. The timestamps betray her.

On a honest flow sheet, you should see variable intervals clustered loosely around two hours: 06:17, 08:34, 10:12, 12:45. On a fabricated flow sheet you'll see 07:00, 09:00, 11:00, 13:00 — clean hourly pegs. Worse, you'll see multiple separate entries (turn to right, turn to back, turn to left, checked brief, applied barrier cream) all stamped with the exact same clock-minute, because they were all entered in a single session and the EHR assigned the current time to each.

In native-format EHR exports you can also see the entry-created timestamp alongside the event-time timestamp. A turn "performed at 07:00" with an entry-created timestamp of 14:52 is a seven-hour backfill. That's not a contemporaneous record under any nursing standard of practice and it's not a business record under most state evidence rules.

Pattern 2: Impossible Staffing Ratios

Pull the unit's assignment sheets in discovery. Pull the daily census. Do the math.

One CNA on a unit with 15 residents on bed rest, each requiring q2h turns, documented pericare, and documented meal assistance, is asked by the flow sheet to perform roughly:

• 15 residents × 12 turns per 24-hour day = 180 turns
• Plus 15 peri-care cycles per shift
• Plus vital signs, ADL charting, meal assistance, incontinence care, linen changes

If one CNA on a single 12-hour shift logs 94 repositioning events, and the standard clinical time per reposition (including hand-washing, glove change, repositioning, documenting) is four to six minutes, she has charted roughly eight hours of pure repositioning in a twelve-hour shift — on top of everything else the flow sheet says she did. That's the math a jury can do on a napkin.

The cleanest version we've seen: one CNA charted 18 repositioning events on eight different residents inside a single 30-minute block. Physically impossible. The staff nurse signed off on it anyway because the sheet was presented to her at end of shift and she didn't do the math. The CNA testified she "must have charted in batches." Exactly.

Pattern 3: Post-Hoc Signature Drift

This one matters most with paper charts and with hybrid systems where paper CNA sheets are scanned into the EHR at end of shift. When a facility re-creates a chart for production — not typical but not rare — signatures tell on the forger.

Lay them side by side. The same CNA's signature on a February 3 flow sheet and her signature on a May 19 flow sheet should look like the same person writing six months apart. Allow for variation. But when you see:

• Pen pressure visibly different (one uniformly light, one heavy) across the same CNA's signatures on charts from the same week
• Slant going from rightward to vertical to leftward on dates that were supposedly charted in real time
• Ink color variation on a single day's chart (blue, black, blue, black) that doesn't match the pattern of pens available at the nursing station (we've seen discovery where the facility only issued black Bic medium for clinical signatures)

…you're looking at entries made at different times, often months apart, by someone reconstructing what a chart "should have looked like." A forensic document examiner becomes a viable expert at that point. Depending on jurisdiction, you may also have the predicate for a spoliation motion and an adverse-inference instruction.

Pattern 4: The Version-Control Delta

This is the pattern that wins cases and that most plaintiff firms don't use. You get the first production in response to your initial subpoena. Months later you get a supplemental production — maybe after a motion to compel, maybe after a change in the facility's law firm, maybe because a new custodian pulled the file. You treat them as the same document.

Don't. Diff them.

In a case we've seen handled by a plaintiff firm in the Mid-Atlantic, the first production showed six turns on the day before the resident's sacral ulcer was first documented as Stage 2. The second production, pulled fourteen months later by a different custodian, showed eleven turns on that same date — on a single CNA's assignment. Five entries added. The added entries were perfect two-hour intervals filling the gap on the first version. Same chart, different contents, both produced under oath as complete copies.

That is a fabrication, it is provable on the face of the two documents, and it is the moment every other weak piece of the case becomes strong. In some jurisdictions it is its own tort; in all of them it is powerful cross-examination material and a strong basis for a sanctions motion under the analysis in Rosenblatt v. Center for Nursing & Rehabilitation, Inc., 2021 NY Slip Op 50166(U) (Sup Ct, Kings County Feb. 23, 2021), where documentation gaps and inconsistencies were enough to defeat summary judgment under New York Public Health Law § 2801-d.

Practical tip: on every case, preserve the original production (hash it, store it, don't let it get overwritten). On supplemental production, run a diff. If you don't have the technical tooling, a paralegal with Acrobat and patience can do it page by page.

Pattern 5: The Payroll Mismatch

The most underused discovery request in nursing-home litigation is the one for CNA payroll and timekeeping records. Facilities keep these in ADP, Kronos, Paylocity, or similar. They have to, for wage-and-hour compliance. They don't forge them because the IRS and state labor boards are looking at them.

Request them by name. Match the data to the flow sheets.

If a CNA's payroll shows her clocked out at 15:00 and her signature is on a repositioning entry at 17:30 on the same day, one of those records is wrong. If the payroll shows she called out sick on March 14 and the flow sheet shows six of her repositioning entries on that date, one of those records is wrong. If the CNA worked only day shift every scheduled week for a year but the flow sheet has her repositioning the resident at 02:15, one of those records is wrong.

In our experience this is where defense counsel stops returning calls and starts returning settlement offers. The facility can explain a pen-pressure difference. It cannot explain how a CNA repositioned a resident three times on a day the ADP record says she was on a plane to Atlanta.

Since 2021, CMS has required certified facilities to report daily staffing through the Payroll-Based Journal (PBJ) system under 42 CFR § 483.70. PBJ data is public. Pull it. Compare it to the flow sheets. When the PBJ says the facility had 0.3 CNA hours per resident day on a given date and the flow sheets say one CNA performed a complete q2h repositioning protocol on 14 residents, the numbers are not compatible.

How Timeline Builder Exposes This in Minutes

Manually, the five-pattern forensic workup takes a good LNC 10–20 hours on a 90-day chart, because the flow sheets come in as PDF exports with timestamps scattered across hundreds of pages, and the payroll comes in as a CSV. You have to normalize, merge, sort, and cross-reference.

Our Timeline Builder ingests both streams and normalizes them to a single chronological view. It automatically flags:

• Same-minute batch entries — any cluster of five or more events within a single clock-minute on one CNA's signature
• Impossible-ratio blocks — any CNA charting more than 30 discrete care events in a 30-minute window
• Entry-created vs. event-time lag — any record where the entry was created more than 90 minutes after the supposed event time (a tunable threshold)
• Payroll-flow mismatch — any flow-sheet entry on a date or time when the CNA's payroll shows she was not clocked in
• Production-delta diff — upload two productions, see every added, removed, or modified entry side by side

The output is a timeline with red flags embedded. You send it to your LNC for confirmation. You show it to opposing counsel at mediation. You show it to the jury as a demonstrative.

The Deposition

The mathematical-impossibility cross is the one that breaks the DON on the record. There is no rehabilitation on redirect.

Q: Director, let's look at Exhibit 47, the CNA flow sheets for March 14. Please identify the CNA by initials whose signature appears in the 14:00 column.
A: T.W.
Q: Who is T.W.?
A: Tanya Williams.
Q: On March 14, how many residents on Unit 3 were on a q2h repositioning protocol?
A: I'd have to look — fifteen.
Q: And Ms. Williams was the only CNA on Unit 3 for the day shift on March 14?
A: Based on the assignment sheet, yes.
Q: Now please look at the 14:00 column on all fifteen of those flow sheets. Each of them shows a turn performed by Ms. Williams at 14:00. Correct?
A: That's what the records show.
Q: A proper repositioning of a bed-bound resident takes at minimum four minutes, correct, per the facility's own care-plan standard?
A: Yes.
Q: Fifteen residents at four minutes each is sixty minutes of repositioning work. Correct?
A: Yes.
Q: Then how did Ms. Williams complete fifteen separate repositions all at 14:00?
A: She must have batch-charted.
Q: "Batch-charted" means charted care she did not perform at the time the chart says she performed it. Correct?
A: It means — it means she entered them together.
Q: So the chart is not a contemporaneous record of care?
A: Not exactly contemporaneous.
Q: Director, Ms. Williams's payroll for March 14 shows she clocked out at 14:22. Can a 14:00 entry for fifteen residents — sixty minutes of repositioning work — be performed by a CNA who clocked out at 14:22?
A: (pause)
Q: I'll take that as a no.

This is the moment. The DON just admitted on the record that the facility's repositioning documentation is not contemporaneous, not reliable, and inconsistent with the facility's own payroll records. Every defense her expert is going to offer about care planning, repositioning schedules, and "unavoidable" ulcer development starts from a set of records the DON has now conceded is fabricated. Your expert doesn't have to do much on direct. The record did it for her.

A Note on Ethics and Sanctions

Chart fabrication is not a rounding error. It is the creation of false business records, it interferes with the administration of justice, and in most jurisdictions it is a sanctionable offense independent of the underlying negligence claim. In the cases we've seen handled where fabrication was proven on the face of the records, courts have imposed adverse-inference instructions, struck expert opinions that relied on the fabricated records, and allowed punitive damages on the basis of post-event concealment rather than the underlying care. One California settlement we've seen reported in the trade press reached $4.2 million after falsified care logs were exposed; another case in the Southeast produced a substantial verdict after the plaintiff's Timeline Builder output was shown to the jury on closing.

Further Reading

• CMS Payroll-Based Journal (PBJ) — 42 CFR § 483.70
• 42 CFR Part 483 — Requirements for Long-Term Care Facilities
• Rosenblatt v. Center for Nursing & Rehabilitation, Inc. (2021)
• Sweeney Law Firm — Nursing Home Documentation Fabrication Patterns

About the author: John Mahoney is the founder of MedLegal AI, which builds AI tools for plaintiff attorneys and legal nurse consultants. This article reflects patterns we see across hundreds of nursing-home charts processed through our Timeline Builder and Records Analyzer. It is not legal advice. For case-specific questions, consult a licensed attorney in your jurisdiction.