Bedsore lawsuit records: how pressure injury cases are proven and defended

A bedsore, a pressure sore, a decubitus ulcer, a pressure ulcer and a pressure injury are the same thing for legal purposes when the main cause is pressure. CMS accepts all of those terms. The National Pressure Injury Advisory Panel (NPIAP) calls every stage a pressure injury. CMS, for the MDS, calls Stage 1 a pressure injury and Stages 2, 3 and 4 pressure ulcers. A charted "Stage 2 pressure injury" is coded as a Stage 2 pressure ulcer, and that is not an inconsistency.

NPIAP defines the injury as localized damage to skin and underlying soft tissue, usually over a bony prominence or under a medical device, caused by "intense pressure, prolonged pressure or pressure in combination with shear." Its 2016 definitions add that tissue tolerance can be affected by microclimate, nutrition, perfusion, comorbidities and the condition of the soft tissue. Every causation fight starts there: plaintiffs point to pressure and time, defendants to perfusion and comorbidities.

Stage 1

Intact skin with non-blanchable redness. It can look different in darkly pigmented skin. Purple or maroon discoloration is not Stage 1; it may be a deep tissue pressure injury.

Stage 2

Partial-thickness skin loss with exposed dermis: a shallow pink or red open wound, or an intact or ruptured serum-filled blister. No fat, no slough, no eschar. NPIAP notes these "commonly result from adverse microclimate and shear."

Stage 3

Full-thickness skin loss. Fat is visible; slough or eschar may be present, and there may be undermining and tunneling. No fascia, muscle, tendon, ligament, cartilage or bone.

Stage 4

Full-thickness skin and tissue loss with exposed or directly palpable fascia, muscle, tendon, ligament, cartilage or bone.

Unstageable

Full-thickness loss where slough or eschar hides the depth. Stable eschar on the heel or an ischemic limb "should not be softened or removed," per NPIAP.

Deep tissue pressure injury (DTPI)

Intact or non-intact skin with persistent non-blanchable deep red, maroon or purple discoloration, or a blood-filled blister. NPIAP: "Pain and temperature change often precede skin color changes." The MDS codes a DTPI with intact skin as unstageable, never as Stage 1.

Medical device related

An injury in the shape of a device used for diagnosis or treatment, staged with the same system.

Mucosal membrane pressure injury

Found on a mucous membrane where a device was in place. NPIAP: "These ulcers cannot be staged." They are not coded in MDS Section M.

MASD

Moisture-associated skin damage, including incontinence-associated dermatitis. CMS excludes it from Stage 2, along with intertriginous dermatitis, adhesive injuries and traumatic wounds.

MDS Section M

The skin conditions section of the Minimum Data Set, the federal resident assessment. It records risk, ulcers by stage, present-on-admission status and recent treatments.

Stage drift, and why the stage never goes back down

Stage drift is when the same wound gets different stages from different staff in the same week, or moves in a direction a wound cannot move. The MDS rule is blunt.

"Do not reverse or back stage."

The reason is physiology. CMS's F686 guidance says pressure ulcers "do not heal in a reverse sequence." A Stage 4 ulcer does not regrow muscle and become a Stage 3; it fills with granulation tissue and scar and stays a healing Stage 4 until it closes. The RAI Manual says an ulcer that was ever at a higher stage keeps that stage until healed, unless it becomes unstageable, and a healed ulcer is documented at its highest stage. Even after closure, the skin has about 80% of normal tensile strength, which is one reason healed Stage 3 and 4 sites break down again.

Our opinion: stage drift is the single most useful signal in a pressure injury chart, for either side. For the plaintiff, Stage 2 from the floor nurse and Stage 4 from the wound consultant 3 days later suggests nobody on the floor was looking. For the defense, the same drift can show a DTPI declaring itself, which is what DTPIs do.

State negligence law decides the lawsuit, but the federal requirement for certified nursing homes, and the CMS guidance surveyors apply, frames how both sides' experts talk about the standard of care.

"A resident receives care, consistent with professional standards of practice, to prevent pressure ulcers and does not develop pressure ulcers unless the individual's clinical condition demonstrates that they were unavoidable."

The second half of the rule, 483.25(b)(1)(ii), covers residents who already have ulcers: they must receive necessary treatment and services to promote healing, prevent infection and prevent new ulcers from developing. A facility can be cited for a wound it did not cause.

The word "unless" puts the burden where lawyers notice it. CMS defines a pressure injury as avoidable if the facility failed to do 1 or more of 4 things: evaluate the resident's condition and risk factors; define and implement interventions consistent with the resident's needs, goals and professional standards; monitor and evaluate those interventions; or revise them as appropriate. It is unavoidable only if the facility did all 4 and the wound developed anyway.

That list is the spine of every pressure injury case. Each step maps to specific documents, and each can fail on paper when the care was reasonable, or look fine on paper when it was not.

2 more lines from the CMS guidance shape how both sides should read the chart. First, being at risk "does not by itself indicate that development of a PU/PI was unavoidable." Second, CMS tells surveyors that "to cite F686, it is not necessary to prove that a PU/PI developed." A facility can fail the standard with intact skin, if the assessments and interventions were missing.

For defense counsel, an unavoidability argument without all 4 elements documented is an argument without evidence. The strongest defense files show each step, date by date, and then a clinical condition, such as multi-organ failure, that explains why the wound developed anyway.

The Braden Scale is the risk tool most nursing homes use, and the RAI Manual names it as a common example. Nancy Bergstrom, Barbara Braden and colleagues reported its reliability and predictive validity in Nursing Research in 1987. It has 6 subscales. 5 are scored 1 to 4 and friction and shear is scored 1 to 3, so the total runs from 6 to 23. Lower is worse.

What cut-off means "at risk"? The 1987 validation used 16: "Using a cut-off point of 16, sensitivity was 100% in both studies. Specificity ranged from 64% to 90%." Facilities set their own thresholds in policy. Ask for the facility's policy and the version of the tool it used rather than assume a number. Inter-rater reliability in that original work was high for RNs (r = .99) and lower for aides and LPNs (.83 to .94), which is a fair point for either side when a score looks wrong and the scorer was not an RN.

Timing is where Braden charts fail most often. CMS does not require a specific tool other than the MDS itself, but its F686 guidance describes the common practice: many clinicians use a standardized tool "upon admission, weekly" for the first 4 weeks after admission, "then quarterly or whenever there is a change in the resident's condition." Those first 4 weeks are when the danger peaks.

CMS also says the clinician may judge risk higher than the tool score and should document why. Read the free-text nursing notes next to every score: they can show risk that was ignored, or risk the nurse caught and acted on.

The risk factors CMS lists are the checklist for the causation debate: impaired mobility; diabetes, end-stage renal disease and thyroid disease; steroid use; impaired blood flow; refusal of care; cognitive impairment; incontinence; undernutrition and dehydration; and a previously healed pressure injury.

A pressure injury case lives in 3 places: the hospital stay before admission, the nursing home stay, and the hospital or hospice care after the wound was found. First productions often leave out the flowsheets where the answers sit, so ask for each piece by name. Our full method for the nursing home side is in the pillar guide on nursing home records, the MDS and CMS-2567 survey reports; this chapter covers what is specific to skin.

2 federal rules help with timing. A resident or their representative can inspect records within 24 hours of an oral or written request, excluding weekends and holidays, and obtain copies on 2 working days' notice, under 42 CFR 483.10(g)(2). Facilities must keep medical records for the period state law requires, or 5 years from discharge when state law is silent, under 42 CFR 483.70. For the process on the hospital side, including HIPAA timelines, see how to get medical records for a lawsuit.

Ask for electronic records in native or export format where you can. A PDF print of a CNA flowsheet often hides the entry time and shows only the care time. If late entries become an issue, the EHR's own access history is a separate request with its own rules; our guide to the EHR audit trail in litigation covers it.

[DATE]

[FACILITY NAME]
Attn: Medical Records / Health Information Management
[ADDRESS]

Re: [RESIDENT INITIALS AND DATE OF BIRTH]; stay from [ADMISSION DATE] to [DISCHARGE OR DEATH DATE]

Please produce the complete medical record for the stay above, in electronic native or export format where the record is kept electronically, including:

1. Admission packet received from [SENDING HOSPITAL], including transfer forms and discharge summary.
2. Admission nursing assessment and all skin assessments, with date and time of entry.
3. All pressure injury risk assessments and the policy that sets the risk threshold.
4. Baseline care plan, comprehensive care plan, and every revision, with revision dates and authors.
5. All MDS assessments, including Section M, and the dates each was completed and transmitted.
6. CNA point-of-care or ADL flowsheets, including turning, incontinence care and meal intake, showing time of care and time of entry.
7. Provider orders and progress notes, treatment and medication administration records.
8. Wound assessments, wound measurement records, wound consultant reports and recommendations.
9. Wound photographs with capture date and time, and the facility's wound photography protocol.
10. Dietitian assessments, weight records, intake records and laboratory results.
11. Orders, delivery records and rental invoices for pressure redistribution mattresses, overlays and cushions.
12. Nursing notes, change of condition notes, and transfer forms sent to [RECEIVING HOSPITAL].

Please identify any item withheld and the basis.

[NAME]
[FIRM OR ROLE]
[CONTACT]

"Turn every 2 hours" is not a CMS rule. Its F686 guidance says frequency is individualized, then gives 2 reference points. For dependent residents who are sitting, or in a bed or recliner with the head raised 30 degrees or more, "many clinicians recommend a position change 'off-loading' hourly." For a reclining dependent resident, the care plan "may include repositioning at least every 2 hours or more frequently depending upon the resident's condition and specific needs." CMS adds that a "microshift," a small change for a short time, "may not be adequate."

Raising the head of the bed to 30 degrees or more creates pressure comparable to sitting. A support surface that "bottoms out," leaving less than 1 inch between the resident and the support material, is unlikely to reduce pressure. Donut-type cushions are not recommended. For residents who already have ulcers on the sacrum, coccyx or ischia, CMS says they should limit sitting to 3 times a day, in periods of 60 minutes or less.

The MDS asks a separate question. Item M1200C records whether the resident was on a turning or repositioning program in the last 7 days. The RAI Manual says such a program must be "organized, planned, documented, monitored, and evaluated," and should name the intervention and its frequency. A ticked MDS box is a claim the flowsheet should support, not proof of turning.

Here is how we read a turning record, whichever side we are on:

1. Find the care plan frequency. Measure against the plan, not a rule of thumb.
2. Count the gaps. Longest interval without a documented position change, per shift. Mark every gap longer than the plan allowed.
3. Compare care time with entry time. If the system records both, ask for both. Clusters of entries within minutes of each other at shift end are late documentation.
4. Look for patterns that repeat too well. Identical entries for weeks, or turns charted while the resident was at dialysis or in the hospital, suggest charting by routine. See charting by exception.
5. Check seated hours. Residents get turned in bed and then left in a wheelchair.
6. Check refusals. A documented refusal is a real defense point if it is followed by education, a care plan update and a retry. A refusal code with nothing after it is not.

A fair warning for plaintiffs: an empty flowsheet cell does not prove a resident was left in a single position. What moves a case is a pattern: many gaps, clustered by shift, lining up with the first skin change. For defense: "we turned him but did not chart it" is a hard sell against weeks of 6 hour gaps.

Once a wound exists, CMS's F686 guidance says that with each dressing change, or at least weekly, and more often when the wound changes, the facility should document the date and at minimum: location and stage; size (perpendicular length and width), depth, undermining and tunneling; exudate; pain; the wound bed and tissue type; and the edges and surrounding tissue. When a wound is present, CMS expects daily monitoring of the site and dressing. Photographs "may be used to support this documentation, if the facility has developed a protocol."

Build the measurement series before anything else

We put every measurement into 1 table, 1 row per assessment, and read down the columns: size and depth over time, the stage at each entry, and who measured.

That ledger is hypothetical, but every problem in it is common. The Day 22 entry matters most. It is either a different wound, a measurement error, or a note written without looking. Defense counsel should find it first: an unexplained "improving" entry 4 days before a Stage 4 hurts every other note.

The TAR is the other half. Match each wound order to its TAR boxes, and look for orders that run unchanged through weeks of no healing. CMS says a wound that "fails to show some evidence of progress toward healing within 2-4 weeks" calls for reassessment of the wound and the resident, and a decision whether to continue or modify the plan. An unchanged order through a growing wound is the clearest revise failure there is. A changed order, with a consultant note explaining why, is the clearest defense exhibit.

Infection is the complication that turns a wound case into a wrongful death case. CMS lists cellulitis, osteomyelitis, septic arthritis, abscess, bacteremia and sepsis, and warns against culturing pus, slough or necrotic tissue. The nursing home wrongful death guide covers what comes after.

The first factual question in almost every case is where the wound started. The answer usually sits in 2 documents written hours apart: the hospital's last skin assessment and the nursing home's first one.

CMS gives both sides something. Its F686 guidance says an at-risk resident "can develop a PU/PI within hours of the onset of pressure," which is why the admission skin check has to happen promptly. It also says a deep tissue injury present on admission can evolve into an unavoidable Stage 3 or 4, and lists pressure that happened before admission: immobility during a hospital stay or surgery, a long ambulance transport, or lying on the floor waiting for help after a fall or a stroke. It adds a point every reviewer should take seriously: erythema is harder to see in darkly pigmented skin, so staff should look for changes in sensation, temperature and firmness.

How the MDS codes present on admission

The RAI Manual's coding examples are the clearest statement of the rules, and both sides should read them before arguing about the MDS. The facility codes from the first skin assessment done on or after admission, as close to admission as possible. "Present on admission" requires the ulcer to be at the same location and to not have increased in numerical stage or become unstageable due to slough or eschar. Worked examples from the manual:

These are coding rules, not liability rules. But the MDS is signed and submitted to CMS, and a facility that coded a wound as not present on admission has a hard time arguing later that it arrived with it. The reverse is also true. If the MDS says present on admission and the nursing notes from the first week describe intact skin, one of those documents is wrong.

The Kennedy terminal ulcer (KTU) is named for K.L. Kennedy, whose 1989 paper in Decubitus reported pressure ulcer prevalence in an intermediate care facility. The broader idea of skin changes at life's end (SCALE) came out of an 18-member expert panel that met in 2008, used a modified Delphi process with 52 international reviewers, and published a final consensus statement in 2009 with 10 statements. The panel took up the proposed concept of skin failure alongside other end of life skin changes.

CMS has taken a position in both manuals:

"Skin changes at the end of life (SCALE), also referred to as Kennedy Terminal Ulcers (KTUs) and skin failure, are not primarily caused by pressure and are not coded in Section M."

The F686 guidance puts the burden on the facility: it "is responsible for accurately assessing and classifying an ulcer as a KTU or other type of PU/PI and demonstrate that appropriate preventative measures were in place to prevent non-KTU pressure ulcers." Its list of KTU features: it appears suddenly and within hours; usually on the sacrum and coccyx, but it can appear on the heels, posterior calves, arms and elbows; the edges are irregular and red, yellow and black as it progresses, often described as pear, butterfly or horseshoe shaped; and it often looks like an abrasion, blister or darkened area and may develop rapidly to Stage 2, 3 or 4.

End of life is a related defense. CMS says an advance directive "does not absolve the facility" from supportive care the directive does not prohibit. But when a facility has put individualized end-of-life care in place in line with the resident's wishes, and kept up routine, less aggressive care such as cleaning, turning and repositioning, a new or worsening pressure injury "may be considered unavoidable." In our view, hospice enrollment on its own does not end the facility's duty to turn, clean and check the skin. See hospice eligibility documentation.

Our opinion, for both sides: the KTU is a real clinical entity and a much overused label. Written 3 weeks after an untreated Stage 2, the label reads as an excuse. Documented the evening it appeared, in a resident on comfort care with turning continued, it reads as honest medicine.

Nutrition shows up on both sides of a pressure injury case. The federal rule at 42 CFR 483.25(g) separately requires the facility to help each resident maintain "acceptable parameters of nutritional status, such as usual body weight or desirable body weight range and electrolyte balance, unless the resident's clinical condition demonstrates that this is not possible or resident preferences indicate otherwise," and to offer "sufficient fluid intake to maintain proper hydration and health." Surveyors cite that as F692.

The plaintiff theme: the resident was losing weight and nobody changed the plan. The defense theme: the resident was dying, and weight loss and non-healing were part of it. CMS supports a version of each. It says continuing weight loss and failure to heal "despite reasonable efforts to improve caloric and nutrient intake may indicate the resident is in multi-system failure or an end-stage or end-of-life condition." The key words are "despite reasonable efforts." The defense needs the efforts in the chart.

Labs are weaker evidence than most people think

Albumin and prealbumin get a lot of attention in depositions. CMS is skeptical of both. The F686 guidance says "no laboratory test is specific or sensitive enough to warrant serial/repeated testing," and that "a low albumin level combined with the facility's lack of supplementation, for example, is not by itself sufficient to cite a nutrition related deficiency." A low number alone does not settle causation for either side.

Supplements have a limit too. The RAI Manual says supplementation above the US Recommended Daily Intake "has not been proven to provide any further benefits for management of skin problems including pressure ulcers/injuries." A facility that did not add megadose vitamins did not fall below the standard for that reason.

What does carry weight is the trend and the response:

1. Weights. Every weight for the stay, and any that stop being recorded.
2. Intake. Meal percentages and fluids. A run of 25% entries is a signal; so are identical entries that never vary.
3. Dietitian assessments. Timing after admission, a new wound or weight loss, and whether recommendations were ordered.
4. Interventions. Supplements, feeding assistance, preference changes, swallowing evaluations. MDS item M1200D records whether a nutrition or hydration intervention to manage skin problems was in place in the last 7 days.
5. Goals of care. If the resident or family declined a feeding tube or aggressive nutrition, that choice should be documented and the care plan should show it.

The nursing home malnutrition and dehydration guide goes deeper on weight loss triggers and intake records.

We built the timeline below from the kind of records described in chapter 4. The question for both sides is the same: what did the facility know, when, and what did it do?

The Day 22 note was amended the next morning:

The amendment is labeled and timed, which is how a correction should look, but nobody measured the wound that week. See altered medical records for the line between a proper late entry and an alteration.

Plaintiff counsel will map these facts onto the 4 CMS elements; defense counsel will cite the transport, the risk factors and the added interventions. Our read: the 41-hour gap and the "per protocol" care plan are the hardest facts for the defense, because they come before the first skin change and they cannot be explained by the resident's condition. The pre-admission transport is the best defense fact, but it needs a hospital skin note that supports a deep injury, and here the hospital charted blanchable skin.

WOUND AND REPOSITIONING REVIEW WORKSHEET
Resident (initials): [  ]   Facility: [  ]   Stay: [ADMIT DATE] to [END DATE]
Wound site: [  ]   Reviewer: [  ]

A. BASELINE
Last hospital skin assessment before transfer: [FINDING] [DATE/TIME] [PAGE]
First facility skin assessment: [FINDING] [DATE/TIME] [PAGE]
Hours between arrival and first skin check: [  ]
First MDS Section M: stage [  ]  present on admission [Y/N]  [PAGE]

B. RISK
Braden scores (date, total, page): [  ] [  ] [  ] [  ] [  ]
Weekly for first 4 weeks? [Y/N]   Policy threshold: [  ]

C. CARE PLAN
Baseline care plan date (due within 48 hours): [  ] [PAGE]
Repositioning frequency named? [Y/N]  Text: [  ]
Support surface ordered [DATE]  delivered [DATE]  [PAGES]
Revisions after skin change (date, change, page): [  ]

D. REPOSITIONING RECORD
Longest documented gap per week: W1 [  ] W2 [  ] W3 [  ] W4 [  ]
Batch or late entries (care time vs entry time): [DESCRIBE, PAGES]
Seated time over plan limits: [DESCRIBE, PAGES]
Refusals and follow-up: [PAGES]

E. WOUND SERIES (1 line per assessment)
Date | Staff role | Stage | L x W x D cm | Undermining | Tissue | Photo | Page
[  ]
Stage drift or back staging: [DATES, PAGES]
Missed weekly assessments: [DATES]
2 to 4 weeks without progress, plan changed? [Y/N, PAGES]

F. NUTRITION
Weights (date, kg): [  ]   Intake trend: [  ]
Dietitian assessments and recommendations ordered? [DATES, PAGES]

G. DEFENSES TO TEST
Pre-admission pressure (surgery, transport, found down): [EVIDENCE, PAGES]
DTPI on admission? [EVIDENCE, PAGES]
KTU or skin failure: last intact check [  ] onset [  ] signs of dying [  ] who documented [  ]

A 60-day stay can produce thousands of pages, with the pressure injury evidence spread thinly across flowsheets, wound modules and the TAR. That is where AI medical record review earns its keep, and where it misleads anyone who trusts it blindly. Some vendors now pitch agentic tools that run multi-step tasks on their own; each unsupervised step is another place for an error to hide unless every step is cited.

OCR and handwritten notes on paper flowsheets

Many facilities still use paper TARs and turning sheets. Optical character recognition (OCR) reads typed text well. Handwritten notes are harder: checkmarks in small grid cells, initials, and times written as "2a" or "0200" are exactly what OCR gets wrong. A good tool shows which cells it read with low confidence. On a turning log, 1 misread time can create or hide a gap; see can AI read handwritten medical records.

Hallucination risk and page-level citations

A general-purpose large language model (LLM) asked to "summarize the wound care" will write fluent prose, and some of it can be invented: a measurement that was never taken, a stage the chart never used, a turning schedule that sounds right. That is hallucination, and in litigation it has consequences. Lawyers sanctioned under Rule 11 in Mata v. Avianca, Inc., 678 F. Supp. 3d 443 (S.D.N.Y. 2023) filed a brief citing cases a generative AI tool had made up. In a medical chronology the same failure is quieter: a false wound measurement in an expert's timeline.

The fix is design, not caution. Tools built on retrieval-augmented generation (RAG) pull the answer from the uploaded pages and cite the page each fact came from. Page-level citation turns every line into something a person can check in seconds. Clinical NLP, natural language processing tuned for medical text, knows that "sacral decub," "PI to sacrum" and "bed sore on bottom" are the same site.

What AI should not decide

No software should decide whether a wound was avoidable, whether a KTU label is justified, or whether a staffing gap caused the injury. An AI medical records summary that says "the facility was negligent" is overstepping. The right model is human-in-the-loop: the software finds and cites, a nurse or physician reviewer checks and interprets, and the attorney decides. For the accuracy question in detail, see is AI accurate enough for court.

Choosing legal AI tools for pressure injury files

Using only the uploaded records, build a chronology of skin and wound care for this resident from [ADMISSION DATE] to [END DATE].

For each entry give: date and time, author role, record type, wound location, stage as written, length x width x depth, undermining or tunneling, tissue type, treatment ordered or given, and the source page.

Include:
1. Every skin assessment and every pressure injury risk score (Braden or other), with subscale scores where recorded.
2. Every care plan entry or revision about skin, repositioning, support surfaces or nutrition.
3. Every repositioning entry, with time of care and time of entry if both are shown.
4. Every weight, meal intake percentage and dietitian recommendation.
5. Every reference to deep tissue injury, Kennedy terminal ulcer, skin failure, hospice or refusal of care.

Then list, with page citations:
A. Any entry that records a lower stage than an earlier entry for the same site.
B. Any week with no wound measurement for an open wound.
C. Any gap between repositioning entries longer than [CARE PLAN INTERVAL].
D. Any wound nurse, consultant or dietitian recommendation, and the date it became an order, if it did.

Do not infer facts that are not written in the records. Where a page is illegible or a value is uncertain, say so and cite the page. Do not give opinions on the standard of care or causation.

Medrecords AI reads the records you upload and returns a chronology, summaries and flags, with every line linked to its source page. On a pressure injury file, the hospital and nursing home records become 1 dated, searchable list. You review, you revise, you sign.

• A cited chronology. The chronology puts every skin assessment, Braden score, wound measurement, care plan change and TAR entry in date order, each linked to its page through source citations. Stage drift and missed weeks show because the entries sit side by side.
• Scans and handwriting. OCR and handwritten record extraction read paper TARs and turning sheets, and mark text they could not read with confidence.
• Questions with cited answers. Q&A answers questions like "every measurement of the sacral wound" or "when was the overlay delivered" with the pages that support the answer.
• Gaps and near-duplicates. Missing records identification points to records the chart refers to that are not in the file. Record alteration detection surfaces near-duplicate pages that differ, such as 2 printed versions of the same wound note.
• Provisional flags. The standard of care deviation detector raises possible departures for a clinician to review. Flags are signals, not verdicts.

What it does not do: Medrecords AI does not retrieve records, sign opinions, score the merit of a case, or give legal advice. It does not decide whether a wound was avoidable. Your nurse reviewer or expert does that, with the cited facts in front of them. The platform is SOC 2 and HIPAA compliant and signs a BAA; details are on the HIPAA and security pages. Self-Service bills 10 cents a deduplicated page, down to 5 cents at volume, duplicates free; Enterprise On-Prem is an annual license. See pricing.

See also the nursing home neglect record review and legal nurse consultant workflows.

Is every bedsore in a nursing home a sign of neglect?

No. Federal rules allow for unavoidable pressure injuries, but only when the facility assessed the resident, put appropriate interventions in place, monitored them and revised them, and the wound developed anyway. CMS also says being high risk does not by itself make a wound unavoidable.

Can a Stage 3 pressure ulcer heal into a Stage 2?

No. CMS says pressure ulcers do not heal in reverse sequence, and the RAI Manual tells facilities not to reverse or back stage. A healing Stage 3 stays a Stage 3 until it closes.

How often should a nursing home turn a resident?

As often as the resident's own care plan requires, based on an individual assessment. CMS describes hourly off-loading for dependent residents who are seated or reclined 30 degrees or more, and repositioning at least every 2 hours or more often for dependent residents who are reclining.

What records prove whether the bedsore started in the hospital or the nursing home?

The sending hospital's last skin assessment and transfer form, the nursing home's first skin assessment with its date and time, and the first MDS Section M. If a deep tissue injury is possible, look for purple or maroon skin, blood blisters, pain or warmth over the site in the first days.

Is a Kennedy terminal ulcer a complete defense?

Only when the facts fit. CMS says the facility must correctly assess and classify an ulcer as a KTU and show that measures were in place to prevent other pressure ulcers. KTUs appear suddenly, within hours, usually in residents who are dying. A label applied weeks after a wound was first charted does not fit that description.

Can AI software decide whether a pressure injury was avoidable?

No, and it should not try. Avoidability is an expert judgment about assessment, interventions, monitoring and the resident's condition. AI software can gather every wound measurement, Braden score and turning entry into a cited timeline so that the expert makes that judgment faster and on complete facts.

Is it safe to upload nursing home records to an AI tool?

Only to a tool built for protected health information, under a signed business associate agreement, with security controls you have reviewed. Never paste resident records into a consumer chatbot.

Can AI read handwritten turning sheets and paper TARs?

Often, but not perfectly. OCR and handwriting extraction handle clear entries well and struggle with tiny grid cells, initials and ambiguous times. A good tool marks low-confidence reads so a person checks them against the page image.

How does an AI medical chronology help defense counsel in a pressure injury case?

The same way it helps plaintiffs: it shows the whole record in date order. Defense counsel can see early whether the facility documented all 4 elements CMS looks for, whether a KTU or pre-admission injury argument has support in the chart, and which entries, such as a back staged note, need an explanation before a deposition.

We read each source below in September 2026 and dropped any figure we could not confirm. The 80% and 96% onset figures are quoted as they appear in CMS guidance, which cites Lyder and Ayello's chapter in the 2008 AHRQ nursing handbook; we could not open that chapter directly. We did not state a single Braden cut-off as a standard, because the only one we verified is the 16 used in the 1987 validation studies. Worked examples are hypothetical.

• 42 CFR 483.25: pressure ulcers and nutrition.
• 42 CFR 483.21: 48-hour baseline care plan.
• 42 CFR 483.20: assessment within 14 days of admission.
• 42 CFR 483.10(g)(2): 24-hour access, copies on 2 working days' notice.
• 42 CFR 483.70: record retention.
• CMS State Operations Manual, Appendix PP, F686 guidance (Rev. 229, 04-25-25).
• CMS MDS 3.0 RAI Manual v1.20.1, October 2025, Section M.
• NPIAP pressure injury stages, 2016
• Bergstrom N, Braden BJ, Laguzza A, Holman V. The Braden Scale for Predicting Pressure Sore Risk. Nursing Research, 1987
• Braden score for early mortality risk assessment in critically ill intensive care patients with multiple organ dysfunction syndrome, Scientific Reports, 2025
• Kennedy KL. The prevalence of pressure ulcers in an intermediate care facility. Decubitus, 1989.
• SCALE: Skin Changes at Life's End. Wounds, 2009
• Sibbald RG, Krasner DL, Lutz J. SCALE Final Consensus Statement, October 1, 2009. Advances in Skin and Wound Care, 2010
• Mata v. Avianca, Inc., 678 F. Supp. 3d 443 (S.D.N.Y. 2023), Rule 11 sanctions for AI-fabricated citations.

Related guides

• Nursing home records, the MDS and CMS-2567 survey reports
• Nursing home malnutrition and dehydration
• Nursing home wrongful death records
• Change in condition in nursing homes
• Nursing home fall lawsuit records
• Assisted living vs nursing home records
• HIPAA compliant AI medical record review