# Nursing home malnutrition and dehydration: records that decide it > The nursing home records behind a malnutrition or dehydration claim: F692, MDS Section K weight math, meal intake logs, dietitian notes, labs and AI review. Canonical page: https://medrecords.ai/guides/nursing-home-malnutrition-dehydration/ --- New — **Missing Records Detection:** flags every visit, provider, and date missing from the file. [See how →](https://medrecords.ai/product/missing-records-identification/) Field guide, September 2026. 14 chapters. ## Nursing home malnutrition and dehydration: the records that prove or defend a weight loss case For nursing home plaintiff attorneys, defense counsel and legal nurse consultants. You walk away with the federal weight loss thresholds and the math behind them, a way to test meal intake records, a decision path for avoidable versus unavoidable decline, a records request and a weight and intake worksheet. By [Ahmed Jemaa](https://medrecords.ai/authors/ahmed-jemaa/), Co-Founder and CEO, Medrecords AI. Published 26 September 2026. Nursing home malnutrition and dehydration cases turn on whether the facility met 42 CFR 483.25(g), cited as F692. The records that decide it are the weight log, MDS Section K, meal intake percentages, fluid intake and output, dietitian assessments, care plans, physician notes and labs. CMS surveyor guidance treats weight loss as avoidable unless the facility shows it assessed, intervened, monitored and coordinated care. A nursing home weight loss case is built from small numbers written down by tired people: a weight on a scale nobody recalibrated, a "75%" ticked at the end of a shift, a fluid total added up after the fact. Federal rules turn those numbers into triggers. The case asks whether anyone noticed, told the physician and dietitian, changed the plan, and checked that the change worked. Built from 42 CFR part 483, CMS State Operations Manual Appendix PP (Rev. 232), the MDS 3.0 RAI Manual v1.20.1 (October 2025), the HIPAA access rule and 5 peer-reviewed studies, each cited where used. Every case example is a labeled hypothetical. Nothing here is legal or medical advice. Whether care met the standard, and what caused an injury or death, are questions for qualified experts under the law of your state. 10 numbers ### Malnutrition and dehydration records in 10 numbers 5% in 30 days weight loss MDS item K0300 counts, against the weight closest to 30 days earlier RAI Manual v1.20.1, K0300 10% in 180 days the second K0300 threshold, against the weight closest to 180 days earlier RAI Manual v1.20.1, K0300 7.5% in 3 months a significant loss under the CMS surveyor parameters; more than 7.5% is severe Appendix PP, F692 4 weeks of weekly weights after admission, then at least monthly Appendix PP, F692 1,500 ml daily fluid intake below which the MDS counts 1 indicator of dehydration; 2 indicators code J1550C RAI Manual v1.20.1, J1550C 14 hours maximum gap from a substantial evening meal to breakfast; 16 with a nourishing bedtime snack and resident group agreement 42 CFR 483.60(f)(2) 8 hours minimum State-approved training before a paid feeding assistant may feed residents 42 CFR 483.160(a) 22% by which staff meal documentation overestimated residents' total intake in 1 direct observation study Simmons and Reuben, 2000 35 to 40 minutes of staff time per resident per meal for feeding assistance that improved intake Simmons and Schnelle, 2006 67 tests for dehydration in older adults reviewed; none reliable enough to use alone Hooper and colleagues, Cochrane, 2015 Chapter 1 Everyone ### The federal rule and why CMS presumes weight loss is avoidable Every nursing home that takes Medicare or Medicaid is bound by 42 CFR part 483. The nutrition and hydration duty sits at [42 CFR 483.25(g)](https://www.ecfr.gov/current/title-42/chapter-IV/subchapter-G/part-483/subpart-B/section-483.25). Based on the comprehensive assessment, the facility must ensure that the resident: - Maintains "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" (483.25(g)(1)). - Is offered sufficient fluid intake to maintain proper hydration and health (483.25(g)(2)). - Is offered a therapeutic diet when there is a nutritional problem and the health care provider orders one (483.25(g)(3)). Paragraphs (g)(4) and (g)(5) cover feeding tubes, cited separately as F693. This guide covers F692, the tag for (g)(1) to (3), plus the food service tags in the F800 series from [42 CFR 483.60](https://www.ecfr.gov/current/title-42/chapter-IV/subchapter-G/part-483/subpart-B/section-483.60). #### 2 exceptions, both written into the rule The defense sits in the same sentence as the duty. The facility need not hold weight steady if "the resident's clinical condition demonstrates that this is not possible" or if "resident preferences indicate otherwise." Most malnutrition cases become an argument about those 2 clauses: care failure on one side, dying, refusing or planned diuresis on the other. The records decide which account holds. #### The presumption surveyors apply CMS gives surveyors interpretive guidance in [State Operations Manual Appendix PP](https://www.cms.gov/Regulations-and-Guidance/Guidance/Manuals/downloads/som107ap_pp_guidelines_ltcf.pdf). The F692 guidance sets the burden this way: "Weight loss, poor nutritional status, or dehydration should be considered avoidable unless the facility can prove it has assessed/reassessed the resident's needs, consistently implemented related care planned interventions, monitored for effectiveness, and ensured coordination of care among the interdisciplinary team." That sentence is written for surveyors, and Appendix PP is not a statute. How far a court lets an expert use survey guidance or a survey finding varies by state, so ask local counsel. Either way, the 4 verbs make a good audit plan for any file: assessed, implemented, monitored, coordinated. Chapter 8 turns them into a decision path. Appendix PP also warns the other way: "Weight loss, abnormal protein and electrolyte lab values, and dehydration are not, by themselves, sufficient to support noncompliance at F692." And a resident need not lose weight for a facility to be out of compliance. #### How bad the survey says it was If a survey cited F692 during the stay, the Form CMS-2567 describes it and the survey record carries a scope and severity rating. Appendix PP's immediate jeopardy examples include "repeated failure to assist a resident who required assistance with meals and drink." Its actual harm examples include "the failure to identify a decrease in food intake" that led to significant unintended weight loss. Severity level 1 does not apply to F692 at all. For how the 2567 fits the wider file, see the [nursing home records pillar guide](https://medrecords.ai/guides/nursing-home-records-mds-cms-2567/). If you remember one thing 483.25(g)(1) carries its own 2 exceptions, clinical condition and resident preference. Both sides argue them against the records of the 4 verbs: assessed, implemented, monitored, coordinated. Chapter 2 Everyone ### Weight records and the math behind every threshold The weight log is the spine of the case and the most error-prone record in the chart. A weight is only as good as the scale, the clothing, the time of day and the person who wrote it down. #### How often a resident should be weighed Appendix PP cites the professional standard of weighing on admission or readmission, "weekly for the first 4 weeks after admission and at least monthly thereafter," and again after a significant change in condition or a persistent drop in intake. Weights may not be indicated for a terminally ill resident who asks for comfort care only. It also expects "a consistent method of weighing a resident (e.g. using the same scale, wearing the same clothes, weighing at the same time of day, adjusting for use of a prosthetic, etc.)," and verified weights when a change shows up. Ask for the weighing policy by name and compare it with the log. #### The 2 sets of thresholds **CMS suggested parameters for significant unplanned weight loss** — 3 intervals — Appendix PP, F692 guidance Each figure is a significant loss; anything greater is severe. The 3 month line is the one K0300 never checks. **The F692 surveyor parameters.** 5% in 1 month, 7.5% in 3 months and 10% in 6 months are significant; anything more is severe. Percent loss = (usual weight minus actual weight) divided by usual weight, times 100. **The MDS item K0300.** The [RAI manual](https://www.cms.gov/files/document/final-mds-3-0-rai-manual-v1-20-1-october-2025.pdf) multiplies the weight closest to 30 days before the current weight by 0.95, and the weight closest to 180 days before by 0.90. If the current weight is at or below either result, the resident has lost 5% or more in 30 days or 10% or more in 180 days. Weights are rounded to whole pounds first. Codes: 0 (no or unknown), 1 (yes, on a physician-prescribed weight loss regimen, including planned diuresis), 2 (yes, not prescribed). The manual says those 2 weights "are the only 2 weights considered for this item," so a slow loss can slip under both snapshots while crossing the 3 month parameter. It also says the facility "should not wait for the 30- or 180-day timeframe to address the problem." A K0300 coded 0 does not mean the resident was fine. **A slow loss the MDS snapshots miss** — hypothetical — Illustration; thresholds from Appendix PP and RAI K0300 At day 90 the resident has lost 8.0% in 3 months, past the severe line, yet both K0300 comparisons come out under threshold. **Hypothetical.** Resident R.T. is admitted at 150 lb, then weighs 149, 148, 147 and 146 weekly, 142 at day 60 and 138 at day 90, with no earlier weight on file. At day 90, 142 times 0.95 is 134.9 and 150 times 0.90 is 135. 138 is above both, so K0300 is coded 0. But (150 minus 138) divided by 150, times 100, is 8.0% in 3 months: a severe loss under Appendix PP. #### How to build the trend yourself 1. **Collect every weight.** Flowsheets, the weight log, dietitian notes, K0200B, hospital, therapy and physician notes. Record date, time, scale and page. 2. **Set the usual weight.** Use pre-admission hospital or physician weights and the admission weight, and note any gap between them. 3. **Flag suspect weights.** A 6 lb swing in 2 days, a scale change, or a large change with no reweigh is a data point to question, not to average in. 4. **Compute both sets.** K0300 at each assessment reference date; the 1, 3 and 6 month parameters on a rolling basis from usual weight. 5. **Put the response next to each crossing.** Date the next dietitian note, physician call, care plan revision and intervention. The interval is often the finding. 6. **Adjust for fluid.** Appendix PP lists edema among the factors that affect a weight. Chart diuretic starts and stops, and edema, on the same timeline. 2 more RAI rules matter. If the last weight is more than 30 days before the assessment reference date, K0200B requires a reweigh. And for an amputation, the manual's worked example adds back the weight of the missing limb before comparing; in that example the loss was still significant. If you remember one thing K0300 looks at 2 weights only. Run the 3 month surveyor parameter yourself from the usual weight, because slow losses cross it while the MDS reports no loss. Chapter 3 Everyone ### MDS Section K, the care area triggers and the assessment clock Section K (swallowing and nutritional status) and item J1550C (dehydration) of the Minimum Data Set drive a malnutrition file. The staff who complete them sign them, so they fix what the facility knew on a given date, and their answers trigger Care Area Assessments the team must work through and then decide, in writing, whether to care plan. #### When assessments happen Under [42 CFR 483.20](https://www.ecfr.gov/current/title-42/chapter-IV/subchapter-G/part-483/subpart-B/section-483.20), a comprehensive assessment is due within 14 calendar days after admission, within 14 days after a significant change, and at least every 12 months, with a quarterly review at least every 3 months. A significant change is "a major decline or improvement" that "will not normally resolve itself," affects more than 1 area of health status, and "requires interdisciplinary review or revision of the care plan." A resident who stops eating and becomes dependent for meals may meet it. Ask whether anyone opened one. #### The items, annotated MDS 3.0, Sections J and K hypothetical quarterly K0100 Swallowing, 7-day look-back C. Coughing or choking during meals — checked1 K0200 Height and weight B. Weight (lb) — 1382 K0300 Weight loss — 0. No or unknown3 K0520 Nutritional approaches C3. Mechanically altered diet, while a resident — checked4 J1550 Problem conditions C. Dehydrated — not checked5 Quarterly: no CAAs required 1. 1 **K0100 swallowing signs** Loss of food from the mouth, pocketing, coughing or choking, or pain swallowing, over 7 days. A checked box should match a speech therapy referral or diet order. 2. 2 **K0200B weight** The most recent weight in the last 30 days, rounded to a whole pound. Match it to the weight log by date. 3. 3 **K0300 weight loss** Only the 2 snapshot weights count. A 0 here can coexist with a severe 3 month loss, as chapter 2 shows. Code 1 needs a documented goal of planned loss or diuresis. 4. 4 **K0520C mechanically altered diet** Pureed food, thickened liquids and similar. The manual says it "should not automatically be considered a therapeutic diet," which is K0520D. 5. 5 **J1550C dehydrated** Checked when 2 of 3 indicators are present (chapter 5). Compare it with the intake and output records for the same 7 days. #### The care area triggers 2 Care Area Assessments in chapter 4 of the RAI manual carry nutrition and hydration cases. **CAA 12, Nutritional Status** , triggers on J1550C checked, body mass index under 18.5 or over 24.9, K0300 or K0310 (weight gain) coded 1 or 2, parenteral or IV feeding, a mechanically altered or therapeutic diet while a resident, or a stage 2 or higher pressure ulcer. That last trigger is why malnutrition claims often travel with [pressure injury claims](https://medrecords.ai/guides/pressure-injury-bedsore-lawsuit-records/). **CAA 14, Dehydration and Fluid Maintenance** , triggers on fever, vomiting, dehydration, internal bleeding, several infections, constipation, IV feeding or a feeding tube. A trigger means the team had to look, not that care was missed. CAAs are required only on comprehensive assessments (admission, annual, significant change and significant correction); on a quarterly the team still reviews and revises the care plan. Section V records whether the team chose to care plan each triggered area, and the care plan is due within 7 days after the CAAs are completed. Pull the CAA worksheets as well as the Section V grid. A triggered CAA with no worksheet, or one that says only "will monitor," is a finding. A worksheet that explains why no new intervention was needed is strong defense evidence. If you remember one thing Every MDS answer should match a source record for the same look-back window, and every trigger should have a CAA worksheet behind it. Chapter 4 Everyone ### Meal intake records and how far to trust them Aides usually record meal intake as a percentage, often in 25% steps, in a point-of-care system or on a paper meal sheet. Some residents also have an intake and output record, and a worried dietitian may order a calorie count of what was served and what was left. Appendix PP lists what surveyors review to judge interventions: "weight records, meal monitors, intake and output logs, nurses' notes, lab values, and physician or dietitian assessments." Use it as a records list. ##### Meal percentage record Who writes itNurse aide, per mealWhat it measuresEstimated share of the tray eatenStrengthContinuous, 3 entries a dayWeak pointEstimates, often entered after the fact ##### Intake and output log Who writes itNursing staff, per shiftWhat it measuresFluids in and out, in mlStrengthNumbers you can total and compare with J1550CWeak pointOften ordered, then kept only for some shifts ##### Calorie count Who writes itNursing staff; dietitian reviewsWhat it measuresItems served and left, turned into calories and proteinStrengthClosest to measured intakeWeak pointShort, and kept only when ordered #### What the research says about meal percentages Studies that compared staff documentation with trained observers, weighed trays or photographs agree on the direction: documentation overstates intake, and it misses the residents who eat least. 22% — staff overestimate of total intake; 56 residents, 9 meals each. Simmons and Reuben, 2000 About 20% — error in staff estimates against weighed trays, mostly overestimates; 27 residents in 3 homes. Pokrywka and colleagues, 1997 42% to 91% — of residents, by facility, for whom staff failed the indicator for identifying low intake; 302 residents in 10 homes. Simmons and colleagues, 2002 In [Simmons and Reuben](https://pubmed.ncbi.nlm.nih.gov/10682952/), tray photographs matched the trained observers closely, so a photo protocol is a credible check. In [Pokrywka and colleagues](https://pubmed.ncbi.nlm.nih.gov/9329485/), almost 1 in 3 at-risk residents was missed. The range in [Simmons and colleagues (2002)](https://pubmed.ncbi.nlm.nih.gov/12242322/) shows that accuracy depended heavily on the facility. About 5 of 10 — 53% Share of residents eating 75% or less at most meals whom staff documentation failed to identify. Simmons and Reuben, J Am Geriatr Soc, 2000. These studies are old and small, and a facility will say its staff do better. That is testable. They do not prove any chart wrong; they show why a meal record needs corroboration before an expert relies on it. #### How to test the intake record in your file Look for entries not made at the meal. Meal intake record, resident R.T., hypothetical — 9 rows | Date and meal | Intake | Fluids (ml) | Entered | Staff | | --- | --- | --- | --- | --- | | 03/10 breakfast | 50% | 240 | 03/10 08:40 | CNA 1 | | 03/10 lunch | 25% | 180 | 03/10 13:05 | CNA 1 | | 03/10 supper | 75% | 240 | 03/10 22:51 | CNA 2 | | 03/11 breakfast | 75% | 240 | 03/11 14:58 | CNA 2 | | 03/11 lunch | 75% | 240 | 03/11 14:58 | CNA 2 | | 03/11 supper | 75% | 240 | 03/11 14:59 | CNA 2 | | 03/11 15:20 nurse note | "Refused lunch, sleepy" | n/a | 03/11 15:24 | LPN | | 03/12 breakfast | Refused | 120 | 03/12 08:55 | CNA 3 | | 03/12 lunch | 25% | 120 | 03/12 12:50 | CNA 3 | Flagged rows: identical values entered in a block, and a supper entry time-stamped before supper was served. The key row is a nurse note that contradicts the lunch entry. 4 signals: identical percentages meal after meal, a whole day entered at 1 sitting, an entry time-stamped before the meal, and a nurse note that contradicts the record. Point-of-care systems keep the entry time apart from the meal time, so ask for both in native format. A block of late entries is a question for the aide, not proof. The aide may have charted late but accurately, and defense counsel should find that out first. For the defense, the comparisons run the other way. Stable weights during recorded good intake, calorie counts that match the percentages, and dietitian notes describing an observed meal all support the record. A resident who ate well but still lost weight points toward disease, medication or fluid shifts rather than care. If you remember one thing Meal percentages are estimates that tend to run high. Test them against entry times, weights and nurse notes before an expert relies on them. Chapter 5 Everyone ### Hydration, dehydration and the labs that carry weight The hydration duty appears twice. Under 483.25(g)(2) the resident must be offered sufficient fluid. Under 483.60(d)(6), cited as F807, the facility must provide drinks "sufficient to maintain resident hydration." The F807 guidance notes that under-hydrated residents "are more susceptible to urinary tract infections, pneumonia, decubitus ulcers, skin infections, confusion and disorientation," which are often the hospital diagnoses in the complaint. #### There is no single number for fluid needs Appendix PP says "there is no reliable calculation to determine an individual's fluid needs." The estimated daily need in the dietitian note is the facility's own target, not a legal standard. Compare it with the intake and output totals. The MDS gives 1 benchmark. Item J1550C is checked when 2 or more of these indicators are present: 1. **Intake under 1,500 ml a day.** Beverages and high-fluid foods such as soups count. The manual notes the figure "has been changed from 2,500 ml to 1,500 ml to reflect current practice standards," which matters when an old policy or expert report uses the higher number. 2. **1 or more clinical signs.** Such as dry mucous membranes, poor skin turgor, thirst, sunken eyes, dark urine, new or increased confusion, fever, or abnormal labs. 3. **Fluid loss greater than intake.** From vomiting, fever or diarrhea, for example. #### Bedside signs are weak on their own A [2015 Cochrane review by Hooper and colleagues](https://pubmed.ncbi.nlm.nih.gov/25924806/) tested 67 signs and simple tests against serum osmolality in people aged 65 and over, counting 295 mOsm/kg or more as water-loss dehydration. It found that fluid intake, urine specific gravity, urine colour, urine volume, heart rate, dry mouth, feeling thirsty and bioelectrical impedance "are not useful, and should not be relied on individually." Only 3 tests showed any stand-alone promise (expressing fatigue, missing drinks between meals, and 1 impedance measure), each from small studies, and no test was consistently useful in more than 1 study. That cuts both ways. A note saying "mucous membranes dry" does not prove dehydration, and "skin turgor good" does not rule it out. The review concludes that individual tests "miss a high proportion of people with dehydration, and wrongly label those who are adequately hydrated." Read the labs, the intake record and the clinical picture together. MythResidents need 2,500 ml of fluid a day, so anything less is neglect. RuleThe MDS indicator is 1,500 ml, and it is 1 of 3 indicators, 2 of which code J1550C. CMS says no reliable calculation of individual need exists. MythLow albumin proves the resident was starved. RuleAppendix PP says low albumin may reflect acute illness unrelated to nutrition and may not improve even with adequate calories and protein. MythGood skin turgor and no complaints of thirst rule out dehydration. RuleThe Cochrane review found such signs not useful on their own in older adults. #### The labs, and which ones to weight Appendix PP says labs such as "electrolytes, BUN, creatinine and serum osmolality" can help "identify, manage, and monitor fluid and electrolyte status," and CAA 14 says diagnosis in older persons relies "primarily" on history and lab testing. Put every metabolic panel on the timeline with its collection time, result time and physician notification time. Protein markers are weaker. The F692 guidance says low albumin "may also be the result of an acute illness for reasons unrelated to nutrition," and that for albumin and pre-albumin "there is no evidence that they are useful for the serial follow-up of undernourished individuals." A falling albumin is weak proof of starvation, and a normal one is weak proof of good nutrition. BUNBlood urea nitrogen. Rises with dehydration, and also with kidney disease, bleeding in the gut and high protein intake. SodiumA high sodium (hypernatremia) is a classic laboratory sign of water loss. Low sodium has many other causes. Serum osmolalityThe concentration of the blood. The Cochrane review used 295 mOsm/kg or more as its reference standard for water-loss dehydration. Intake and output (I&O)A nursing record of fluids taken in and lost, in ml, usually totaled by shift and by day. If you remember one thing For dehydration, lean on the chemistry panel and intake totals, not bedside signs. For malnutrition, lean on weights and intake, not albumin. Chapter 6 Everyone ### The dietitian, the care plan and the physician Weights and intake show what happened. The dietitian notes, care plan and physician record show what the facility did about it. #### The dietitian Under 483.60(a), the facility employs a qualified dietitian or other clinically qualified nutrition professional, full-time, part-time or as a consultant; the training rules include at least 900 hours of supervised dietetics practice. For a part-time or consultant dietitian, ask for the contract and the visit log, and find out who covered the weeks between visits. Therapeutic diets must be prescribed by the attending physician, who may delegate that task to a registered or licensed dietitian "to the extent allowed by State law" (483.60(e)). A dietitian recommendation often still needs an order. #### Tracing each recommendation For every dietitian recommendation, follow it through 4 documents: the recommendation, the order, the delivery record, and the follow-up. Step 1 — **Recommendation** — Dietitian note: what, how much, how often, and why Step 2 — **Order** — Physician or delegated order, with date and time signed Step 3 — **Delivery** — MAR or treatment record entries for supplements, with amount taken Step 4 — **Follow-up** — Next weight, next dietitian note, and whether the plan changed Gaps show up at the joins: a supplement recommended on day 62 and ordered on day 71; "refused" at most passes with no care plan response, though Appendix PP expects declined interventions to be described in the care plan; a supplement signed as given with no amount consumed. #### The note that did not change Copy-forward text becomes a problem when it contradicts the data around it. **2 quarterly dietitian notes, 85 days apart** — hypothetical — Illustration **Admission note** — signed day 5, 10:14 Nutrition assessment, resident R.T. Wt 150 lb. Usual wt 150 lb. Intake 75 to 100% most meals. Regular diet, thin liquids. Wt stable. Skin intact. Continue current plan. RD to follow quarterly. **Quarterly note** — signed day 90, 09:47 Nutrition assessment, resident R.T. Wt 138 lb. Usual wt 150 lb. Intake 75 to 100% most meals. Regular diet, thin liquids. Wt stable. Skin intact. Continue current plan. RD to follow quarterly. Only the weight line changed. "Wt stable" and "continue current plan" were carried forward next to an 8.0% loss. For a plaintiff, the right-hand note shows the dietitian recorded the loss and did not act. For the defense, it is a documentation problem that needs context: a separate progress note that week, or a call to the physician. Check the entry metadata and nearby nutrition entries before either side characterizes it. See [altered medical records](https://medrecords.ai/guides/altered-medical-records/) for how copied notes are proved. #### The care plan and the team Under [42 CFR 483.21](https://www.ecfr.gov/current/title-42/chapter-IV/subchapter-G/part-483/subpart-B/section-483.21), a baseline care plan is due within 48 hours of admission and a comprehensive care plan within 7 days after the comprehensive assessment. The team must include the attending physician, a registered nurse and a nurse aide with responsibility for the resident, and a member of food and nutrition services staff. Check the signature page. A useful nutrition care plan names a measurable goal, specific interventions and who does each. A care plan last revised before the loss began is often the clearest document in the file. #### The physician Under 483.10(g)(14), cited as F580, the facility must immediately inform the resident, consult the physician and notify the representative of a significant change in status or a need to alter treatment significantly. A weight crossing, a sustained drop in intake or an abnormal sodium may qualify, depending on the facts. Find the notification, the response and any new order. The [change in condition guide](https://medrecords.ai/guides/nursing-home-change-in-condition/) covers these records. Under [42 CFR 483.30(c)](https://www.ecfr.gov/current/title-42/chapter-IV/subchapter-G/part-483/subpart-B/section-483.30), the physician sees the resident at least every 30 days for the first 90 days after admission, then at least every 60 days; a visit is timely within 10 days after it was due. A visit note saying "eating well" during a documented loss is a problem for the physician too. If you remember one thing Date every step from dietitian recommendation to order, delivery and follow-up. The days between a weight crossing and the first real change in the plan are often the key number. Chapter 7 Everyone ### Feeding assistance, meal timing and diet texture Many residents who lose weight are not refusing food. They eat too slowly, fall asleep, or the tray leaves before anyone sits with them. The records here are staffing, assistance levels and diet orders. #### Help at meals takes real time In a [6 home study of 91 residents, Simmons and Schnelle (2006)](https://pubmed.ncbi.nlm.nih.gov/16776786/) found that feeding assistance that improved intake took an average of 35 to 40 minutes of staff time per resident per meal, for residents needing only verbal cues as well as dependent ones. That supports a simple capacity check. Item — Minutes — Basis Residents on the unit who need help at supper — 8 — hypothetical Staff time per resident per meal — 35 — low end, Simmons and Schnelle, 2006 Assistance time needed for the meal — 280 — 8 times 35 Aide time available: 2 aides for a 60 minute service — 120 — hypothetical assignment sheet Shortfall for the meal — 160 — before call lights, toileting or other duties That arithmetic does not prove any resident went unfed. The study measured long-stay residents with low intake who responded to a research feeding protocol, so its figure may not fit every resident on a unit. It tells the plaintiff which records to request (assignment sheets, staffing postings, dining assignments, residents coded as needing help with eating) and the defense what to gather (extra staff at meals, family who fed at supper, paid feeding assistants). #### Paid feeding assistants A facility may use paid feeding assistants under 483.60(h), cited as F811. The rules are specific: - The assistant must first complete a State-approved course of at least 8 hours under [42 CFR 483.160](https://www.ecfr.gov/current/title-42/chapter-IV/subchapter-G/part-483/subpart-D/section-483.160), covering feeding and hydration techniques, communication, safety and emergency procedures including the Heimlich maneuver, infection control, resident rights, and reporting changes to the supervising nurse. The facility keeps a record of each assistant who completed it. - The assistant works under the supervision of an RN or LPN, and in an emergency must call a supervisory nurse for help. - The assistant may help only residents with "no complicated feeding problems," which the rule says include "difficulty swallowing, recurrent lung aspirations, and tube or parenteral/IV feedings." Selection must rest on the interdisciplinary team's assessment and the latest assessment and care plan. If an assistant fed a resident with a K0100 swallowing sign or a texture-modified diet, compare the selection with the MDS and the speech therapy record, and ask for that assistant's training record. #### Meal timing and equipment Under 483.60(f), residents get at least 3 meals daily at regular times, with no more than 14 hours between a substantial evening meal and breakfast, or 16 with a nourishing bedtime snack and resident group agreement. Under 483.60(g), residents who need special eating equipment get it, and help using it. A built-up spoon that was ordered but never reached the tray is a fact juries understand. #### Swallowing and diet texture orders Swallowing problems link malnutrition to aspiration. The records are the K0100 signs, the speech-language pathology evaluation and any instrumental study, the diet order and K0520C. Under 483.60(d)(3), cited as F805, food must be prepared "in a form designed to meet individual needs." CMS guidance is cautious about texture. It says modifying consistency "may unnecessarily decrease quality of life and impair nutritional status by affecting appetite and reducing intake," that a swallowing abnormality "alone does not necessarily warrant dietary restrictions or food texture modifications," and that "no interventions consistently prevent aspiration." So texture cases run both ways: a missed swallow evaluation before an aspiration, or a downgrade to puree against the resident's wishes, followed by falling intake, severe weight loss and irreversible decline, which Appendix PP gives as an immediate jeopardy example. The defense answer in both is the evaluation, the documented risk discussion and the resident's choice. If you remember one thing Match the care plan assistance level to who actually fed the resident, and the diet order to what was on the tray. Chapter 8 Deciding ### Avoidable or unavoidable: hospice, refusals and the other defenses Advanced dementia, cancer, heart failure and kidney failure cause weight loss that good care may slow but cannot stop. The question is whether the record shows the facility recognized the decline, evaluated it, and made a documented, individualized decision about it. #### What the guidance accepts Appendix PP accepts that "weight stability, rather than weight gain, may sometimes be the most pertinent short-term or long-term objective," and that weight may "stabilize at a lower level, sometimes indefinitely." Then it sets a condition: "There should be a documented clinical basis for any conclusion that nutritional status or significant weight change are unlikely to stabilize or improve (e.g., physician's documentation as to why weight loss is medically unavoidable)." If a practitioner wrote at the time that the loss was expected from a named condition, the defense has a contemporaneous record. If the explanation first appears in a retained expert's report, the plaintiff will call it hindsight. ##### What the plaintiff looks for Signs the decline was treatable and untreated - Weight crossings with no dietitian or physician response for days or weeks - Dietitian recommendations never ordered, or ordered and not given - A care plan not revised after the loss began - No significant change assessment despite a major decline - An end-of-life explanation that first appears after the resident died or the claim was made ##### What the defense looks for Signs the decline was expected and managed - A physician note, written at the time, explaining why loss was unavoidable - Hospice election, with a coordinated hospice and facility care plan - Documented refusals, with the risks explained and alternatives offered - A physician-ordered diuresis or weight loss goal, supporting K0300 code 1 - Interventions tried, monitored and changed when they failed #### Hospice and end-of-life decline Hospice does not end the facility's duties. Under 42 CFR 483.70(n), the written agreement sets out each provider's services and "a communication process, including how the communication will be documented," and the facility must still "meet the resident's personal care and nursing needs in coordination with the hospice representative." Ask for the agreement, both care plans and the hospice's separate records. The [hospice eligibility documentation](https://medrecords.ai/solutions/hospice-eligibility-documentation/) page covers certification. The election date carries the argument. Hospice after months of unaddressed loss invites the claim that the care caused the decline. Hospice early in the course, with comfort feeding care planned and a documented family discussion, points the other way. #### Refusals and preferences Residents may refuse food, supplements, texture changes and weights. Appendix PP expects the facility to discuss "the risks and benefits associated with that decision and offer alternatives," and to describe declined interventions in the care plan. A single charted "refused" is not a refusal defense. A pattern with a documented discussion, alternatives and a care plan entry is. 2 more points: a do not resuscitate order "does not by itself indicate that the resident is declining other appropriate treatment and services," and appetite stimulants "are not a substitute for appropriate investigation" of modifiable causes. #### Planned loss and fluid shifts Loss from a physician-ordered diuresis or diet is coded K0300 = 1 and needs a documented goal. Weight lost as edema resolves is not malnutrition, and fluid retention can hide loss of body mass, so chart diuretics and edema beside the weights. #### A decision path for "was it avoidable?" 1 **Was the resident's nutrition and hydration risk assessed on admission and after each change?** YesGo to 2. Cite the assessment pages and dates. NoAn assessment gap. Note when it was due. 2 **Did the care plan name a goal and specific interventions, and were they carried out?** YesGo to 3. Tie each intervention to delivery records. NoAn implementation gap. List interventions with no delivery record. 3 **Were weights and intake monitored on schedule, and the plan changed when they worsened?** YesGo to 4. NoA monitoring gap. Measure the days from each crossing to the next change. 4 **Were the physician, dietitian, nursing and, if involved, hospice working from the same plan?** YesGo to 5. NoA coordination gap. Find the unanswered recommendation or call. 5 **Is there a contemporaneous, documented clinical basis for calling the loss unavoidable, or a documented resident choice?** YesThe record supports the exception in 483.25(g)(1). Experts decide if it holds. NoThe file lacks the documentation CMS guidance asks for. Experts decide what that means for causation. If you remember one thing An end-of-life or refusal defense is only as strong as the note written at the time. Date the first written statement that the loss was expected or chosen, and compare it with the date the loss began. Chapter 9 Everyone ### A worked example: 1 resident, 90 days Here is the chapter 2 resident as an expert would want the file: every event dated and cited, gaps marked. The details are invented. **Hypothetical.** Resident R.T., 84, with moderate dementia and chronic kidney disease, is admitted for rehabilitation after pneumonia. She eats independently on admission. By day 90 she is dependent for meals, has lost 12 lb, and goes to the hospital with a sodium of 152 mmol/L and acute kidney injury. **Resident R.T.: weight, intake and response** — hypothetical — Illustration 1. Day 0 **Admission weight 150 lb** Baseline care plan: "regular diet, independent with meals." Hospital weight 151 lb. Admission nursing assessment p. 3; hospital discharge summary p. 2 2. Day 5 **Dietitian admission assessment** Estimated fluid need 1,800 ml a day. No supplement. "RD to follow quarterly." Nutrition assessment p. 41 3. Day 7 to 28 **Weekly weights 149, 148, 147, 146** Same chair scale. Meal intake charted 50% to 100%. Weight log p. 88; meal record pp. 120 to 131 4. Day 34 **Aide charts "needs cueing, falls asleep at meals"** ADL record changes eating from independent to supervision. Care plan not revised. ADL flowsheet p. 140; care plan p. 60 5. Day 35 to 59 **No weights recorded** Policy calls for monthly weights. No dietitian note. Weight log p. 88 (no entries); facility weight policy p. 3 6. Day 60 **Weight 142 lb, 5.3% below admission** No reweigh charted. Nurse note: "wt down, RD aware." No physician notification found. Weight log p. 89; nurse note p. 212 7. Day 62 **Dietitian recommends supplement 240 ml 3 times a day** Also weekly weights for 4 weeks and a 3 day calorie count. Dietitian progress note p. 44 8. Day 62 to 71 **Recommendation not ordered** Supplement order signed day 71. Calorie count and weekly weights never ordered. Physician orders p. 18; MAR pp. 300 to 310 9. Day 80 to 88 **Intake and output started** Daily fluid totals 900 to 1,100 ml. Supplement charted "refused" at 11 of 27 passes. No care plan entry about refusals. I&O sheets pp. 330 to 338; MAR pp. 311 to 320 10. Day 88 **Sodium 148, BUN rising** Lab faxed 14:10. Physician notified day 89, 09:30. Order: encourage fluids. Lab report p. 402; telephone order p. 22 11. Day 90 **Weight 138 lb; transfer to hospital** Quarterly MDS codes K0300 = 0 and J1550C not checked. Dietitian quarterly note says "Wt stable." Hospital sodium 152. Weight log p. 89; MDS pp. 510 to 540; nutrition note p. 46; ED record p. 3 The timeline shows where the record has gaps. Whether the care behind them fell below the standard, and whether it caused the admission, is for the experts. The plaintiff reads an eating change on day 34 that nobody care planned, a 32 day gap between weights, a 9 day supplement delay, fluid totals under both the MDS benchmark and the dietitian's target, and an MDS and dietitian note that contradict the weight log. The defense reads an 84 year old with dementia and kidney disease after pneumonia, a dietitian who responded within 2 days of the day 60 weight, a supplement offered and refused, and a physician notified the morning after the lab. Both stories come from the same pages. The experts decide whether 1,800 ml suited a resident with kidney disease, whether cueing would have held her weight, and whether the admission followed from dehydration, the underlying illness or both. If the resident died, the same file feeds the [nursing home wrongful death records](https://medrecords.ai/guides/nursing-home-wrongful-death-records/) analysis. If you remember one thing Build 1 timeline of weights, intake, fluids, labs, notes, orders and care plan revisions, and mark every gap with the rule that says a record should exist. Chapter 10 Deciding ### Where AI helps with weight and intake records, and where it does not A 90 day stay produces about 270 meal entries, fluid totals, supplement passes, weights, MDS assessments and a stack of notes, often with paper meal sheets mixed in. The facts that matter are small and scattered: a flowsheet cell, a "refused" on a MAR, "wt down, RD aware" in a narrative note. That is where AI medical record review helps, and where it can go wrong quietly. #### What software does well here - **Finding every weight.** Medical record search and question answering tools can list weights from flowsheets, notes, K0200B and hospital records, each with a page-level citation. - **Reading scans.** OCR and handwriting recognition make paper meal sheets and handwritten notes searchable. Good tools flag low-confidence pages for a human read. - **Building the timeline.** AI medical chronology tools put weights, labs, orders, notes and MDS dates in 1 cited sequence, which turns chapter 9 into a draft you check. - **Spotting near-duplicates.** Side by side comparison surfaces copied-forward text like the dietitian note in chapter 6. - **Reading coded data.** Clinical natural language processing and ICD-10 extraction find codes such as E86.0 (dehydration), R63.4 (abnormal weight loss), E87.0 (hyperosmolality and hypernatremia) and E43, E44.0 or E46 (protein-calorie malnutrition), and show when each first appears. #### What it should not do No software should decide whether loss was avoidable, whether care met the standard, or what caused a hospitalization or death. Generative AI tools built on a large language model (LLM) can also produce fluent statements no page supports, which is what hallucination means in practice. A made-up weight in a chronology is worse than a gap, because it looks like evidence. Courts have sanctioned lawyers under Rule 11 for AI-invented citations, most famously in *Mata v. Avianca, Inc.*, 678 F. Supp. 3d 443 (S.D.N.Y. 2023). Every line needs a cite, and a person needs to check it. Manual review Weights — Found page by page; easy to miss 1 in a therapy note Meal records — Read as a stack; patterns in entry times rarely checked Copied notes — Caught only if the reader remembers the earlier note Judgment — Reviewer and expert With AI, human in the loop Weights — Pulled into 1 list, each cited to its page, then checked Meal records — Searchable; blocks of identical entries easier to spot Copied notes — Near-duplicates shown side by side Judgment — Still the reviewer and the expert #### How to choose a tool for this work These checks are in priority order. The first 2 are pass or fail. 1 — ##### A signed BAA and SOC 2 Nursing home records are protected health information. The vendor should sign a business associate agreement, hold SOC 2 and not train models on your files. 2 — ##### A citation on every line Every weight, lab, order and quote should link to its source page. Grounded citations are what let an expert rely on the output and let you check it quickly. 3 — ##### Handles flowsheets and handwriting Test it on a meal record grid, an intake and output sheet and a handwritten nurse note from your own file. Ask how it flags pages it cannot read. 4 — ##### Answers questions you can check Ask "list every weight with date and page" or "every supplement pass charted refused." The answer should be a list you can verify, not a summary you have to trust. 5 — ##### Exports your team can use Hyperlinked exports let an expert click from a timeline line to the page. That matters more here than agentic features. 2 newer questions belong in discovery. Were any chart notes drafted with an ambient AI scribe or other generative AI documentation tool, and how were they reviewed? If the other side used AI document review for its production or chronology, how was the output checked? If you remember one thing Use legal AI tools to find and cite every weight, intake entry and note, then check the cites. Judgments about avoidability and causation stay with people. Chapter 11 Building ### Records to request and 3 templates #### Getting the records A resident or representative has 2 overlapping rights. Under [42 CFR 483.10(g)(2)](https://www.ecfr.gov/current/title-42/chapter-IV/subchapter-G/part-483/subpart-B/section-483.10), the facility gives access within 24 hours, excluding weekends and holidays, and copies within 2 working days of advance notice, for a reasonable cost-based fee. Under [45 CFR 164.524(b)(2)](https://www.law.cornell.edu/cfr/text/45/164.524), a covered entity must act within 30 days, with 1 extension of up to 30 days. After a death, [45 CFR 164.502(g)(4)](https://www.law.cornell.edu/cfr/text/45/164.502) treats the executor, administrator or other person with authority under applicable law as the personal representative. Records are kept for the period state law requires or, absent one, 5 years from discharge (42 CFR 483.70(h)). See [how to get medical records for a lawsuit](https://medrecords.ai/guides/how-to-get-medical-records-for-a-lawsuit/). Name the records. A request for "the chart" often returns printed progress notes without the flowsheets, point-of-care data or MDS worksheets. The core set: | Record | Where it usually lives | What it answers | | --- | --- | --- | | Weight log and vitals flowsheet | EHR flowsheet; sometimes a paper weight book | The trend, the scale used, reweighs | | Meal intake records | Point-of-care aide documentation | Intake by meal, with entry times if produced natively | | Intake and output sheets | Nursing flowsheet or paper | Daily fluid totals against J1550C and the dietitian target | | MAR and treatment record | EHR medication module | Supplements given, refused and amounts taken | | MDS assessments and CAA worksheets | MDS software, separate from progress notes | K0100, K0200B, K0300, K0520, J1550C, triggers and care planning decisions | | Dietitian assessments and notes | EHR notes; consultant reports | Estimated needs, recommendations, follow-up | | Care plans, all versions | EHR care plan module with revision history | Goals, interventions, revision dates, team signatures | | Physician orders and visit notes | EHR orders; practitioner notes | Diet, texture, supplement and lab orders; visit timing | | Labs | Lab interface or faxed reports | Sodium, BUN, creatinine, osmolality; result and notification times | | Staffing and assignment records | Scheduling system; daily staffing postings | Who was on the unit at meals | ##### 1. Records request to the facility Use for a resident, a representative, or the personal representative of a resident who has died. Attach proof of authority. [DATE] [FACILITY NAME] Attn: Medical Records / Health Information Management [ADDRESS] Re: Request for records of [RESIDENT NAME], DOB [DATE] Dates of stay: [ADMISSION DATE] to [DISCHARGE OR DEATH DATE] I am [the resident / the resident representative / the personal representative of the estate], and request a complete copy of the records below under 45 CFR 164.524 and, for a current or former resident, 42 CFR 483.10(g)(2). Proof of authority is attached. Please produce electronic records in native or electronic format, including entry dates and times, and paper records as scanned copies: 1. Weight log and all vitals flowsheets, with scale type if recorded 2. Meal intake documentation for every meal, with entry date and time 3. Intake and output records 4. Medication and treatment administration records, including nutritional supplements and amounts consumed 5. All MDS assessments, CAA worksheets and Section V 6. All dietitian and nutrition assessments, notes and reports 7. All versions of the care plan, with revision history 8. Physician and practitioner orders, telephone orders and notes 9. All laboratory reports, with collection and result times 10. Speech therapy evaluations and diet order history 11. ADL documentation for eating 12. Nurse progress notes and change in condition notes 13. Hospice agreement and hospice communication records, if any 14. The facility's weight, hydration and dining policies in effect during the stay If any category does not exist, please say so in writing. [NAME] [CONTACT DETAILS] ##### 2. Weight and intake worksheet Fill in as you review. The response column is the one experts read first. RESIDENT: [INITIALS] USUAL WEIGHT: [ ] lb (source, page) ADMISSION WEIGHT: [ ] lb (page) HEIGHT: [ ] in (date) WEIGHT LOG Date | Weight | Scale | Reweigh? | Source page | Note THRESHOLD CHECKS (from usual weight) 1 month: loss [ ]% 5% line = usual x 0.95 = [ ] lb 3 months: loss [ ]% 7.5% line = usual x 0.925 = [ ] lb 6 months: loss [ ]% 10% line = usual x 0.90 = [ ] lb Percent loss = (usual minus actual) / usual x 100 MDS K0300 AT EACH ASSESSMENT ARD | Current wt | Wt nearest 30 d x 0.95 | Wt nearest 180 d x 0.90 | Code RESPONSE TO EACH CROSSING Crossing date | Physician notified | Dietitian note | Order | Care plan revised | Days INTAKE AND FLUIDS (weekly) Week | Meals under 50% | Refusals | Avg daily fluid ml | Target ml | Labs GAPS LOG Missing record | Rule or policy requiring it | Page showing it should exist ##### 3. Deposition topics for the dietitian or director of nursing Works for both sides. Plaintiffs test the response; defense counsel prepares the witness on the same list. 1. Weighing procedure: scale, clothing, time, reweigh trigger 2. Who reviews weights, how often, and how a loss is escalated 3. How meal percentages are estimated and when they are entered 4. The resident's estimated calorie, protein and fluid needs: basis 5. Each recommendation made: date, to whom, order date, follow-up 6. Supplement delivery: who gives it, how intake is recorded 7. Refusals: what was explained, alternatives offered, care plan entry 8. Assistance at meals: level ordered, who provided it, staffing 9. Diet texture: who ordered it, swallow evaluation, resident choice 10. MDS K0300 and J1550C coding on each assessment: source records 11. Any statement that the loss was unavoidable: who wrote it, when 12. Hospice coordination, if any: agreement, shared care plan If you remember one thing Ask for records by name and in native format. Meal entries, care plan revisions and lab notifications all carry times that disappear in a printed chart. Chapter 12 Publisher ### What Medrecords AI does with a malnutrition and dehydration file Medrecords AI is medical chronology software. It works on the records you upload; it does not request or retrieve records from facilities. On a nursing home nutrition file it does 5 things: - Builds a [cited chronology](https://medrecords.ai/product/chronology/) of weights, labs, orders, dietitian notes, MDS dates and nurse notes, with [a citation on every line](https://medrecords.ai/product/citations/), so the chapter 9 timeline starts as a draft. - Lets you [search the whole file](https://medrecords.ai/product/medical-record-search/) and [ask questions](https://medrecords.ai/product/qa/) such as "list every recorded weight with its date," each answer cited to its page. You run the threshold math. - Reads scanned and handwritten pages through [OCR](https://medrecords.ai/product/ocr/) and [handwritten record extraction](https://medrecords.ai/product/handwritten-medical-record-extraction/), and flags low-confidence pages for a person to read. - [Flags records the file implies](https://medrecords.ai/product/missing-records-identification/) but does not contain, such as a dietitian note that references a calorie count nobody produced. - Shows near-duplicate pages side by side through [record alteration detection](https://medrecords.ai/product/record-alteration-detection/), which is how a copied-forward note like the one in chapter 6 surfaces for review. It does not decide whether weight loss was avoidable, whether care met the standard, or what caused a hospitalization or death, and it does not score a case. Flags are signals for a reviewer, not findings. It runs under SOC 2 and HIPAA with a signed BAA (see [security](https://medrecords.ai/security/) and [HIPAA](https://medrecords.ai/hipaa/) ). Self-Service bills 10 cents a deduplicated page, down to 5 cents at volume, duplicates free; Enterprise On-Prem is an annual license. More: [nursing home neglect record review](https://medrecords.ai/solutions/nursing-home-neglect-record-review/) and [MDS documentation validation](https://medrecords.ai/solutions/mds-pdpm-documentation-validation/). The offer #### See a cited weight and intake timeline from your own file. Book a demo on a nursing home file you are working, then run your first case free on us. Every line comes back cited to its source page. You review, you revise, you sign. [Book a demo](https://medrecords.ai/demo/) [See the cited chronology](https://medrecords.ai/product/chronology/) Scheduling only. No records move from a public page. Chapter 13 Everyone ### Frequently asked questions What federal regulation covers malnutrition and dehydration in nursing homes?42 CFR 483.25(g), cited by surveyors as F692. It requires the facility to help each resident maintain acceptable nutrition parameters and offer enough fluid, unless the resident's clinical condition makes that impossible or the resident's preferences indicate otherwise. Food service rules are at 42 CFR 483.60. How much weight loss is significant in a nursing home?CMS surveyor guidance treats 5% in 1 month, 7.5% in 3 months and 10% in 6 months as significant, and anything more as severe. The MDS item K0300 counts 5% or more in 30 days or 10% or more in 180 days, using only 2 snapshot weights. How often should nursing home residents be weighed?Appendix PP cites weighing on admission or readmission, weekly for the first 4 weeks, and at least monthly after that, plus extra weights after a significant change or a persistent drop in intake. Are meal intake percentages reliable?Research says they tend to overstate intake. In 1 direct observation study, staff documentation overestimated total intake by 22% and missed more than half of residents eating 75% or less at most meals. Does low albumin prove malnutrition?No. CMS guidance says low albumin can come from acute illness unrelated to nutrition, may not improve even with adequate intake, and is not useful for following undernutrition over time. Weights and intake are better evidence. Is weight loss in hospice or at the end of life a defense?It can be, if the record shows it. CMS guidance asks for a documented clinical basis, such as a physician note explaining why the loss is medically unavoidable. Hospice does not end the facility's nursing and personal care duties. Can AI calculate nursing home weight loss from the records?AI tools can find every recorded weight across flowsheets, notes and the MDS and list each with its page citation. A person should check the list and run the threshold math, because a single misread weight changes the result. Can ChatGPT summarize a nursing home chart for a malnutrition case?A general chatbot gives no page citations, can invent facts, and usually has no business associate agreement for protected health information. For work an expert will rely on, use a HIPAA compliant tool that cites every line to its source page. Can software decide whether a resident's weight loss was avoidable?No. Software can organize and cite the evidence and flag gaps. Whether the loss was avoidable, whether care met the standard, and what caused an injury or death are questions for qualified experts. Chapter 14 Everyone ### Sources and method Regulations were read on eCFR in September 2026. CMS guidance was checked against the full text of Appendix PP (Rev. 232, July 2025, in which the F692 guidance carries its April 2025 Rev. 229 text) and the MDS 3.0 RAI Manual v1.20.1 (October 2025). The replacement manual, v1.20.11, takes effect October 1, 2026 and keeps the same text for the Section J, K and V items and CAAs cited here; study findings against the published abstracts; ICD-10-CM titles against the National Library of Medicine clinical tables service. Quotes are verbatim. The resident, chart, staffing figures and intake log are hypothetical. - [42 CFR 483.25(g)](https://www.ecfr.gov/current/title-42/chapter-IV/subchapter-G/part-483/subpart-B/section-483.25), assisted nutrition and hydration. - [42 CFR 483.60](https://www.ecfr.gov/current/title-42/chapter-IV/subchapter-G/part-483/subpart-B/section-483.60), food and nutrition services. - [42 CFR 483.160](https://www.ecfr.gov/current/title-42/chapter-IV/subchapter-G/part-483/subpart-D/section-483.160), paid feeding assistant training. - [42 CFR 483.20](https://www.ecfr.gov/current/title-42/chapter-IV/subchapter-G/part-483/subpart-B/section-483.20), resident assessment; [483.21](https://www.ecfr.gov/current/title-42/chapter-IV/subchapter-G/part-483/subpart-B/section-483.21), care plans; [483.30(c)](https://www.ecfr.gov/current/title-42/chapter-IV/subchapter-G/part-483/subpart-B/section-483.30), physician visits. - [42 CFR 483.10](https://www.ecfr.gov/current/title-42/chapter-IV/subchapter-G/part-483/subpart-B/section-483.10), (g)(2) record access and (g)(14) notification of changes; [483.70](https://www.ecfr.gov/current/title-42/chapter-IV/subchapter-G/part-483/subpart-B/section-483.70), (h) medical records and (n) hospice services. - [CMS State Operations Manual, Appendix PP](https://www.cms.gov/Regulations-and-Guidance/Guidance/Manuals/downloads/som107ap_pp_guidelines_ltcf.pdf): F692, F805, F807 and F811 guidance. - [MDS 3.0 RAI Manual v1.20.1](https://www.cms.gov/files/document/final-mds-3-0-rai-manual-v1-20-1-october-2025.pdf), October 2025: Sections J, K and V; CAAs 12 and 14. - [Hooper L and colleagues](https://pubmed.ncbi.nlm.nih.gov/25924806/), Cochrane Database of Systematic Reviews, 2015: 67 tests for water-loss dehydration in older people ([full text](https://pmc.ncbi.nlm.nih.gov/articles/PMC7097739/) ). - [Simmons SF, Reuben D](https://pubmed.ncbi.nlm.nih.gov/10682952/), J Am Geriatr Soc 2000;48:209 to 213: staff documentation against direct observation. - [Pokrywka HS and colleagues](https://pubmed.ncbi.nlm.nih.gov/9329485/), J Am Geriatr Soc 1997;45:1223 to 1227: staff estimates against weighed trays. - [Simmons SF, Babineau S, Garcia E, Schnelle JF](https://pubmed.ncbi.nlm.nih.gov/12242322/), J Gerontol A 2002;57:M665 to M671: low intake identification across 10 homes. - [Simmons SF, Schnelle JF](https://pubmed.ncbi.nlm.nih.gov/16776786/), J Am Geriatr Soc 2006;54:919 to 924: staff time for feeding assistance. - [45 CFR 164.524](https://www.law.cornell.edu/cfr/text/45/164.524), HIPAA right of access, and [45 CFR 164.502(g)(4)](https://www.law.cornell.edu/cfr/text/45/164.502), personal representatives of deceased individuals. - *Mata v. Avianca, Inc.*, 678 F. Supp. 3d 443 (S.D.N.Y. 2023): Rule 11 sanctions for fabricated AI-generated citations. Related guides: [nursing home records, the MDS and the CMS-2567](https://medrecords.ai/guides/nursing-home-records-mds-cms-2567/), [pressure injury lawsuit records](https://medrecords.ai/guides/pressure-injury-bedsore-lawsuit-records/), [nursing home change in condition](https://medrecords.ai/guides/nursing-home-change-in-condition/), [charting by exception](https://medrecords.ai/guides/charting-by-exception/), and [record review for legal nurse consultants](https://medrecords.ai/solutions/legal-nurse-consultants/). ### More guides - [**Nursing home records: the MDS, the CMS-2567 and every record a case needs** — MDS nursing home assessments, care plans, CNA flowsheets, MAR and TAR, CMS-2567 survey reports and staffing…](https://medrecords.ai/guides/nursing-home-records-mds-cms-2567/) - [**Nursing home wrongful death: proving the chain from care to cause of death** — Nursing home wrongful death cases turn on the chart: the death certificate, rare autopsies, hospice and DNR…](https://medrecords.ai/guides/nursing-home-wrongful-death-records/) - [**Nursing notes: what they contain and how to read them after the fact** — What goes in nursing notes and flowsheets, the federal rules behind them, charting formats, a worked example, a…](https://medrecords.ai/guides/nursing-notes/) - [**Operative report: what it contains and how to read it against the rest of the surgery** — What goes in an operative report, the records written around it, the CMS rules, what studies show about op…](https://medrecords.ai/guides/operative-report/) - [**Parts of a medical record: what each section is, what the law requires and how to tell if a file is complete** — The parts of a medical record, the legal medical record vs the designated record set, federal content rules…](https://medrecords.ai/guides/parts-of-a-medical-record/) - [**Pathology report: what each part means and how to read it after the fact** — What each part of a surgical pathology report means, amended and addended reports, CLIA rules, diagnostic…](https://medrecords.ai/guides/pathology-report/)