How to Read a Nursing Home Inspection Report

What a nursing home inspection report actually is

It is a form: the CMS-2567 Statement of Deficiencies and Plan of Correction. When state surveyors inspect a Medicare- or Medicaid-certified nursing home for the federal government, every rule they find broken is written up on it. The left column holds the surveyor's findings. The right column holds the facility's response — what it says it will do about each one. Both columns are public, and both are worth your time.

The document exists because certification is conditional: a nursing home participates in Medicare and Medicaid only by meeting federal requirements, and the survey is how that is checked. CMS's own consumer guidance points families to inspection information as a core part of evaluating a home 1Ref 1Centers for Medicare & Medicaid Services (2025).Finding a Nursing Home.That CMS directs families evaluating a nursing home to use inspection information as part of the process — establishing the inspection report as a core consumer resource rather than an internal regulatory artifact.. This is the raw material behind everything else — the star ratings, the summaries, the marketing.

The star rating is a summary of this document. This document is not a summary of anything. When they disagree, the report is the primary source.

What makes it valuable also makes it hard: it is written by an inspector for a regulator, not by a journalist for you. It uses tag numbers instead of headings, letters instead of adjectives, and a house style that describes a catastrophe in the same flat register it uses for a mislabeled food container. The skill is not reading comprehension. It is knowing where the signal hides. Finding inspection reports online is a separate short exercise, and the document is free.

The three-year window and the kinds of inspection

Before reading any single citation, orient yourself in time and type. The public record carries roughly the last three years of inspections for each facility — health surveys, fire-safety surveys, infection-control surveys, and complaint investigations — along with the dates each occurred and how many citations each produced 2Ref 2Centers for Medicare & Medicaid Services (2026).Survey Summary (Nursing homes including rehab services dataset).That the public per-facility record covers roughly the last three years of health, fire-safety, infection-control, and complaint inspections, including the dates of each inspection and counts of citations — the window and the inspection types a family is reading across.. A report is a snapshot of a few days. Three years of reports is a pattern, and the pattern is what you are actually shopping for.

The kinds are not interchangeable, and families routinely read one thinking it is another.

• The standard health survey. The comprehensive, unannounced inspection of resident care, on a recurring cycle. This is the main event and carries the most weight in how a facility is scored.
• The fire-safety survey. Conducted against the Life Safety Code — sprinklers, alarms, exits, doors. Serious, but a different question from whether anyone turns a bedbound resident.
• Infection-control surveys. A focused look at infection prevention.
• Complaint investigations. Triggered when somebody — a family, a staff member, a resident — reported something specific. Targeted, not comprehensive.

The complaint reports are where families under-read. A standard survey tells you how a home performs when an inspection is due. A substantiated complaint tells you what a real person found bad enough to report, and that a regulator confirmed it. A home with clean standard surveys and a thick complaint file is telling you something about the gap between how it performs and how it presents. Check the dates first, always: a glowing report from thirty months ago describes a building that may have changed hands, administrators, and its entire nursing staff since.

F-tags: the code that says what was broken

Every citation opens with an F-tag — an F followed by three digits. It is a pointer to the specific federal regulation the surveyor found violated, and it is the report's table of contents. The tag tells you the subject: which requirement was broken. It tells you nothing about how bad the breach was. That is the letter's job, and confusing the two is the most common beginner error.

The tags cover the full territory of a nursing home's obligations — resident rights, freedom from abuse and neglect, care planning, quality of care, nursing services, pharmacy, food, infection control, physical environment, administration. A single tag can be cited for something trivial or something fatal, which is why the tag alone is never a verdict. Understanding what a nursing home deficiency is at the level of the underlying rule is useful background, and the tag is where you look it up.

Read the tags as a shape rather than a list. Which tags repeat across three years is more informative than any individual entry.

• Repeat tags are the loudest thing in the document. The same tag cited in consecutive cycles means the home was told, agreed to fix it, and did not. That is a management signal, and it survives changes in administrator and marketing.
• Clusters mean more than counts. Several tags across care planning, nursing services, and quality of care in one survey usually describe a single underlying problem — often too few hands — refracted into separate rule violations.
• Abuse, neglect, and resident-rights tags deserve their own read. Whatever the letter says, read every word of those narratives.

Scope and severity: the letter that carries the meaning

Beside each F-tag sits a single letter from A to L, and it is the most information-dense character in the document. It plots the citation on two axes at once: severity — how much harm was done or risked — and scope — how many residents were affected. Learning to read that one letter is roughly ninety percent of learning to read the report, and it takes about two minutes.

Severity runs in four bands, and the jump that matters is from potential to actual.

Scope runs across three columns: isolated (A, D, G, J), pattern (B, E, H, K), and widespread (C, F, I, L). So the letter names both facts at once. A D is a one-off with potential for harm — the most commonly cited letter and rarely a reason to walk away. A G means somebody was actually hurt. A J means immediate jeopardy: the most serious finding a surveyor can make, requiring the home to act at once.

The line between F and G is not one step. It is the line between what could have happened and what did happen to a real person.

A subset of citations in certain care-related regulatory areas is additionally classified as substandard quality of care, which triggers extra consequences for the facility. If you see that phrase, it is not boilerplate. The full scope-and-severity grid rewards a closer look, but the four bands above will carry you through almost any report. What you are looking for is the shape of the worst entries: any G or above, anything in the J-to-L band, and whether the serious ones cluster around one theme or one unit.

The narrative is the part that matters

Under each tag and letter sits a block of prose describing what the surveyor observed, and this is where the document stops being bureaucracy and starts being evidence. It is the only place the actual event is recorded. Two citations carrying the identical tag and the identical letter can describe completely different homes, and the only way to know which is to read the paragraph.

Surveyors write in a specific deadpan: what they saw, what the record said, what staff told them, with residents anonymised as "Resident #1" and staff as "Employee A." There are no adjectives. The horror, where there is horror, arrives in the form of a timestamp.

What to read for.

• The gap between the record and the observation. "The care plan directed repositioning every two hours. Observation on 3/14 from 0900 to 1300 revealed the resident was not repositioned." That sentence pattern — the plan said, the surveyor saw — is the engine of most serious citations.
• What staff said when asked. Interview quotes are the most revealing lines in the document. A charge nurse saying she was unaware of the care plan, or that there was nobody available to help, is describing the system rather than the shift.
• Time. Four hours unrepositioned. A call light unanswered for fifty minutes. A wound not assessed for six days. The numbers in the narrative are the story.

The payoff is that the narrative converts an abstraction into a question you can ask out loud. You do not walk into a tour saying "I see you have an F689." You say: "I read that a resident here wasn't repositioned for four hours in March. What changed after that?" Nobody expects the second question, and the answer tells you a great deal — this is where a nursing home red flags list stops being generic and becomes about this building.

The plan of correction, and why it is not proof of a fix

The right-hand column is the facility's answer, written by the facility. A plan of correction is required for every citation, it is not an admission of anything, and it is emphatically not evidence that the problem went away. It is a promise, filed on a deadline, by the party with the strongest possible incentive to make the paperwork close. Read it that way and it becomes useful.

A plan is supposed to say four things: what was done for the resident affected, how the facility identified others at risk, what systemic change prevents recurrence, and how it will monitor. Reading a facility's plan of correction well is mostly asking whether the last two answers are real.

• "The staff member was re-educated." The most common sentence in the genre and the least reassuring. It locates the fault in one person and changes nothing structural. If understaffing produced the failure, education does not staff the floor.
• "The Director of Nursing will audit weekly for four weeks." Ask what happens in week five. Monitoring that expires on a schedule is monitoring designed to end.
• A specific systemic change — a new process, a changed assignment, a checkpoint that catches the failure — is what a serious plan looks like.

A plan of correction is a promise, not a receipt. The evidence that it worked is whether the same tag appears in the next survey cycle — and that is in the next report, not this one.

The federal record on this is sobering. CMS runs a Special Focus Facility program for the poorest-performing homes, with heightened inspections and enforcement until they improve enough to graduate. The HHS Office of Inspector General found that within three years of graduating, 64% of those facilities received a serious deficiency 3Ref 3U.S. Department of Health and Human Services, Office of Inspector General (2025).CMS's Special Focus Facility Program for Nursing Homes Has Not Yielded Lasting Improvements.The OIG finding that 64% of facilities received a serious deficiency within three years of graduating from CMS's Special Focus Facility program (2013-2022) — evidence that documented correction under intensive federal oversight frequently does not hold, which is why a plan of correction is not proof of a fix.. Most slid back despite the most intensive corrective attention the federal system offers. Against that, a single promise to re-educate a staff member deserves the weight it earns.

Reading a whole record instead of one citation

One report is an anecdote. Three years of reports is a character study, and the character is what you are choosing. Read the stack chronologically, with an eye on trajectory: is this home getting better, getting worse, or oscillating? A facility whose serious citations are three years old and whose recent cycles are clean is a very different proposition from one whose bad year was last year.

That instinct matches how the federal scoring itself works. CMS's methodology for the health-inspection portion of the star ratings weights findings by both scope-and-severity and recency, so more recent surveys count more heavily than older ones, and complaint findings are folded in alongside standard surveys 4Ref 4Centers for Medicare & Medicaid Services (2025).Design for Care Compare Nursing Home Five-Star Quality Rating System: Technical Users' Guide.That the health-inspection domain is scored by weighting deficiencies according to scope-and-severity and by recency, with more recent surveys counting more heavily than older ones and complaint findings included alongside standard surveys — the principle behind reading a record for recency and severity rather than raw citation count.. You do not need the arithmetic. You need the principle: recent and severe outweighs old and minor, and the raw citation count is close to meaningless on its own.

Four questions that get you through any stack:

• What is the worst letter in the last three years, and what happened? Start at the top of the severity scale and work down. If there is a G or above, that narrative is your first read, not your last.
• Does anything repeat? The same tag across cycles is the single most predictive thing in the document.
• Which direction is it moving? Order the surveys by date and look at severity, not volume.
• Do the complaint findings contradict the standard surveys? When they do, believe the complaints.

What the report cannot tell you. It only captures what a surveyor happened to observe in a few days, and says nothing about whether the staff are kind. It lags — a home that changed ownership last quarter has a record describing a previous regime. And it is silent on whether your specific parent, with their specific needs, will be all right here. Vetting a nursing home properly means reading the report alongside the staffing data, the ownership record, and your own eyes.

What to do with what you found

The report's value is realised in the room, not on the screen. Take the two or three worst narratives from the last three years, write the dates on a card, and bring them to the tour. Every serious citation is a question with a name and a date attached, and a facility's response to being asked is itself data — sometimes better data than the citation was. Medicare's official visit checklist covers what to observe in the building 5Ref 5Centers for Medicare & Medicaid Services (2022).Questions to Ask When You Visit a Nursing Home (Nursing home checklist).That Medicare publishes an official checklist of questions to ask and items to observe during a nursing home visit — the in-person counterpart to the inspection record, used here as the tool a family carries alongside the report.; the report tells you what to ask about it.

How to ask without turning it into an ambush. Curious beats prosecutorial; the point is to hear how they think, not to win. "I read the March survey. Can you walk me through what changed?" is enough. Then listen for which of three answers you get:

• A specific account — what happened, what they changed, how they know it held. This is the good answer, and it is not rare.
• A deflection — every home gets cited, the surveyor was aggressive, it was paperwork. Sometimes true. Rarely the whole story, and never when it is the entire answer.
• Surprise. An administrator who does not know their own survey history is telling you something they did not mean to.

Get a second reader. Every state runs a Long-Term Care Ombudsman program that advocates for residents of nursing homes, board-and-care, and assisted living, and works to resolve complaints about their health, safety, welfare, and rights 6Ref 6Administration for Community Living (HHS) (2025).Long-Term Care Ombudsman Program.That every state operates a Long-Term Care Ombudsman program advocating for residents of nursing homes, board-and-care, and assisted living, and working to resolve complaints about their health, safety, welfare, and rights — the free, independent reader a family can bring an inspection report to.. Ombudsmen are in these buildings regularly, they are free, they do not work for the facility, and they are who to call when a report says something you cannot interpret — before a move and after one.