🧰 SAMEKTRA SERVICES →
← Healthcare Compliance
HEALTHCAREJOINT COMMISSIONCMS

Safe All Year, or Safe for Survey Week?
What happens to safety when nobody is watching

Hospitals get measurably safer the week the surveyors arrive — and then drift back. Fire does not check the survey calendar. Here is the research, and how to make survey week boring.

By Stanislav Samek, Samektra · 9 min read · Last updated October 15, 2026(Today)
Headshot of Stanislav Samek, founder of Samektra Safety Management & Training
WRITTEN BY · FOUNDER & EDITOR

Stanislav Samek

Founder of Samektra Safety Management & Training in Gwinnett County, Georgia, and the writer and editor behind LifeSafetyWiki. Has walked plenty of hospitals the week before a survey — and plenty of them on an ordinary Tuesday.

Be honest: when was your building safest this year?

If you work in healthcare, you know the moment. The email lands at 7 a.m., someone says the words “The Joint Commission is here,” and the building transforms. Carts vanish from corridors. Wedges disappear from under fire doors. The extinguisher with the blank tag gets a fresh one. Somebody finds the generator log. By lunch, the place looks like the brochure.

So here is the uncomfortable question: if your building can look like that by lunch, why didn’t it look like that yesterday?

I wanted to know whether that feeling holds up, so I did some research online — medical studies, CMS reports to Congress, NFPA fire data — and found some really interesting statistics. I’ve pulled the best ones out below as “Did you know?” facts, each with its source, so you can check them for yourself.

The short answer: survey readiness and everyday safety are not the same thing. One is a performance for an audience. The other is what happens at 2 a.m. on a holiday weekend with half the usual staff. Fire, a patient fall or a power failure only ever shows up for the second one.

The survey-week effect is real — and measurable

This is not a hunch. Researchers have actually measured what happens when the inspectors walk in.

DID YOU KNOW? · #1
Fewer deaths
Patients admitted during a Joint Commission survey week had lower 30-day mortality (7.03%) than those admitted in the weeks around it (7.21%). At major teaching hospitals it fell from 6.41% to 5.93%. Same hospitals, same staff — different week.
DID YOU KNOW? · #2
3×
Hand-hygiene rates were about three times higher when an auditor in a white coat was visible, measured by electronic sensors that never left. That is the Hawthorne effect: we perform better when watched.
DID YOU KNOW? · #3
98%
More than 98 percent of surveyed hospitals pass. Passing is the expectation. The real question is what the building looks like on the days nobody is grading it.

The researchers in the mortality study could not link the improvement to any specific item the surveyors check. Their best explanation was simple: everyone was paying closer attention. Which raises the obvious point — if paying attention saves lives for one week, it would save them for the other fifty-one too.

Why “survey mode” exists at all

DID YOU KNOW? · #4
2006
Until 2006 — the year Twitter launched — hospitals were told in advance when The Joint Commission was coming. A whole culture of “getting ready for the survey” grew up around known dates. Surveys have been unannounced for nearly 20 years, but the habit never fully died.
Source: Joint Commission policy change, effective January 1, 2006
DID YOU KNOW? · #5
18–36 mo.
Surveys typically land somewhere between 18 and 36 months after the last full one. In practice that means you are always inside the window. There is no “off season.”
Source: Survey window described in coverage of the JAMA study and TJC policy

Survey readiness made sense when you knew the date. Once surveys went unannounced, the only strategy that still works is being ready on any random day — which is just another way of saying being safe all the time.

Passing the survey doesn’t mean nobody missed anything

DID YOU KNOW? · #6
~46%
CMS sends state surveyors back in after some accreditation surveys to double-check. For hospitals, they found serious problems the accreditor had not cited in roughly 45–47% of those validation surveys. The physical environment — including fire safety — was among the areas missed most.
Source: CMS reports to Congress, FY2016–2018 hospital data
DID YOU KNOW? · #7
1 in 4
The most-cited life safety tag in nursing homes is K-353, sprinkler maintenance and testing — cited at about a quarter of providers. Number two is K-918, the emergency power system (think generator testing). Corridor doors (K-363) are in the top tier too.
Source: AHCA, nursing facility life safety data, FFY 2023

Look at that list again. None of the top findings are about money or design. They are about routine: a test that wasn’t done on time, a log that wasn’t filled in, a door that didn’t latch. These are exactly the things that slip when the building runs on survey adrenaline instead of daily habit — and exactly the things a fire will find.

The wedge that vanishes when the surveyor arrives. A propped fire door does nothing in a fire, and fires don’t send a 7 a.m. email.

Corridor storage: gone on survey day, back by Friday. 📸 Photo: Kyle Brown.

The worst survey on record wasn’t one bad day

I went looking for the worst survey outcome I could find — and I’m deliberately not naming the hospital. I learned that lesson first-hand. When I wrote about a positive Joint Commission survey at my own organization, it was flagged, and I had to rewrite it as a much more generic article. Something was lost in that rewrite: the personal, first-hand detail is what makes a story teach. But I understand why, and I respect the organizations, the people who work in them, and the readers who come here to learn rather than to judge.

So the point of this case isn’t to pile on anyone. The hospital stays anonymous here, and the public record is linked at the bottom of the case file for anyone who wants to read the full story.

CASE FILE · A LARGE U.S. PUBLIC HOSPITAL, 2004–2007
  • It lost its Joint Commission accreditation first.
  • In the summer of 2006 it failed a federal (CMS) inspection and lost its eligibility for Medicare — about $200 million a year, roughly half its budget.
  • To survive, it cut inpatient beds from about 250 to 48 and closed its physician-training programs. CMS kept funding it on a short leash while it tried to improve.
  • It didn’t hold. After another multi-day inspection in 2007, CMS revoked its federal funding on August 10, 2007, citing repeated surveys that had found serious health and safety violations. Its emergency department closed that same evening.
  • The detail that matters most: officials said the hospital had been out of compliance with federal standards since January 2004. More than three and a half years.

The worst survey ever wasn’t a surprise and it wasn’t one bad day. It was years of findings that got fixed for the inspection and came back afterwards — survey mode, at its most expensive.

Read the record: the 2006 failed inspection · the funding extension · the 2007 termination (California Healthline)

Fire doesn’t check the survey calendar

DID YOU KNOW? · #8
~5,750
U.S. fire departments responded to about 5,750 fires a year in healthcare facilities. That is roughly 16 a day — on survey weeks and every other week.
Source: NFPA estimates, 2011–2015 (via ECRI)
DID YOU KNOW? · #9
4%
Only about 4% of those fires spread beyond the room where they started. That is not luck. It is closed doors, working sprinklers and trained staff — the everyday stuff.
Source: NFPA estimates, 2011–2015 (via ECRI)
DID YOU KNOW? · #10
8,330 → 1,080
Hospital and hospice fires dropped from 8,330 in 1980 to about 1,080 in 2015. Codes and enforcement get a lot of the credit — but codes only work if they are followed on the days nobody is looking.
Source: NFPA estimates, 1980 vs 2015 (via ECRI)
DID YOU KNOW? · #11
74 lives
The deadliest hospital fire in U.S. history killed 74 people in 1949. The building had no sprinklers, no detection and no alarm, and combustible ceiling tiles. Much of the healthcare code you work under today was written in response.
Source: 1949 St. Anthony's Hospital fire, Effingham, Illinois
DID YOU KNOW? · #12
31 deaths
Two nursing home fires in 2003, in Hartford and Nashville, killed 31 residents. Neither building was required to have sprinklers. CMS then required sprinklers in every nursing home by August 13, 2013 — and more than 900 homes still weren’t fully sprinklered after the deadline passed.
Source: GAO-04-660; CMS final rule, August 13, 2008

Every rule you get cited on was paid for by somebody. The code isn’t a test you study for. It’s a list of lessons, most of them learned the hard way.

Quick self-check: survey mode or always-ready?

SELF-CHECK · 0/8 ANSWERED

For each statement: is this true all year, or mostly around survey time? Be honest — nobody is grading this but you.

  1. 1. Fire doors in our corridors close and latch on a random Tuesday at 2 a.m., not just on the morning of the survey.
  2. 2. Nobody "cleans out the corridors" when the surveyor email arrives — there is nothing to clean out.
  3. 3. Sprinkler, fire alarm and generator test records are filed the week the test happens, not assembled the week of the survey.
  4. 4. Our environment-of-care rounds happen on a schedule we do not tell the units about in advance.
  5. 5. Frontline staff can explain RACE and PASS without a laminated card in front of them.
  6. 6. Fire drills are run on every shift, including nights and weekends, when staffing is thinnest.
  7. 7. When a deficiency is found, it is tracked to a documented fix, even if no surveyor will ever ask.
  8. 8. We would be comfortable if a surveyor walked in today, unannounced, right now.
Answer all 8 to see your result.

Seven ways to make survey week boring

The goal isn’t a better survey. It’s a survey that looks exactly like every other day.

  1. Round on random days. Short environment-of-care walks on days the units don’t know about — including nights and weekends.
  2. Use observers people don’t recognize. The hand-hygiene research says familiar auditors get a performance. Swap rounders between units, or bring in a peer from another site.
  3. File records the week they happen. Sprinkler, fire alarm and generator tests (the K-353 and K-918 tags) go in the binder when done, not when the surveyor asks.
  4. Fix the boring stuff first. Doors that latch, clear corridors, current extinguisher tags. Cheap to fix, and the most cited.
  5. Drill every shift. Nights and weekends are when staffing is thinnest and the drill matters most.
  6. Track every finding to a documented fix — even the ones no surveyor will ever see.
  7. Celebrate the ordinary day. Thank the unit whose corridor is clear on a random Tuesday, not the one that cleaned up fastest when the email came.

The test that matters: if a surveyor walked in right now, unannounced, would anything need to move? If the answer is no, you are not survey-ready. You are safe — and that’s better.

Try it on yourself

Practice spotting the everyday findings surveyors write up — propped doors, blocked panels, missing tags — in Spot the Violation. The Clinical EOC track is built for nurses and unit leaders.

OBSERVATION: During an unannounced Tuesday evening environment-of-care round, the cross-corridor fire door on the 4th floor medical/surgical unit was held open with a wooden wedge, and two equipment carts were stored in the corridor. The unit had been free of both conditions during the previous accreditation survey. FIELD BASIS: NFPA 101-2012 requires corridor and smoke-barrier doors to close and latch, and limits what may be kept in a healthcare corridor; CMS cites these conditions under the corridor door and means-of-egress K-tags (e.g., K-363). The requirement applies continuously, not only during surveys. RISK: A wedged door cannot contain smoke or fire, and corridor obstructions slow the horizontal evacuation that defend-in-place depends on. Conditions that only disappear during surveys indicate a habit, not a one-time lapse. RECOMMENDED CORRECTIVE ACTION: Remove the wedge and relocate the carts. Re-educate unit staff on why the door must close. If the door must stay open for workflow, evaluate a listed hold-open device released by the fire alarm. Add the unit to the random-round schedule and document the repeat check.

SUGGESTED PROMPT

“How can our hospital stay ready for a Joint Commission or CMS life safety survey all year instead of scrambling when the surveyors arrive? What are the most commonly cited life safety findings and how do we prevent them day to day?”

Frequently Asked Questions

Does The Joint Commission still give hospitals notice before a survey?
No. Since January 1, 2006, regular Joint Commission accreditation surveys have been unannounced. Before that, hospitals knew the dates in advance, which is where the culture of "survey preparation" grew up. Today the organization typically learns a survey is starting that morning. Survey cycles generally fall within a window of roughly 18 to 36 months after the previous full survey, so in practice you are always somewhere inside the window.
Is it really true that patients do better during survey weeks?
A large 2017 study in JAMA Internal Medicine found a small but measurable drop. Looking at Medicare admissions at 1,984 hospitals between 2008 and 2012, 30-day mortality was 7.03% for admissions during survey weeks versus 7.21% in the surrounding weeks, and the effect was larger at major teaching hospitals. The researchers could not tie it to any specific inspected item and suggested heightened vigilance under observation as the likely explanation. It is an observational study, so it shows an association rather than proof of cause, but the timing of unannounced surveys makes it a strong natural experiment.
What is the "Hawthorne effect" and why does it matter for life safety?
It is the tendency for people to perform better when they know they are being observed. A 2014 study using electronic monitoring of hand-hygiene dispensers found hand-hygiene rates were about three times higher when an auditor was visible. For life safety the lesson is the same: a corridor, a fire door or a storage room looks its best when someone is watching. The only reliable measure is what the building looks like when nobody is.
If we pass our survey, doesn’t that prove we are safe?
It proves you met the standard on the days the surveyors looked. CMS runs validation surveys after accreditation surveys, and for hospitals it reported that state surveyors found serious (condition-level) problems the accreditor had not cited in roughly 45 to 47 percent of validation surveys in fiscal years 2016 to 2018 — with the physical environment, which includes fire safety, among the areas missed most. A pass is a snapshot, not a guarantee.
What life safety findings are cited most often?
The boring ones. In nursing home survey data for federal fiscal year 2023, the most frequently cited life safety tag was K-353 (sprinkler system maintenance and testing), cited at about a quarter of providers, followed by K-918 (essential electrical system, which includes generator testing), with K-363 (corridor doors) also in the top tier. Hospitals are not identical, but the same tags appear on hospital lists year after year. None of these require a capital project; they require someone doing the routine work on time and writing it down.
How do we stay ready all year without burning people out?
Make readiness a routine instead of an event. Run short environment-of-care rounds on random days, including nights and weekends. Use observers staff do not recognize for some of them. File test records the week they happen. Track every deficiency to a documented fix even if nobody will ask. Run drills on every shift. When these are habits, survey week stops being a fire drill of its own and becomes an ordinary day.
FROM SAMEKTRA
Need a professional set of eyes on this?

Our own team handles hospital EOC/TJC readiness walk-throughs, plan review, and OSHA program development — Samektra staff on-site, not a vendor referral.

References

1. Barnett ML, Olenski AR, Jena AB. Patient Mortality During Unannounced Accreditation Surveys at US Hospitals. JAMA Internal Medicine 2017;177(5):693-700. Medicare admissions at 1,984 surveyed hospitals, 2008-2012: 30-day mortality 7.03% during survey weeks vs 7.21% in non-survey weeks (adjusted difference -0.12%); major teaching hospitals 6.41% to 5.93%, a 5.9% relative decrease. See also STAT News coverage.

2. Srigley JA, Furness CD, Baker GR, Gardam M. Quantification of the Hawthorne effect in hand hygiene compliance monitoring using an electronic monitoring system. BMJ Quality & Safety 2014;23(12):974-980. PubMed 25002555 — hand-hygiene rates about threefold higher when an auditor was visible.

3. The Joint Commission moved regular accreditation surveys to unannounced on January 1, 2006 (approved March 2003, piloted 2004-2005). Trade coverage: HME News; Relias Media.

4. CMS reports to Congress on accrediting-organization performance (validation surveys). Hospital disparity rates of 46%, 45% and 47% for FY2016-FY2018 on CMS QCOR; physical environment, including fire safety, among the most-missed areas — PSQH summary and PSQH.

5. American Health Care Association, Life Safety Code tips (FFY 2023 nursing facility citation data): K-353 cited at 25.70% of providers (rank 1), K-918 at 19.50% (rank 2), K-363 at 14.40% (rank 6). Skilled nursing data, not hospital-specific.

6. NFPA estimates for U.S. health care facilities, 2011-2015 (via ECRI): about 5,750 structure fires per year; cooking 65% of fires; 4% spread beyond the room of origin; hospital and hospice fires down from 8,330 in 1980 to about 1,080 in 2015. See also HFM: codes and enforcement credited with decline in fires.

7. 1949 St. Anthony’s Hospital fire, Effingham, Illinois — 74 deaths (some sources say 77); no sprinklers, detection or alarm; combustible wall and ceiling finishes. Healthcare Facilities Today.

8. GAO, Nursing Home Fire Safety: Recent Fires Highlight Weaknesses in Federal Standards and Oversight (GAO-04-660) — the 2003 Hartford and Nashville fires (31 deaths combined). CMS final rule, 73 Fed. Reg. (August 13, 2008), requiring automatic sprinklers in all long-term care facilities by August 13, 2013; more than 900 homes were still not fully sprinklered after the deadline (Georgia Health News).

9. California Healthline coverage of the case file: failed CMS inspection (2006); funding extended through March 2007; federal funding terminated, August 10, 2007. Bed and program reductions as summarized on Wikipedia.

DISCUSSION
Be the first to contribute.

Open the discussion panel to comment, flag an inaccuracy, add field experience, or ask a question. Approved contributions earn SRP and may be incorporated into the article.