Claims ledger

The six labels

The vocabulary is closed. There are six labels and no others, and a row carries exactly one. The full definitions are published at Editorial standards; what follows is the working form.

• Traced

  We obtained the document and it contains the claim.

• Corroborated

  Two or more independent primary sources, obtained, agree.

• Scope mismatch

  The source is real and states the number, about a different population than the claim asserts.

• Chain broken

  The cited source was obtained and does not contain the claim.

• Chain unobtainable

  The source is named and specific but could not be retrieved — dead URL, non-public report, paywall.

• No chain found

  The claim circulates with no citation we could find anywhere. No row on this page carries this label today.

A label describes the state of the paper trail on the date in the row. It is not a verdict on the number. Where a claim is also contradicted by better evidence, that is a separate sentence in the row with its own citation.

How a chain was walked

A hop is one document. For each hop we record what the document is, what our client got when it asked for it, the date we asked, and what the document did or did not contain. Where a publisher's server refused us, the refusal is recorded as a fact about the retrieval and not as a fact about the document: an HTTP 403 means we were blocked, not that the text is missing. Where we read a document from an Internet Archive capture rather than from the publisher, the row says so and gives the capture date.

Absence is scanned, not asserted. Where a row says a document does not contain something, there is a keyword scan on this page with the character count of the text scanned, the terms scanned, and the date. Anyone can repeat it. The full method, and where it could be wrong, is at Method.

This ledger covers figures in circulation about surgical and medical waste and its regulation. It is not a register of products, manufacturers or device clearances, and it will not become one.

“CDC says only 3–5% of hospital waste requires disposal as regulated medical waste”

The chain

The same sentence also appears in Practice Greenhealth's Less Waste How-to Guide and on its live page about reducing regulated medical waste, in both cases with no citation at all.16 Practice Greenhealth's HTML pages returned 403 to our client on 2026-08-06; the pages were read during the research behind this ledger on the same date.

The scan

What the cited document contains instead

What the record does support

Those three sentences, which are CDC's and are quotable.2 The third is the one worth carrying into a policy argument: CDC says the determination is a matter of judgment because no reasonable standards for making it have been developed. That is a stronger and more useful statement than the orphan percentage, and it is in the document. For a figure about red-bag contents specifically, the traced number is 11–15%, with the limits stated in that row.

This would change if

Someone sends us a CDC publication — an MMWR report, a HICPAC guideline, a NIOSH document — containing a 3–5% figure for the share of hospital waste requiring disposal as regulated medical waste, with a page or section number. We will print it and change the label the same week. An archived copy of the trade page at the address Hsu cited, as it stood in 2019, showing what it said and what it cited, would also move this row. [email protected]

“Up to 85% of red bag waste is not infectious”

The chain

What the cited document says

The subject of that sentence is everything a health system throws away, everywhere. The subject of the claim in circulation is what is inside a red bag in a US hospital. Those are not the same denominator, and the difference is the whole of this row.

What the record does support

WHO's sentence, as WHO wrote it and with its scope attached: about all waste from health-care activities, globally, and unsourced on WHO's own page.11 For red bags in the United States, the figures that survive tracing are Hsu 2020 and Sharma 2024. Note that “about 85%” is right for red bags only as the arithmetic complement of Hsu's 14.9% at a single site — that is the same study restated, not independent confirmation of the WHO number, and it should not be cited as though it were.

This would change if

WHO publishes the source of its 85%, which would let the figure be traced rather than merely quoted; or a study opens red bags in a US operating room or inpatient unit and reports a percent-correct figure, which would give the claim in circulation a document of its own. We would add either the day we saw it.

“The operating room generates 20–33% of a hospital's waste”

The chain

Both citations are real papers by real authors on adjacent subjects. Neither measured a hospital's total waste, so neither could report the operating room's share of it. A composition study tells you what OR waste is made of; it does not tell you how much of the hospital's waste the OR produces.

What the record does support

That the operating room's share of hospital waste is not an established number. The field says so itself, in the introduction of a 2023 scoping review:

A reader who needs a citable statement about OR waste can cite that range, with its breadth intact, instead of a point estimate whose two sources do not contain it.9

This would change if

Someone sends a study that measured OR waste and total hospital waste in the same facilities over the same period and reported the ratio, or an accessible copy of the 2009 commentary showing what it measured and what it cited. Either would give this figure a document. Until then the honest form of the sentence names the range and its uncertainty.

“60% of a hospital's regulated medical waste comes from the operating room”

The chain

The chain does not break here so much as stop. If the 2001 report were obtained tomorrow, what it would establish is the share of regulated medical waste attributable to the operating theatres of one military hospital, measured a quarter of a century ago. That may be a true number about that building. It is not a benchmark, and the module does not present it with the caveats that would make it one.

What the record does support

Nothing about the operating room's share of a hospital's regulated medical waste, from this chain. What is available from the same publisher, and is retrievable, is its benchmark series — see the benchmark row — which reports regulated medical waste as a share of a hospital's total waste, not the operating room's share of the regulated stream. Those are different quantities and substituting one for the other is the error this row exists to prevent.

This would change if

A copy of the June 2001 IERA report reaches us. It is a US government contractor report, it was written to be circulated, and somebody has one. A multi-site study reporting the operating room's share of a hospital's regulated medical waste would move the label further and faster.

“Regulated medical waste should be no more than 15% of the waste stream”

The chain

Three different thresholds for the same quantity have been published by the same organisation, and each can be pointed at a document. The 2011 module's footnote 3 gives 15% or less. The orphan 3–5% figure appears in the same publisher's Less Waste How-to Guide without a citation — see the first row on this page. The live page on regulated medical waste says “less than 8 percent”.16 Meanwhile the organisation's own benchmark tables put the median at about 6% and the best decile at about 3%.1415 In our assessment those numbers are not reconcilable with one another as targets, and a reader quoting any of them should say which document it came from and what year the data behind it were collected.

What the record does support

The benchmark tables, which are retrievable and dated — see the benchmark row. They report what a self-selected group of hospitals achieved in a stated collection year. That is a different kind of statement from a target, and it is the one the record can carry.

This would change if

A copy of the 2010 Sustainability Benchmarking Report reaches us, from a library, a conference bag or an old hard drive. We would publish what its footnote 3 actually rests on and change the label accordingly.

“OSHA requires a sharps container to be replaced when it is three-quarters full”

The chain

One hop. We retrieved the full codified text of 29 CFR 1910.1030 from the eCFR on 2026-08-06 (HTTP 200; title 29 issue date 2026-08-04) and scanned it. There is no percentage in it.

What the standard does contain

Four container criteria and two duties: replace routinely, and do not allow the container to overfill. “Routinely” and “overfill” are the operative words, and neither is a number.

What the record does support

The criteria and the two duties above, quoted, which is what a facility is actually held to under the federal standard.1 Where three-quarters comes from is a fill line moulded into the container and printed in the manufacturer's instructions for use — manufacturer documentation, which is a different kind of obligation with a different enforcement route. That mechanism is the subject of a separate article and is not a claim this row makes.

This row says nothing about whether three-quarters is the right place to stop filling a container. The finding is about the citation, not the practice. Someone defending a fill policy at a survey is better off knowing which instrument they are standing on.

This would change if

Anyone points to a percentage in the codified text of 1910.1030, or to an OSHA standard, directive, compliance instruction or letter of interpretation that sets a fill percentage for sharps containers. We will print it verbatim and change the label.

“OSHA prohibits pouring blood down the drain”

The chain

One hop, the same retrieval as the row above: the full codified text of 29 CFR 1910.1030 from the eCFR, 2026-08-06, HTTP 200. The standard has no disposal-destination provision to break, which is the finding.

What the standard says about disposal

That is one sentence and it is the whole of OSHA's codified position on where regulated waste goes. It is a deferral: it points at federal, state and local law and stops. The claim in circulation asserts a prohibition the text does not contain — and so does the opposite claim, that OSHA permits discharge to a sewer. Both invent a provision. The standard does neither.

What the record does support

The sentence above, quoted.1 Because it defers, the question is decided further down: by the state medical-waste rule, then by the sewer-use ordinance of the publicly owned treatment works receiving the discharge, then by facility policy. Those are the subject of two maintained registers on this site — Register A, state rules on liquid blood to the sanitary sewer, and Register B, sewer-use ordinances of large publicly owned treatment works.

This would change if

OSHA amends 1910.1030 to address disposal destinations, or anyone points to an OSHA standard, directive or letter of interpretation that prohibits or permits discharge of liquid blood to a sanitary sewer. We treat letters of interpretation as interpretation and not as the standard, and a row would say which it was.

The OSHA letter of interpretation behind the fluid-waste engineering-controls argument

The chain

This row is labelled Traced because the document is real, is OSHA's, and is on the subject it is cited for. What the record adds is who asked. The letter is a reply to an inquiry from a manufacturer of the category of equipment the letter is read as favouring, and the secondary literature that quotes it does not say so.

Two things follow, and only two. The first is a correction to the citation: the letter is dated 20 July 2000, and sources that date it from the osha.gov URL slug are citing the address rather than the document. The second is OSHA's own standing position, from a letter written five years earlier to a different vendor who had asked the agency to confirm that a product label complied with the standard:

What the record does support

That the letter of 20 July 2000 is a genuine OSHA letter of interpretation and may be cited as one, at its correct date.12 That the correspondent was a manufacturer of enclosed fluid-waste systems is a fact about how a framing entered the field; it is not evidence that the framing is wrong, and this row does not say it is. And that OSHA states in writing that it does not approve or endorse products.13 A reader holding a vendor's compliance claim can cite that sentence directly.

This would change if

OSHA publishes the underlying inquiry, which would show what question the agency was answering; or a second agency statement to the same effect emerges that was not written in reply to an interested party, which would make the provenance point moot.

“About 11–15% of what goes into a red bag actually meets regulated-medical-waste criteria”

The chain

The limits, which travel with the figure

Both US studies are single-site, small-n and short in duration: one department for one day, and one clinic for thirty. Neither was conducted in an operating room. Hsu's paper prints its own figure two ways — 14.9% in the abstract, and in the results “only 15% (7.45 kg) of the waste disposed in red bags met the criteria for RMW”, that second form excluding sharps containers, which the authors state were not individually audited for safety reasons. We reproduce both rather than choosing one.3

Sharma's 11% is a pre-intervention baseline measured in a practice that then improved to 69.4% after two education sessions, so it is a starting point rather than a steady state.4 And in the searches behind this ledger we found no US operating-room or inpatient study that opened red bags and reported the share of contents meeting regulated-medical-waste criteria. That is a real gap in the literature, it is the setting most readers of this site work in, and any use of 11–15% in an operating-room argument is an extrapolation across settings.

What the record does support

Two independent US teams, in different care settings, arriving at 11% and 14.9%.34 That is genuine corroboration and it is the best-evidenced anchor available for the red-bag question, provided the limits above are carried with it every time it is used. The UK audit points the same direction under a different legal definition and should be described as consistent rather than confirmatory.5

This would change if

A US operating-room or inpatient red-bag audit is published — that would close the setting gap in either direction. A multi-site study with a sampling frame, rather than a convenience site, would move the label's strength. So would a well-conducted audit that found a materially different share, which we would publish next to these two rather than in place of them.

“Regulated medical waste is a few per cent of a hospital's waste by weight and about a third of its waste disposal cost”

The chain

One hop, twice. Practice Greenhealth publishes the figures itself, in report PDFs that served our client normally on 2026-08-06 even though the site's HTML pages returned 403. We read two of them in full and the table below is transcribed from those two documents.

There is a naming trap in this series and it has to be stated before the numbers are used. Practice Greenhealth titles each report by its award year, not by the year the data describe. The report published as 2023 Sustainability Benchmark Data says in its methods section:

So the columns below are labelled by collection year, not by report title. Anyone citing “the 2023 benchmark report” for a 2023 figure is off by a year.

The research behind this ledger also read the reports covering collection years 2021, 2023 and 2024, which put the median regulated-medical-waste share at 6.5%, 6.2% and 6.1% and the 90th percentile at 3.0%, 2.5% and 3.1%. Those reports are not in the table because we could not link a retrievable PDF for each of them on this date, and a table cell on this site carries the document it came from. Practice Greenhealth itself attributes the CY2023 declines to data collection challenges resulting from a change to a waste contractor's reporting system, so a CY2023 figure should be treated as possibly artifactual wherever it is used. The per-adjusted-patient-day and per-staffed-bed benchmarks were discontinued after the CY2023 report.

What the record does support

That in two separate cohorts of award-applicant hospitals, in collection years 2020 and 2022, regulated medical waste was about 6% of waste by weight and about a third of waste disposal cost, and the best decile reached about 3% by weight.1415 That is the best-sourced framing available for the cost argument, and it is the one figure in this ledger a reader can hand to a value-analysis committee with a document behind it — provided the collection year and the self-selected award-applicant cohort are stated in the same sentence, every time.

This would change if

A census-based dataset appears — a state or federal reporting requirement covering all hospitals rather than volunteers — which would let the cohort caveat be dropped. Or Practice Greenhealth restates the CY2023 figures after the reporting-system problem, which would tell us how large the artefact was.

What is deliberately not on this page

Four further claims were chased during the research behind this ledger and none of them has a row. In two of them the document the claim is attributed to was retrieved and does not contain the number. In the other two the trail ends at a publication reporting a figure from a source we could not obtain in any form.

We are not printing them, not even as examples of what a broken chain looks like. Printing an untraceable sentence in order to label it untraceable puts the sentence back into circulation with our name attached to it, and a reader who six months later remembers the number and not the label is worse off than one who never met it. That is a judgment about harm rather than about evidence, and we are stating it as ours. If you have run into a figure you cannot source and want to know whether we chased it, write and ask.

There is also no row here about any device, any manufacturer or any regulatory clearance. This ledger is about numbers in circulation, and it is not the place to relitigate a product.

Documents named above that we could not link

Every entry in the Sources list below has a URL and an access date. These do not, and are listed separately rather than being papered over with a link to a page about them:

• Practice Greenhealth, Greening the OR module (2011) — read during the research behind this ledger; no public URL our client could retrieve on 2026-08-06.
• Practice Greenhealth, Sustainability Benchmarking Report, 2010 — not located in any public or archived form.
• US Air Force IERA, Medical Waste Incinerator Waste Management Plan — Malcolm Grow Medical Center, Building 1056, Andrews Air Force Base, MD, June 2001 — the URL printed in the citing document is dead; no copy located.
• Healthier Hospitals, Less Waste How-to Guide — read during the research behind this ledger; no retrievable public URL on 2026-08-06.
• Esaki & Macario, 2009 — a Medscape commentary behind a registration wall; not obtained.

If you hold any of these and can send a copy, or a working address for one, that is the single most useful thing anyone could do for this page. [email protected]