The Operative Sentence

Six figures that anchor the surgical-waste business case, traced to the documents they came from

Four of the six chains end at a document that does not contain the figure, or at one nobody can obtain; two figures survive.

In brief

• The claim that only 3–5% of hospital waste requires disposal as regulated medical waste is attributed to CDC. We found no percentage of any kind in the section of the CDC guideline it is cited to: zero occurrences of “%” and zero of “percent” in 20,791 characters.2
• The 85% figure is real and is WHO's, and it is about a different denominator: all waste from health-care activities, worldwide. We found no source cited for it on the fact sheet.7 The field's own systematic review carries it forward as a statement about United States hospitals.8
• The operating room's share of hospital waste may lie anywhere in a wide band: the two papers cited for the familiar 20–33% range are a 27-case composition study and a breathing-circuit recycling study, and a 2023 scoping review puts the published spread at 20% to 70%.12
• Two figures survive tracing. In the two US audits we retrieved, 11–15% of red-bag contents met regulated-waste criteria — low-certainty evidence, both single-site and small;315 and Practice Greenhealth's benchmark tables report regulated medical waste at about 6% of hospital waste by weight against about a third of waste disposal cost, in stated collection years and for a self-selected cohort.1819
• We found no US operating-room or inpatient study that opened red bags and reported a percent-correct figure. One audit team recorded in its limitations that it could not open the containers to check.17

How a chain was walked, and what a break means

For each figure we started where a reader meets it, opened the document it points at, and kept opening documents until the figure appeared or the trail stopped. Each hop names the document, what our client got, and the date it asked. A refusal is a fact about the retrieval, not about the text: an HTTP 403 means we were blocked, not that a sentence is missing. Where this page says a document does not contain something, there is a keyword scan with the terms, the character count and the date, so anyone can repeat it. Every quotation was pulled from the retrieved file and read; none came from a search result or a summariser.

A broken chain is a finding about the record, not about the world. Nothing below says a figure is false. A number we could not trace may still be correct, and several of these probably are. What we report is that the evidence is not where the citation says it is — a smaller statement, and the one a person defending a number in a meeting actually needs.

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

The usual proximate source is Hsu 2020, a 24-hour physical waste audit of an emergency department. Its reference 20 is not a CDC publication.3

While our waste audit revealed that over 10% of total ED waste was disposed of as RMW, the Centers for Disease Control and Prevention (CDC) suggests that only 3–5% of hospital waste requires disposal as RMW.²⁰

20. Sustainability Roadmap for Hospitals. Reduce Regulated Medical Waste (RMW) Generation. [Accessed December 6, 2019]. Available at: http://www.sustainabilityroadmap.org/pims/42

Read the full paper

That address is a hospital-association trade webpage. Asked for it on 2026-08-06, our client received a 301 to a general American Hospital Association sustainability page, where the cited content is not present.4 The Internet Archive holds the page as it stood on 2019-04-13, the nearest capture before Hsu's stated access date of 6 December 2019. It does contain the sentence — with no citation attached and no reference list anywhere on the page.

Case studies prove that with comprehensive education, hospitals can realistically aim to decrease RMW waste to a mere 6-10% of their waste stream and the Centers for Disease Control (CDC) suggests that only 3-5% of hospital waste truly needs to be disposed of as RMW.

Archived copy

The same sentence, spelled out in words, appears in Practice Greenhealth's Less Waste How-to Guide, a 2015 PDF with no reference section anywhere in it.5 The trail therefore reaches the CDC guideline carrying no page, no section and no report number. We retrieved that guideline in full and scanned it.

This figure is attributed to CDC; it does not appear in the cited CDC document, and we could not find it in any CDC publication. What the cited section does contain is more useful, and is quotable.

No epidemiologic evidence suggests that most of the solid- or liquid wastes from hospitals, other healthcare facilities, or clinical/research laboratories is any more infective than residential waste.

Although any item that has had contact with blood, exudates, or secretions may be potentially infective, treating all such waste as infective is neither practical nor necessary.

Archived copy of the PDF

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

A real number attached to the wrong subject. It is WHO's, it opens the key facts on WHO's health-care waste fact sheet, and WHO cites no source for it.7

The swap is not confined to vendor material. It is in the field's systematic review of healthcare waste audits, which renders the global figure as a United States figure and cites the WHO fact sheet for it.8

Safe management of medical waste can be extremely expensive, but in the United States, up to 85% of waste from hospitals does not need to be treated as infectious according to regulatory definitions (WHO, 2018).

Read the full review

There is a trap inside the trap. About 85% is roughly right for red bags in exactly one place: it is the arithmetic complement of the 14.9% Hsu measured at a single site.3 That is the same study restated, not independent confirmation of WHO's number, and quoting the two together as though they agree double-counts one audit.

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

The best-known statement of this figure is in a 2012 CMAJ review, which gives it exactly two citations.9

Although operating rooms occupy a proportionally smaller physical area of hospitals, they are estimated to generate 20%–33% of total waste produced in hospitals.^13,14

Read the full review

Reference 13 is Goldberg 1996, a feasibility study of recycling single-use anaesthesia breathing systems, built on a questionnaire to 413 anaesthesiology departments in three states and a time-and-cost analysis of dismantling circuits.10 Reference 14 is Tieszen and Gruenberg 1992, a case series of the waste from 27 operations across five procedure types at one hospital, reporting composition by weight and volume — disposable linens 39% of weight, plastic 26%, paper 7%.11 Neither indexed abstract contains a figure for the operating room's share of a hospital's waste, and neither design produces one: a composition study says what operating-room waste is made of, a recycling feasibility study says what one stream costs to divert. We could not clear either publisher's paywall, so that is a statement about the abstracts we read and the designs they describe, not about every sentence in the two papers.

A second, independent chain for the same figure is no better. Practice Greenhealth's 2011 Greening the OR module states it and footnotes it to a Medscape commentary; the address printed in that footnote is a section landing page rather than an article address, and medscape.com returned 403 to our client on 2026-08-06.13 We did not obtain the commentary.

What the record supports is a range with its width intact, and the field states it itself.

Estimates of operating room (OR) contributions to this waste vary widely, ranging from 20% to 70%, with a quantifiable impact on greenhouse gas emissions.

Read the full review

4. “60% of a hospital's regulated medical waste comes from the operating room”

This figure and the next come from the same page of the same 2011 module, which served normally to our client on 2026-08-06.13

A 2001 study by Malcolm Grow Medical Center estimated that approximately 60% of the hospital’s RMW was generated by its ORs.² National benchmarks tell us that RMW should be no more than 15% of the total waste stream,³ with the best performers driving down to well below 10%.

2 United States Air Force IERA. Medical Waste Incinerator Waste Management Plan-Malcolm Grow Medical Center, Building 1056, Andrews Air Force Base, MD. June 2001. Accessed on Feb 13, 2011.

3 Practice Greenhealth. Sustainability Benchmarking Report. 2010.

Read the full module

The address in endnote 2 does not answer: the host resolves in DNS, our client's connection timed out on 2026-08-06, and the Internet Archive holds no capture of it. The report is catalogued nonetheless. Its NTIS record gives accession ADA393684, a page count of 13, a corporate author of Pacific Environmental Services, Inc. of Herndon, Virginia, contract F41624-95-D-9017, and an abstract describing a compliance document.14

State of Maryland’s regulations require that any person who owns or operates a hospital, medical and infectious waste incinerator prepare a waste management plan that identifies the feasibility and approach to solid waste segregation or material substitution to reduce the amount of toxic emissions. This waste management plan has been written to meet these requirements for the Malcolm Grow Medical Center at Andrews AFB, MD.

Read the catalogue record

The Defense Technical Information Center holds the same accession, and its PDF endpoint redirected our client to a site maintenance page on 2026-08-06. We did not obtain the document and cannot say whether a 60% figure is in it. What the catalogue record establishes is that the document at the end of this chain is a thirteen-page waste-management plan, credited to a private environmental services firm under an Air Force contract and written to satisfy a state incinerator rule — not, as the module describes it, a study by the medical center. In our assessment that distinction matters more than the age of the number: a compliance plan for one building is nobody else's benchmark, and if the report arrived tomorrow it would still be one military hospital measured a quarter of a century ago.

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

Endnote 3 of the same module points at Practice Greenhealth, Sustainability Benchmarking Report, 2010, giving no address beyond the organisation's homepage.13 We located that report in no public form and in no archive we searched. It is also the source of two per-ton cost figures still quoted in this field.

The interesting part is not the missing document but what has been published around it. Several different numbers for the same quantity each point at a page from the same organisation. The 2011 module says 15% or less.13 The Less Waste How-to Guide tells a facility to reduce regulated medical waste to less than 10% of total waste or less than 8 tons per operating room per year, and on the same page repeats the uncited 3–5% attribution to CDC.5 The live topic page says something different again.6

Though it can make up less than 8 percent of a hospital’s total waste production, it can cost more than 40 percent of their waste management budgets to handle.

Archived copy

A further set of numbers, from a different publisher, sits at the address Hsu cited. The archived AHA and ASHE Sustainability Roadmap page tells a facility to set basic, intermediate and advanced goals and to “aim to reduce the RMW rate to 15, 9, and 5 percent of the total waste stream”, and on the same page to aim for 6–10%. It cites nothing for any of them.4

Practice Greenhealth's own benchmark tables, retrievable and dated, put the median at about 6% and the best decile at about 3% — see the last section of this article.1819 We think these published thresholds are not reconcilable with one another as targets, and that anyone quoting one should say which document it came from and what year the data behind it were collected. That is not an accusation of error. The older of them were probably fair statements of somebody's aspiration when written. They are simply still in circulation without their dates.

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

This chain has one hop. We retrieved the full codified text of 29 CFR 1910.1030 from the eCFR on 2026-08-06, including Appendix A, and scanned it.1

There is no percentage in the standard. What it sets out are four container criteria and two duties.

(1) Contaminated sharps shall be discarded immediately or as soon as feasible in containers that are: (i) Closable; (ii) Puncture resistant; (iii) Leakproof on sides and bottom; and (iv) Labeled or color-coded in accordance with paragraph (g)(1)(i) of this standard.

(2) During use, containers for contaminated sharps shall be: (i) Easily accessible to personnel and located as close as is feasible to the immediate area where sharps are used or can be reasonably anticipated to be found (e.g., laundries); (ii) Maintained upright throughout use; and (iii) Replaced routinely and not be allowed to overfill.

Read the full section

“Routinely” and “overfill” are the operative words and neither is a number. This says nothing about whether three-quarters is the right place to stop filling a container; the finding is about the citation, not the practice, and a fill line moulded into a container by its manufacturer is a real instruction with an enforcement route of its own. Which instrument a facility is standing on is the subject of a separate article on this site.

What survives tracing

Two things do, and they are why this article is not a list of complaints.

Red-bag contents: 11–15%, in two US settings, neither of them an operating room

Two independent United States teams, in different care settings, opened regulated-waste containers and weighed what was inside. Hsu 2020 audited 24 hours of waste from the emergency department of an urban, tertiary-care academic medical center and Level I trauma center with about 110,000 encounters a year, collecting 671.785 kg in total.3 Its abstract reports that 14.9% of waste disposed of in red bags met regulated-waste criteria; its results section gives the same finding as 15% and 7.45 kg, excluding sharps containers, which were not opened for safety reasons. We report both figures and have not reconciled the denominator between them.

Sharma 2024 audited a private outpatient dermatology practice in California daily for 30 days and found 11% of the contents of regulated-waste bins appropriately placed before any intervention, rising to 69.4% in the final ten days after two thirty-minute staff education sessions.15 One methodological detail travels with that number: the study defined regulated waste using the definitions published by Vermont and New York State, not California's.

A United Kingdom audit points the same way without being comparable. Runcie 2018 spot-checked 80 clinical waste bins at two NHS trusts over two months, recording visible contents, and found 41% of items met the hospitals' own disposal guidelines.16 Different legal definitions, different bins, and a visible-contents method rather than sorting and weighing.

The honest form of the finding is narrow. In the two US audits we retrieved, 11–15% of red-bag contents met regulated-waste criteria; both are single-site, small and short, and neither was in an operating room. We found no United States operating-room or inpatient study that opened red bags and reported a percent-correct figure. One audit team recorded the reason it could not in its limitations.

We were unable to open and sort RMW, PHI, and sharps waste to determine correct categorization due to safety and privacy concerns.

Read the full audit

Practice Greenhealth's benchmark tables, with their collection year attached

The benchmark reports are the one dated, retrievable, normalised series in this field. Their weakness is stated by their publisher: the data come from Environmental Excellence Award applicants, hospitals that chose to enter an environmental award, so the figures describe high performers and not United States hospitals generally. Each report covers the calendar or fiscal year before its award cycle, which is why the numbers below carry a collection year rather than a report year.181920

Two cautions belong on the same screen as those numbers. The first is the publisher's own, printed under its CY2023 regulated-waste table: some values that year declined against previous years, and the report attributes this to data collection difficulties arising from a waste contractor's system update, which may have limited hospitals' ability to report fully.19 The CY2023 declines should not be read as improvement. The second is ours: the cohort changes size between reports — 370 hospitals in the CY2022 tables against 429 in the CY2023 tables — so differences of a few tenths of a percentage point between columns carry no weight.1819

With those attached, the best-sourced framing available to a reader building an argument is this: among hospitals that entered an environmental award, regulated medical waste was about 6% of waste by weight and roughly a third of waste disposal cost, best decile about 3%, in the collection years shown. That is a smaller and duller claim than any of the six above. It is also the one with a document behind it, a date on it, and a cohort you can describe out loud.

What would change this answer

• 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 reference. We will print it and say so.
• A copy of the June 2001 Air Force report accessioned as ADA393684. It is a government contractor report written to be circulated, the Defense Technical Information Center holds the accession, and its PDF endpoint was in maintenance on the day we asked. Whether the 60% figure is inside it is genuinely unknown to us.
• A copy of Practice Greenhealth's 2010 Sustainability Benchmarking Report, from a library, a conference bag or an old hard drive. It is the document behind the 15% target and behind two per-ton cost figures still in circulation.
• A study that measured operating-room waste and total hospital waste in the same facilities over the same period and reported the ratio. That is the unresolved question in this literature: the published spread runs from 20% to 70% and nobody has closed it.12
• A United States operating-room or inpatient waste audit that opens red bags and reports a percent-correct figure. Until one exists, the 11–15% finding cannot honestly be carried into an operating room.

Where this could be wrong

The scans are the part most likely to be attacked, and they should be. A keyword scan proves a string is absent, not that an idea is. We scanned for “percent” as well as “%” for that reason and found neither in the cited section, but a paraphrase we did not anticipate would survive the scan. We scanned the document the claim is attributed to and looked for the figure elsewhere in CDC's published output without finding it; that is not the same as establishing that no such CDC sentence exists anywhere.

Two chains rest partly on abstracts. We read the indexed MEDLINE abstracts of Goldberg 1996 and Tieszen 1992 in full and could not clear either publisher's paywall, so our statement that neither reports a hospital-share figure rests on those abstracts and on the designs they describe. A full text containing an offhand sentence in a discussion would change that section, and we would say so.

The benchmark table is a synthesis and is ours to defend. We assembled it from three separate PDFs, and the CY2021 column is missing because we did not retrieve that report, not because it does not exist. Putting three columns side by side invites exactly the trend reading we tell the reader not to make; we judged that showing the years with their gaps and caveats beats printing one year and letting it look permanent, but a reasonable editor would disagree.

One bias worth naming. Tracing citations rewards the tracer for finding breaks, and four breaks out of six should make its author suspicious of himself. Three of these figures may well be approximately true. Nothing on this page is evidence that they are not.