Surgical Sponges: How to Read Radiopacity and Count-Pack Claims

Under 21 CFR 878.4450, nonabsorbable gauze for internal use is "a device made of an open mesh fabric intended to be used inside the body or a surgical incision or applied to internal organs or structures, to control bleeding, absorb fluid, or protect organs or structures from abrasion, drying, or contamination." The same identification says the device is woven from material that is not less than 50 percent by mass cotton, cellulose or a simple chemical derivative of cellulose, and that it contains x-ray detectable elements. The classification paragraph places it in Class I (general controls) and exempts it from premarket notification, subject to the limitations in 21 CFR 878.9.

FDA's product classification database ties that regulation to product code GDY, "gauze/sponge, internal, x-ray detectable": device class 1, submission type 510(k) exempt, General and Plastic Surgery panel. The database also lists a separate code, EFQ, "gauze/sponge, internal", as an unclassified pre-amendment device with a 510(k) submission type. The two codes meet in at least one record: 510(k) K990584, a Broadline Medical Industries "laparotomy sponge, X-ray detectable" found substantially equivalent in April 1999, carries EFQ as its classification product code and GDY as its subsequent product code.

For a buyer, the practical reading is narrow. The exempt GDY path is written around an internal gauze that has x-ray detectable elements, so a product without them does not match that identification and cannot simply borrow its exemption. If a supplier lists an internal-use sponge under EFQ, ask for the 510(k) number. And 21 CFR 878.9 says the exemption only reaches devices with the existing or reasonably foreseeable characteristics of the generic type: a device intended for a different use, or one that operates using a different fundamental scientific technology, needs a premarket notification first. Gauze sold on an active hemostatic claim is a different buying decision, covered in our guide to verifying hemostatic gauze labels; gauze for external wounds is covered in wound dressing classification.

The fields below come straight from the regulation and FDA's database records. Use them to read a quotation or a GUDID record before you compare prices.

AORN's Guideline for Prevention of Unintentionally Retained Surgical Items (effective December 9, 2021) states that surgical sponges are the most commonly retained items and that retained items remain one of the most common sentinel events reported to The Joint Commission. It also says the true incidence is unknown, partly because retained items may go undetected for months or years. AORN describes counting radiopaque soft goods as one method of accounting for items used on the field, warns of a significant potential for inaccurate manual counts, and gives guidance on using adjunct technology during manual counting.

The most useful buyer-facing dataset is a descriptive study by Steelman and colleagues (Patient Safety in Surgery, 2018) of 319 retained surgical sponge events voluntarily reported to The Joint Commission from 2012 to 2017. The sponge type was identified in 159 reports: 83 laparotomy sponges (52.2%), 54 sponges measuring 4 x 4 or 4 x 8 inches (34.0%), 11 towels (6.9%) and 5 cottonoids (3.1%). Of the reports that gave a location, 50.2% involved the abdomen or pelvis and 23.9% the vagina.

The count data are the reason purchasing matters. A sponge count was reported as performed in 77.4% of events (247 of 319), and when it was performed it was reported as correct 80.6% of the time (199 of 247), even though a sponge was retained. The authors also cite earlier research finding intraoperative radiographs only 67% sensitive for detecting retained items, and they estimate that RF detection used correctly may have prevented up to 97.2% of the reported retained sponges. These are voluntary reports, not incidence rates, but they show where a specification can help and where it cannot.

The same paper names the items that no detection feature could have caught. RF technology would not have prevented 9 of the retained sponges because the technology could not detect them: 5 cottonoids, 1 peanut sponge, 2 pieces of Kerlix gauze and 1 non-radiopaque sponge from an anesthesia kit. On Kerlix, the authors are blunt:

It is worth noting that kerlix is a dressing, not a surgical sponge, has no radiopaque marker, and its use as a sponge is not recommended.

Steelman et al., Patient Safety in Surgery (2018)

AHRQ's PSNet primer on retained surgical items (Gibbs and Romano, February 2025) adds the system view. Barcode-label, RF detection and RFID systems are adjuncts and must be used in addition to manual counts, and all surgical sponges must also contain a radiopaque barium marker that can be identified by x-ray should the manual or electronic count fail. The primer reports that no studies have compared these systems against one another in the same OR settings, so purchasers choose on preference, and that in one Veterans Health Administration study, acquiring count technology alone failed to improve retained-item rates because problematic practices and communication issues persisted.

So the purchasing job is to protect a count process the buyer does not run: one marker construction the OR has approved, one retrieval feature the count sheets assume, a pack configuration whose printed number staff can check, and a clear line between counted radiopaque sponges and non-counted dressing sponges. The AST counts standard draws that line too: radiopaque counted sponges should not be used as dressings, and dressing sponges should be kept separate from counted sponges so that non-counted, non-radiopaque items do not get mixed in.

"X-ray detectable" tells you a marker is present. It does not tell you what the marker is, how it is attached, or whether it works with an adjunct system. The catalogs reviewed for this guide describe three different constructions, plus one lap sponge listing that describes none:

• Woven barium sulfate strands (gauze sponges). Medline's lap sponge and O.R. towel brochure (dated 2020) says its X-ray detectable gauze sponges have dual X-ray detectable barium sulfate strands layered into the sponge "in a unique figure 8 design". That description sits under the 4 x 4 and 8 x 4 inch gauze line, not the lap sponges.

• Barium strip sewn into the seam (O.R. towels). The same brochure describes X-ray detectable O.R. towels with a barium strip sewn into the seam, yellow stitching to differentiate them from standard towels, and green labeling to help prevent incorrect item picking.

• RF detectable plus X-ray detectable (Situate). Medtronic's Situate code guide (08/2025) labels each sponge code XRAY (X-ray detectable) and RFD (RF detectable). Steelman and colleagues describe the RF approach as a chip in the sponge that is detected by a handheld wand or by wands built into an underbody mat. The RF feature only does its job if the facility has the matching detection system.

• Lap sponges described only as X-ray detectable. Medline's lap sponge listing gives the material (100% cotton, pre-washed, delinted), a 7-inch (17.8 cm) loop and the words X-ray detectable, but no marker construction. When a listing stops there, ask for the construction in writing and inspect a sample.

Construction matters because the reported retained sponges split along it. In Steelman's table, all 83 retained laparotomy sponges and all 5 cottonoids were radiopaque, as were 44 of the 54 retained 4 x 4 or 4 x 8 sponges (81.5%). None of the 11 retained towels, neither piece of Kerlix gauze and not the single 2 x 4 sponge had a radiopaque marker. Across all 319 reports, 62.4% of the retained sponges were radiopaque. The items an x-ray could not find were the ones that had not been bought, stocked or used as counted radiopaque goods.

Use the table to translate label wording into a construction you can write into a tender.

Ordering by size alone, such as "18 x 18 lap sponge", leaves out the part of the product that the count process handles. The same size is catalogued with different retrieval and counting features, and suppliers sell each one as a separate item:

• Sewn loop. Medline lists a 7-inch (17.8 cm) loop on its lap sponges. If loop length matters to your counter bags or count practice, put the length in the specification.

• Plastic ring. Medline sells an 18 x 18 lap sponge with a ring (MDS251518RNG), Medtronic's code key defines RING as a plastic ring (L1818-04P01C-R), and the 2019 Medical Action recall listed 18 x 18 and 12 x 12 X-ray detectable lap sponges with rings.

• No loop. Medtronic lists an 18 x 18 Situate lap sponge coded NOLOOP (L1818-04P01C-L).

• Strung specialty sponges. Tonsil sponges are sold double strung (Medline and Medtronic) or tape strung (Medtronic). Medline also lists a peanut sponge with a tail and says all of its specialty sponges include a counting mechanism.

• Neuro sponges with a counting card. Medline's neuro sponges all have strings, all are X-ray detectable, and each pack includes a counting card to assist with counting protocol. Its listed sizes run from 0.5 x 0.5 inch to 1 x 3 inches, 10 per pack.

The AST counts standard says sponges should not be altered, for example by cutting them to a smaller size or cutting off the strings on tonsil or neuro sponges, because altering a sponge raises the chance of an incorrect count or of the cut portion being retained. For purchasing, that becomes a rule: if a size or retrieval feature does not suit a procedure, the answer is a different SKU, never a modified one. Which variant a service line uses is a perioperative decision; procurement's job is to keep that decision fixed in the item master.

Watch the code grammar. In Medtronic's guide, most of the sterile lap sponge, gauze and O.R. towel codes have a non-sterile twin made by replacing "C" with "CN" in the reorder code, and the loop, ring and no-loop variants differ only in the final suffix. Medline's brochure prints cross-reference tables against other brands' item numbers; treat those as a starting list for evaluation, not proof of equivalence, because they are point-in-time marketing documents. Your supplier quality agreement is the place to require notice before any change to the marker, retrieval feature, pack count or sterility claim, and to bar unapproved substitutions.

A purchase order line such as "sterile lap sponges, 5 per pack, 40 packs per case" reads like a guarantee of 200 identical, pre-counted units. The clinical record says otherwise. AHRQ PSNet's primer notes that ten 4 x 4 sponges are supposed to be in each pack by industry standards, but that sponges are packaged and sold by weight, not by the number of sponges; nursing staff frequently find packs with 9 or 11 sponges, and by policy those "bad packs" are not used. The primer's table of latent factors lists an incorrect number of sponges in a package, and defective or missing radiopaque markers, under manufacturers and distributors.

The primer's case example shows the cost. During a heart operation, a second package of 4 x 4 sponges added mid-case contained eleven sponges instead of ten. The team did not separate them, called the count correct at twenty, and a sponge was left by the heart. A radiologist recognized its radiopaque marker on review of the ICU images that evening, and the patient was taken back to the operating room to remove it.

The AST Recommended Standards of Practice for Counts (effective 2006 and still published) set out the clinical side of the rule:

• Each sponge should be separated to confirm that all sponges are present, are not sticking together, and were packaged in the correct number by the manufacturer.

• The surgical technologist and circulator should never assume a manufacturer's prepackaged item is correctly counted; a count is incorrect if there are too many sponges as well as too few.

• A sterile package with an incorrect number is removed from the sterile field, bagged, labeled as incorrect with the number of sponges, and kept in the OR until the procedure is complete.

• The binding strip of packaged sponges should only be broken when the sponges are to be counted and used.

• All sponges used during the procedure, except dressing sponges, must be radiopaque.

None of that is a purchasing procedure, and buyers should not write clinical counting steps into a tender. What a tender can do is treat the printed count as a labeled claim the OR will verify. Ask each supplier how pack counts are controlled (by weight or by piece), how incorrect-count packs should be reported, what information it needs (lot, REF and a photo of the bagged pack), and how it investigates and credits them. Then make sure OR staff have a route to report bad packs to purchasing with the lot number, so complaints can be matched across sites.

The configurations below are taken from the two catalogs, the openFDA GUDID records and the 2019 recall record. They show how many distinct line items can sit behind one sponge size.

Sterility and pack data should come from records, not catalog adjectives. Through FDA's openFDA device UDI API, GUDID records for GDY lap sponges show the fields a buyer can check before award:

• Device description. For example, an 18 x 18 inch, 4-ply lap sponge at 5 pieces per pack and 40 packs per carton, or a 12 x 12 inch lap sponge described as X-ray detectable, 5 count.

• Packaging hierarchy. A primary device identifier for the pack and package-level identifiers with their quantities (40 packs in one record; 10 packs per box and 4 boxes per carton in another). See what receiving teams should capture from UDI labels.

• Sterility. A sterile flag and, where the labeler entered it, the sterilization method (ethylene oxide in two of the records reviewed).

• Use and material flags. Single-use status, natural rubber latex labeling and commercial distribution status.

GUDID entries are manufacturer-entered and uneven. In the records reviewed, one 12 x 12 inch lap sponge was marked sterile with no sterilization method listed, the 18 x 18 inch records were flagged both as sterile and as requiring sterilization before use, and one GDY 18 x 18 inch description did not mention x-ray detectability at all. Use GUDID to check consistency with the label and the quotation, not as a substitute for either.

FDA's recall database then shows how these products have actually failed. Three records define the receiving risk set for this family:

1. 2016: breach of the sterile barrier. Covidien's recall event 73427 (initiated March 3, 2016; Class 2; terminated June 15, 2017) gave the reason as "compromised sterility due to breach of barrier", with an FDA-determined cause of packaging process control. The event covered Curity and Kerlix X-ray detectable laparotomy sponge lines as well as gauze. Its first record, Z-1847-2016, is a Curity 16-ply 4 x 4 inch gauze sponge, sterile 10s in a soft pouch, with 713,600 units in commerce in the US and abroad. Customers were told to stop using the product and return all inventory. For documentation at the dock, see recording sterile-pouch damage at goods receipt.

2. 2019: potential lack of sterility assurance. Recall Z-1068-2019 (Class 2; initiated March 1, 2019; terminated April 6, 2021) covered Medical Action Industries X-ray detectable lap sponges under GDY in 18 x 18, 4 x 18, 12 x 12, 8 x 36 and 17 x 26 inch sizes, including ring variants: 228,300 packs of 5, distributed across US states, Canada and Saudi Arabia. FDA's cause field read "under investigation by firm". End users were told to examine inventory, quarantine affected product and return it.

3. 2024: the wrong item inside a sealed 5-pack. Recall Z-1330-2024 (Class 2; initiated February 21, 2024; open and classified when accessed) covers the Situate Laparotomy Sponge, RF and X-ray Detectable, REF L1818-04P01C-1, in three lots. FDA's record says the 5-packs may contain an incorrect product with a blue loop, an X-ray detectable feature and ethylene oxide treatment that was not steam pre-treated before sterilization and has no RF tag; if such a sponge were in the scannable area, the system would display "CLEAR". The FDA-determined cause is labeling mix-ups, with 755,800 units in commerce. Matching recall notices to lots is covered in tracing mixed-lot consumables during a recall.

The 2024 case is the clearest lesson for buyers: a correctly labeled outer pack can hold the wrong sponge, and receiving cannot see it without breaking sterility. The controls that work are lot capture at receipt, fast recall routing by lot, and an OR reporting path for packs whose contents do not match the label.

Keep storage conditions in view as well; labeled storage and transport controls cover the evidence to request from distributors.

Towels go into the same procedures as lap sponges, and they are where the detectability assumption fails most often. In Steelman's data, none of the 11 retained towels had a radiopaque marker. The authors still counted those 11 towels among the retained items that correctly used RF technology may have prevented, and RF-tagged towels do exist: Medtronic's Situate O.R. towels are coded XRAY and RFD.

The AST counts standard treats towels placed in the wound as counted items. Miscellaneous items that will be used on or placed within the patient should have a radiopaque marker and be included in the count, and towels used to pack away bowel should be counted and kept separate from non-counted items.

Catalogs offer both kinds side by side. Medline sells standard, deluxe, heavy, premium and natural O.R. towels without a marker, and a separate X-ray detectable line with a barium strip in the seam, yellow stitching and green labeling. Medtronic's Situate O.R. towels come 4 per carton. Detectable and standard towels are both sold in blue, green and white, so towel color tells you nothing about detectability.

The buying rule follows. If surgical services allows towels inside the incision, stock the X-ray detectable towel under its own item number in sterile OR storage, keep standard towels on separate item numbers for uses outside the wound, and block substitution between them in the item master and the distributor agreement. The same separation applies to non-radiopaque dressing sponges and rolled gauze: AST keeps dressing sponges apart from counted sponges, and Steelman's authors note that Kerlix is a dressing whose use as a sponge is not recommended.

Before award, a value analysis team or buyer should be able to fill in every line below for each SKU and each supplier:

1. Classification. The GUDID record shows GDY under 21 CFR 878.4450, or, if a SKU sits under EFQ, the supplier has given you the 510(k) number.

2. Marker construction. Stated in writing (woven strand, sewn strip, RF tag with X-ray detectability, or another design), approved on a sample by the perioperative team, and consistent with any adjunct detection system you own. RF-detectable sponges are bought together with the matching scanner fleet.

3. Retrieval or count feature. Loop (with its length), ring, no loop, double or tape string, tail, or counting card, matched to the count sheets and counter bags already in use.

4. Pack configuration. Pieces per pack, packs per case, and pouch, rigid tray or band, with sterile and non-sterile item numbers kept apart.

5. Sterility evidence. Sterile claim, sterilization method, expiry and single-use status on the label and consistent with GUDID. Non-sterile banded sponges go only to a kit assembler that documents sterilization.

6. Lot and UDI capture. Package-level identifiers and lot numbers captured at receipt so recall notices can be matched by lot.

7. Supplier commitments. How pack counts are controlled; how marker and loop attachment is tested; how bad-pack and missing-marker complaints are handled; and advance notice of changes to the marker, retrieval feature, pack count or sterility claim, written into the quality agreement.