Left to dry: unseen risks lurking on reusable medical devices

Abstract

In this study, the intricate relationship between the complexity of reusable medical devices and their cleaning efficacy is explored to the point of cleaning failure. The results identified complex device features that require manual cleaning regardless of soil drying time, as well as device designs (i.e., features) that are appropriate for a fully automated cleaning process. All features were evaluated for cleaning performance under conditions where the soil was not allowed to dry, as well as replacing manual cleaning steps with non-manual methods, (i.e., substituting manual brushing and flushing with soaking and sonication). Findings revealed that soil drying on reusable medical devices can increase the cleaning challenge specific to the device feature. Devices engineered with visually accessible challenging geometries, as opposed to those with enclosed or concealed features, successfully met the requirements for soil drying times and semi-automated cleaning evaluations. However, device features with occluded geometries such as ball bearings, leaf springs, threaded screws, and mated surfaces proved more challenging to clean effectively. Since device features vary in their cleaning complexity, each device should be assessed based on their features. By recognizing the risks associated with soil drying on reusable devices, healthcare facilities can implement appropriate practices that streamline more effective cleaning processes that will inform patient safety along with paving the way for future automation.

Introduction

As the bright lights blur and consciousness fades during the countdown to surgery, most people are unlikely to be pondering the odds of an unsterilized instrument being used^1,2. The US FDA³ noted that “Infections can occur through multiple situations including inadequate cleaning, disinfection and sterilization of reusable medical devices”. For example, an estimated risk of infection transmitted by endoscopy is 1 per 1.8 million procedures, and infectious agents such as Helicobacter pylori, Salmonella spp., Pseudomonas aeruginosa, Strongyloides sterocoralis, and hepatitis B and C viruses have been attributed to gastrointestinal endoscopy⁴. However, due to a number of factors such as inadequate or no surveillance and low occurrence or absence of clinical symptoms, the true risk of infection is difficult to estimate⁵. Hospital-Acquired Infections (HAIs) are infections that occur 48 h or more after a patient is admitted to a healthcare facility, which were neither present nor incubating at the time of admission¹. It was observed that on any given day, about one in 31 patients has at least one HAI⁶. Moreover, in the ‘2015 HAI Hospital Prevalence survey’, the CDC reported that about 72,000 of the estimated 687,000 (or 10.48%) patients with HAIs died during their acute hospital stay in the US⁷.

The occurrence of HAIs has a direct medical cost impact on hospital finances and many researchers have performed analysis to understand the economic benefits of investing in an infection prevention and control program^8,9. Additionally, the risk of contracting an HAI can be mitigated by effective device processing². However, the complexity of device features and the importance of implementing appropriate cleaning processes are majorly underappreciated^2,10. For example, the US Food and Drug Administration (FDA) posted 35,039 adverse event reports related to outbreaks, injuries, and reprocessing failures associated with medical devices in 2024¹¹. This is corroborated by ‘Sedgwick’s 2025 US State of the Nation Recall Index report’¹² that noted the medical device sector recorded an 8.6% increase in recall events, reaching 1,059 in 2024. This report noted that device failures lead recall causes, accounting for 11.1% of medical device recalls, the highest rate in over five years. Moreover, Ofstead and co-workers¹³ recently noted that “processing breaches in endoscopes are common and often involve profound errors in multiple steps”. In summary, there are notable published reports highlighting that improperly reprocessed medical devices caused HAIs^14–25.

Devices used in healthcare facilities must undergo processing before their initial or subsequent use(s), as outlined by the World Health Organization (WHO)²⁶. This cycle should be managed under a quality system to ensure devices are safe and effective for every use²⁷. Devices requiring cleaning, disinfection, and/or sterilization between patient exposures include those used directly during surgery (e.g., forceps, endoscopes), items with minimal patient contact (e.g., blood pressure cuffs), and even non-patient contacting equipment that is just located in the same room (e.g., monitors)²⁷. However, the ability to effectively and safety clean, disinfect and/or sterilize reusable medical devices remains extremely challenging and is often overlooked^2,10.

The device processing cycle (Fig. 1) begins with patient use, where the device is in a ready state and is safe for patient use. Immediately following patient use, the preparation for cleaning should begin at the point of use^26,28,29. This treatment includes the removal of the visible soil on the device, disassembly (if required), flushing lumens, and brushing hard to reach areas on the device. To facilitate the effective removal of residual soil, instruments should be handled in a way (even if simply covered with a wet cloth) to keep them moist and minimize soil from drying. Immediate treatment at the point of use is a crucial first step in preventing a more challenging cleaning process, potential device damage, or the growth of microorganisms during the waiting period before cleaning³⁰.

Fig. 1.

Typical processing cycle for critical, reusable medical devices.

The processing cycle for reusable medical devices is often depicted as a standalone loop (Fig. 1); but the full supply chain is far more complex, involving multiple transfers of responsibility to ensure patient safety (Fig. 2). The process begins with the manufacturer, who is responsible for producing devices with the required microbiological quality, as well as providing IFUs throughout the lifecycle. Once the healthcare facility acquires the device, it assumes responsibility for cleaning, disinfecting, and sterilizing the product after each patient use, ensuring it is ready for reuse¹. For certain surgical kits, ownership may be impractical due to cost. To address this, manufacturers or external companies often provide loaner programs. After patient use, these loaner kits should be cleaned, disinfected, inspected and packaged before being returned to a distribution center or transported to another user. While distribution centers may process the kits for reuse, best practices dictate that receiving facilities ultimately process the kits themselves to ensure they are ready and fully traceable³⁰.

Fig. 2.

End to end device processing cycle.

The complexity of the end-to-end device processing cycle and soil drying is a risk that requires special attention, which can be used to improve the appropriateness of manufacturer’s IFUs¹. The guidance provided to healthcare facilities should emphasize minimizing the time between use and decontamination to prevent soil from drying. Recent studies have investigated the effect of soil drying time on reusable medical devices, which yielded results that show prolonged drying can influence subsequent ability to safely process and/or sterilize. An experiment involving a flat stainless steel coupon and a stainless steel hemostatic clamp revealed that prolonged drying of soil increases the difficulty of cleaning, indicating soil solubility decreased as drying time increased²⁸.

Kimble et al.³¹ explored the chemistry of the protein changes over time and found that the matrix of albumin, which is a primary protein in clinical soils, creates a water insoluble barrier that may present an increased cleaning challenge to remove residual proteins. In order for soil to re-solubilize following drying, water (or the cleaning solution) must integrate into the spaces between the molecules and aggregate to lessen the attractive and cohesive forces holding them together. Then again, as water evaporates from the soil during drying polarity changes, caused by tertiary structural changes, in the proteins can reduce the wetting effectiveness of water, and make rehydration and re-solubilization more difficult. Introducing water into a dried proteinaceous soil initially leads to wetting, followed by swelling of the proteins and aggregates, ultimately ending with dissolution facilitating their removal. However, water alone may not be sufficient to facilitate removal of dried protein and other bodily soils³¹. Hoover et al.³² investigated the concept of rehydrating dried soil by evaluating whether an extended soak in a cleaning agent could reverse the chemical changes in proteins caused by drying and restore the soil to its original wet chemical composition. Among the cleaning agents tested, only an alkaline enzymatic detergent achieved successful solubility results on the flat surface but demonstrated a decrease in solubility with the added complexity of a device feature (i.e., box hinge). These results suggest that the design features of the device could significantly impact the drying and subsequent removal of soil, making it likely that a standardized drying time for use in healthcare facilities may not be feasible.

In this study, the focus was placed on how the complexity of device design influences the cleaning procedures required for reusable medical devices. The aim was to identify specific device features that necessitate manual cleaning efforts, which become increasingly challenging when soil drying occurs on the device. In the experiment, all aspects of cleaning performance were assessed under conditions that prevented soil from drying. This setup facilitated a comparison between traditional manual cleaning methods—such as brushing and flushing—and innovative non-manual techniques, including soaking, sonication, and automated washing. This investigation highlights the significant impact that intricate design features have on the efficacy and feasibility of cleaning processes, particularly in the face of soil drying.

Experimental investigation

Test article sample preparation

Using the device feature approach, 23 of the most common reusable medical device characteristics (Table 2) were tested in accordance with AAMI ST98 Cleaning Validation of Health Care Products – Requirements for Development and Validation of a Cleaning Process for Medical Devices²⁷. In each experiment, the independent variable was the duration allowed for soil to dry on the device, while all other variables were held at the most challenging conditions. All devices (i.e., test articles) underwent seven simulated use cycles. Each cycle included soiling with Defibrinated Blood Soil (DBLSO)^33,34, followed by drying, rinsing with tap water, and soaking in Valsure Enzymatic Detergent (STERIS). The test articles were brushed with an M-16 nylon bristle brush to remove blood and debris, with lumens thoroughly brushed as part of the procedure. The test articles were then subjected to mechanical washing in a washer-disinfector under defined parameters. After cleaning, the articles were double-wrapped and sterilized using moist heat at 132 °C for 4 min. This process conditioned the devices into a “used” state prior to further testing.

Table 2.

Device soiling and extraction criteria.

Feature name	Surface area (cm2)	Soil volume (mL)	Extraction volume (mL)	Extraction container	Lumen flush volume
Ball detent/ball bearing	163.87	0.40	400	52 mm × 229 mm 4 mil Whirl-Pak bags with a 710 mL capacity	First flush—200 mL Second flush—100 mL Final flush—100 mL
Ball seal springs	6.31	0.02	15	65 mm × 90 mm 4 mil Whirl-Pak bags with a 90 mL capacity	NA
Blind slot	89.80	0.25	200	100 mm × 200 mm 4 mil Whirl-Pak bags with a 500 mL capacity	First flush—150 mL Second flush—25 mL Final flush—25 mL
Button w/spring	37.15	0.10	90	100 mm × 200 mm 4 mil Whirl-Pak bags with a 500 mL capacity	NA
Buttons-exposed springs	82.31	0.20	200	152 mm × 229 mm 4 mil Whirl-Pak bags with a 710 mL capacity	NA
Captured screw	62.88	0.15	150	152 mm × 229 mm 4 mil Whirl-Pak bags with a 710 mL capacity	NA
Hinges, joints, pivot points	9.93	0.03	20	65 mm × 90 mm 4 mil Whirl-Pak bags with a 90 mL capacity	NA
Leaf springs	73.85	0.20	180	152 mm × 229 mm 4 mil Whirl-Pak bags with a 710 mL capacity	NA
Mated surfaces	215.35	0.50	500	10 in × 15 in 4 mil Whirl-Pak bags with a 2721 mL capacity	NA
Mated surfaces-small clearance	211.63	0.50	500	10 in × 15 in 4 mil Whirl-Pak bags with a 2721 mL capacity	NA
O-rings, external O-ring	13.53	0.04	30	65 mm × 90 mm 4 mil Whirl-Pak bags with a 90 mL capacity	NA
O-rings, internal O-ring	18.48	0.05	45	100 mm × 200 mm 4 mil Whirl-Pak bags with a 500 mL capacity	NA
Rough surface	60.22	0.15	150	152 mm × 229 mm 4 mil Whirl-Pak bags with a 710 mL capacity	NA
Screws-threaded rod/threaded thru hole	52.42	0.20	130	52 mm × 229 mm 4 mil Whirl-Pak bags with a 710 mL capacity	NA
Screws-threaded rod/threaded blind hole	67.72	0.20	150	152 mm × 229 mm 4 mil Whirl-Pak bags with a 710 mL capacity	First flush—25 mL Final flush—25 mL
Sliding shaft-short	21.10	0.10	50	100 mm × 200 mm 4 mil Whirl-Pak bags with a 500 mL capacity	NA
Sliding shafts-long	72.13	0.20	180	10 in × 15 in 4 mil Whirl-Pak bags with a 2721 mL capacity	NA
Smooth blind lumen	4.57	0.02	11	65 mm × 90 mm 4 mil Whirl-Pak bags with a 90 mL capacity	First flush—5 mL Second flush—3 mL Final flush—3 mL
Smooth through lumen	54.36	0.25	130	15 in × 20 in 4 mil Whirl-Pak bags with a 5441 mL capacity	First flush—100 mL Second flush—15 mL Final flush—15 mL
Snap rings	10.6	0.03	25	65 mm × 90 mm 4 mil Whirl-Pak bags with a 90 mL capacity	NA
Spring-internal	151.78	0.34	350	152 mm × 229 mm 4 mil Whirl-Pak bags with a 710 mL capacity	NA
Threads-blind hole	31.71	0.10	80	100 mm × 200 mm 4 mil Whirl-Pak bags with a 500 mL capacity	First flush—40 mL Second flush—20 mL Final flush—20 mL
Through slot	95.33	0.25	200	100 mm × 200 mm 4 mil Whirl-Pak bags with a 500 mL capacity	First flush 150 mL Second flush 25 mL Final flush 25 mL

Before each experiment, test articles were thoroughly cleaned to remove residual analytes following simulated use. The cleaning process began with a 1-minute rinse under RO/DI water, followed by brushing to remove all visible soil. Each article was then placed in a container and soaked in an alkaline NeoDisher cleaning solution (Dr. Weigert), prepared at a concentration of 10 mL/L, for 60 min. During the soak, the articles were brushed until visibly clean, and any movable parts were actuated. The articles were subsequently submerged in a freshly prepared alkaline NeoDisher solution to ensure full coverage and sonicated for 15 min in a Branson 8800 Ultrasonic Cleaner operating at 40 kHz. After sonication, the articles were rinsed under running Critical Water as defined by ANSI/AAMI ST108³⁵. Lumens were flushed three times with 20 mL of Critical Water using a 25 mL syringe fitted with a 20G needle. Finally, the test articles were allowed to dry completely and inspected for cleanliness using a borescope or 10x lighted magnification.

Cleaning evaluation

An extraction method validation was conducted for each feature to determine the test soil extraction efficiency using the additive extraction method²⁸. A high extraction efficiency demonstrates the test system is effective at removing residual soil and measuring for the test analytes, protein and total organic carbon (TOC). For each experiment, test articles were soiled using a combination of two application methods: immersion in 100 mL of modified coagulated blood soil^32,33 and direct inoculation, where 0.20 mL (200 µL) of soil was pipetted into lumens and complex features. The articles were rotated to ensure complete surface coverage, and the minimum soil volume was verified by weight. The test article was then allowed to dry under the specified condition for the experiment (e.g., 2–72 h at 22 °C and 50%RH). For the semi-automated experiment, once the soil was applied, the device was covered with a soaked OR towel before being dried for 2 h at 22 °C and 50% RH.

Test articles were cleaned as outlined in Table 1.

Table 1.

Cleaning steps per experiment.

Cleaning step	Experiment details
	2-h Dry	72-h Dry	Extended soak	Semi-automated
Extended soak			60 min
Pre-rinse	X	X	X	X
Manual soak	5 min	5 min	5 min	15 min
Manual cleaning	X	X	X
Intermediate rinse				X
Ultrasonic cleaning	10 min	10 min	10 min	15 min
Intermediate rinse	X	X	X	X
Final rinse	X	X	X
Manual dry	X	X	X
Automated cleaning				X

1. Extended soak: Test articles were soaked in 10mL/L Neodisher MediClean forte alkaline cleaning agent (Dr. Weigert) for 60 min.

2. Pre-rinse: Test articles were manually brushed with soft bristle brushes (Sklar, Key Surgical) to remove debris. Lumens and tight crevices were flushed with water using a syringe.

3. Manual soak: Test articles were soaked for 5 min in Valsure Enzymatic cleaning agent (STERIS) at 7.9 mL/L in critical water (~ 20 °C).

4. Manual cleaning: While soaking, all surfaces were brushed with soft non-metallic bristles, lumens were thoroughly cleaned using twisting motions with brushes (Key Surgical), and movable parts were actuated to expose surfaces. Syringes were used to flush intricate parts and tight clearances.

5. Rinse: Test articles were rinsed 3X by submersing in Critical Water³⁴.

6. Ultrasonic cleaning: Test articles were ultrasonically cleaned for 10 min at 40 kHz in Valsure Enzymatic solution (3.9 mL/L at ~ 25 °C). Lumens, blind holes, and intricate areas were flushed to prevent air pockets.

7. Intermediate rinse: Test articles were immersed in < 40 °C tap water for at least 1 min, rinsed until visibly clean, and flushed thoroughly with a 50 mL syringe. Movable parts were actuated to ensure thorough rinsing.

8. Final rinse: A 15-second rinse was performed using < 40 °C critical water.

9. Drying: Test articles were dried with a clean, lint-free cloth or compressed air, ensuring lumens and articulated areas were fully dried. Movable parts were actuated to remove fluid from threads, ratchets, and hinges.

10. Automated cleaning: Test articles were loaded into the washer so that each basket was accessible to the washer arms. The following washer disinfector program was run in the automated washer disinfector (Getinge):

    1. Pre-wash 1–02:00 min w/ cold tap water.
    2. Wash 1–01:00 min w/ 43 °C Valsure Enzymatic 1.9mL/L (STERIS).
    3. Wash 2–05:00 min w/ 66 °C Valsure Enzymatic 1.9mL/L (STERIS).
    4. Rinse 1–02:00 min w/ hot tap water.
    5. Rinse 2–00:15 min w/ ambient critical water³⁵.

Extraction and residual analyte determination

Post-cleaning, the test articles were extracted using the validated additive extraction method: an initial water extraction followed by SDS extraction. The water extract was analyzed for protein residues. Each experiment included at least five samples; experiments with results < 6.4 µg/cm² were further tested for a total sample size of 10.

An appropriately sized Whirl-Pak bag was prepared with the extraction eluent (Table 2), calculated using Eq. 1 with the protein acceptance criteria of 6.4 µg/cm² and the method LOQ of 2.5 µg/mL:

1

Extraction eluents included ‘Purified Water’ (< 50 ppb TOC, generated using an ELGA water filtration system) and ‘2% Alkaline SDS’ (prepared with sodium dodecyl sulfate (Thermo Scientific) and purified water, pH adjusted to 10 ± 0.9 using 1.00 N sodium hydroxide). For test articles requiring lumens or flushing, one-third of the extraction volume was flushed through the device; otherwise, the full eluent volume was added to the container. The container was sonicated in a Branson 8800 Ultrasonic Cleaner at 40 kHz for 15 minutes. Lumens were flushed with extraction fluid using a syringe to ensure complete coverage, if applicable. Sealed containers were shaken at 150 RPM for 30 minutes, and lumens were flushed again per Table 2.

The following experimental controls were used for each experiment as defined by ANSI/AAMI ST98: positive sample control, positive device control, negative sample control, and negative device control. Water and SDS extractions were analyzed for protein residuals, with water extractions also tested for total organic carbon (TOC). Protein was evaluated for both the water-soluble proteins and the non-water-soluble proteins using the BCA standard addition method²⁷. Analyte concentrations were corrected by subtracting the negative sample control concentration. If the control result was negative, the method LOD was used instead. Dilution factors were applied as needed. SDS extraction efficiency was calculated as the percentage of protein extracted by water (1:20) relative to the total protein extracted, including SDS. This calculation assesses the total protein remaining, accounting for non-water-soluble proteins. A one-sample t-test was used to calculate the 99% confidence intervals of the sample set. Utilizing the high value of the 99% confidence interval as the worst-case result to evaluate the sample set against the acceptance criteria serves as a robust strategy to mitigate the risk of error in statistical analyses and decision-making processes.

Results

A passing result for cleaning result as defined by ISO 15883-5, Washer-disinfectors Part 5: Performance requirements and test method criteria for demonstrating cleaning efficacy is ≤ 6.4 µg/cm² protein residuals, while the TOC limit is ≤ 12 µg/cm^238,39. These levels were established by the ISO committee to reflect appropriate cleaning endpoints to ensure patient safety in a clinical setting during device use⁴⁰. If the 99% confidence interval’s upper bound exceeded these criteria, the feature failed the cleaning process. The experiment results per device feature for protein and TOC residuals can be found in Tables 3 and 4 respectively.

Table 3.

Experiment protein results per device feature.

Device feature	Extraction efficiency (%)	Cleaning 72 h dry-water (µg/cm2)	Cleaning 72 h dry-additive protein (µg/cm2)	Cleaning 2 h dry water (µg/cm2)	Cleaning 2 h dry additive-protein (µg/cm2)	Extended soak-water (µg/cm2)	Extended soak-additive protein (µg/cm2)	Semi-automated-water protein (µg/cm2)	Semi-automated-additive protein
Ball detent/ball bearing	67.310	40.710	33.300	1.171	3.139	45.52	104.55	0.05816	0.1139
Ball seal springs	55.170	15.010	15.340	0.379	4.113	201.7	12,246	6.754	6.735
Blind slot	70.540	89.900	242.900	9.885	24.890	NA	NA	9.226	10.167
Button w/spring	66.830	23.610	20.00	2.833	3.462	NA	NA	1.763	1.812
Buttons-exposed springs	45.750	0.885	2.935	0.058	4.881	NA	NA	3.678	3.885
Captured screw	49.620	4.867	9.124	3.603	3.545	14.882	13.605	0.05882	0.1139
Hinges, joints, pivot points	79.900	112.400	202.700	5.565	8.530	166.4	257.7	4.863	4.994
Leaf springs	92.161	36.030	65.760	1.342	10.570	NA	NA	3.708	3.646
Mated surfaces	79.020	25.570	32.920	4.701	4.574	NA	NA	0.058	0.111
Mated surfaces small clearance	85.259	0.058	0.113	1.825	18.590	NA	NA	8.7	5.56
O-rings– external O-ring	87.320	0.742	1.323	2.104	5.498	NA	NA	0.05498	2.472
O-rings–internal O-ring	68.180	2.815	67.700	7.550	9.520	NA	NA	4.987	5.044
Rough surface	96.133	5.382	9.364	4.676	4.756	NA	NA	3.088	4.528
Screws threaded rod/threaded thru hole	90.634	54.490	87.900	1.338	1.299	NA	NA	2.034	1.944
Screws threaded rod/threaded blind hole	97.145	4.943	7.580	4.496	5.757	NA	NA	0.05196	0.1026
Sliding shaft short	92.192	0.056	0.110	1.135	13.100	NA	NA	2.283	2.495
Sliding shafts long	92.100	0.059	0.115	15.070	15.310	NA	NA	8.811	8.756
Smooth blind lumen	90.330	2.391	24.660	15.500	39.310	NA	NA	3.215	24.1
Smooth through lumen	91.980	9.600	35.130	126.900	125.700	NA	NA	4.139	7.111
Snap rings	73.070	0.060	0.755	2.481	8.050	NA	NA	5.496	15.26
Spring internal	64.510	37.590	44.990	17.610	17.410	NA	NA	4.154	4.074
Threads blind hole	93.633	15.900	15.630	11.820	12.260	NA	NA	0.06023	2.483
Through slot	33.120	121.900	151.330	4.474	12.420	NA	NA	3.475	4.501

Table 4.

Experiment TOC results per device feature

Device feature	Extraction efficiency (%)	Cleaning 72 h dry-water (µg/cm2)	Cleaning 2 h dry water (µg/cm2)	Extended soak-water (µg/cm2)	Semi-automated-water protein (µg/cm2)
Ball detent/ball bearing	96.3465	19.45	1.436	102.5	2.037
Ball seal springs	96.5754	20.97	0.08453	269.9	1.951
Blind slot	90.4004	114.9	8.09	NA	5.502
Button w/Spring	95.0360	48.92	0.4148	NA	0.0861
Buttons-exposed springs	95.0360	1.18	0.12534	NA	0.5966
Captured screw	96.4285	8.1	0.0882	4.752	0.08882
Hinges, joints, pivot points	90.1588	152.3	9.95	279.4	0.4353
Leaf springs	99.2599	50.12	0.886	NA	0.08591
Mated surfaces	95.8175	NA	0.12292	NA	0.9962
Mated surfaces small clearance	96.6463	1.835	0.3362	NA	8.39
O-rings–external O-ring	95.0237	1.385	0.8149	NA	0.08149
O-rings–internal O-ring	91.0928	115.2	1.114	NA	1.121
Rough surface	99.2233	1.34	0.3118	NA	0.1883
Screws threaded rod/threaded thru hole	96.0339	1.188	1.428	NA	0.6185
Screws threaded rod/threaded blind hole	96.7571	3.576	0.536	NA	1.529
Sliding shaft short	98.3958	0.08427	0.101	NA	1.239
Sliding shafts long	54.7300	1.159	2.513	NA	0.8582
Smooth blind lumen	98.7278	0.3004	4.813	NA	8.339
Smooth through lumen	93.1769	12.88	290.4	NA	4.513
Snap rings	93.5469	0.855	0.4888	NA	0.08791
Spring internal	93.0076	58.89	7.086	NA	0.08345
Threads blind hole	95.9214	6.422	5.373	NA	5.949
Through slot	92.7300	105.5	13.481	NA	4.6152

Features designed to expose all challenging geometries, including buttons-exposed springs and external O-rings, passed both test conditions of soil drying times and semi-automated cleaning tests. While the buttons-exposed springs showed an extraction efficiency of 45.75%, suggesting potential fluid dynamics issues, cleaning studies confirmed that the complex geometry could be cleaned without manual brushing or flushing. Device features with occluded geometries proved more challenging to clean. The following features passed the 2-hour soil dry test but failed the 72-hour test: ball detent/ball bearing, button w/ spring, leaf springs, mated surfaces, screws threaded rod/threaded thru hole, screws threaded rod/threaded blind hole, and sliding shaft short. Additionally, leaf springs and sliding shaft short failed the additive extraction, indicating that their fluid dynamics allow soil to dry enough to alter protein chemistry, decreasing the soil solubility.

Some features—captured screw, hinge joints, pivot points, internal O-ring, and rough surfaces—showed variable results in soil drying experiments but passed semi-automated cleaning. Failures suggest that even brief soil drying increases cleaning difficulty, requiring more manual brushing and flushing. Conversely, passing semi-automated results highlight the reduced cleaning effort when soil remains moist before cleaning. The spring internal, threads blind hole, and through slot features failed both soil drying tests but passed semi-automated cleaning. This suggests they are more challenging to clean than features with variable results but can still be effectively cleaned without manual intervention.

The final group includes features that showed variable results in soil drying experiments and failed semi-automated cleaning: mated surfaces small clearance, sliding shafts long, smooth blind lumen, smooth through lumen, snap rings, ball seal springs, and the blind slot. Their failure in semi-automated cleaning indicates that manual brushing and flushing are necessary regardless of soil drying. The variability in soil drying suggests that their fluid dynamics promote soil drying and restrict fluid movement (e.g., water, cleaning agents), limiting the dissolution and removal of residual soil, thereby increasing cleaning difficulty. Furthermore, all features tested with an extended 60-minute soak failed to meet acceptance criteria, leading to the discontinuation of further testing for additional features that failed the 72-h soil dry experiment.

Discussion

Possibly the greatest misconception in sterile processing is assuming that all reusable medical devices pose the same level of difficulty when it comes to cleaning (Fig. 3). Infection risks increase when devices are designed in a way that allows soil to dry while preventing water and cleaning agents from effectively reaching, lifting, and removing contaminants. This ‘one-size-fits-all’ mindset not only compromises our current cleaning processes; but, it also creates a major obstacle to automation. Without rethinking our current approach, it will be extremely challenging to appropriately advance sterile processing into an automated future.

Fig. 3.

Feature ranking by cleaning challenge.

Medical device manufacturers invest significant resources in validating the processing instructions outlined in the IFU¹. These validations are time consuming and are not designed to determine the point of cleaning failure². Instead, they are structured to pass, often using cleaning steps that may be onerous for achieving cleanliness or include test conditions that do not accurately reflect the true worst-case scenarios in healthcare settings. For example, prolonged processing times or lengthy procedures can disrupt proper device handling prior to cleaning, such as the instructions in the IFUs that recommend covering the device with a wet OR towel to prevent soil from drying. However, the actual drying time in a clinical environment may differ from the conditions used in validation testing. As a result, some medical device manufacturers³⁶ validate two sets of instructions—one assuming soil does not dry and another to address the added challenge of dried soil.

Sterile processing professionals are urging medical device manufacturers to simplify instructions and provide more flexibility to align with standardized cleaning processes¹. However, standardization assumes that all devices can be effectively cleaned using the same instructions, regardless of initial soiling conditions, an assumption that does not hold true. Given this challenge, it is unlikely that most manufacturers will offer solutions that support standardization in sterile processing, as it does not align with their interests or provide a clear advantage for them.

Healthcare facilities worldwide are rethinking their approach to this challenge, recognizing that when responsibility for a reusable device shifts to the facility, so does the obligation to verify that their cleaning process is robust enough to ensure even the most challenging device features are properly cleaned^2,27. Leading institutions are taking a proactive approach by grouping devices into processing families and testing the most difficult-to-clean device within each group³⁷. This process verification ensures that their cleaning protocols are effective in delivering clean, reusable medical devices, as defined by ISO 15883-5.

Developing cleaning families for thousands of unique devices can be a daunting task, but a quantitative approach to grouping can provide clarity. In their publications, the authors in references^2,38 outline a method to establish cleaning families by assessing risk through key factors such as device features and patient exposure. This risk assignment is further supported by this research (Tables 3 and 4), which evaluated the impact of soil drying on cleaning effectiveness.

Kremer’s cleaning classification^39,40 simplifies the organization of reusable medical devices into cleaning families, allowing healthcare facilities to compare cleaning IFUs from different manufacturers and develop standardized cleaning instructions for greater efficiency. The effectiveness of these processes can then be verified by testing the most challenging device within each processing family to ensure thorough soil removal. Additionally, findings from this features-focused study also supports future sustainable waste management for reusable medical devices⁴¹.

Certain device features may support fully automated cleaning, eliminating the need for human intervention. Automation in medical device development and real-time monitoring systems will rely on high quality actual data being collected where the findings of this study aligns with meeting this expectation⁴². However, it is important to note that cleaning process automation is not typically included in the requirements set by medical device manufacturers, which means that its implementation rests primarily with the healthcare facility. By recognizing the increased cleaning difficulty when soil is left to dry on a device, healthcare facilities can establish policies to prevent this issue. Utilizing Kremer’s cleaning classification, they can create cleaning product families that minimize the need for manual intervention (e.g., wiping, brushing) and support the transition toward a fully automated sterile processing department.

Conclusion

Soil drying on reusable medical devices should be avoided, as it significantly complicates the cleaning process. Given the varying complexities of device features, cleaning processes in the healthcare facility should be designed to accommodate the variability in device design and verified to demonstrate efficacy. Given the challenges posed by soil drying, healthcare facilities should implement robust measures to optimize cleaning processes and minimize risks to patient safety. Establishing standardized protocols, such as covering instruments with a wet operating room towel at the point of use, can significantly improve the cleaning effectiveness of reusable medical devices and decrease the likelihood of inadequate decontamination.

This integrated study further indicated that certain features can be effectively cleaned without manual intervention. As a result, devices incorporating these features may qualify for a fully automated decontamination process. While the complete automation of cleaning procedures has not yet been achieved, this research establishes a foundation to support, enable and advance future developments in automated processes. Initiatives such as combining tacit knowledge and know-how from subject-matter experts from academia and industry will further advance design thinking for appropriate end-to-end cleaning and sterilization, which will also help inform pipeline innovation for effective processing with a lens on automation and patient safety⁹.

Acknowledgments.

Author contributions

All authors (T.K., G.McD., J.F., C.R., C.H.R, and N.J.R.) wrote the main text. T.K., J.F., C.H.R., and C.R. conducted research studies and prepared figures and tables. T.K., G.McD., and N.J.Rowan developed the methods and it’s conceptualization. G.McD. and N.J.R. mentored the study. All authors reviewed the manuscript.

Data availability

Data availability : The authors confirm that the data supporting the findings of this study are available within the article.

Declarations

Competing interests

The authors declare no competing interests.