There have been recent widespread reports of residual biological material such as dried blood, bone, and tissue in cleaned reusable devices. The presence of such biological material can prevent the effective sterilization or high level disinfection of devices, potentially exposing patients to harmful microbes. Manufacturers of reusable medical devices perform cleaning validation experiments to ensure that their cleaning instructions are adequate to remove biological material. The methods used to assess residual debris may not accurately quantify biological matter such as bone and tissue.
Therefore, we developed a filter-weighing method to determine the total mass of insoluble debris using partially-insoluble test soils. This type of assay may be particularly useful for quantification of particulate test soils that cannot be accurately measured using conventional methods for measuring soluble protein. After simulated soiling and cleaning of model and actual medical devices, residual debris on these devices was quantified by extracting in a liquid, filtering the liquid, and determining the total mass of debris captured by the filter. This method can be used as an adjunct to traditional methods to quantify contamination when devices are predicted to encounter particulate debris in a clinical setting.
We developed a filter-weighing method to determine the total mass of insoluble debris using partially-insoluble test soils.
Reprocessing reusable medical devices is a multistep procedure designed to maximize patient safety.1 Cleaning the device is a critical first step. A correctly cleaned device should be free of contaminants such as biological material (such as blood, tissue, and mucus), or nonbiological material (such as lubricants, detergents, and brush bristles). The presence of such material can potentially compromise the effectiveness of sterilization or disinfection processes,1–12 and transmit pathogens to subsequent patients. Several published reports have documented the transmission of infectious material from improperly reprocessed devices.13–15
Manufacturers of reusable medical devices provide users with cleaning instructions, which should be scientifically validated for efficacy in removing residual soil.16 A general protocol for such validation is to soil the devices through simulated or actual clinical use, clean the devices as specified in the instructions, and finally assess the cleanliness of the device.
In the past, an acceptable assessment for cleanliness was a visual examination of the device to ensure it was free of contaminants.1 However, modern medical devices are increasing in complexity, with internal design features that are not visible to the naked eye, including narrow opaque lumens, hinges, and stopcocks. Therefore it has become necessary to adopt more objective and quantitative assessments of a “clean” device.
A relatively simple method to quantitatively assess residual soil on a device is to extract the soil onto a carrier such as a liquid or a swab, and then assess these carriers for markers of soil. There are several caveats to this approach. Measurements of contaminants in the liquid will be expressed in terms of concentrations, and the volume of liquid may dilute the contaminant to a level below the limit of detection of the assay. Liquid extractions do not provide information on the location of debris within the device itself, but that information is important when device manufacturers consider improvements in design to increase cleanability of the device.
The goal of this study was to develop a method that would accurately quantify any residual particulate material.
Additionally, particulate residual contamination (such as bone or coagulated blood) is not fully soluble. Particulate matter in a liquid extract may cause sampling error when only a small fraction of the sample is assayed, and there is a random chance that the particulate contamination is measured. In addition, for accurate measurement, particulate matter is generally solubilized or digested prior to chemical analysis.17
Insoluble material was recently found in arthroscopic shaver handles, an orthopedic device used to shave the cartilage and bone of the knee.18 The interior of these devices had residual bone, cartilage, and tissue, which would not be accurately measured by conventional methods such as protein analysis. Given that the clinical soil found in these devices was largely particulate biological material, such material should be incorporated into test soils used during cleaning validation of these devices. The major challenge in using a test soil with particulates is that conventional methods of soil measurements would be unable to accurately quantify particulates such as bone or tissue.
The goal of this study was to develop a method that would accurately quantify any residual particulate material. A quantitative filter-weighing method to assess the total mass of residual debris in complex devices was developed. Model devices of different internal complexities and an actual arthroscopic shaver handle were used in this study.
Method and Results
Verification of Filter-Weighing Approach to Quantify Particulate Debris
In order to use the filter-weighing method in cleaning validation studies, we needed to establish that the mass measurements accurately reflected the debris present, and that the filter captured the test soil. The test soils chosen were a coagulated protein test soil, which consisted of purified hemoglobin, albumin, fibrinogen, and thrombin,19 as well as a bone meal, a plant fertilizer made of crushed bovine bones. Both the bone meal and the coagulated protein test soils appeared at least partially insoluble in water over the course of eight hours at room temperature (data not shown).
Varying amounts of test soils were dispensed into clean tubes, 50 ml of filtered water were added to extract the soils, the sample was passed through a 0.2 μm nylon filter, and the total weight of the debris captured by the filter was then determined. Control experiments with the coagulated protein test soil indicated that the mass of the test soil corresponded with the calculated mass based on the concentration of protein in the test soil and the volume of test soil dispensed.
The amount of bone meal captured increased linearly with the amount of bone meal dispensed into the tube (Figure 1A), indicating the filter was able to capture much of the bone meal. The percentage of bone meal captured by the filter is plotted in Figure 1B. With low levels of bone meal, the percentage of debris captured is variable due to small losses of the sample during the experiment. Sample losses have a greater impact on small samples. For bone meal levels above 5000 μg, the percentage of bone meal captured by the filter approached 80%.
To determine the extent to which proteins in the bone meal dissolved in the extraction water, an aliquot of the liquid extract was removed and analyzed for protein concentration by a conventional Bradford assay. For all bone meal samples, the amounts of soluble protein were below the limit of detection of the assay.
The amount of coagulated protein test soil captured by the filter also increased with the amount of test soil dispensed in the tube (Figure 2A), indicating that the filter could also capture this test soil. Bradford assays, performed concurrently, revealed that approximately 50% of the total coagulated protein test soil had dissolved in the extraction water (data not shown), explaining why the percentage of test soil captured by the filter was much lower for the coagulated protein test soil than the bone meal test soil (~30%; see Figure 2B).
For both test soils, ~80% of the test soil could be accounted for with either the filter-weighing approach or a combination of the filter-weighing approach and a protein assay. The 20% of unaccounted material may be due to sample loss during extraction, or loss of soil material that passed through the filter.
Quantification of Debris From Model Devices
In order to demonstrate that the filter-weighing method to assess residual debris can quantify the debris found in reusable medical devices, several model lumen devices of differing complexity were used in mock cleaning validation experiments. In general, we observed a trend towards increasing amounts of debris with increasing model complexity (Figure 3), which supports the hypothesis that complex devices are more likely to retain debris than devices with less complex internal design features.
These experiments also demonstrated the usefulness of the filter-weighing approach to quantify residual debris in a cleaned device. Concurrent Bradford assays revealed measurable amounts of protein only with device F (the actual arthroscopic shaver handle). Protein readings ranged from 0.05 – 0.1 mg/ml.
Bone meal test soil, consisting of ground bovine bone, was purchased from Maestro-Gro (Hamilton, TX). Varying amounts of bone meal were measured into a disposable weigh boat made of aluminum foil. The weigh boat was dropped directly into a 50 ml centrifuge tube, shaken to remove the contents on it, and removed (Figure 1). The coagulated protein test soil was described elsewhere19 with some modification.
Briefly, two components are prepared separately. Component A consists of 400 mg hemoglobin (MP Biomedicals, Solon, OH), 400 mg bovine albumin (Sigma, St. Louis, MO), and 60 mg fibrinogen (Sigma), dissolved in 8.6 ml of 0.9% sodium chloride solution. Component B consists of 400 mg hemoglobin, 400 mg albumin, and 25 NIH units of bovine thrombin (Sigma), dissolved in 8 ml of 0.9% NaCl and 5mM CaCl2 solution. Equal parts of Component A and Component B were mixed immediately before use (Figure 2).
Model devices A through E in Figure 3 were made of aluminum, and were designed and created at FDA facilities in Silver Spring, MD. Aluminum was chosen as a softer metal that would allow the models to be machined using appropriately sized available tools. In order to remove any grooves left from the machining process, the models were subjected to glass bead blasting that resulted in a uniform surface texture.
Model A had a single narrow lumen, with an internal diameter of 5 mm, and a length of 18 cm.
Model B had a lumen with two different connected diameters. The dimensions of the lumens were 5 mm diameter x 9 cm length, and 10 mm diameter x 9 cm length.
Model C had a lumen with varying diameters: 5 mm diameter x 6 cm length, 10 mm diameter x 6 cm, and 5 mm diameter x 6 cm length.
Model D was based on the design of an arthroscopic shaver handle, with the suction lumen diameter of 5 mm and length of 12 cm, a 45° angle to the stopcock region, with a diameter of 5 mm and a length of 2 cm, leading to the burr region, with a 2 cm diameter and 6 cm in length.
Model E was identical to Model D, with the exception of a 90° angle between the suction lumen and the burr region.
The arthroscopic shaver handle in Figure 3 (device F) and suction tips in Figure 4 (devices 1-4 and 6-7) were generously provided by Jahan Azizi, University of Michigan Hospitals. The arthroscopic shaver handle in Figure 4 (device 5) was purchased from a dealer of used medical devices. The make and model of device 5 in Figure 4 differs from that of device F in Figure 3.
Application of Test Soil on Model Devices
To quantify residual soil in model devices, 2 ml of the coagulated protein test soil (1 ml of Component A, and 1 ml of Component B) was mixed with 10 mg of bone meal and dispensed directly into the lumen to model devices. The volume of coagulated protein test soil was sufficient to completely coat the interior of the model devices. The ends of the devices were stoppered, and the devices were inverted five times, and then laid horizontally for 10 minutes. Excess test soil was shaken out of the devices, and residual test soil remaining in the devices was allowed to dry overnight (16–20 hours) at room temperature.
The cleaning method was designed to simulate the worst case cleaning of medical devices in a clinical setting. The devices were rinsed under tepid tap water for 10 seconds, and the exteriors of the devices were wiped with clean paper towels. The interiors of the devices were brushed 5 times with brushes ranging in diameter from 5mm to 10mm, depending on the diameter of the lumen (Spectrum Surgical Instruments, Stow, OH), followed by a 10 second tap water rinse. After cleaning, the exterior of all devices were visibly clean.
Each cleaned device was placed individually in a plastic bag containing 100 ml of filtered distilled water. The devices were placed on a rotating platform at 120 rpm for 2 hours at room temperature.
Millipore nylon filters (0.2 μm pore size) were purchased from Fisher Scientific (Pittsburgh, PA) and weighed on a calibrated Sartorius MEF-5 microbalance (Göettingen, Germany). Daily control measurements over the course of two weeks indicated the microbalance had an error of ±5 μg when weighing a metal gasket. Weight measurements for the nylon filters had greater day-to-day variability, possibly due in part to humidity changes; therefore an unused filter was used to normalize for the variability. One ml of sample was set aside for a protein assay, and the remainder of the liquid extract was passed through the nylon filter. Filters were allowed to dry for at least 18 hours at room temperature, and re-weighed to quantify debris captured by the filter.
Protein dissolved in the liquid extract was measured using the Bradford Assay. A sample (50 μL) was added to 1.5 ml of Bradford reagent (Sigma). Protein concentration was determined by spectrophotometry, as compared with a BSA standard linear curve. Under these standard conditions, the limit of protein quantification was 0.050 mg/ml.
The soil used in these experiments was in excess of the amount needed to coat the interior lumen of the devices, and in order to facilitate drying of the soil on the devices, extra soil was shaken out 10 minutes after soiling. Positive control experiments suggested the amount of residual debris measured on the devices reflected the inherent cleanability of the devices, and not the amount of debris that initially adhered to the device during the soiling process.
Devices that were soiled but not cleaned had between 1.6 and 3.4 mg of particulate soil, as measured by filtration (data not shown); and the amount of soil did not correspond to the median trends of residual debris from cleaned devices observed in Figure 3. A negative control experiment was also conducted in which the unsoiled devices were cleaned, and residual debris was extracted. No debris was captured from the devices.
Quantification of Clinical Soil From Used Devices
To determine if this method can quantify residual soil on clinically used devices, a used arthroscopic shaver and six clinically used suction tips, which are used to remove excess fluids and small particles, were tested. The exterior of all devices appeared visibly clean. Two of the suction tips (Devices 6 and 7) were cut open, and the residual soil was visible, trapped around the outlet port. The unknown variables with these devices include the age of the devices, number and types of procedures performed with the devices, types of clinical soil encountered by the devices, and the cleaning methods used on them.
The amount of debris extracted from the devices and captured by the filter varied from 0 to 145,000 μg (145 mg, see Figure 4) of debris. Protein levels in all the samples were below the limit of quantification. A chunk of clinical soil from Device 7 that was crushed into smaller pieces and was dropped directly into Bradford reagent failed to elicit the colorimetric shift typical of proteins. The lack of a positive response to the Bradford reagent suggests that the clinical soil found in this suction tip had very little protein, or that the protein was not accessible to the Bradford reagent.
Discussion and Conclusion
Clinical Soil and Test Soil
This work was initiated in response to reports of residual tissue and bone remaining on cleaned arthroscopic shaver handles. A literature search revealed similar reports of solid or insoluble material on cleaned, clinically used devices, such as instrument sets,20 laparoscopic and conventional instrument sets,21 and endoscopes.22 In some reports, the clinical soil has been described as either proteinaceous or as nonproteinaceous crystals.20,23 The nonproteinaceous crystals found in those reports may be similar to the clinical debris found in Device 7 (Figure 4). Additional work will be needed to determine the components of this nonproteinaceous clinical soil. It has been noted that a major contaminant for surgical instruments will be body fluids and particulate matter, such as tissue or bone.24
The exterior of all devices appeared visibly clean ... cut open, the residual soil was visible.
To simulate clinical conditions, the cleaning validation studies of devices that are predicted to encounter insoluble patient material should include a test soil that incorporates some amount of solid or insoluble material. To facilitate reproducibility from experiment to experiment, the components of the soil should be standardized, which may prove to be a challenge.
The coagulated protein test soil used in this work is partially insoluble, and is composed of readily available purified biochemical components. However, much of this test soil dissolves in water (Figure 2B), and may not adequately simulate particulates such as tissue. Although the bone meal test soil employed in this work was commercially available, it is not standardized, and there may be lot-to-lot variability in the particle size and content of the bone meal. Also, the bone meal test soil does not aggregate to the same extent as the clinical soil. Additional work will be needed to develop standardized insoluble test soils.
Quantifying Soil by Mass
Our results have shown that insoluble test soil or clinical soil can be quantified by mass from medical devices. The use of mass to quantify total debris has been described elsewhere.25,26 In previous reports, solid or insoluble residual patient material on clinically used devices has been described qualitatively by visualization20,21,23 or discarded.22 Mass measurements have also been used to quantify biofilm27 and manufacturing residues from metallic medical components.28
Additionally, the entire mass of a device with debris was measured to determine the mass of debris on the device; however, it is not clear if the debris on the device was water-insoluble. Quantification of miniscule amounts of particles is now possible with microbalances that have a precision level of 1 μg or less. These highly sensitive instruments are often used to assess air quality by measuring particle mass.29
The entire sample is assayed, reducing sampling error.
The assay is useful with a variety of soils with particulate components, regardless of the chemical composition.
The assay is sensitive.
The assay is quantitative.
Filtrate can be saved to measure the concentration of specific solutes.
The method requires few pieces of specialized equipment.
The assay does not require extensive expertise or training.
It may also be possible to characterize the debris on the filter, either by submersing the filter in a particular detection reagent, or by utilizing an image analysis technique such as Fourier transform infrared spectroscopy.
Additional research will be needed to address several remaining questions. It is unclear what amount of residual particulate debris can remain in the device for the device to still be considered “clean.” Further testing will be needed to determine a safe level (if any) of residual particulate debris, as well as the lowest levels of residual particulate debris that can be reasonably achieved in a clinical setting. There will also need to be additional work to determine the components of clinical soil found in different reusable medical devices, so that appropriate test soils can be used with different devices. Additionally, it is unclear what impact different materials may have on the accumulation of residual debris.
In conclusion, quantifying the mass of residual debris in medical devices will provide valuable information when used in conjunction with assays to measure traditional markers of contamination, such as protein, total organic carbon, and other organic and inorganic materials. This approach is particularly valuable for quantifying insoluble or low-protein soils.
We thank Jahan Azizi for generously providing devices for testing, and Anne Lucas and Brian Haugen for helpful discussions.
About the Authors
Shani Haugen, PhD, is a premarket medical device reviewer with the Office of Device Evaluation, Center for Devices and Radiological Health, U.S Food and Drug Administration (CDRH/FDA). E-mail: firstname.lastname@example.org
Nandini Duraiswamy, PhD, is a research biomedical engineer with the Office of Science and Engineering Laboratories, CDRH/FDA. E-mail: email@example.com
Victoria M. Hitchins, PhD, is a research microbiologist with the Office of Science and Engineering Laboratories, CDRH/FDA. E-mail: firstname.lastname@example.org