Introduction

A number of concerns have been raised over the efficacy of instrument decontamination in sterile services departments,1,2 endoscopy reprocessing units3,4 and general dental practice.5 A critical control point within the decontamination cycle is the efficacy of the cleaning process. A large observational study has demonstrated that in general dental practice the cleaning of dental instruments is poorly controlled6 and insufficiently managed.7 For some dental devices that are difficult to clean, such as matrix bands and endodontic files, this may result in visible blood contamination remaining even after reprocessing,8,9 with consequent recommendations for these items to be classed as single use only.10,11 In addition to blood contamination, concerns have been raised that the proteinaceous residues derived from previously treated patients may represent a risk of transmission of vCJD.12 These concerns are heightened by reduced susceptibility of the infectious agent to be removed and/or inactivated by conventional cleaning and sterilisation processes.13

Within general dental practice the level of risk of cross-infection associated with poor instrument decontamination is unclear. Anecdotal evidence suggests that transmission of viruses such as hepatitis B14 and bacteria such as Staphylococcus aureus can occur in dental practice.15,16,17 Since no systematic surveillance of post-operative infections following dental procedures is undertaken, it is unclear how frequently these events take place in reality. It would seem prudent, therefore, that the reprocessing of dental instruments follows the route used in the reprocessing of other surgical instruments identified in European standards18,19,20 and national guidelines21,22 and under the management of an appropriate quality management system.23

While a number of studies have been undertaken of instruments reprocessed in sterile service departments,24,25,26 little work has concentrated on the range of processes used to clean commonly used dental instruments, particularly in a real-life setting. The aim of this study was to investigate the efficacy of cleaning dental instruments by measuring residual protein following a) manual cleaning only, b) manual plus ultrasonic cleaning, and c) automated washer disinfector (AWD) cleaning undertaken as part of the routine dental surgery reprocessing schedule in dental practices.

Methods

Selection of dental surgeries

Dentists located in the South West of England were selected from local primary care trusts' published lists of practices holding NHS contracts. The surgeries were contacted in writing inviting them to participate in a dental instrument decontamination study. If practitioners wished to take part in the study they were asked to indicate which instrument cleaning process was used in their surgery: manual cleaning only, manual plus ultrasonic cleaning or automated washer disinfector (AWD). From this initial list, ten surgeries from each instrument cleaning group were randomly selected to provide instruments for analysis. No instruments submitted for analysis in this study were returned to the surgeries; all were replaced with new items funded by the study. No information was available regarding either the instrument or the number of times each instrument had been used and reprocessed. Instruments were collected over the period December 2005 to October 2007.

Selection of dental instruments

Six different types of dental instrument, selected to represent a range of complex surfaces and degree of invasiveness, were analysed from each surgery. The instruments were extraction forceps, sickle scalers, diamond and steel burs, matrix bands (Siqveland) and the associated matrix band retaining clips.

In order to provide a source of reference for the extent of protein removal following each cleaning process, a collection of used but uncleaned instruments was also assayed for protein content. These instruments were obtained from a similar cohort of dental surgeries in the South West of England. All these instruments were steam sterilised through a 134°C cycle before analysis to enable safe handling.

Decontamination equipment and process data collection

For each dental surgery, an observer collected information by direct observation of the cleaning process and equipment and, where appropriate, reviewed relevant documentation on a standardised data collection form.5

Visual assessment of cleaned instruments

All instruments were assessed for visual contamination under a binocular microscope and scored between 0 (no visible debris) and 3 (high levels of visible debris) at HPA, CEPR, by three independent operators. The scoring system was based on that previously used and published9 for the visual scoring of endodontic files. Instruments were viewed through an Olympus SZ40 microscope and captured on a Nikon D50 Digital SLR camera with macro lens (SIGMA 50mm 1:2.8 DG MACRO D). Captured images and scores were stored using an image database system (Image Access Standard 5).

Protein analysis

Residual protein on uncleaned and cleaned instruments was extracted by sonicating the working end of each instrument at 32-38 kHz for 2x60 minutes in freshly prepared 0.05% aqueous Decon 90 (Decon Laboratories Ltd, Sussex, UK) at room temperature using a Medisafe digital PC ultrasonic bath and an Ultrawave model QS3. The wash liquids were assayed for protein concentration using the o-Phthalaldehyde/N-acetylcysteine assay as previously described27 with a limit of quantification of 0.3 μg of protein per instrument. The protein values from both washes were combined to obtain the total residual protein on each instrument. Instruments with two working ends were sonicated twice in the same wash solution.

Statistical analysis

Continuous data were expressed as median and interquartile ranges (IQR). Two proportion and Mann-Whitney pairwise analysis was used to compare the significance of relative residual protein levels between processes (p <0.008 with a Bonferroni correction applied) using MiniTab (version 15). Spearman's rank correlation coefficient (rho) and ROC curve regression models were both applied independently to assess the correlation of visual score and corresponding residual protein level.

Results

Decontamination equipment and processes used

Where manual cleaning of instruments was observed, 7/14 surgeries used no detergent and 7/14 used surgical handwash. Where ultrasonic cleaning of instruments was reported, the detergent used in the ultrasonic bath was either neutral detergent or the manufacturer's recommended brand in 4/8 sites. The remaining 4/8 sites used a disinfectant/detergent combination. In surgeries undertaking ultrasonic cleaning of instruments the range of time for emptying the ultrasonic bath varied from three to 40 hours. In none of the eight surgeries was the ultrasonic bath subjected to any cleaning or ultrasonic efficacy testing.

The machine in all ten surgeries operating AWDs was from the same manufacturer (Medisafe), with the same model (Pico), using the same programme and the same detergent (3E-Zyme). The automated cleaning process was by spray action only; no channel irrigation of lumened devices (handpieces) was observed. No surgery undertook checks to ensure correct loading of each carrier before processing and no records of cleanliness failures were kept. A small number of surgeries (3/10) had undertaken some form of performance testing of the AWDs. There was insufficient information available at the surgeries to determine whether any machine had been tested for cleaning efficacy. No test records had been independently audited by an Authorised Person (Sterilisers) for any surgeries using an AWD.

Visual assessment

The median visual scores across all the different instrument types for the range of process used were as follows: manual only (0.5), manual plus ultrasonic (0.25) and AWD (0.25). The median visual scores for each type of instrument for all cleaning processes showed that the steel burs scored highest (0.5) followed by the forceps (0.25), matrix band retaining clips (0.25), matrix bands (0.25), diamond burs (0.25), with scalers having the lowest median visual score of 0.

The maximum median visual score seen for any instrument type across the different cleaning processes was 0.75; this was observed on matrix band retaining clips cleaned by a manual process only, and steel burs cleaned by a manual plus an ultrasonic process and also by AWD.

Effect of cleaning on instrument residual protein levels

For all instruments subjected to a cleaning process (n = 1,304), 72% had detectable residual protein contamination. The median amount across all the instruments tested was 10.25 μg with a range from 0.3 μg to 3.85 mg. To provide an estimate of the protein load on instruments before cleaning, 177 used instruments were analysed and found to have a wide range in level of protein contamination (Table 1). The forceps had the highest median values of 462 μg (interquartile range of 285–759 μg), with the diamond and steel burs having the lowest median protein recovered: 0.6 μg (0.1–2 μg) and 0.4 μg (0–3 μg) respectively.

Table 1 Recovery of total protein per instrument per cleaning process. Values expressed as the median total μg of extractable protein per instrument (interquartile range). *Values below the measurable range of the assay were reported as 50% of the Limit of Quantification (LOQ)

Analysis of the effect of the three cleaning processes on the six different types of instrument is shown in Table 1. Results varied both within and between the different processes and instrument types. Compared to uncleaned sickle scalers, the median residual protein levels on cleaned sickle scalers using manual or manual and ultrasonic cleaning were higher. This phenomenon was also observed for steel and diamond burs with the exception of diamond burs in the AWD. For the matrix bands and matrix band retaining clips, the lowest level of residual protein was found in those cleaned in the AWD. The lowest levels of residual protein from extraction forceps was found in those subjected to a manual cleaning process.

Statistically, the recovery of residual protein from the combined manual and ultrasonic process across all instrument types was significantly higher than the other two processes (p <0.008). Although use of an AWD was not statistically better overall than manual cleaning alone, the automated method did significantly reduce the number of instruments with residual protein levels above 50 μg per instrument (Figs 1a–c).

Figure 1: a, b, c Bar charts showing the distribution of protein residue measurements for different instruments and cleaning processes.
figure 1

The protein residues for instruments cleaned by manual only (a), manual with ultrasonic (b) or automated washer-disinfector (c) were measured. Data, expressed as μg of protein per instrument, is provided to demonstrate the greater distribution of data at the lower end of the measurable range (shown in 20 μg classes) and to provide an indication of the distribution of measurements at the higher ranges

Effect of instrument type on cleanability

Comparison of the residual protein contamination following cleaning of all instrument types demonstrated that the extraction forceps had the highest median level of residual protein of 28.4 μg (IQR 5–84 μg). This was followed by the matrix band retaining clip, scalers, steel burs and matrix bands.

Correlation between visual scoring and protein contamination

Comparative analysis of the visual score data with residual protein data showed no correlation as determined by a Spearman's rho analysis. All correlations between processes and instruments showed r-values of less than 0.7, meaning no correlation between the visual score assigned by the operator and the actual residual protein present on the instrument. This was regardless of process used or instrument type.

Discussion

This study is the first to provide detailed analytical findings on the efficacy of manual and automated cleaning processes carried out under general dental practice conditions on a large number of dental instruments. Observations on the processes, cleaning chemicals and management control are similar to earlier reports from a larger observational study6 demonstrating multiple shortcomings. Our findings add to the literature by providing data on AWDs, which before 2005 were not in widespread use in dental practice in the UK. These observational data are important when interpreting the results of our study since the efficacy of the cleaning process will be strongly influenced by a number of interrelated factors such as cleaning chemicals, water quality, physical energy used (manual, ultrasonic, water jets), cleaning time, cleaning temperature, instrument set-up and design. The manual plus ultrasonic method was significantly less efficient than the other processes as determined by Mann-Whitney statistical test (p <0.008 with a Bonferroni correction applied). The use of inappropriate cleaning chemicals, for example, Hibiscrub, would have compromised the cleaning process. Similarly, failure to change the ultrasonic bath water frequently allows the build-up of contaminants that will increase detectable protein residues on instruments exposed to these liquids as opposed to reducing them. These observations may explain the higher levels of protein found on instruments compared to uncleaned instruments following the use of ultrasonic baths. Additionally, the efficacy of the ultrasonic baths may have been compromised since no periodic testing of functionality of these devices was undertaken. Similarly, further improvements in the cleaning efficacy of the AWDs beyond the level seen in this study are likely if equipment is validated and tested according to the required European standards. The differences in protein levels on forceps processed by manual-only cleaning and in the AWD may be explained by incorrect loading of the AWD, since no checks were made on the AWD loading pattern. The use of an enzymatic detergent (3E-zyme) containing protein may have contributed to the residual protein measurements for the AWD samples, if inadequately rinsed. Nevertheless, use of an AWD gave the lowest median levels of residual protein for four of the six instrument types.

Analysis of residual protein levels on the different instrument types cleaned by the three processes demonstrates the innate complexities of cleaning dental instruments. Under general dental practice conditions, no single process was universally most effective at cleaning all instrument types. In addition, our data showed that there was no single instrument type which proved to have consistently either the highest or lowest levels of residual protein following cleaning. Manual cleaning of dental instruments is subjective and not reproducible, with variations, for example, in the type of detergent, water temperature, brush type, and in the number and strength of strokes used in the cleaning process, with minimal opportunity for control or validation. However, when carried out as a sole method of cleaning for some types of instrument, the process can be highly effective as demonstrated by the low levels of protein recovered from forceps (0.3 μg, IQR 0.3–38 μg). This may well reflect the targeted cleaning of joints and serrations by a dedicated staff member when compared to forceps which may have been incorrectly loaded into an AWD that has not been validated.

Manual cleaning was not as effective for the remaining instrument types although overall statistical analysis does not demonstrate a significant difference compared to the AWD used under the conditions in this study. Proportional analysis of the four larger instrument types (scalers, matrix bands, retaining clips and forceps) demonstrated that use of the AWD showed a significantly greater number of instruments with <50 μg protein recovered than the manual process (p <0.118). This difference may well be enhanced further in inappropriately installed, tested and operated equipment. It should be noted that this study analysed the effect of cleaning in only one model of AWD using the detergent and wash cycle as specified by the manufacturer. Other models of AWD, chemistries and cleaning cycle parameters may produce different results in terms of cleaning efficacy.

The lack of correlation between visual scoring systems and protein residues has been reported previously,9,24,25,26 demonstrating that visual assessment even under high magnification has a low predictive value for determining residual protein levels. However, the examination of cleaned instruments under magnification and controlled lighting is highly specific in detecting gross soil deposits and material defects in instruments and forms an important part of overall instrument decontamination quality control. Further work is necessary to develop tests that are more sensitive and specific for determining a quantifiable end point in the cleaning process and which can be used in general dental practice.

The ortho-phthaldialdehyde assay only reacts with exposed primary amines on proteins and there are no data currently available that correlate the values detected on instruments post-cleaning in this study with potential infectivity from TSE infectious agents,28 although it does provide useful data for assumptions on the likely reduction in risk through improvements in cleaning processes.29 Our findings are also useful to put other instruments and processes into context; for example, using a similar methodology, the median residual protein value on cleaned endodontic files was 5.4 μg per file (range 0.5-63.2 μg).27 For other surgical specialities it has been demonstrated that protein residue levels in the range 0.1 μg to >1.0 mg per instrument are not uncommon.24,25,26

No threshold value has been defined as representing an acceptable residual level of protein on any form of surgical instrument after reprocessing, with guidance provided as to achieving 'best practice' rather than specific performance levels. Further work should focus on defining achievable baseline levels of residual protein on surgical and dental instruments, for different cleaning processes. This could form the basis for improved decontamination practices. Such a threshold would enable policy decisions to be made on which of the alternative cleaning processes would remain part of acceptable practice and drive improvements to instrument installation, commissioning and routine use. In particular, the use of ultrasonic baths needs to be reviewed and improved guidelines established. While the expectation is that improved procedures would result in reduced levels of residual protein contamination, there is little evidence in the literature to support this conclusion and further studies would be invaluable.

The findings described here are especially relevant since the instruments were reprocessed using cleaning practices in common use and were compared to instruments cleaned in AWDs. Despite inadequate set-up and use, the AWDs demonstrated improved consistency in the cleaning of instruments. No doubt further improvements in efficacy could be made with closer attention to the relevant standards and guidelines for their operation and management. Some instruments, such as extraction forceps, may benefit from a manual clean before loading into an AWD. Further improvements in cleaning efficacy could also be made with appropriate education and training of dental staff in cleaning parameters. Practitioners should also receive more logistical support to assist with the technical challenges posed by inadequate commissioning and testing of equipment by manufacturers and suppliers.

In conclusion, this study provides additional data from large numbers of dental instruments on the levels of residual protein following cleaning in general dental practice. The study highlights areas where improvements in cleaning efficacy should be achievable with enhanced training and technical support for equipment installation, operation and routine testing.