Slower drill rotation could reduce COVID spread during dental procedures
Researchers have found that careful selection and operation of dental drills can minimise the spread of COVID-19 through aerosols
Researchers at Imperial College London and King’s College London have measured and analysed aerosol generation during dental procedures leading to suggested changes that could improve safety for both patients and the practice workforce.
Dental procedures can pose a high risk of viral transmission because the tools that are used often produce aerosols, which can contain high numbers SARS-CoV-2 virions; copies of the virus causing COVID-19.
The aerosols are generated when saliva mixes with water and air streams used in dental procedures. As a result, access to routine dentistry continues to be limited during the COVID-19 pandemic.
Dental practices have had to introduce new room decontamination processes and personal protective equipment measures which have dramatically reduced the number of patients that can be treated in a single day.
In particular, dentists need to leave long intervals between treatments, leaving rooms unoccupied to allow aerosols to dissipate. This is limiting patient access and challenging financial feasibility for many dental practices worldwide.
The researchers suggest that dentists avoid using dental drills that use a mixture of air and water as the abrasion coolants, and carefully select and control drill rotation speeds for those instruments that only use water as a coolant. Parameters have been identified that would allow some procedures such as dental fillings to be provided whilst producing 60 times fewer aerosol droplets than conventional instrumentation.
Lead author Dr Antonis Sergis of Imperial’s Department of Mechanical Engineering said: “Aerosols are a known transmission route for the virus behind COVID-19, so, with our colleagues at King’s, we have tested suggested solutions that reduce the amount of aerosols produced in the first place. These could help reduce the risk of transmission during dental procedures.”
Co-author Professor Owen Addison of King’s College London’s Faculty of Dentistry, Oral & Craniofacial Sciences said: “This important work describes the basic mechanisms that lead to the features of dental aerosols that we currently consider to be high risk. This has enabled us to choose drill parameters to keep our patients and the dental team safe at this difficult time. Although we cannot provide every procedure, because slowing our drills is much less efficient, we now have the basis to do more than we have done in the last 6 months.”
The results are published today in the Journal of Dental Research and are already being included as evidence in guides for dental practices in the UK during the pandemic. The collaborative research used the engineering expertise at Imperial and clinical expertise at King’s College London’s Faculty of Dentistry, Oral & Craniofacial Sciences.
The researchers used dental clinical rooms at Guy’s Hospital in London to test how aerosols are generated during procedures such as decay removal, applying and polishing fillings and adjusting prostheses. They measured the aerosol generation using high speed cameras and lasers. They then used these findings to suggest modifications.
They found that using air turbine drill types, which are the most common type of dental drill, creates dense clouds of aerosol droplets that spread as fast as 12 metres per second and can quickly contaminate an entire treatment room. Just one milliliter of saliva from infected patients contains up to 120 million copies of the virus, each having the capacity to infect.
They tested a different type of drill, known as high torque electric micromotor, with and without the use of water and air streams. They found that using this drill type at low speeds of less than 100,000 rpm without air streams produced 60 times fewer droplets than air turbine drill types.
In addition, they found that aerosol concentration and spread within a room is dependent on the positioning of the patient, presence of ventilation systems, and the room’s size and geometry. It is also influenced by the initial direction and speed of the aerosol itself, which can be affected by the type of cutting instrument (burr), and the amount and type of cooling water used.
The researchers say that by understanding how to reduce the amount of aerosol generated in the first place, their suggestions could help dentists practice more and help patients get the treatment they need.
They also note that patients should still not attend dental appointments if they have symptoms of COVID-19.
Professor Owen Addison from King’s said: “Because of the COVID-19 pandemic, dentistry has become a high-risk practice – but the need for treatments hasn’t gone away. Our suggestions could help begin to open up dentistry to patients once again.”
Their suggestions have been included in the evidence appraisal in dentistry document entitled “Rapid Review of Aerosol Generating Procedures in Dentistry”, published by the Scottish Dental Clinical Effectiveness Programme (SDCEP). The results from the study have also been considered by an expert task force convened by the Faculty of General Dental Practice (FGDP (UK)) and the College of General Dentistry and published in their guide entitled “Implications of Covid-19 for the safe management of general dental practice.”
Co-author Professor Yannis Hardalupas of Imperial’s Department of Mechanical Engineering said: “The impact of the results is significant. For example, the risk categorisation for dental procedures included in the FGDP (UK) document was certainly influenced by our work.”
The team’s research is ongoing. They are currently better assessing the risk of infection by quantifying the amount of saliva mixed into the generated aerosols by dental instruments.
The work is supported by LaVision, Photron, the Engineering and Physical Sciences Research Council (part of UK Research and Innovation), British Endodontic Society and Guy’s and St Thomas’ NHS Foundation Trust.
Figure 1. (A) is a still frame from high speed imaging of an unobstructed spray from an air turbine possessing a high velocity spray, a turbulent shear layer at the periphery of the core and recirculation regions close to the burr tip. (B) is the same image false coloured with the central spray core in orange, a shear layer region located outside of the central core region in yellow and recirculation regions in blue. A group of high droplet velocity straight trajectories is shown by the red streaks towards the top of the image. Figure 1 (C) When the same instrument is placed inside a simulated oral cavity (palatal to the maxillary central incisors) a turbulent fine mist of a reduced but significant velocity is produced (principle direction indicated by the yellow arrow).
Figure 2 shows an instantaneous image of the spray formed by an air turbine unobstructed running in a steady state (A), the probability distribution of the spray droplet concentration based on >2000 images (B) and its standard deviation (C). For interpretation in (B) pixels that are red indicated a 100% chance of encountering a droplet at any point in time, pixels that are black 0%. In (C) standard deviation is plotted on an equivalent scale.
Figure 3. Probability density function maps of droplet concentration for a micromotor with a 5:1 speed increasing handpiece run with no ‘chip air’ rotating between 200,000 and 20,000 rpm. Probability density function maps are based on >2000 images for each modality with the instrument running unobstructed in steady state. At 200,000 rpm, most of the droplet velocity is ~1.4m/s which is an order of magnitude less than an air turbine. PDFs of droplet concentration (top line) and standard deviation of the concentration fluctuations (middle line) for a micromotor run with no ‘chip air’ rotating at decreasing speed with the burr tip in contact with wet enamel. Distributions are again based on >2000 images for each modality. Spray distribution and trajectories are modified by tooth contact. There is evidence at higher speeds that atomization near the tooth surface occurs; however with decreasing speed the coolant largely streams over the tooth surface with a limited number of low velocity droplets being deposited within the imaging field of view. Scale bars are equivalent to 50 mm in all images.
Figure 4. (A) Proportion of image field where the probability density function of droplet concentration is greater than 1% plotted against micromotor speed, for the entire imaging field in (B) (black coloured bars) and the left side of the imaging field (grey coloured bars), which is ~90 mm from the burr tip. For comparison, equivalent values for an air turbine are plotted as diamonds at the right of the histogram. We observe that reduction of micromotor speed to 60,000 rpm results in: at a distance of >90 mm away from the burr only 0.1% of the imaged pixels have >1% chance of encountering a droplet. This represents a ~280-fold reduction compared with an air turbine. (C-E) show representative sprays with an elimination of radial atomization below 100,00 rpm. F and G are histograms showing droplet particles sizes associated with non- pre-misted micromotor speed (rpm) compared with ambient baseline (BL) and air turbine with pre-misted coolant.
Figure 5. Representative images showing spray formation when mechanisms identified to cause atomization are mitigated, achieved here by running an electric micromotor with 5:1 speed increasing handpiece with ‘chip air’ blocked and revolutions restricted to 60,000 rpm. The burr was held in contact with the palatal surface of right maxillary central incisor (A), lingual surface of right mandibular central incisor (B), and the occlusal surface of the right mandibular first molar (C).