*4.1. Setting*

A 24-month investigation was carried out in a Hospital Odontostomatology Clinics (HOC) and in three Private Dental Clinics (PDC) housing 13 and 6 dental units, respectively.

A validated checklist was used to identify the risk factors associated with contamination of DUWLs. The checklist included all the necessary items to create a schematic diagram showing the layout of the building's water systems and DUWLs, as well as the information related to the compliance with good practices in controlling *Legionella* colonization.

The DUs examined here varied in model and year of installation: in the HOC, 8 Puma ELI, Castellani S.p.A., were installed in 2010 and 5 Skema 4, Castellani S.p.A. were installed in 2005; in the PDCs 2 C4+, Sirona Dental Systems GmbH, were installed in 2014, 2 LINEA 90, World Health Organization (WHO) OMS, were installed in 2007; and, finally, 2 Puma EVO 5, Castellani S.p.A., were installed in 2010.

All devices were supplied through an independent system that received water from a 1-liter polypropylene bottle containing sterile water, manually filled. During the dental care activity, the municipal water fed the DUWLs by activating a switch when the bottle was running out of water. Contrastingly, the DU in the operating room was completely disconnected from the municipal water supply. In the hospital, the drinking water was softened and treated with a secondary chlorine dioxide disinfection (0.2–0.3 mg/L), while in private dental clinics the water was only softened.

In almost all DUs (Castellani and Sirona Dental Systems GmbH), at the end of the day, an automatic disinfection cycle was activated. According to the manufacturer's recommendations, a disinfectant product that generated peracetic acid (0.2%) was applied. Only in one PDC, housing 2

OMS dental units, was disinfection performed manually on a daily basis through the application of 2.5% quaternary ammonium compounds. In the hospital, a procedure was written and shared with the dental clinic sta ff that was trained for the correct adoption of the DU managemen<sup>t</sup> procedure. According to the procedure, all DUWLs were flushed for at least two minutes at the beginning and end of the day and after any significant period of inactivity. In addition, flushing was performed for at least 20–30 s between patients. No private clinic had a procedure where the DUs' maintenance protocol was reported.

Following the results obtained from the microbiological investigations, a control program was started with more restrictive measures on the managemen<sup>t</sup> of the quality of the supplied water, the periodic maintenance of the dental units (for example, the lubrication of the quick coupling seals on the handpiece, self-contained water bottles disinfection, etc.), the compliance with the flushing frequency, and the introduction of a shock disinfection procedure for the DUWL, where necessary.

A shock disinfection treatment was applied under the recommendations of the di fferent dental unit manufacturers, when pathogens were isolated from the DUWL (the presence of *P. aeruginosa, Legionella* spp. or other pathogens). In particular, treatments were applied in all DUs except for the one of the operating room, which was disconnected from the water supply and fed only with sterile water.

#### *4.2. Samplings and Microbiological Tests*

Before the disinfection treatment, every three months, for each DU, 1.5 L of water was collected from the inlet, spittoon and handpiece. To assess the e ffectiveness of the treatment, sampling was carried out after 30 days. Moreover, 1.5 L of tap water was collected from a sink located near the DU. All water samples were collected without flushing before the start of professional activity.

The total microbial count at 22 and 37 ◦C, *Legionella* spp., *P. aeruginosa* and coliform bacteria were determined in all water samples, in accordance to the Italian and French guidelines [20,21]. One more liter of water was collected from the same devices for the detection of free-living protozoa (FLA) according to international standards [42].

Water temperatures were tested in all samples; residual chlorine concentrations were determined only in water sampled from taps and dental units fed from the municipal water network.

*Legionella* spp. was detected as described by ISO11731:2017 [43]. One-liter samples were filtrated through a 0.2-μm membrane (Millipore, Billerica, MA, USA). The membrane was immersed in 10 mL of water and sonicated for 5 min to allow the bacterial cells to separate from the membrane. The resulting suspension was divided into three aliquots, the first one was brought to 50 ◦C to select *Legionella* spp. over the other bacterial species non heat resistant, the second one was added with 9 volumes of HCl-KCl acid solution, and the last one was plated with no treatment. A total of 0.1 mL of each aliquot was plated on Bu ffered Charcoal Yeast Extract (BCYE) Agar and Glycine Vancomycin Polymyxin Cycloheximide (GVPC) agar plates (Oxoid, UK), and incubated at 37 ◦C for 7–10 days in jars under an atmosphere containing 2.5% CO2. The suspect *Legionella* colonies were tested for species and serogroup by polyvalent agglutination latex test (Legionella latex test—Oxoid, UK).

The total microbial count was performed by inclusion in Plate Count Agar (Oxoid, UK) according to ISO 6222:1999 [44]. Coliform bacteria and *Pseudomonas aeruginosa* detection was performed by filtrating 100 mL of water through 0.45-μm membranes (Nalgene, Rochester, NY, USA). The membranes were laid on Endo Agar Les plates (Liofilchem, Italy) and Cetrimide Agar (Oxoid, UK) according to ISO 9308-1:2014 [45] for coliforms growth and ISO 16266:2006 [46] for *Pseudomonas aeruginosa* growth. The species' confirmation of suspect colonies was obtained by Mini API galleries (bioMeriéux, Marcy-l'Étoile, France).

One-liter samples were filtrated through a 0.2-μm pore size of membrane (Millipore, Billerica MA, USA) for free-living protozoa search. The membrane filters were minced in 10 mL of sterile phosphate-bu ffered saline pH 7.2 (PBS), homogenized by vortex for 5 min and centrifuged at 1200 *g* for 15 min. A total of 200 μL of pellet was inoculated on Non-Nutrient Agar (NNA) with a lawn of heat-inactivated *Escherichia coli* in Page's Amoeba Saline solution (PAS) and incubated at 37 ◦C. The

presence of FLA was investigated by examining the NNA culture plates by inverted microscope Dmi1 (Leica Wetzlar, Germany) using 20× and 40× objectives.

From all positive samples, the growing amoebae were harvested from culture plates, placed in Eppendorf tubes and washed two times with PBS, pH 7.4, before molecular procedure. DNA extraction was performed by using the QIAamp DNA Micro Kit (Qiagen, Milan, Italy). To identify FLA species, the 18S rDNA amplification with primers P-FLA-F and P-FLA-R was performed, according to the protocol published by Cristina ML et al. [18].

#### *4.3. Shock Disinfection Treatment*

Shock disinfection treatments were applied in DUs housed in PDC and HOC, when contaminated by pathogens. Subsequently, the treatment was carried out every three months until an acceptable result was obtained. The effectiveness of disinfection was assessed after one month and in the subsequent sampling, scheduled every three months.

In PDC, dental units the disinfection was performed with a galenic formulation of 3% v/v hydrogen peroxide applied in dental units' system for 1 h and followed by water flushing.

In HOC, a shock disinfection was performed by using a galenic formulation of 3% and 6% v/v hydrogen peroxide and a second treatment was applied with a solution containing 4% v/v hydrogen peroxide and surfactants (Green Line Hydrogen Peroxide Cleaner 4%, Corcraft Product, Comstock, NY, USA).

All disinfection treatments were carried out in a similar way, by using a biofilm removing system and a dye tracing to detect the disinfectant passage. The method of use was the following:


Membrane filters, 0.2 μm in size (Pall Corporation, New York, NY, USA), were set at the inlet of each dental unit to ensure a good microbial quality of the entrance water. The filters' installation was done simultaneously to the first disinfection of the dental unit waterline in the PDC where they were installed after the treatment with 6% v/v hydrogen peroxide in HOC.

**Author Contributions:** B.C., A.B., G.P. conceived and designed the experiments; B.T., B.C. and M.T. performed the water sampling; B.T., M.T. and B.C. performed laboratory tests; B.C., B.T., D.D.C., M.L.C., A.M.S., S.P. performed data elaboration and wrote the paper. All authors have read and agree to the published version of the manuscript.

**Funding:** This research received no external funding.

**Conflicts of Interest:** The authors declare no conflict of interest.
