**3. Discussion**

The dental unit consists of a complex water pipeline network connected to multiple pieces of equipment. Such an environment implicates a high risk of microbial contamination and transmission, especially through the contamination of water and bioaerosol by dental instruments, which are placed very close to patients and medical staff during dental treatments [31].

Even though the evidence associating DUWLs with risks for patients and staff is contradictory [32–35], exposing patients or dental healthcare personnel to water of uncertain microbiological quality is inconsistent with general accepted infection control principles [19]. Since patients and dental staff are regularly exposed to water and aerosols generated from the dental units, the microbial quality of the water in the DUWLs is extremely important. It is not acceptable that a DUWL should not meet the drinking water standards (<100 CFU/mL at 22 ◦C and <10 CFU/mL at 36 ◦C) [20,21].

To reduce the risk coming from contaminated DUWLs, it is necessary that dental healthcare workers routinely apply suggested infection prevention strategies [13]. The Italian guidelines sugges<sup>t</sup> many approaches to reduce the microbial contamination and/or biofilm formation, including both non-chemical (using the anti-stagnation device, flushing, supplying the circuit with sterile solutions, antimicrobial filter installation, etc.) and chemical methods, which provide for the use of disinfectants continuously or in periodic shock treatment, following the manufacturer's instructions [36]. Not an available single method or device will eradicate the bio-contamination of DUWLs or exclude the risk of cross-infection. To reduce contamination risks, a combination of methods is desirable [36]. Where delivered water quality is in doubt, dental practice should consider adopting continuous dosing systems or shock disinfection, if permitted by the manufacturer's instructions [19]. Output water from DUs continuously treated with disinfection products is more compliant with the recommended standards (heterotrophic bacteria load) and it is reported to be remarkable in preventing the contamination by *Legionella* and *P. aeruginosa* [8,36,37]. The efficacy of the adopted measures depended on the strict adherence to the planned protocols [36].

The level of *Legionella* contamination in DUs is not established in the Italian Legislation. Nevertheless, all control actions to reduce the risk of contamination, biofilm formation in DUWLs and a risk assessment based on patients and clinical practices are required [13].

Besides the technical-practical measures and disinfection protocols, an integrated approach for microbial risk managemen<sup>t</sup> in a dental health care setting should also include regular microbiological monitoring. Environmental surveillance for *Legionella* is useful not only to assess the efficacy of preventive measures but also as a guide for the choice of corrective strategies, under the principles of the internal control plan [38].

Our study confirms the literature evidences with the finding of *Legionella* or *P. aeruginosa* contamination in a large part of our samples before disinfection: 32% (6/19) and 68% (13/19) respectively. *P. aeruginosa* inside DUWLs may be related to the water quality or to the retro-contamination at the outlet of dental units. The competitive advantage of *P. aeruginosa* in the colonization of water lines is because of its ability to inhibit the growth of other bacteria by producing bacteriocins [39,40].

Frequent switchover from reverse osmotic to drinking water resulting from the intensive dental care activity in the hospital clinic, may have caused the higher *Legionella* spp. and *P. aeruginosa* isolation observed in the hospital DUWLs compared to private dental clinics. In large hospital buildings, drinking water quality may affect the DUWLs microbial contamination (more complex plumbing systems, large water storage tanks, multiple dead legs, cold water over 20 ◦C, etc.).

As reported by Lizzaro et al., the possibility of switching between two different water flows (municipal water and sterile water) reduces the risk of circuit contamination, but a mixed water supply is not recorded in water safety plans [31].

In the Hospital Dental Clinic (HDC), a Water Safety Plan (WSP), a maintenance and a control program were constantly applied to the building's water and aeration systems but there was a low adherence to good practices in DU management, despite the training activity of the dental healthcare

sta ff on the correct adoption of the hospital procedure (exclusive use of sterile reverse osmosis water, flushing between patients, self-contained water bottles disinfection, etc.). A low adherence to the best practices guidance may also contribute to biofilm proliferation. On the contrary, a good compliance with manufacturer's instructions for DU managemen<sup>t</sup> and the use of biocides was observed in PDCs, although the sta ff was not informed on either on water risk assessment and managemen<sup>t</sup> or on good practices guidance. The medical sta ff's and manufacturers' poor knowledge about water quality and the role of biofilm formation was suggested as one of the main problems related to dental unit contamination [31].

In only one Private Dental Clinic was shock disinfection with hydrogen peroxide 3% v/v and filters simultaneously installed at the inlet of each dental unit proven e ffective in eliminating pathogens, although the bacterial loads remained too high. These results were in line with the study conducted by Ditommaso et al., that showed an increase to unacceptable levels of bacterial loads in the DUWL of a dental chair treated with HP 3%, getting a significant reduction of counts only after 9 months of treatment [41].

On the other hand, in the hospital dental clinic, the first shock disinfection with HP 3% v/v applied in 12 dental units reduced positive samples for *Legionella* to 8% (1/12 DUs), and after the second treatment with HP 6% v/v and filters installation *Legionella* was removed. Conversely, after HP 6% v/v disinfection, *P. aeruginosa* proved to be more resistant and was detected again after two cycles of treatment, giving no evidence of substantially decreasing (100%, 10/10). Only after a third cycle of HP 4% v/v and surfactants were the samples proved to be negative.

DU disconnected from the water supply and fed only with sterile water, as with the DU in the operating room of the HDC, are less contaminated and safer to use in invasive dental practices.

In conclusion, we assert that filter installation and shock disinfection with a solution of 4% v/v hydrogen peroxide and surfactants appear to be a promising alternative for decreasing *Legionella* colonization in DUWLs, although further field studies in other healthcare and community settings are required to confirm its e ffectiveness and its long-term e fficacy in reducing biofilm.

#### **4. Materials and Methods**
