Summary
Three water-treatment protocols for dental unit waterlines were compared under routine clinical conditions at the University Dental Clinic Vienna. The one that won was then run across the whole dental unit network for three years, which is what makes this study unusual: most disinfection evidence stops at the trial.
Why dental waterlines are a hard case
Dental unit waterlines are narrow, carry a high surface area relative to their volume, and stand still for long periods — overnight, over weekends, and through clinic shutdowns. That combination favours biofilm, and biofilm shelters opportunistic waterborne pathogens. Thermal disinfection, the usual answer in hospital water systems, is not available here: tubing, seats and connectors are not heat-resistant. Chemical control is therefore the primary strategy, and it has to work without damaging a wide range of materials. A meta-analysis cited by the authors puts contamination prevalence in dental waterlines at roughly 12% for Legionella species and 8% for Pseudomonas species.
How it was done
| Setting | University Dental Clinic Vienna — 103 dental treatment units in daily use. Nine units, three per arm, entered the protocol comparison. |
|---|---|
| Equipment | Hypochlorous acid generated electrolytically on site from source water, salt and electricity, dosed continuously into the water supply. The generator was a third-party system, not EcoloxTech equipment. |
| Protocol A | Continuous HOCl dosing to a target free chlorine concentration of 0.3–0.6 mg/L |
| Protocol B | The same, at an increased HOCl concentration |
| Protocol C | HOCl supplemented with 1.41% hydrogen peroxide in the unit water tanks |
| Analysis | Membrane filtration and culture to ISO 11731 and 11731-2 for Legionella pneumophila, ISO 16266 for Pseudomonas aeruginosa, ISO 6222 for heterotrophic plate counts, ISO 19458 for sampling and ISO 7393-2 for free chlorine |
| Sampling | Two evaluation periods in 2022 chosen to stress the protocols — high summer, and a two-week clinical shutdown with the water standing. 14 sampling series, twice daily morning and afternoon, 28 sampling events, 250 mL from the second and fourth instrument lines and the air-water syringe. |
Result
- Protocol A was selected. It gave the best balance of microbiological efficacy, material compatibility and operational feasibility.
- More was not better. Raising the HOCl concentration produced no relevant additional microbiological benefit and brought technical impairments with it.
- Adding hydrogen peroxide made it worse. The combined protocol showed lower microbiological performance, and the cartridge system behind it also meant continuous replenishment, higher running cost and additional packaging waste.
- The three-year outcome. With Protocol A combined with automated flushing, daily free-chlorine monitoring and routine microbiological surveillance, suppression of L. pneumophila was sustained across the dental unit network from 2023 to 2025, with only isolated residual findings in a single unit.
- Heterotrophic plate count is not a sufficient indicator. Low HPC did not consistently exclude the presence of opportunistic pathogens. The authors conclude that targeted pathogen surveillance is needed alongside routine HPC monitoring.
What this establishes, and what it does not
- The negative finding may be the more useful one. A clean heterotrophic plate count is not evidence that Legionella or Pseudomonas are absent. Any water safety programme resting on HPC alone is measuring the wrong thing.
- Dose discipline, not dose maximisation. Two of the three arms were attempts to do more, and both were worse. That is a useful corrective to the assumption that a higher concentration is a safer margin.
- This is a very low concentration, continuously applied. 0.3–0.6 mg/L free chlorine in a closed, continuously dosed water system is a different problem from surface disinfection, and the figures do not transfer to one.
- It is a single-centre study under routine conditions, not a controlled laboratory trial, and the three-year follow-up describes a network running one standardised programme without a parallel untreated arm.
- Not an EcoloxTech study and not a product claim. Another manufacturer’s generator was used. EcoloxTech generators are regulated by the EPA as devices and carry no efficacy claims; HOCl-AS200 (EPA Reg. No. 101112-1) may be used only according to its accepted label.
What it means in practice
A low, continuously maintained free chlorine concentration combined with automated flushing held Legionella down for three years, while higher concentrations and a hydrogen peroxide addition both performed worse. The study also found that low heterotrophic plate counts did not reliably exclude opportunistic pathogens, so pathogen-specific surveillance is needed alongside routine HPC monitoring.
Conditions tested
| Parameter | Value |
|---|---|
| Concentration | 0.3 to 0.6 mg/L free chlorine, dosed continuously |
| Contact time | Continuous dosing; three-year follow-up, 2023 to 2025 |
| Measured result | Sustained suppression of Legionella pneumophila across 103 dental units, with isolated residual findings in a single unit |
| Organisms | Biofilms, Legionella pneumophila, Pseudomonas spp., Total microbial count |
| Relevant sectors | Dental, Healthcare, Public health, Water Treatment |
| Publication | Frontiers in Oral Health, 2026 |
| Authors | Alexandra Wolf, Tanya Ezat Abdullah Abdullah, Andreas Moritz, Apostolos Georgopoulos |
Reading this responsibly
Published figures are obtained under controlled conditions, usually on clean surfaces or in defined matrices. Organic load, surface condition, temperature and application coverage all affect real-world performance. Use these numbers to design a protocol, then validate it on your own site with environmental swabbing.
Applying it
Concentrations in this range are within the adjustable output of every EcoloxTech commercial system, so implementing this protocol is a settings question rather than a procurement one. Where the claim needs to sit on a registered label, see HOCl-AS200, EPA Reg. No. 101112-1.