Biweekly versus triweekly hydrogen peroxide shock treatment for controlling dental unit waterline contamination: a single-center pilot randomized clinical study

Introduction

During dental treatment, both patients and oral healthcare providers are exposed to microbial risks from dental unit waterlines (DUWLs) via contact with contaminated water and inhalation of aerosols and splatter (1, 2). DUWLs consist of narrow plastic tubing supplying water to high-speed handpieces and air–water syringes, whose small diameter and complex configuration favour nutrient deposition, microbial adhesion, and biofilm formation (3, 4). Although the American Centers for Disease Control and Prevention (CDC) recommends that dental treatment water should contain ≤ 500 colony-forming units per milliL (CFU/mL) (5), actual contamination levels frequently exceed this threshold (6–8), with detection of pathogens such as Pseudomonas aeruginosa and Legionella spp (9, 10)., thereby increasing the risk of healthcare-associated infections, particularly in immunocompromised patients (11, 12). Following the Coronavirus Disease 2019 (COVID-19) pandemic, concerns regarding airborne transmission in dental settings have further intensified (13). Reflecting this, the Chinese Stomatological Association (CSA) mandated a limit of ≤ 100 CFU/mL for dental treatment water in China (14).

Current DUWLs control strategies typically combine physical and chemical measures (15). Physical approaches (e.g., anti-retraction valves, routine flushing, and optimized tubing materials) can reduce microbial proliferation but cannot reliably remove established biofilms in narrow-bore lines (16–18). Consequently, chemical disinfectants such as sodium hypochlorite, hypochlorous acid, and hydrogen peroxide (H₂O₂) are widely used (19, 20). H₂O₂ offers broad-spectrum activity, rapid action, low toxicity, and good environmental compatibility, but its intrinsic instability limits long-term efficacy (21). Professional guidelines therefore advocate combining continuous low-level dosing with periodic high-concentration shock disinfection to achieve more sustained control of DUWLs contamination (22, 23). To improve the stability and performance of H₂O₂, compound formulations such as Dentosept® S have been developed. In vitro, continuous low-concentration Dentosept® S with the manufacturer-recommended four-week shock interval reduced biofilm bacterial counts ≤ 100 CFU/mL (3), and another study reported bacterial levels < 200 CFU/mL in over 91% of water samples (24). However, a clinical study in China using this standard regimen achieved only 89.9% compliance after 25 weeks, with bacterial counts regularly increasing during the third week of each cycle (25). These findings suggest that a four-week shock interval may be insufficient under domestic clinical conditions, and that the optimal shock frequency remains unclear.

To address this gap, our previous pilot study compared different shock intervals under continuous low-concentration H₂O₂ dosing (26). Weekly and biweekly H₂O₂ shock disinfection both achieved 100% water-quality compliance, whereas the triweekly regimen achieved only 68.42%. These preliminary findings suggested that shorter intervals may be more effective than the commonly used triweekly schedule. However, the pilot study had a limited sample size and short follow-up, which were insufficient to support an immediate change in routine practice. In addition, although weekly shock disinfection was effective, it is difficult to implement in busy dental clinics because of the added workload and cost. In real-world practice, the relevant unresolved question is therefore whether shortening the shock-treatment interval from three weeks to two weeks could improve water-quality compliance under routine clinical conditions. Accordingly, the present pilot study compared biweekly and triweekly H₂O₂ shock disinfection, combined with continuous low-concentration H₂O₂ dosing, to generate preliminary comparative clinical data on water-quality compliance in Chinese dental clinics.

Materials and methods

Dental chair unit selection

This study was conducted between May and August 2024 at a dental teaching hospital in China. The sample size was determined by the number of eligible DCUs available for the pilot study. Thus eight dental chair units (DCUs) (Intego Pro, Dentsply Sirona, USA) were selected as research subjects. Four units were from the Department of Prosthodontics and four were from the Department of General Dentistry. All had been in clinical use for 10 months to 4 years, had not undergone chemical disinfection previously, and had not had their DUWLs replaced. All DCUs were equipped with automatic flushing and disinfection systems, and filtered municipal water was used as the input source. To minimize confounding, the eight DCUs were allocated equally to Group A (n = 4) and Group B (n = 4) by department-stratified randomization. Within each departmental stratum, the four DCUs were allocated in a 1:1 ratio to Group A or Group B using a computer-generated random permutation containing two assignments to each group. Randomization was completed before implementation of the disinfection protocols. After the initial 6-week disinfection and evaluation phase, one unit from each group, designated Group A1 and Group B1, was followed for an additional 6 weeks of exploratory observation of total viable counts (TVCs), and these data were analyzed descriptively.

Baseline biofilm removal protocol

Prior to the implementation of the disinfection protocols, a standardized biofilm removal procedure was performed using the Biofilm-Removing-Set (BRS®, ALPRO MEDICAL GMBH, Germany) to restore the DUWLs to a near-original state per ISO 16954-2015 (27) standard test method. The BRS® includes the BRS® PreCleaner, BRS® Remover, and BRS® Activator. On a Saturday afternoon, 10 g BRS® PreCleaner was dissolved in 2 L of 50 ℃ water and flushed through the lines using a vibrating piston pump for a period not exceeding 15 min. This was followed by a solution of 160 g BRS® Remover and 20 mL BRS® Activator dissolved in 2 L of 50-60 ℃ water, which was also flushed through the lines. Subsequently, 1 L of the blue solution Bilpron® was introduced into the built-in water tanks and allowed to dwell in the DUWLs for 24 h. On the following Sunday afternoon, the lines were flushed with municipal water until no blue color remained in the water, followed by an additional 30-second flush. Water samples were collected immediately after flushing to assess the total bacterial count, and if the count exceeded 100 CFU/mL, the biofilm removal protocol was repeated.

Continuous disinfection protocol

During the study period, continuous low-concentration disinfection was performed using Dentosept® S (ALPRO MEDICAL GMBH, Germany), a H₂O₂-based disinfectant composed of 1.41% H₂O₂, 98.5% water, and less than 0.1% stabilizers. According to the manufacturer's instructions for continuous DUWLs disinfection, the DCU automatic disinfection system mixed municipal input water with the 1.41% H₂O₂ concentrate at a ratio of 100:1, yielding a constant treatment-water concentration of 0.0141% H₂O₂. Before study commencement, engineers tested the performance of all DCUs. This ensured that the automatic systems could accurately execute the commands for disinfectant dilution and flushing. In accordance with the guideline (14), the automatic system performed a 2-minute DUWL flush followed by a drying cycle at the start and end of each clinical session. In addition, to reduce the risk of cross-contamination, all high-speed handpieces were run idle for 30 s between patients.

Shock disinfection protocol

In addition to continuous disinfection, periodic shock treatments with undiluted Dentosept® S were carried out every 3 weeks in Group A and every 2 weeks in Group B. At the end of clinical sessions on Friday afternoon, the DCU automatic disinfection system filled the DUWLs with 1.41% H₂O₂ solution for a 24-hour dwell period. On Saturday afternoon, the system automatically flushed the lines with municipal water and completed a drying cycle to prepare the units for subsequent clinical use. To ensure water safety, an additional shock disinfection cycle was performed before reuse if a DCU had remained idle for more than 3 days.

Water sample collection

424 water samples were collected from the outlets of high-speed handpieces and air-water syringes across three study phases. Phase I established baseline levels and the efficacy of the BRS® biofilm removal procedure by sampling all eight DCUs for three consecutive days prior to the BRS® implementation and once immediately following the procedure. Phase II spanned a 6-week disinfection period, during which water samples were obtained from all eight units every Monday, Wednesday, and Friday morning before the start of clinical operations. Phase III involved an additional 6-week observation of two randomly selected units, representing one from Group A and one from Group B, designated Group A1 (n = 1) and Group B1 (n = 1), following the same sampling schedule.

For each sample, 10 mL of water was collected into a sterile tube after a 2-minute flush and surface disinfection of the outlet with a 75% ethanol swab. To ensure analytical accuracy, each sample was immediately mixed with 30% sodium thiosulfate solution at a ratio of 3:1 to neutralize residual H₂O₂.

Tubing-segment collection for qualitative SEM assessment

To provide qualitative morphological information on DUWL surface deposits, tubing segments were excised from high-speed handpiece waterlines. Because SEM sampling required destructive tubing excision and the air-water syringe tubing in the clinical DCU had limited spare length, SEM sampling was restricted to high-speed handpiece tubing to avoid affecting subsequent clinical use. One DCU was randomly selected from Group A and Group B, respectively, and 1-cm segments were harvested using sterile scissors at three predetermined time points: prior to BRS® implementation, immediately following the BRS® procedure, and upon completion of the 6-week disinfection period. All segments were immediately transported to the laboratory for analysis. Samples that could not be processed immediately were maintained at 4 °C and analyzed within 24 h to preserve surface-associated material and morphology.

Biofilm sample preparation and observation

Biofilm architecture on the internal surfaces of the high-speed handpiece tubing was examined using a JSM-5600 scanning electron microscope (SEM) (JEOL, Japan) as a qualitative assessment. To eliminate potential contamination, 0.1 cm was trimmed from each end of the 1-cm tubing segment with sterile scissors, and the segment was then slit longitudinally to expose the inner lumen. The specimens were fixed in 2.5% glutaraldehyde at room temperature for 1 h and dehydrated through a graded ethanol series (10%, 30%, 50%, 75%, 90%, and 100%) for 5 min at each concentration, with the 100% step repeated three times. After air-drying, the samples were mounted on SEM stubs with conductive adhesive and sputter-coated with gold to improve conductivity. The prepared inner surfaces were scanned and photographed to qualitatively document changes in surface-associated deposits and morphology.

Determination of TVCs

TVCs were used to quantify the total burden of culturable bacteria in output water but did not identify microbial taxa or non-culturable organisms. 1 mL of the neutralized water sample was spread evenly in a sterile petri dish and mixed thoroughly with 20 mL of nutrient agar medium (Aoboxing Biotech Co., Beijing, China). After solidification, the plates were inverted and incubated at 36 ± 1 ℃ for 48 h. Colonies were then counted and recorded. According to the CSA guideline, dental treatment water was considered compliant when the TVC did not exceed 100 CFU/mL (14). The compliance rate was calculated as the number of compliant samples divided by the total number of samples collected in each group during a specific study phase.

Blinding of sample collection and outcome assessment

Only the personnel responsible for implementing the scheduled shock disinfection were aware of group allocation. Sample collectors were blinded to the assigned disinfection interval, and all water and tubing samples were labeled with coded identifiers that did not disclose treatment allocation. Two laboratory assessors, both blinded to group allocation, performed and cross-checked TVC colony counting and qualitative SEM assessment using the coded identifiers. Any discrepancies were reviewed and resolved by consensus before the results were recorded.

Statistical analysis

SPSS version 26.0 (Chicago, IL, USA) was used for descriptive and inferential analyses. Normally distributed variables are presented as mean ± standard deviation (SD), and between-group comparisons were performed using the independent-samples t test. Non-normally distributed data are presented as median (Q1, Q3), and between-group comparisons were performed using the Mann–Whitney U test. Categorical variables are presented as frequencies and percentages. Water-quality compliance was analyzed as the primary binary outcome. Generalized estimating equations (GEE) were used to compare compliance between groups while accounting for repeated measurements within DCUs. DCU was specified as the clustering unit, with an exchangeable working correlation structure and robust standard errors. Group was included as the main explanatory variable, and outlet site and sampling day were included as covariates. A two-sided P value < 0.05 was considered statistically significant.

Results

Baseline status and BRS® efficacy

Both Group A (n = 4) and Group B (n = 4) operated 5.5 working days weekly, with respective average weekly patient volumes of 226 and 223. Prior to the intervention, all DCUs demonstrated heavy microbial contamination. Before the BRS® procedure, TVCs at high-speed handpiece and air-water syringe outlets showed no statistically significant difference between groups (high-speed handpiece: Z = −0.289, P = 0.773; air-water syringe: Z = −0.866, P = 0.149).

After the BRS® procedure, the microbial load in all DUWLs decreased to compliant levels. No significant between-group differences were observed at either the high-speed handpiece or air-water syringe outlets (high-speed handpieces: Z = −1.461, P = 0.144; air-water syringes: Z = −0.289, P = 0.773). The highest bacterial count recorded across all units was 58 CFU/mL, indicating that the BRS® procedure effectively restored the waterlines to a near-original clean state before the disinfection phase (Table 1).

Table 1. Total viable counts in output water from DCUs [median (Q1, Q3)].

CFU/mL, colony-forming units per milliliter; HSHP, high-speed handpieces; AWS, air-water syringe; Group A, underwent shock treatment every 3 weeks; Group B, underwent shock treatment every 2 weeks.

Bold values indicate statistically significant between-group differences (P < 0.05).

6-week disinfection efficacy

During the 6-week disinfection period, each group contributed 144 water samples, collected from four DCUs at two outlet sites over 18 sampling time points. Group B (biweekly shock treatment) achieved a 100% compliance rate (144/144), with TVCs in all high-speed handpieces and air-water syringe samples consistently remaining below 100 CFU/mL (Figure 1). In contrast, Group A (triweekly shock treatment) exhibited a lower overall compliance rate of 78.47% (113/144). Because repeated samples from the same DCUs were not independent, the primary compliance outcome was analyzed using a GEE-based linear probability model, with DCU specified as the clustering unit, an exchangeable working correlation structure, and robust standard errors. After adjustment for outlet site and sampling day, the difference in compliance between the two shock treatment intervals remained significant (Wald χ² = 29.34, P < 0.001). Compared with triweekly shock treatment, biweekly shock treatment increased the probability of compliance by 21.53 percentage points (95% CI: 13.74–29.32, P < 0.001). Notably, TVCs in Group A remained compliant during the first two weeks following a shock treatment but increased sharply during the third week of each cycle (Figure 1). Significant between-group differences in TVCs emerged as early as Day 17 (high-speed handpieces: Z = −2.021, P = 0.043; air-water syringe: Z = −2.309, P = 0.021). Specifically, air-water syringe samples showed significant differences at multiple intervals (P < 0.05), including Days 8, 15, 17, 22, 33, 36, 38, and 40. For high-speed handpieces samples, significant divergence was observed on Day 17, 24, 29, 33, 36, 38, and 40 (Table 1).

Descriptive 12-week observations

The additional 6-week follow-up involved only one DCU from each group and was therefore analyzed descriptively. Over the 12-week period, Group B1 consistently maintained TVCs below 100 CFU/mL in all sampled water from both high-speed handpieces and air-water syringes (Figure 2). Notably, a gradual upward trend in TVC was observed during the final sampling period (Day 75–82) although all values remained compliant. In contrast, beginning with the second disinfection cycle (defined as one 3-week interval), Group A1 showed a recurrent increase in microbial counts during the third week of each cycle, with TVCs exceeding 100 CFU/mL. Specifically, TVCs surpassed the compliant limit between Days 29–40 during the second cycle, Days 59–61 during the third cycle and again around Days 80–82 during the fourth cycle. Interestingly, the maximum observed TVC values appeared lower in later cycles than in earlier cycles.

Qualitative SEM observations of high-speed handpiece tubing surfaces

Figure 3 shows SEM images of high-speed handpiece waterlines from randomly selected DCUs in Groups A and B at three time points: before BRS® treatment (Figures 3a,3d), immediately after the BRS® procedure (Figures 3b,3e), and after 6 weeks of disinfection (Figures 3c,3f). All images were captured at 5000x magnification in secondary electron mode with an acceleration voltage of 5.0 kV.

Before any intervention, the inner tubing surfaces showed extensive surface-associated material. Dense aggregates of rod-shaped structures were tightly attached to the tubing surface (Figure 3a), and some structures appeared embedded within continuous matrix-like material (Figure 3d). These features were morphologically compatible with biofilm-like accumulation; however, their biological identity could not be confirmed by SEM alone. After the BRS® biofilm-removal procedure, the sampled tubing segments showed marked reduction in surface-associated material and cleaner inner surface as observed in SEM images (Figures 3b,3e).

After 6 weeks of disinfection, the sampled segment from Group A showed renewed accumulation of dense rod-shaped and matrix-like structures (Figure 3c). In contrast, no obvious dense biofilm-like structure was observed in the examined field from Group B, although scattered irregular particles were present (Figure 3f). Because standard SEM morphology alone cannot determine whether these particles were biological or non-biological, their biological nature could not be confirmed. These qualitative SEM findings were consistent with the differences in water-quality compliance observed between the two groups, but were not used to determine the biological nature of the observed particles.

Discussion

This single-center pilot randomized clinical study showed that, after baseline biofilm removal with BRS®, a biweekly 1.41% H₂O₂ shock-disinfection schedule combined with continuous 0.0141% H₂O₂ treatment maintained 100% compliance with the CSA standard for DUWL water quality during the 6-week. By contrast, the triweekly schedule was associated with a lower compliance rate and recurrent TVC rebound during the third week of each cycle. These findings suggest that, after initial BRS® biofilm removal, shortening the maintenance shock-treatment interval from three weeks to two weeks may improve the consistency of DUWL water-quality compliance under the conditions of this single-center pilot study.

Our findings differ from those of Zemouri et al. (3), who suggested that a 4-week shock-disinfection interval might be adequate. Several factors may explain this discrepancy. First, our study was conducted in a real-world clinical setting rather than under in vitro conditions. As such, it captured practical factors that may promote recolonization, including handpiece backflow, intermittent water stagnation during shutdown periods, and variation in clinical workload. Second, we used the stricter CSA microbiological threshold of ≤ 100 CFU/mL rather than the CDC benchmark of < 500 CFU/mL. Third, the Dentosept® S formulation used in this study is silver-ion-free, which may have reduced its residual antimicrobial effect and thereby increased the need for more frequent shock treatment (28, 29). Taken together, these factors may explain why the manufacturer-recommended interval appeared insufficient in our clinical context.

The present study extends our preliminary pilot work (26) by using a randomized design, including a larger number of dental chair units, and involving both prosthodontic and general dentistry departments. These design features strengthen the clinical relevance of the findings. In particular, evidence from sequencing-based studies suggests that microbial communities in DUWLs may vary across dental specialties (30), which may in turn influence the effectiveness of decontamination strategies. The consistency between the pilot study and the present trial therefore provides more preliminary support for further evaluation of a biweekly shock-disinfection interval in larger clinical studies.

An important finding of this study was the recurrent third-week increase in TVCs observed in the triweekly group during the 6-week. This pattern was consistent with the report by Kramer et al. (31) and suggests that, under the conditions of this pilot study, the third week after shock treatment may represent a vulnerable period for loss of TVC-based water-quality compliance. In parallel, qualitative SEM observations showed renewed accumulation of dense rod-shaped and matrix-like structures in the sampled segment from the selected triweekly DCU (Group A) after 6 weeks, whereas no obvious mature biofilm-like structure was observed in the examined field from the selected biweekly DCU (Group B). These observations were consistent with the recurrent TVC rebound in Group A. However, because SEM cannot distinguish biological particles from non-biological deposits, these images should be interpreted as qualitative morphological evidence rather than direct proof of biofilm regeneration. The descriptive 12-week extension also showed recurrent third-week non-compliance in Group A1, but this observation was based on only one DCU per group and should be regarded as hypothesis-generating. The apparent decline in rebound peaks over successive cycles in Group A1 may reflect unit-specific variation, stochastic fluctuation, or changes in residual culturable bacterial burden under repeated shock treatment, and requires confirmation in larger longitudinal studies. From the perspective of hospital infection prevention and control, although the biweekly regimen was associated with lower TVCs and more consistent compliance with the CSA threshold, these findings cannot be interpreted as evidence that opportunistic pathogens were absent or reduced. The clinical significance of the observed differences in TVCs therefore requires further validation using pathogen-specific testing and molecular microbiological analyses.

Another noteworthy aspect of the study is the use of BRS® pretreatment to standardize the baseline condition of the DUWLs. Previous studies have shown that contamination in DUWLs may develop rapidly after commissioning and that once established, biofilms can be difficult to eradicate completely (32–35). Without baseline standardization, residual biofilm burden could confound comparisons between maintenance protocols. In our study, qualitative SEM observations showed a marked reduction in surface-associated biofilm-like material after BRS® pretreatment, consistent with previous reports (36). This methodological step improves confidence that the observed between-group differences were less likely to be explained by unequal initial contamination levels. However, it also limits the scope of interpretation. Our findings apply primarily to DUWLs after initial biofilm removal and baseline standardization. They should not be generalized directly to DUWLs that have not undergone baseline biofilm removal or to units with persistent established biofilm.

These pilot findings provide preliminary clinical evidence to guide the design of larger studies evaluating optimized DUWL shock treatment intervals, particularly when stringent water-quality standards are applied. After initial BRS® biofilm removal, a biweekly shock-disinfection schedule combined with continuous low-level H₂O₂ treatment was associated with more consistent compliance in busy dental settings. Nevertheless, the biweekly regimen should be considered a candidate strategy requiring further validation rather than a definitive replacement for existing protocols. Future implementation studies should evaluate workload, disinfectant consumption, feasibility, material compatibility, and long-term sustainability in routine practice.

Several limitations should be acknowledged. First, this single-center pilot study involved a limited number of DCUs, which restricts the statistical power and generalizability of the findings. Data were collected in Beijing from May to August 2024, and DUWL contamination and H₂O₂ performance may vary with source water quality, water hardness, residual municipal disinfectants, ambient temperature, DUWL design, patient volume, and local maintenance protocols. Second, the extended 12-week observation included only one randomly selected DCU per group and was therefore descriptive rather than representative of group-level long-term performance. Third, the study relied on TVCs and qualitative SEM observations. TVCs quantified only culturable bacterial burden and could not identify opportunistic pathogens, detect viable-but-nonculturable organisms, or characterize microbial community changes. SEM provided only qualitative morphological information from limited high-speed handpiece tubing segments and could not quantify biofilm biomass, determine biofilm dynamics, or distinguish biological from non-biological particles. Future studies should include larger multicenter sampling, longer follow-up, paired biofilm assessment of both outlet types, microbial identification, pathogen-targeted qPCR, 16S rRNA sequencing, quantitative biofilm assays, and implementation evaluations of cost, workload, material compatibility, and long-term sustainability.

Conclusions

Within the limitations of this single-center pilot study, continuous disinfection with 0.0141% H₂O₂ combined with biweekly 1.41% H₂O₂ shock treatment was associated with more consistent compliance with the CSA TVC threshold than triweekly shock treatment after initial BRS® biofilm removal. These preliminary findings suggest that the biweekly regimen is a candidate interval for further evaluation, but they are not sufficient to support a definitive change in routine clinical protocols. Larger multicenter, year-round studies incorporating longer follow-up, microbial identification, microbial community analysis, and quantitative biofilm assessment are needed.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Anas Ahmed Shamala, University of Science & Technology, Yemen

Reviewed by: Ravinder Saini, King Khalid University, Saudi Arabia

Patrick Ampofo, University of Ghana, Ghana