Evaluation of Legionella pneumophila Decrease in Hot Water Network of Four Hospital Buildings after Installation of Electron Time Flow Taps

Round 1

Reviewer 1 Report

I believe that the manuscript describes a quite novel strategy for Legionella decontamination, which is very important concerning public health.

Besides that, I also believe that it needs some improvement, in order to be published.

Below are some comments and changes needed to be done:

Line 45, 216: “ancient hospital water networks and high volume hot water storage tanks provide optimal conditions”, change the word ancient.

 How many days for decontamination before the TFT installation? Which decontamination method was used?

You refer to the Water Safety Plan, but you did not give any details about it. For example, which are the disinfection methods used, or how you choose  the sites where TFTs where placed, the decontamination procedure e.t.c. Please explain.

I noticed that in 3 out of the 4 hospitals four TFTs were installed, besides the different number of levels and beds. Was it a coincidence? Please explain.

Line 112: “Legionella spp. Research”. You should use another word.

Remove all the correction trucks throughout the text.

Which method was used for chlorination? Was the same before and after the implementation of TFTs?

Was the implementation of 7 TFTs for 5 blocks enough, in order to reduce the Legionella spp. Colonization in hospital 3?        

Why do you believe that TFTs and continuous chlorination were not able to the reduce of  legionella contamination in hospital 3? You refer that maybe was the low consumption of hot water, but the same flow rate was achieved (192L/day) was applied. So there must be another reason Please explain.

The headline for figure 3 must change. There are 7 TFTs, not 4, as it is mentioned.

Put a citation for statistical analysis.

Author Response

Line 45, 216: “ancient hospital water networks and high volume hot water storage tanks provide optimal conditions”, change the word ancient.

The word “ancient” has been removed.

 How many days for decontamination before the TFT installation? Which decontamination method was used?

In the text we added the following section.

“Continuous disinfection with chlorine dioxide was applied in water networks. This method was used one day before TFTs installation. In hospital water stations a chemical feed pump, a chlorine solution tank, a chlorine residual monitoring kit were installed.   ”.

This method has been applied in all the hospitals.

You refer to the Water Safety Plan, but you did not give any details about it. For example, which are the disinfection methods used, or how you choose  the sites where TFTs where placed, the decontamination procedure e.t.c. Please explain.

Some details about the WSP have been added in the text.

“The hospital WSP is a self-control manual aimed to manage the microbiological water risk from waterborne pathogens, as Legionella spp. This protocol is a technical guide, which describes the control critical points of the water systems (dead end branches, pipeworks materials; devices, ect); the water samplings for microbiological tests; the disinfection procedure of water networks and all responsibilities related to the plants management”.

Disinfection method has been explained in the same section.

I noticed that in 3 out of the 4 hospitals four TFTs were installed, besides the different number of levels and beds. Was it a coincidence? Please explain.

It is a coincidence. Despite the different number of levels or beds, the number of dead ends detected hot water networks were the same. This sentence has been written in Discussion section.

Line 112: “Legionella spp. Research”. You should use another word.

The word “research” has been changed word “detection”

Remove all the correction trucks throughout the text.

All correction have been removed

Which method was used for chlorination? Was the same before and after the implementation of TFTs?

The chlorination system was not present Hospital 1,2,4 before TFTs installation. As added in the text, a continuous disinfection with chlorine dioxide was applied in water networks. This method was used one day before TFTs installation. In hospital water stations a chemical feed pump, a chlorine solution tank, a chlorine residual monitoring kit were installed.

In Hospital 3, a continuous chlorination system was present before TFTs installation.

After TFTs installation the continuous disinfection system was the same.

 

Was the implementation of 7 TFTs for 5 blocks enough, in order to reduce the Legionella spp. Colonization in hospital 3?

            During the study of the water network plant we detected seven dead end branches. Considering the good results achieved in this study, it was not necessary to increase the number of TFTs in hospital 3.

Why do you believe that TFTs and continuous chlorination were not able to the reduce of  legionella contamination in hospital 3? You refer that maybe was the low consumption of hot water, but the same flow rate was achieved (192L/day) was applied. So there must be another reason Please explain.

As described in the text, only in Hospital 3 the continuous chlorination system was already present in the hot water network. (statement has been re-written in the text)

Despite this treatment Legionella was however present the water samples.

The lack of Legionella colonization was achieved only after TFTs installation.

In hospital 3 the only chlorination system was not enough to achieve Legionella counts decrease, but the synergy of this control measure with TFTs installation eases a better chlorination circulation and the absence of the bacteria in hot water.

The headline for figure 3 must change. There are 7 TFTs, not 4, as it is mentioned.

Headline has been changed.

Put a citation for statistical analysis.

A citation has been added.

Author Response File: Author Response.docx

Reviewer 2 Report

The manuscript presents the result of a follow up study in four hospitals, where electronic time flow taps were installed and disinfectant was added. This is important work and results are interesting. However, the discussion is not sufficient and it does not take into account other important factors such as the increase in temperature observed after the installation of the TFTs, or the position of the non-TFT sampling points relative to the newly installed TFTs.

General comments

The structure is repetitive, enumerating the 4 hospitals one after the other. Some of the results should be presented combined for all hospitals.

In the material and methods, many details are provided regarding the location of the different units in the hospital; however it is not used further in the manuscript. For example, there is no information on the location of the sampling points and no discussion on the impact of the medical units in the water circulation and requirement for TFT faucets.

There are several revisions required regarding choice of words or typos, language revision would be required.

Legionella should be in italic everywhere in the manuscript

Specific comments

Abstract

L19 – …was detected or was investigated? Searched does not seem like the right word.

L21 – Decontamination implies complete eradication. Loss of culturability would be more specific.

Materials and methods

L68-70 - Too detailed, especially since the location of TFTs is not specified and the information is not used in the discussion.

L76-77 – “After the identification of all the dead-end branches present in the hot water distribution system, from September 2017 four TFTs were installed in correspondence with each one.” Do you mean each one of the dead-end identified? There were only four dead-end branches within the entire hospital building?

L109 – Detection instead of research

L129 – Do you mean incubators?

A general sampling methodology section applying to all hospitals would help to avoid repetition. This section should include how the samples were collected (first flush?), if there was prior controlled stagnation, if sodium thiosulfate was used, when was the temperature and disinfectant concentration measured, …

Two point of use without TFTs – where were they located in comparison to TFTs? Are they on the same pipeloop as a TFT? A schematic showing the location of the non-TFT vs TFT would be helpful to understand why the temperature increase is also observed on the non-TFT.

Results

L150 – When was the temperature measured? On first flush, after 1 minute or after reaching maximum temperature? This should be included in the materials and methods section.

L154 – all the samples were still positive…

Table 1 (and tables for other hospitals) – what is the number of replicates before and after installation of TFTs?

Figure 1 – Since each individual results are not discussed, it would be helpful to show the Legionella pneumophila results as boxplot over time, with each hospital shown as a different box. This would allow to combine Figure 1 to 4 in one graph, and see the overall trend.

In general, the presentation of the results for the four hospitals successively generates a lot of repetition, with the exact same sentences from one section to the next. The use of a general section would help streamline the presentation.

It would be interesting to show a figure with the mean temperature before and after for all 4 hospitals in a boxplot format. There seems to be at least 10°C increase after the implementation of TFTs. Was the temperature at the boiler changed?

There has been a recent publications showing the importance of the hydraulics in controlling Legionella pneumophila (Galvada et al. 2019). Could improved circulation be achieved without TFTs? Could the same effect be achieved without chloride dioxide?

Discussion

L272 – “A reduction of stagnant water reduces biofilm growth …” This could be discussed further. Manuel et al. 2010 showed the impact of unsteady flow and stagnation on the drinking water quality. Other studies also showed the release of bacteria from the biofilm into the water after only 1-hour stagnation (Bédard et al. 2018, Fish et al. 2017).

There should be more explanation/discussion on the decrease observed at the boiler and in the recirculation. What role did the TFTs played compared to the disinfectant or the temperature increase?

 

 

References:

Gavaldà, L., Garcia-Nuñez, M., Quero, S., Gutierrez-Milla, C. and Sabrià, M. (2019) Role of hot water temperature and water system use on Legionella control in a tertiary hospital: An 8-year longitudinal study. Water Research 149, 460-466.

Bédard, E., Laferrière, C., Déziel, E. and Prévost, M. (2018) Impact of stagnation and sampling volume on water microbial quality monitoring in large buildings. PlosOne 13(6), e0199429.

Fish, K., Osborn, A.M. and Boxall, J.B. (2017) Biofilm structures (EPS and bacterial communities) in drinking water distribution systems are conditioned by hydraulics and influence discolouration. Science of the Total Environment 593-594, 571-580.

Author Response

The structure is repetitive, enumerating the 4 hospitals one after the other. Some of the results should be presented combined for all hospitals.

Figure 1 – Since each individual results are not discussed, it would be helpful to show the Legionella pneumophila results as boxplot over time, with each hospital shown as a different box. This would allow to combine Figure 1 to 4 in one graph, and see the overall trend.

According to you, method and discussion sections have been re-written. A further figure (Figure 5), highlighting Legionella spp. presence, chlorine concentrations and temperature values in all the hospitals, has been included in Discussion section.

In the material and methods, many details are provided regarding the location of the different units in the hospital; however it is not used further in the manuscript. For example, there is no information on the location of the sampling points and no discussion on the impact of the medical units in the water circulation and requirement for TFT faucets.

In Material and Methods section we added the following details.

“TFTs were installed close to dead end branches of the water networks (storage rooms, kitchens and lockers rooms), while P1 and P2 were chosen from medical stays or surgery areas”.

In Discussion section we added the following sentences.

“The choice of TFTs installation points depends on the kind of subjects (patients, healthcare workers, ect) present in those areas. We avoided TFTs installation in stays areas, surgeries and medical offices, in order to limit the exposition to contaminated aerosol during TFTs activities. These devices were mostly placed in storage rooms, kitchens and lockers rooms”.

There are several revisions required regarding choice of words or typos, language revision would be required.

Legionella should be in italic everywhere in the manuscript

Corrections have been applied

Specific comments

Abstract

L19 – …was detected or was investigated? Searched does not seem like the right word.

Correction has been applied

L21 – Decontamination implies complete eradication. Loss of culturability would be more specific.

Correction has been applied

 

Materials and methods

L68-70 - Too detailed, especially since the location of TFTs is not specified and the information is not used in the discussion.

Details have been removed

L76-77 – “After the identification of all the dead-end branches present in the hot water distribution system, from September 2017 four TFTs were installed in correspondence with each one.” Do you mean each one of the dead-end identified? There were only four dead-end branches within the entire hospital building?

No, but we chose the most critical points. Sentence has been edited.

L109 – Detection instead of research

Correction has been applied

L129 – Do you mean incubators?

Yes, sentence has been edited.

A general sampling methodology section applying to all hospitals would help to avoid repetition.

This section has been re-written.

This section should include how the samples were collected (first flush?), if there was prior controlled stagnation, if sodium thiosulfate was used, when was the temperature and disinfectant concentration measured, …

Sampling details have been added in the text.

Two point of use without TFTs – where were they located in comparison to TFTs? Are they on the same pipeloop as a TFT? A schematic showing the location of the non-TFT vs TFT would be helpful to understand why the temperature increase is also observed on the non-TFT.

In Material and Methods section we added the following details.

“TFTs were installed close to dead end branches of the water networks (storage rooms, kitchens and lockers rooms), while P1 and P2 were chosen from medical stays or surgery areas”.

This choice is due to the lack of Legionella risk evaluation in high risk settings.

In fact, in Discussion section we added the following sentences.

“The choice of TFTs installation points depends on the kind of subjects (patients, healthcare workers, ect) present in those areas. In fact, we avoided TFTs installation in stays areas, surgeries and medical offices, in order to limit the exposition to contaminated aerosol during TFTs activities. These devices were mostly placed in storage rooms, kitchens and lockers rooms.

TFTs and non-TFTs are located on the same pipeloop. Considering the maintenance works on the hot water networks performed during TFTs installation, we achieved a temperature increase in all the hot water networks.

Results

L150 – When was the temperature measured? On first flush, after 1 minute or after reaching maximum temperature? This should be included in the materials and methods section.

Sampling details have been added in the text.

L154 – all the samples were still positive…

Sentence has been edited

Table 1 (and tables for other hospitals) – what is the number of replicates before and after installation of TFTs?

As added in Materials and Methods section “All physical-chemical parameter were measured in triplicate before and after TFTs installation”.

In general, the presentation of the results for the four hospitals successively generates a lot of repetition, with the exact same sentences from one section to the next. The use of a general section would help streamline the presentation.

Results section has been re-written in order to avoid the repetitions.

It would be interesting to show a figure with the mean temperature before and after for all 4 hospitals in a boxplot format. There seems to be at least 10°C increase after the implementation of TFTs. Was the temperature at the boiler changed?

In Discussion section we added the following sentences.

“in all hospitals the increase of temperature values was obtained from T0 to T1. This is due to the maintenance works on the hot water networks performed during TFTs installation. Considering the low hot water temperatures detected at T0 (≤40°C), we increased this value, in order to achieve boiler water temperature at least of 50°C”.  

In Figure 5 (C) we showed, all 4 hospitals, the mean temperature values detected before and after TFTs installation.

 

There has been a recent publications showing the importance of the hydraulics in controlling Legionella pneumophila (Galvada et al. 2019). Could improved circulation be achieved without TFTs? Could the same effect be achieved without chloride dioxide?

TFTs and the chemical disinfection represent a synergistic method aimed to reduce Legionella colonization from pipeworks, mostly in old and large water networks with dead end branches. In fact, in this kind of plants, the only chemical disinfection is often insufficient. The low water consumption allow water stagnation and biofilm growth in critical point of the pipeworks. On the other hand the application of TFTs without the chemical chlorination does not eliminate bacterial colonization because it is not a biocidal treatment. According to Galvada study, the assessment of hot water temperature plays an important role for Legionella spp. control in water networks. In fact, in our study we performed technical hydraulic works aimed to increase the hot networks temperature.

This statement has been added in Discussion section.

Discussion

L272 – “A reduction of stagnant water reduces biofilm growth …” This could be discussed further. Manuel et al. 2010 showed the impact of unsteady flow and stagnation on the drinking water quality. Other studies also showed the release of bacteria from the biofilm into the water after only 1-hour stagnation (Bédard et al. 2018, Fish et al. 2017).

Sentence has been argued in Discussion section following your suggested references.

There should be more explanation/discussion on the decrease observed at the boiler and in the recirculation. What role did the TFTs played compared to the disinfectant or the temperature increase?

 Temperature increase was due to technical maintenance works on hot water networks, while the presence of chlorine at point of use was also due to the increase of water flushing in pipeworks (TFTs).

 

In Discussion section we described:

“Water flushing eases the chlorine dioxide efficacy in water networks…”

and

“in all hospitals the increase of temperature values was obtained from T0 to T1. This is due to the maintenance works on the hot water networks performed during TFTs installation. Considering the low hot water temperatures detected at T0 (≤40°C), we increased this value, in order to achieve boiler water temperature at least of 50°C”.  

 

 

References:

Gavaldà, L., Garcia-Nuñez, M., Quero, S., Gutierrez-Milla, C. and Sabrià, M. (2019) Role of hot water temperature and water system use on Legionella control in a tertiary hospital: An 8-year longitudinal study. Water Research 149, 460-466.

Bédard, E., Laferrière, C., Déziel, E. and Prévost, M. (2018) Impact of stagnation and sampling volume on water microbial quality monitoring in large buildings. PlosOne 13(6), e0199429.

Fish, K., Osborn, A.M. and Boxall, J.B. (2017) Biofilm structures (EPS and bacterial communities) in drinking water distribution systems are conditioned by hydraulics and influence discolouration. Science of the Total Environment 593-594, 571-580.

References have been added in Discussion section.

Author Response File: Author Response.docx

Reviewer 3 Report

The problem of contamination of hot water distribution system with Legionella bacteria is current. The purpose of the research was to analyze this type of systems in hospital buildings, in that were applied a continuous disinfection treatment and the installation of time flow taps (TFTs). The method used to combat Legionella has been shown to be highly effective.

Comments:

In the summary, please complete the type of disinfectant/chlorination system used (line 18). Line 22 - Chlorine concentration was detected in the range between 0.23 and 0.36 mg/l - does it mean mg Cl2/L or mg ClO2/L? Line 104 – “while” - ? Line 113 and 121 – chlorine concentration - does it mean the content of free, total chlorine or chlorine dioxide? Line 46, 47, 54, 127, 129, 156, 168, 185, 204, 255…. Legionella - should be in italics, like all Latin names in scientific texts (Legionella spp.; Legionella…). In Table 1, 2, 3 and 4 complete the results (BEFORE TFTs INSTALLATION). No SD value. The authors state the concentration of total chlorine. The content of total chlorine consists of free chlorine (providing protection against secondary contamination) and combined chlorine (which as a combination of nitrogen and chlorine forms can be an additional impurity and has a low disinfecting effect). Why did the authors measure total chlorine? After all, chlorine dioxide was dosed into the system. Line 171, 196, 224 – of? Line 207 – the ? On the figures 1-4 instead of T0, T1 ... I suggest to give specific times (24 hours, 15 days ...). This will allow for easier inference and the impact of TFT application time on CFUs. Explanation of the method of chemical water disinfection (chlorination) in the analyzed installations is necessary for proper interpretation of the results presented in Tables 1-4. Reading the text shows that disinfection was carried out with chlorine dioxide. Was the chlorine dioxide concentration converted to total chlorine? The entries in Tables 1-4 should be clarified and supplemented. Line 77-79: Why was it programmed to obtain hot water flow of just 192 L/day from each TFT in Hospitals. How was this flow determined (measured)? Regardless of the size and range of the installation (in Hospital 1,2,3,4), has a flow rate of 192 L/ day been established?

Author Response

In the summary, please complete the type of disinfectant/chlorination system used (line 18).

Disinfectant has been added

Line 22 - Chlorine concentration was detected in the range between 0.23 and 0.36 mg/l - does it mean mg Cl2/L or mg ClO2/L?

Cl2. It has been added.

Line 104 – “while” - ?

It has been removed

Line 113 and 121 – chlorine concentration - does it mean the content of free, total chlorine or chlorine dioxide?

Total chlorine concentration. It has been added in the text

Line 46, 47, 54, 127, 129, 156, 168, 185, 204, 255…. Legionella - should be in italics, like all Latin names in scientific texts (Legionella spp.; Legionella…).

Legionella has been written in italics.

In Table 1, 2, 3 and 4 complete the results (BEFORE TFTs INSTALLATION). No SD value. The entries in Tables 1-4 should be clarified and supplemented.

As suggested by other reviewers, this part of result has been edited and some specific details have been included in Tables.

Tables have been completed with SD values (before TFTs installation)

The authors state the concentration of total chlorine. The content of total chlorine consists of free chlorine (providing protection against secondary contamination) and combined chlorine (which as a combination of nitrogen and chlorine forms can be an additional impurity and has a low disinfecting effect). Why did the authors measure total chlorine? After all, chlorine dioxide was dosed into the system.

We detected all type of chlorine (free, combined, total). We chose to express data related to total chlorine because they were very similar to free chlorine values.

Line 171, 196, 224 – of? Line 207 – the ?

Corrections have been done.

On the figures 1-4 instead of T0, T1 ... I suggest to give specific times (24 hours, 15 days ...). This will allow for easier inference and the impact of TFT application time on CFUs.

Specific times have been included

Explanation of the method of chemical water disinfection (chlorination) in the analyzed installations is necessary for proper interpretation of the results presented in Tables 1-4.

Some details about chemical disinfection have been added in Material and Methods section.

Reading the text shows that disinfection was carried out with chlorine dioxide. Was the chlorine dioxide concentration converted to total chlorine?

Yes, all values regarded the total chlorine.

Line 77-79: Why was it programmed to obtain hot water flow of just 192 L/day from each TFT in Hospitals. How was this flow determined (measured)?

In the text we added this sentence.

“They present an adjustable flowmeter in order to measure the scheduled water flow”.

Regardless of the size and range of the installation (in Hospital 1,2,3,4), has a flow rate of 192 L/ day been established?

According to the hospital facility management office, we chose a standard water flow rate ensuring an economy saving. Considering our data, this flow rate has not modified.

 

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

Thank you for the thorouh revision work that was done. The manuscript is greatly improved.

One minor comment: L 328, at the end of the sentence, I would rephrase to  add disinfectant levels throughout the system. "... a good cost-effective method which helps maintain disinfectant levels throughout the system."

 

Author Response

One minor comment: L 328, at the end of the sentence, I would rephrase to  add disinfectant levels throughout the system. "... a good cost-effective method which helps maintain disinfectant levels throughout the system."

Sentence has been rephrased.

Author Response File: Author Response.docx

Reviewer 3 Report

I accept the corrections and additions.

One more note:

The entry in table 1-4 regarding pH should be corrected. It should be pH (mean ±SD).

Author Response

One more note:

The entry in table 1-4 regarding pH should be corrected. It should be pH (mean ±SD).

Corrections have been applied

Author Response File: Author Response.docx