Advertisement

Failure of a hollow-fibre shower filter device to prevent exposure of patients to Pseudomonas aeruginosa

Open AccessPublished:August 29, 2022DOI:https://doi.org/10.1016/j.jhin.2022.08.007

      Summary

      Background

      Pseudomonas aeruginosa in hospital water is a risk for invasive infection. Point-of-use (POU) filters are used to reduce patient exposure to the organism, and hollow-fibre filters are becoming more popular. However, retrograde colonization of the filter mechanism may contaminate the effluent.

      Aims

      To assess the efficacy of POU filter head (polysulfone; hollow-fibre matrix) shower filters in preventing the exposure of high-risk patient groups to P. aeruginosa.

      Methods

      Pre-flush (opening the outlet and collecting the first 100 mL of water) samples were analysed to measure P. aeruginosa contamination from 25 shower outlets (∼21% of all showers on the six wards), with and without a hollow-fibre filter. P. aeruginosa was measured in a subset of outlets harbouring P. aeruginosa (sampling period 19th August 2019 to 10th January 2020).

      Findings

      Water from all 25 showers was heavily colonized [>300 colony-forming units (cfu)/mL] with P. aeruginosa at the showerhead. P. aeruginosa was found in 32% (8/25) of post-filter shower water effluent samples with a geometric mean of 4x106 cfu/mL (N=4) (6.8x104–2x108). Filters were sampled at 15–150 days of use (median 15 days), with 26% (6/23) of filter units becoming colonized before the expiry date.

      Conclusion

      POU filter showerhead units may not be effective in preventing exposure of vulnerable patients to P. aeruginosa in hospital water due to retrograde contamination (external contamination of the showerhead passed back to the filter cartridge itself) or failure of the hollow-fibre filter matrix. Reliance should not be placed on the use of hollow-fibre filters to protect patients from exposure to P. aeruginosa without repeated microbiological monitoring.

      Keywords

      Introduction

      Pseudomonas aeruginosa is an opportunistic Gram-negative bacterial pathogen that causes hospital-acquired infection of surgical wounds, blood, the respiratory tract and the urinary tract, particularly in patients in haematology wards and intensive care units (ICUs) [
      • Ducel G.
      • Fabry J.
      • Nicolle L.
      Prevention of hospital-acquired infections. A practical guide.
      ,
      • Trautmann M.
      • Lepper P.M.
      • Haller M.
      Ecology of Pseudomonas aeruginosa in the intensive care unit and the evolving role of water outlets as a reservoir of the organism.
      ]. P. aeruginosa commonly colonizes hospital water systems, and has been associated with outbreaks of infection in vulnerable patients [
      • Loveday H.P.
      • Wilson J.A.
      • Kerr K.
      • Pitchers R.
      • Walker J.T.
      • Browne J.
      Association between healthcare water systems and Pseudomonas aeruginosa infections: a rapid systematic review.
      ]. By analysing the relatedness of P. aeruginosa strains from patient infections and hospital water, studies have suggested that water systems, outlets and wet environments are implicated in serious infection in immunosuppressed patients [
      • Loveday H.P.
      • Wilson J.A.
      • Kerr K.
      • Pitchers R.
      • Walker J.T.
      • Browne J.
      Association between healthcare water systems and Pseudomonas aeruginosa infections: a rapid systematic review.
      ,
      • Aumeran C.
      • Paillard C.
      • Robin F.
      • Kanold J.
      • Baud O.
      • Bonnet R.
      • et al.
      Pseudomonas aeruginosa and Pseudomonas putida outbreak associated with contaminated water outlets in an oncohaematology paediatric unit.
      ]. The mode and direction of transfer between the patient and the clinical environment is difficult to demonstrate, but many studies have provided evidence of an association between P. aeruginosa infection and colonized taps and shower water [
      • Aumeran C.
      • Paillard C.
      • Robin F.
      • Kanold J.
      • Baud O.
      • Bonnet R.
      • et al.
      Pseudomonas aeruginosa and Pseudomonas putida outbreak associated with contaminated water outlets in an oncohaematology paediatric unit.
      ,
      • Venier A.G.
      • Leroyer C.
      • Slekovec C.
      • Talon D.
      • Bertrand X.
      • Parer S.
      • et al.
      Risk factors for Pseudomonas aeruginosa acquisition in intensive care units: a prospective multicentre study.
      ]. As such, installation of filters on water outlets has been recommended when disinfection fails to eradicate the organism.
      P. aeruginosa tends to become established in distal parts of a water system, such as sinks, taps and showers [
      • Loveday H.P.
      • Wilson J.A.
      • Kerr K.
      • Pitchers R.
      • Walker J.T.
      • Browne J.
      Association between healthcare water systems and Pseudomonas aeruginosa infections: a rapid systematic review.
      ]. Showers are liable to develop P. aeruginosa biofilm due to the materials used, low flow rate and operating water temperature of 25–40 °C which favour the growth of this pathogen [
      • Sasahara T.
      • Ogawa M.
      • Fujimura I.
      • Ae R.
      • Kosami K.
      • Morisawa Y.
      Efficacy and effectiveness of showerheads attached with point-of-use (POU) filter capsules in preventing waterborne diseases in a Japanese hospital.
      ]. The aerosol droplets produced can be inhaled by patients or can contaminate intravenous line insertion sites and damaged mucous membranes, posing a risk, particularly to patients following chemotherapy.
      In the UK, remedial actions to mitigate the risks posed by P. aeruginosa contamination in water systems are described in Health Technical Memorandum (HTM) 04–01 [
      Department of Health
      Part A: Design, installation and commissioning.
      ,
      Department of Health
      Part C: Pseudomonas aeruginosa – advice for augmented care units.
      ]. Where efforts to reduce the number of P. aeruginosa using mechanical (shearing by flushing water) and chemical (e.g. chlorine dioxide, silver/copper ion etc.) methods fail, a physical barrier approach, such as a point-of-use (POU) membrane filter unit, may be implemented if water pressure is adequate.
      The two main types of POU filter units used on taps and shower outlets in the healthcare setting are membrane filters (disposable or reusable) and hollow-fibre filters. Depending on the manufacturer, standard membrane filter units comprise of a double layer membrane with a pore size of 0.1–0.2 μm which prevents the passage of P. aeruginosa, and a pre-filtration layer that retains larger particulates and organic matter [
      • Sasahara T.
      • Ogawa M.
      • Fujimura I.
      • Ae R.
      • Kosami K.
      • Morisawa Y.
      Efficacy and effectiveness of showerheads attached with point-of-use (POU) filter capsules in preventing waterborne diseases in a Japanese hospital.
      ]. Hollow-fibre filter units consist of a sealed chamber, and the incoming water must pass through 0.1-μm-diameter pores spanning the length of a matrix of hollow fibres before exiting the outlet. Standard membrane and hollow-fibre filter units operate as pass-through water filtration systems, and are prone to biofouling and bioscaling with organic debris and inorganic salts (e.g. calcium/magnesium carbonate). The ability of these filters to sequester P. aeruginosa effectively depends on the duration and frequency of use, as well as water quality. The efficiency of membrane POU water filters has been demonstrated, but some studies have reported that P. aeruginosa contamination can occur within the recommended term of use given by the manufacturer [
      • Trautmann M.
      • Halder S.
      • Hoegel J.
      • Royer H.
      • Haller M.
      Point-of-use water filtration reduces endemic Pseudomonas aeruginosa infections on a surgical intensive care unit.
      ,
      • Florentin A.
      • Lizon J.
      • Asensio E.
      • Forin J.
      • Rivier A.
      Water and surface microbiologic quality of point-of-use water filters: a comparative study.
      ].
      Hollow-fibre filters are gaining popularity as they allow greater flow of water, especially when water pressure is low [
      • Totaro M.
      • Valentini P.
      • Casini B.
      • Miccoli M.
      • Costa A.L.
      • Baggiani A.
      Experimental comparison of point-of-use filters for drinking water ultrafiltration.
      ,
      • Smith C.M.
      • Hill V.R.
      Dead-end hollow-fiber ultrafiltration for recovery of diverse microbes from water.
      ]. The advantage of hollow-fibre filters against conventional flat membrane filters is the attainment of a large membrane surface within a limited volume as the membrane is in the form of hollow-fibre bundles [
      • Schmittel A.
      • Basagni M.
      • Gaulle E.
      • Keller T.
      ]. Polysulfone and polyethylene are commonly used materials in hollow fibres, with average pore diameter ranging from 0.25 to 1.5 μm and from 0.5 to 2 μm, respectively [
      • Schmittel A.
      • Basagni M.
      • Gaulle E.
      • Keller T.
      ,
      • Kamo J.
      • Hirai T.
      • Takahashi H.
      • Kenji K.
      Porous polyethylene hollow fiber membrane of large pore diameter, production process thereof, and hydrophilized porous polyethylene hollow fiber membranes.
      ]. Hollow fibres provide structural strength, and therefore increase the average life of a membrane. In addition, they increase water permeability due to their hydrophilic properties [
      • Totaro M.
      • Valentini P.
      • Casini B.
      • Miccoli M.
      • Costa A.L.
      • Baggiani A.
      Experimental comparison of point-of-use filters for drinking water ultrafiltration.
      ]. To determine whether these POU filters continue to prevent egress of P. aeruginosa during the manufacturer usage period, this study investigated the efficacy of 25 historically used polysulfone hollow-fibre shower filter units (medical shower filter; 0.1-μm pore size; polysulfone body; antimicrobial silver-impregnated; in-use lifecycle expiry of 92 days) in patient bathrooms in augmented and non-augmented care wards.

      Methods

      Clinical setting and selection criteria

      Twenty-five patient bathrooms were selected at random from six wards with patients requiring augmented care (haematology, elderly care, adolescent haematology/oncology and infectious diseases) at a 700-bed multi-storey building teaching hospital in London, UK. Each ward was a single floor of the hospital building. The bathrooms selected were en-suite for single isolation rooms or those serving shared-occupancy bed bays (rooms with four to six beds). Apart from elderly care, cases of P. aeruginosa bacteraemia had occurred in all of the wards in the preceding 2 months. All of the bathrooms had a POU hollow-fibre filter integrated showerhead.

      Shower water sample collection and assay by membrane concentration

      Prior to sample collection, the showerheads were disinfected by wiping the entire outer surface with a sterile alcohol wipe (70% isopropyl alcohol) and allowed to air dry (∼15 s).
      The opening of a water sample collection bag (sterile grade) was placed over a showerhead and secured to capture a water sample. An incision was made aseptically to the bottom corner of the bag to create a second opening via which water could be channelled. The shower valve was opened and an aliquot of at least 100 mL of water was collected using the water collection bag into a sample container (pre-dosed with 1 mL of neutralizer solution; composition: 1 g/L sodium thiosulphate, 30 mL/L Tween 80 and 3 g/L lecithin in phosphate-buffered saline). The showerhead was then removed aseptically and placed on a pre-sterilized tray. A second 100 mL water sample was collected in the same manner into a second sample container. These two samples represented ‘with/without POU filter’ sample arrays, respectively. The showerhead was then re-attached, and the entire surfaces of the showerhead and hose were wiped with a sterile alcohol wipe prior to reinstating the shower. This process was repeated for 25 individual showers within the hospital. This represented 21% (25/119) of showers on the test wards. The sampling period was 19th August 2019 to 10th January 2020. Follow-up sampling was performed for two of the showers 24 days after the first water collection.
      Water samples were transferred to a refrigerator (2–8 °C) within 2 h of collection and processed within 24 h. Shower samples (100±5 mL) were concentrated by vacuum filtration (max 65 kPa pressure) through a 47-mm nitrocellulose membrane of pore size 0.45 μm, followed by plating the membrane on to a Pseudomonas C–N agar plate. Plates were incubated aerobically at 37 °C for 48 h prior to counting the colonies. Water sampling and the subsequent procedures were performed in line with HTM guidelines recommended by NHS England [
      Department of Health
      Part B: Operational management.
      ].

      Confirmation of P. aeruginosa isolates

      Suspect colonies were distinguished by colony morphology (blue–green/green–yellow/red–brown) on selective agar (Pseudomonas C–N), harvested for subculture on milk-cetrimide agar (MCA) and nutrient agar in parallel, and incubated at 37 °C for 24 h. Colonies growing on nutrient agar were tested for oxidase reaction, while hydrolysis on MCA was noted.
      Isolates demonstrating oxidase-positive reactions and/or hydrolysis of casein on MCA plates were further confirmed by matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF-MS) analysis (Bruker Daltronics) in line with HTM guidelines. MALDI-TOF-MS analysis was performed as an additional confirmatory step [
      Department of Health
      Part B: Operational management.
      ].

      Measurement of P. aeruginosa

      The upper reading/counting limit of samples analysed using the membrane concentration assay technique was 300 colony-forming units (cfu)/100 mL.
      A subset of four showers, selected at random, was assayed further by taking a 1-mL aliquot from the original sample and performing serial 1/10, 1/100 and 1/1000 dilutions before plating 100 μL on to Columbia blood agar from the neat, 1/10, 1/100 and 1/1000 arrays. Confirmation of P. aeruginosa was done as described previously.

      Shower water pressure measurements

      Water pressure measurements were performed with a pressure gauge (Bourdon Pressure Gauge 0–4 bar, RSRS Components, UK) on 74 showers from 10 wards. Showerheads were dismantled from the hose, and the screw thread of the pressure gauge was fitted directly to the end of the shower hose. The outlet was opened fully to allow maximum water and pressure values, recorded in bar units, once the gauge stabilized (∼5 s). The pressure gauge was dismantled from the shower hose, and its end was disinfected by immersion into absolute ethanol (70% solution) for 2–3 s and then wiping the excess with an alcohol wipe. The showerhead was replaced on to the corresponding hose end, and further disinfected by wiping all external surfaces with a sterile alcohol wipe.

      Validation of 70% ethanol sterilization protocol

      The efficacy of the ethanol spray/wipe protocol for disinfection of the showerhead prior to sampling was validated in-house using representative shower types and a stainless steel control surface, inoculated with up to 106 cfu/cm2 of P. aeruginosa. After spraying with 70% (v/v) ethanol solution and a manual wipe at 10 s, surfaces were sampled by a bead washing technique. Reductions of 6-log10 were achieved (publication pending; data available upon request).

      Statistical analysis

      Chi-squared test with Yates' correction was performed to examine the difference between days of use of the showers that filtered the bacterial load effectively and the showers that failed to filter effectively.

      Results

      P. aeruginosa was found in the effluent from eight (32%) showers, despite the filter being in place (Table I). Six of those eight showerheads were found to have high bacterial counts (>300 cfu/100 mL). One filter (Shower #16) reduced the P. aeruginosa load in the effluent from >300 cfu to 8 cfu, while another (Shower #17) reduced the count to ∼100 cfu. These eight showers had been in use for a mean of 60.87 (95% confidence interval 15.3–106) days. Shower #16 and Shower #17 were sampled on the 15th day of use. At the second sampling (39th day of use), these two showers showed P. aeruginosa 100 and >300 cfu/mL in the effluent, respectively, with the shower filter in place.
      Table IPresence of Pseudomonas aeruginosa in the effluent of hospital shower water fitted with a point-of-use (POU) filter unit at various durations of use
      POU shower filter detailsEffluent water quality (presence of P. aeruginosa)
      Counts depicted as 0 cfu are below the detection limit (1 cfu).
      Shower ref. numberWard ref.Ward specialtyLocation of corresponding shower (bay/SIR)Age of filter (days in use)
      Expiry date of POU filter units is 92 days from date of installation (manufacturer specifications). Numbers of P. aeruginosa present in shower water with and without a POU filter unit were determined by membrane concentration assay.
      Without POU filter (cfu/100 mL)With POU filter in place (cfu/100 mL)
      1Ward EHaematologySIR15>300>300
      2Ward EHaematologySIR15>3000
      3Ward EHaematologySIR15>3000
      4Ward EHaematologySIR15>3000
      5Ward EHaematologySIR15>3000
      6Ward EHaematologySIR15>3000
      7Ward EHaematologySIR15>3000
      8Ward EHaematologySIR15>3000
      9Ward FHaematologySIR15>3000
      10Ward FHaematologySIR15>3000
      11Ward FHaematologySIR15>3000
      12Ward FHaematologySIR15>3000
      13Ward FHaematologySIR15>3000
      14Ward FHaematologySIR15>3000
      15Ward FHaematologySIR15>3000
      16Ward FHaematologySIR15>3008
      17Ward FHaematologySIR15>300100
      18Ward BElderly careBay45>300>300
      19Ward CAdolescent haematology/oncologyBay45>300>300
      20Ward CAdolescent haematology/oncologyBay47>3000
      21Ward DOncology (adult)Bay47>3000
      22Ward AInfectious diseasesBay47>3000
      23Ward FHaematologySIR52>300>300
      24Ward CAdolescent haematology/oncologyBay150>300>300
      25Ward CAdolescent haematology/oncologySIR150>300>300
      SIR, single isolation room; cfu, colony-forming units.
      a Expiry date of POU filter units is 92 days from date of installation (manufacturer specifications). Numbers of P. aeruginosa present in shower water with and without a POU filter unit were determined by membrane concentration assay.
      b Counts depicted as 0 cfu are below the detection limit (1 cfu).
      The remaining 18 showers filtered out the P. aeruginosa bioburden effectively, despite its presence at high numbers (i.e. >300 cfu/100 mL). The duration of use of the POU filters screened averaged 20.65 (standard deviation 12.57) days. There was no significant difference in the days of use between the showers that filtered the bacterial load effectively and the showers that failed to filter effectively (P=0.075).
      P. aeruginosa was quantified in four of eight showers that had P. aeruginosa >300 cfu/100 mL with a filter showerhead in place. There was a geometric mean of 4x106 cfu/100 mL (6.8x104–2x108) (Table II).
      Table IIQuantification of Pseudomonas aeruginosa bioburden to determine water quality of the effluent from four showers
      Shower description and detailsEffluent water quality (presence of P. aeruginosa)
      Shower ref. numberWard referenceWard specialtycfu/100 mL

      without POU filter
      16Ward FHaematology6.8x104
      17Ward FHaematology1.45x107
      23Ward FHaematology1.6x106
      25Ward CAdolescent haematology/oncology teenage cancer2.02x108
      cfu, colony-forming unit; POU, point of use.
      In total, 74 shower water pressure measurements were taken from 10 floors of the hospital, with values averaging 2.94 bar (range 0.3–3.9). Pressure measurements of the four wards tested in this study were:
      • Ward C: 10 shower water pressure measurements, mean 2.43 bar (range 2.3–2.8);
      • Ward D: eight shower water pressure measurements, mean 1.8 bar (range 1.6–2.2);
      • Ward E: eight shower water pressure measurements, mean 1.17 bar (range 1.1–1.25); and
      • Ward F: six shower water pressure measurements, mean 0.83 bar (range 0.8–0.9).

      Discussion

      Exposure to shower water colonized with P. aeruginosa is a potential risk for the development of bacteraemia in immunosuppressed patients [
      • Loveday H.P.
      • Wilson J.A.
      • Kerr K.
      • Pitchers R.
      • Walker J.T.
      • Browne J.
      Association between healthcare water systems and Pseudomonas aeruginosa infections: a rapid systematic review.
      ,
      • Aumeran C.
      • Paillard C.
      • Robin F.
      • Kanold J.
      • Baud O.
      • Bonnet R.
      • et al.
      Pseudomonas aeruginosa and Pseudomonas putida outbreak associated with contaminated water outlets in an oncohaematology paediatric unit.
      ]. In this study setting, the use of hollow-fibre shower filters did not provide assurance of safety for patients in the shower environment. Although not necessarily due to a failure of the filter itself, external contamination and growth inside the showerhead had a similar effect, exposing some patients to high levels of organisms, with a risk of serious subsequent infection in immunosuppressed individuals. Without repeated monitoring, clinical teams may be unaware of the potential source of pseudomonas bacteraemia in vulnerable patients.
      The hollow-fibre POU filter showerheads were in situ for 3 months before the sampling survey commenced; they replaced showers with non-filtration antimicrobial-impregnated showerhead/hose units.
      The hollow-fibre technology was selected due to the high-capacity filtration via the 0.1-μm-diameter pores in the filter matrices, and the long shelf-life of 92 days (manufacturer communications). The POU filters were subjected to routine surveillance to assure efficacy against P. aeruginosa during the period of use.
      Although the POU filters were effective in removing P. aeruginosa from the effluent in the majority of cases, the organism was found distal to the filter in one-third (8/25) of showers. While this study did not explore the sources of contamination, the isolation of P. aeruginosa from filter-treated water was likely due to retrograde contamination from external reservoirs, or failure of the filter matrices in sequestering bacteria.
      In this study, a high bacterial burden (>106 cfu/100 mL) in the pipework proximal to the filter may have overwhelmed the efficacy of the hollow-fibre filter matrix. However, a study using a 0.1-μm porous polyethylene hollow-fibre filter demonstrated >log 6 reduction when challenged with Klebsiella terrigena [
      • Hydreion L.
      Microbiological testing of the Sawyer 7/6B filter.
      ]. Retrograde contamination of taps, and even proximal piping, from drains despite POU filters has been reported [
      • Bédard E.
      • Prévost M.
      • Déziel E.
      Pseudomonas aeruginosa in premise plumbing of large buildings.
      ].
      POU filters are an alternative to chemical disinfection using chlorine dioxide, hydrogen peroxide or copper–silver ionization, and are effective when endemic potential pathogens cannot be eliminated [
      • Sasahara T.
      • Ogawa M.
      • Fujimura I.
      • Ae R.
      • Kosami K.
      • Morisawa Y.
      Efficacy and effectiveness of showerheads attached with point-of-use (POU) filter capsules in preventing waterborne diseases in a Japanese hospital.
      ]. In a surgical ICU, POU filters were associated with elimination of tap water contamination, and a reduction of pseudomonas colonization and infection in patients by 95% and 56%, respectively [
      • Trautmann M.
      • Halder S.
      • Hoegel J.
      • Royer H.
      • Haller M.
      Point-of-use water filtration reduces endemic Pseudomonas aeruginosa infections on a surgical intensive care unit.
      ]. Use of 0.2-μm filters in wards in Japan removed all Gram-negative bacterial contamination in water for up to 2 months [
      • Sasahara T.
      • Ogawa M.
      • Fujimura I.
      • Ae R.
      • Kosami K.
      • Morisawa Y.
      Efficacy and effectiveness of showerheads attached with point-of-use (POU) filter capsules in preventing waterborne diseases in a Japanese hospital.
      ]. Studies in ICUs and bone marrow transplant units found that installation of filters reduced nosocomial pseudomonas infections [
      • Barna Z.
      • Antmann K.
      • Paszti J.
      • Banfi R.
      • Kadar M.
      • Szax A.
      • et al.
      Infection control by point-of-use water filtration in an intensive care unit – a Hungarian case study.
      ,
      • Cervia J.S.
      • Farber B.
      • Armellino D.
      • Klocke J.
      • Bayer R.L.
      • McAlister M.
      • et al.
      Point-of-use water filtration reduces healthcare-associated infections in bone marrow transplant recipients.
      ].
      However, external contamination can affect the efficacy of POU filter devices, and represents an indefinite revenue commitment for replacements. In the present study, the hollow-fibre filters adopted had a specified lifespan of approximately 3 months. Nevertheless, 26% (6/23) of the POU filters became colonized before the expiry date of the device had elapsed. Two of the filters screened in this study were in situ beyond the expiry date, and were decommissioned from use immediately by the hospital estates and facilities management. Membrane filter devices are an alternative to hollow-fibre filter units, but contamination with P. aeruginosa has been demonstrated to occur within the recommended duration of use [
      • Trautmann M.
      • Halder S.
      • Hoegel J.
      • Royer H.
      • Haller M.
      Point-of-use water filtration reduces endemic Pseudomonas aeruginosa infections on a surgical intensive care unit.
      ]. A study from France reported P. aeruginosa contamination 4 and 5 weeks after installation [
      • Florentin A.
      • Lizon J.
      • Asensio E.
      • Forin J.
      • Rivier A.
      Water and surface microbiologic quality of point-of-use water filters: a comparative study.
      ]. Although the contamination level may be low initially, P. aeruginosa can proliferate quickly, presenting a risk for cross-contamination. Polysulfone or polyethylene hollow-fibre filters have practical utility over standard membrane filters in low-pressure water systems, where water output would otherwise be severely attenuated [
      • Totaro M.
      • Valentini P.
      • Casini B.
      • Miccoli M.
      • Costa A.L.
      • Baggiani A.
      Experimental comparison of point-of-use filters for drinking water ultrafiltration.
      ,
      • Smith C.M.
      • Hill V.R.
      Dead-end hollow-fiber ultrafiltration for recovery of diverse microbes from water.
      ]. However, they are susceptible to the same problems of external contamination within a few weeks of installation. A laboratory study of experimental contamination of pristine hollow-fibre filter devices (0.2-μm pore size) before placing on uncontaminated taps and showers found that hollow-fibre shower filters were effective in removing P. aeruginosa [
      • Totaro M.
      • Valentini P.
      • Casini B.
      • Miccoli M.
      • Costa A.L.
      • Baggiani A.
      Experimental comparison of point-of-use filters for drinking water ultrafiltration.
      ]. However, despite a recommended use time of 31 days, hollow-fibre tap filters showed early growth of P. aeruginosa, in one case from day 16. There was no back contamination after filters were removed.
      The mains water supply of the hospital was screened at the incoming site to the hospital, and found to be free of P. aeruginosa (data upon request). In the present study, the water proximal to the filters harboured P. aeruginosa 106 cfu/100 mL. In cases where P. aeruginosa was isolated post-filtration, it could not be ascertained whether the contamination originated from retrograde contamination (e.g. aerosolized droplets from shower trays/drains), translocation through the filter matrix by high-pressure water flow, or as a consequence of perforation of the POU filter cartridge within the body of the showerhead. The pressure of water flow in the test building was below the upper tolerance (5 bar; manufacturer product specification) of the POU filter cartridge. Further exploratory and destructive analysis of the filter device, including microbiological and molecular characterization, is required. Low pressures present another risk because patients may then remove the showerheads and expose themselves to unfiltered shower water colonized by P. aeruginosa. Low shower pressures averaged 0.83 bar on Ward F; a haematology area occupied by immunosuppressed patients. In some cases, showerheads had already been removed by the patients when showers were inspected, despite warnings by nurses, ward sisters and wall posters not to do so.
      An audit conducted after this study screened patients for rectal colonization between 24th January 2020 and 13th May 2020 (110 days). Six hundred and six samples (groin/rectal swabs) were collected from 155 patients. Four patients were P. aeruginosa negative in the first sample, but acquired P. aeruginosa during their stay (unpublished data).
      Various devices are marketed on the premise of delaying retrograde biofilm formation, but efficacy in use against Pseudomonas spp. has not been demonstrated in peer-reviewed studies (e.g. copper inserts for tap outlets and silver-impregnated hoses). Although it is important to demonstrate the source of contamination, investigation of all possible routes of transmission is difficult. Hollow-fibre medical filter devices may be useful in preventing exposure of patients to P. aeruginosa from colonized shower water for short periods of use. However, application of POU shower filter units should be complemented with regular water testing, daily cleaning and internal disinfection of filtered water outlets in augmented care wards, especially when growth of P aeruginosa persists.

      Acknowledgements

      The authors wish to thank Estelle Caine who undertook a 70% ethanol sterilization validation study, which is outside the scope of this paper but is subject to a future publication.

      Conflict of interest statement

      None declared.

      Funding sources

      Özge Yetiş is funded by Republic of Turkiye Ministry of National Education for her doctoral studies. Peter Wilson was funded, in part, by the National Institute for Health Research , University College London Hospitals Biomedical Research Centre .

      References

        • Ducel G.
        • Fabry J.
        • Nicolle L.
        Prevention of hospital-acquired infections. A practical guide.
        2nd ed. World Health Organization, Geneva2002
        • Trautmann M.
        • Lepper P.M.
        • Haller M.
        Ecology of Pseudomonas aeruginosa in the intensive care unit and the evolving role of water outlets as a reservoir of the organism.
        Am J Infect Control. 2005; 33: S41-S49
        • Loveday H.P.
        • Wilson J.A.
        • Kerr K.
        • Pitchers R.
        • Walker J.T.
        • Browne J.
        Association between healthcare water systems and Pseudomonas aeruginosa infections: a rapid systematic review.
        J Hosp Infect. 2014; 86: 7-15
        • Aumeran C.
        • Paillard C.
        • Robin F.
        • Kanold J.
        • Baud O.
        • Bonnet R.
        • et al.
        Pseudomonas aeruginosa and Pseudomonas putida outbreak associated with contaminated water outlets in an oncohaematology paediatric unit.
        J Hosp Infect. 2007; 65: 47-53
        • Venier A.G.
        • Leroyer C.
        • Slekovec C.
        • Talon D.
        • Bertrand X.
        • Parer S.
        • et al.
        Risk factors for Pseudomonas aeruginosa acquisition in intensive care units: a prospective multicentre study.
        J Hosp Infect. 2014; 88: 103-108
        • Sasahara T.
        • Ogawa M.
        • Fujimura I.
        • Ae R.
        • Kosami K.
        • Morisawa Y.
        Efficacy and effectiveness of showerheads attached with point-of-use (POU) filter capsules in preventing waterborne diseases in a Japanese hospital.
        Biocontrol Sci. 2020; 25: 223-230
        • Department of Health
        Part A: Design, installation and commissioning.
        in: Health Technical Memorandum 04-01: Safe water in healthcare premises. DoH, London2016
        • Department of Health
        Part C: Pseudomonas aeruginosa – advice for augmented care units.
        in: Health Technical Memorandum 04-01: Safe water in healthcare premises. DoH, London2016
        • Trautmann M.
        • Halder S.
        • Hoegel J.
        • Royer H.
        • Haller M.
        Point-of-use water filtration reduces endemic Pseudomonas aeruginosa infections on a surgical intensive care unit.
        Am J Infect Control. 2008; 36: 421-429
        • Florentin A.
        • Lizon J.
        • Asensio E.
        • Forin J.
        • Rivier A.
        Water and surface microbiologic quality of point-of-use water filters: a comparative study.
        Am J Infect Control. 2016; 44: 1061-1062
        • Totaro M.
        • Valentini P.
        • Casini B.
        • Miccoli M.
        • Costa A.L.
        • Baggiani A.
        Experimental comparison of point-of-use filters for drinking water ultrafiltration.
        J Hosp Infect. 2017; 96: 172-176
        • Smith C.M.
        • Hill V.R.
        Dead-end hollow-fiber ultrafiltration for recovery of diverse microbes from water.
        Appl Environ Microbiol. 2009; 75: 5284-5289
        • Schmittel A.
        • Basagni M.
        • Gaulle E.
        • Keller T.
        Point-of-use water purifier with polysulfone hollow fibres. Vol. 1. US Pat Appl Publ, 2017 (p. 1–15)
        • Kamo J.
        • Hirai T.
        • Takahashi H.
        • Kenji K.
        Porous polyethylene hollow fiber membrane of large pore diameter, production process thereof, and hydrophilized porous polyethylene hollow fiber membranes.
        2017 (p. 1–2)
        • Department of Health
        Part B: Operational management.
        in: Health Technical Memorandum 04-01: Safe water in healthcare premises. DoH, London2016
        • Hydreion L.
        Microbiological testing of the Sawyer 7/6B filter.
        Report No S05-03. 2005
        • Bédard E.
        • Prévost M.
        • Déziel E.
        Pseudomonas aeruginosa in premise plumbing of large buildings.
        Microbiologyopen. 2016; 5: 937
        • Barna Z.
        • Antmann K.
        • Paszti J.
        • Banfi R.
        • Kadar M.
        • Szax A.
        • et al.
        Infection control by point-of-use water filtration in an intensive care unit – a Hungarian case study.
        J Water Health. 2014; 12: 858-867
        • Cervia J.S.
        • Farber B.
        • Armellino D.
        • Klocke J.
        • Bayer R.L.
        • McAlister M.
        • et al.
        Point-of-use water filtration reduces healthcare-associated infections in bone marrow transplant recipients.
        Transpl Infect Dis. 2010; 12: 238-241