Dear Sir,

Legionellosis poses a significant risk in our 1700-bed tertiary hospital with a large immunocompromised patient population from oncology, hematology, transplant, and intensive care units ¹ . The plumbing infrastructure is old and complex, with municipally treated potable water supplied by Singapore’s Public Utilities Board and stored in an on-site central reservoir before distribution to seven multi-storey inpatient blocks and associated clinical areas through a series of holding tanks. Much of the campus was redeveloped in the 1980s and has remained in continuous use, resulting in aging and complex water distribution infrastructure. Average water age is approximately 4 days. Further, the incoming cold-water temperature is persistently high (>25°C) due to year-round tropical weather. Hot-water lines run from basement tanks at each block, supplying only showerheads in the inpatient areas and operating theaters. All showerheads are fitted with thermostatic mixing valves (TMVs) to prevent scalding.

In April 2022, routine surveillance bulk water samples from showerheads at the burns unit were analyzed by accredited laboratories in accordance with ISO 11731:2017 standards ² and detected L. pneumophila (non-serogroup 1) and/or other Legionella spp. in 4/22 (18.2%) outlets (range: 0.1–126 CFU/mL), including values exceeding 1 CFU/mL.

Despite initial remedial measures, including showerhead replacement and short-duration flushing of 5–10 minutes, five outlets remained persistently positive on repeat sampling. Four were located in high-dependency rooms where shower utilization was low due to extensive dressings or limited mobility. At that time, as part of an institutionwide review by the Water Management Committee (WMC), at least alternate-day flushing was implemented hospitalwide to mitigate stagnation at low-usage showerheads ³ .

WMC also recommended localized thermal disinfection at our location because of the vulnerable patient population. The Facilities Management and Engineering team coordinated operational planning and technical execution. Ward staff were informed in advance to facilitate patient relocation and communication. On 29 April 2022, affected rooms were vacated, and water supply to all involved outlets was isolated. Heated air from industrial hot-air blowers was applied directly to approximately 30 cm of exposed distal outlet pipework to augment heat transfer. Existing showerheads and TMVs were removed and discarded prior to heating, and residual water within the distal pipework was drained. Pipe surface temperatures were raised to levels exceeding 70°C at the distal pipe segment. Surface temperature was monitored using a calibrated infrared thermometer to confirm that the desired temperature and duration were achieved. This threshold served as a surrogate for bactericidal thermal exposure in the distal pipe segment, in the absence of reliable internal water temperature measurements. Following heating, full-volume flushing of both the hot- and cold-water lines was performed for five minutes per outlet. New showerheads and TMVs were installed, water supply was restored, and outlet temperatures were verified to be within safe range before release for use. Environmental services staff performed terminal cleaning of the bathrooms with sodium hypochlorite prior to room reoccupation. All five persistently positive showerheads underwent similar treatment. The intervention resulted in a 3-hour service disruption per affected area.

Intervention effectiveness was defined as the sustained clearance of all Legionella bacterial types, indicated by concentrations below laboratory detection threshold (<0.1 CFU/mL) for 3 months following the intervention. In the second month, there was a single detection of non-pneumophila Legionella (0.1 CFU/mL) at one outlet that was not reproducible on repeat testing and did not require remediation. No cases of healthcare-associated legionellosis were observed during the monitoring period.

Following the success of this pilot, targeted blower-assisted superheat-and-flush intervention was implemented hospitalwide as an escalation strategy for persistent Legionella contamination at water outlets. In the ensuing three years (August 2022 to December 2025), 110/1388 (7.93%) routine surveillance cultures from showerheads across our hospital were positive for any type of Legionella (median: 0.3 CFU/mL; range: 0.1–139 CFU/mL) despite at least alternate-day flushing. Affected showerheads underwent blower-assisted superheating at the terminal ends of the hot and cold-water inlets followed by controlled flushing. Repeat testing 48 hours postremediation was negative in 97/110, representing an 88.2% success rate in clearance of contamination after one round of superheating. Of the remaining 13, 9 resolved after a second round, while 2 needed a third round for resolution. One was from a location that closed shortly after for renovation. Information was incomplete for the last outlet. We noted consecutive episodes of contamination over two years (2022–2023) from one showerhead located in a suite room, despite successful remediation after each individual case. We attribute the repeated occurrence to very low room occupancy rates (<5 admissions per year). Increasing frequency to daily flushing was associated with the absence of further recurrence. For hospital-wide implementation, we streamlined the work process to approximately 1 hour by eliminating the need for patient relocation. For locations where accessible pipes, valves, and showerheads were contained in the same bathroom, superheating was carried out with the bathroom cordoned off. At locations where this was not possible, a 0.2 μm point-of-use filter was installed at the showerhead as an interim measure, and definitive superheating was carried out upon patient discharge.

In summary, targeted blower-assisted thermal disinfection achieves sustained clearance of persistent distal outlet Legionella contamination and provides a feasible escalation strategy between routine outlet management and systemwide remediation in complex healthcare water systems.

This work was conducted as a quality improvement initiative and did not require CIRB approval.