Reversing and controlling microbial proliferation in the water system of a high-risk hospital ward after extended closure and reconstruction

Opportunistic premise plumbing pathogens occur naturally in water but can pose a health risk to hospital patients who are more vulnerable due to illness or treatment. Ward closure periods can lead to microbial proliferation within water systems, posing a challenge to hospital estates and infection control teams. Following the 3-year closure of our paediatric haemato-oncology ward, water testing showed high total viable counts (TVCs) in over 20 % of samples and elevated counts of numerous Gram negative bacterial species (GNBs) in 73 % of samples, despite daily flushing and continuous chlorine dioxide dosing. We aimed to determine the extent of microbial proliferation, measure the impact of three sequential interventions (system disinfections with chlorine, with silver stabilised hydrogen peroxide, and then tap replacement), and assess the long-term performance of this water system. By sampling systematically across spatial and temporal scales, and using a range of microbiological tests (TVCs, Legionella spp., Pseudomonas spp., Gram negative bacteria, atypical mycobacteria and fungi), we showed that microbial proliferation was confined to the closed ward. Chlorine treatment had no significant impact on TVCs, but both silver stabilised hydrogen peroxide and tap replacement resulted in significant decreases (p < 0.01). Similarly, the three Gram negative species that were enriched following the reconstruction period (Cupriavidus pauculus, Sphingomonas paucimobilis, and Acidovorax temperans) were less impacted by chlorine than by the other interventions. Following these interventions, fewer than 1 % of samples exceeded our strict local TVC threshold of 10 CFU/ml and GNBs were detected in 7 % of samples. Since the ward reopened to patients in 2022, there has been no return of the high microbiological counts observed immediately after reconstruction. Gram negative bacteria have been detected only sporadically, and the taxa found in samples collected through 0.2 μm point-of-use filters shifted towards species associated with humans. Our systematic approach was successful in returning this hospital water system to a safe state, and once microbial proliferation within the system itself was rectified, further positive results were likely attributable to the interactions of users with the outlets. Distinguishing between possible sources of microbial counts in water is crucial to selecting the most suitable interventions and helping ensure provision of safe water to patients.