This study confirms that current methods of routine cleaning are inadequate, reporting 28.1% of high-touch ED patient stretcher rails to exceed contamination threshold of 2.5 CFU/cm2 despite routine cleaning. This finding is consistent with the widely acknowledged deficiency of traditional disinfection methods, an alarming cause for concern considering the established relationship between contaminated hospital surfaces, HAI, and patient safety [8, 9, 25,26,27].

Our study advances the solution to ineffective disinfection by answering the call to develop complementary cleaning technology [11, 25, 26]. As a double-blind, placebo-controlled, randomized trial, our project demonstrates that an antimicrobial surface coating significantly reduced bacterial contamination of a high-touch ED surface. Our study builds on the sporadic existing observations, including only one other trial reporting the promise of emerging antimicrobial surface coating technology in the hospital setting [16,17,18, 28]. This represents an early but crucial step towards innovation in healthcare environmental hygiene, a global agenda towards improving patient safety [11].

Contamination of antimicrobial rails at 24 h was significantly lower (0.61 vs. 1.01 CFU/cm2, median difference 0.40 CFU/cm2, 95% CI 0.01 to 1.01 CFU/cm2) than placebo rails. While a widely-accepted benchmark for disinfection efficacy has yet to be established, an absolute reduction of 0.40 CFU/cm2 is likely to be important considering the target threshold for cleanliness is 2.5 CFU/cm2 and often exceeded. Although contamination reduction is not statistically significant at 7 days (1.15 vs. 1.50 CFU/cm2, median difference 0.35 CFU/cm2, 95% CI -0.64 to 1.28 CFU/cm2), the effect size remains consistent and considerable. Indeed, to further convey the clinical meaning of the treatment effects, stretcher rails were classified as “contaminated” if they exceeded the 2.50 CFU/cm2 threshold, and we observed that antimicrobial rails had approximately 43% lower odds of exceeding this threshold than did placebo rails at both timepoints (Table 4).

The potentially short-lived antimicrobial action, when coupled with the significantly lower bacterial contamination seen in the subgroup of rails with visibly intact antimicrobial surface coating, suggests that the lost antimicrobial effect is attributable to stripping of the coating by adhesive-wear. Durability of surface coatings is known to be important for sustained effectiveness of antimicrobial coatings, and wear-resistance to factors like friction, heat, and alcohol solvents are routinely tested [29, 30]. However, testing is limited to in vitro laboratory settings, and to our knowledge durability against adhesive stripping has not been studied. In the context of high-touch frequency healthcare surfaces that are frequently adapted to patient-needs, as in the case of adhering plastic bags to stretcher rails, our results highlight a new potential coating property that needs to be improved.

As the first ED-based study, our finding of short-lived antimicrobial effect of stretcher rail coatings suggests activity- and department-specific considerations when developing and applying antimicrobial surface coatings. Although customizing disinfection protocols to account for different healthcare settings is a known requirement, specific recommendations for self-disinfecting technology is still lacking [31]. In most EDs, high patient-turnover may require more frequent re-application than routine recommendations by antimicrobial coating manufacturers. Properties of antimicrobial coatings may also need to be modified to withstand ED conditions. Furthermore, time between touch instances is particularly short in the ED, posing particular challenges for chemical-based antimicrobials that rely on contact-free time for microbe killing [32]. Taking into account the difficulty of routine cleaning in EDs, developing and studying specific antimicrobial technology is especially important.

MRSA was detected on 20 out of 96 (20.8%) post routine-cleaning, pre-patient-use stretcher rails, with a median of 0.00 CFU/cm2 (IQR 0.00 to 0.00 CFU/cm2). MRSA was detected on 50 out of 143 (35.0%) pre-routine-cleaning, post-patient-use, non-intervention stretcher rails, sampled across three time-points, with a median of 0.00 CFU/cm2 (IQR 0.00 to 0.02 CFU/cm2). Similar rates of environmental MRSA contamination have been shown in other studies. However, most focus on the immediate environment of patients in whom MRSA was already detected, making our undifferentiated, pre-detection research setting unique [33,34,35]. Given the acknowledged conundrum of MRSA carriers contaminating their environment even before detection and precaution can be instituted, our study calls further attention to disinfection in hospital settings with high turnover of patients and equipment like the ED.

The novel methods in this study improves on limitations in previous projects identified in a recent review [6]. While most prior studies used before-and-after study designs, we implemented a control arm to ensure that any observed reduction in bacterial contamination we found was attributable to the antimicrobial coating intervention instead of other inadvertent change in conditions within the study period. Additionally, our ED study setting randomizing individual stretchers instead of ward-and room-based methods used in most studies allowed direct comparison of contamination between each surface. This avoids the limitation of intervention and non-intervention surfaces mixing when individual items move between randomized rooms. Blinding of both coating applicators and microbiological samplers by using an individually unique color-coded system rather than a two-arm method of labelling population groups A or B was also unique. We tracked each stretcher at all sampling points and scheduled sampling strictly after patient-use and before routine cleaning to ensure an accurate assessment of antimicrobial efficacy by treatment coating without interference from routine cleaning.

This study identifies areas for improvement in future research. Trials using alternate antimicrobial surface coatings in the ED are needed given the unique nature of this healthcare setting. Advancement in and testing of coating mechanical properties for greater wear-resistance against adhesive stripping is important, especially for surfaces in which repeat coating is infeasible. Studies comparing different antimicrobial surface coatings may generate valuable information on ideal coating properties. Larger studies that evaluate patient-oriented outcome measures will ultimately be required to demonstrate actual reduction in HAIs, rather than relying on surrogate microbiological outcomes.

Our study had several limitations. Firstly, sample size was not calculated and instead based on all 96 available patient stretchers, likely underpowering the study to show significant differences in contamination. Despite this small sample size, contamination at 24 h was significantly lower in the antimicrobial rails. A pre-specified sample size estimation is likely to demonstrate the full extent of antimicrobial effect. Secondly, as a single-centre trial investigating only ED stretcher rails based on touch-frequency, our findings may not be generalizable to other surfaces or settings [36]. Additionally, sampling between 07:00 to 19:00 h omits after-hour conditions which may affect contamination.

Stretchers were occupied by patients for varying periods of time before sampling was performed, resulting in unequal opportunity for contamination. We ensured that patients occupied stretchers for at least 30 min to represent significant use. Nevertheless, the association between contamination and occupancy-time was not investigated in this study. The subgroup of intervention stretchers with visibly intact coating was acquired post hoc, as adhesive-degradation of intervention coating was unforeseen. However, the blinding that was originally planned was maintained within this subgroup since the coating remained indistinguishable, minimizing potential detection bias. The mechanism of action stated in our methods was provided by the manufacturer and, especially given the short-lived antimicrobial action with observed adhesive-wear, the study was unable to separate an active biocidal effect of the contained ingredients from passive restriction of bacterial surface binding conferred by the coating. Lastly, the secondary outcomes did not include important pathogens like VRE, against which in-vitro experiments using NOMOBAC demonstrated adequate antimicrobial activity.