Urethral catheters (UCs) are pliable plastic tubes carefully inserted into the urinary bladder through the urethra, serving the purpose of collecting urine in a drainage bag (Chuang and Tambyah, 2021; Newman et al., 2018). As indispensable devices in patient care, catheters rank among the most extensively utilized medical tools in hospital settings. They find application in patients with urinary obstruction, both preceding and following specific surgical procedures. Additionally, catheters play a crucial role in aiding various treatments, such as ensuring precise measurement of urine output, managing incontinence, or facilitating the direct administration of particular medications into the bladder.

Despite their valuable utility, prolonged use of these devices comes with the risk of serious complications, notably the occurrence of catheter-associated urinary tract infections (CAUTIs), recognized as the most prevalent cause of healthcare-associated infections (HAIs) (Teixeira-Santos et al., 2022; Tenke et al., 2017). Alarmingly, approximately 15%–25% of hospitalized patients undergo catheterization, underlining the widespread usage and the associated risks. In both Europe and the United States, the annual incidence of CAUTIs surpasses one million cases, contributing to escalated morbidity and mortality rates (Cai et al., 2023; Öztürk and Murt, 2020).

In this context, rates of CAUTIs are significantly higher in Latin American hospitals than in industrialized countries (Yin et al., 2023). For example, CAUTIs in Chile have shown a notable and consistent increase. Prevalence studies conducted by the Ministry of Health (Unidad de Infecciones Intrahospitalarias, 2020) highlight CAUTIs as the most frequently reported infections across hospitals of varying complexity levels nationwide. Noteworthy is the significant prevalence of CAUTIs observed from 2014 to 2019, primarily attributed to the extensive use of UCs within a substantial patient population. Among the six medical devices examined in the HAIs surveillance report, UCs emerged as the most widely employed, boasting the highest utilization percentage compared to oral parenteral nutrition, other probes, and mechanical ventilation (Unidad de Infecciones Intrahospitalarias, 2020).

In that sense, catheterization could be considered a contraindication as it is closely associated with a heightened probability of CAUTIs (Chuang and Tambyah, 2021). These infections result from the formation of biofilms on UCs, consisting of bacteria that exhibit reduced sensitivity to antibiotics. Additionally, CAUTIs are associated with increased mortality rates, prolonged hospitalization times, and higher hospital expenses (Gray et al., 2023; Hollenbeak and Schilling, 2018). In some cases, this CAUTIs incidence was worsened by the impact of the COVID-19 pandemic due to the large volume of patients requiring advanced medical care and subsequent depleted resources (Hyte et al., 2023; Mitra et al., 2021).

Gram-negative bacteria are the most common causative agents of urinary tract infections (UTIs), including Escherichia coli, Proteus mirabilis, Klebsiella pneumoniae, and Pseudomonas aeruginosa (Majumder et al., 2018). Although Gram-positive bacteria are generally less prevalent (Alshomrani et al., 2023), recent studies have reported a growing incidence of S. aureus in cases of asymptomatic bacteriuria and complicated UTIs, particularly among elderly individuals, recently hospitalized patients, and those with indwelling urinary catheters (Paudel et al., 2023). Furthermore, Candida species can thrive and proliferate in the urinary tract environment (Yao et al., 2022). All these CAUTI-associated pathogens exhibit multiple virulence factors including fimbrial adhesins, urease activity, biofilm formation, and immune evasion proteins. These traits facilitate catheter colonization and persistence. Ciprofloxacin (CFX) targets bacterial DNA gyrase but may be ineffective against sessile, biofilm-embedded cells. Conversely, silver (Ag) exerts membrane-disruptive and metabolic effects independent of active replication, offering complementary protection. Understanding the interplay between drug action and pathogen virulence is key to designing effective coatings (Govindarajan and Kandaswamy, 2022; Sivaramalingam et al., 2024).

In that sense, different strategies have been used to develop antimicrobial surfaces for UCs, which can be grouped into five strategies (Teixeira-Santos et al., 2022): 1) release of antimicrobial compounds (Srisang et al., 2020; Lin et al., 2021; Miao et al., 2023), 2) contact-killing (Gao et al., 2020; Tailly et al., 2021; Bai et al., 2023), 3) anti-adhesive (Yu et al., 2017; Zhang et al., 2019; Prateeksha et al., 2021), 4) disruption of biofilm architecture (Durgadevi et al., 2020; Swidan et al., 2022; da Silva et al., 2024), and 5) benign biofilms to inhibit pathogen colonization (Zhu et al., 2015; Carvalho et al., 2021).

Studies have reported antimicrobial coatings using silver nanoparticles (AgNPs) combined with antibiotics. However, many of these strategies face challenges such as nanoparticle aggregation, burst drug release, and limited control over release kinetics under physiological conditions (Kadirvelu et al., 2024). Our study advances this field by introducing a drug delivery system based on an antibacterial layer-by-layer (LbL) coating applied to commercial silicone urinary catheters (UCs) of two sizes (14Fr and 20Fr). The LbL architecture, fabricated by sequentially assembling carboxymethylcellulose (CMC) with the antibacterial chitosan-silver complex (CHI-Ag), allows modular control of coating thickness, drug loading, and silver integration. CFX was incorporated as a model drug to investigate controlled release performance. This dual-action system combines contact-killing properties with the sustained release of antimicrobial compounds. The coated UCs were subjected to morphological and physicochemical characterization, and their antibacterial efficacy was evaluated against both gram-negative E. coli and gram-positive S. aureus strains. Biological in vitro assays were performed using pristine, coated, and CFX-loaded coated samples. To elucidate the mechanisms of CFX loading and release, we implemented a mathematical model based on unidirectional diffusion governed by Fick’s second law. Additionally, molecular dynamics (MD) simulations were employed to investigate the interactions between CFX and the coating matrix, using catheter-relevant geometries (14Fr/20Fr) to better represent real-world performance.

In this work, UCs of varying diameters (14Fr and 20Fr) underwent a coating process involving a thin film of CMC/CHI-Ag, assembled through a layer-by-layer technique (LbL). The synthesis process is outlined schematically in Figure 2. Given that polydimethylsiloxane (PDMS) serves as the basis for the silicone UCs (Zare et al., 2021), the surface of these devices is inherently hydrophobic. Thus, prior to coat the samples, a surface pretreatment involving O 2 plasma oxidation, succeeded by PEI grafting was conducted. This well-established strategy ensures a more uniform and controlled synthesis of LbL films (Hernandez-Montelongo et al., 2017). Subsequently, a ten-cycles assembly process was executed, culminating in immersion in an ascorbic acid solution to facilitate the reduction of Ag + ions to Ag 0 within the coating matrix (Naveas et al., 2023).

Figure 3A depicts UCs before and after the LbL coating, evident from the light brown appearance. Following the LbL treatment, the wettability of UC samples transitioned to a hydrophilic behavior, registering a change from 100 ± 2° to 69 ± 5° for the 14Fr catheter and from 103 ± 2° to 68 ± 7° for the 20Fr catheter. This shift is attributed to the presence of carboxylic groups in CMC, ammonium groups in CHI, and Ag in the coating. In Figure 3B, contact angle images of the 14Fr (hydrophobic) and 14Fr-LbL (hydrophilic) samples are presented. Concurrently, the roughness experienced a significant alteration due to the LbL coating. Both 14Fr and 20Fr samples saw a substantial increase in roughness. 14Fr nearly doubled its roughness from 0.46 ± 0.13 μ m to 0.93 ± 0.33 μ m, while 20Fr increased from 0.70 ± 0.22 μ m to 1.18 ± 0.42 μ m. These values indicate the formation of a micro-structured coating on the surface.

UV–Vis spectroscopy served as a tool to indirectly identify the carboxylic groups of CMC (negatively charged) by utilizing the oppositely charged stain methylene blue (MB) (positively charged) (Eltaweil et al., 2020). Likewise, to indirectly detect the ammonium groups of CHI (positively charged), the stain Rose Bengal (RB) (negatively charged) was applied (Bayraktutan, 2020). Figure 3C illustrates the absorbance peaks of MB and RB for both the 14Fr and 14Fr-LbL samples. The pristine 14Fr sample displayed low absorbance values: 0.089 for MB and 0.020 for RB. In contrast, the coated 14Fr-LbL sample exhibited significantly higher absorbance peaks, measuring 0.476 for MB and 0.635 for RB. A similar trend was observed in the 20Fr and 20Fr-LbL samples.

The chemical analysis of the samples was conducted using ATR-FTIR to directly identify the surface modifications (Figure 3D). The bending vibrations of the fingerprint functional groups of PDMS were discerned, including Si-( C H 3 )₂ (786 c m − 1 ), symmetric bending of Si-CH₃ (1,263 c m − 1 ), stretching of CH (2,960 c m − 1 ) from CH₃, and Si-O-Si stretching at 1,065 c m − 1 and 1,007 c m − 1 , observed in the spectra of both control samples (14Fr and 20Fr) and coated samples (14Fr-LbL and 20Fr-LbL) (Bodas and Khan-Malek, 2006). Furthermore, a broad band at 3,400 c m − 1 corresponding to O-H stretching of CMC and CHI was detected. N H 3 + bending vibration at 1,624 c m − 1 of CHI and -C=O antisymmetric stretching vibration at 1,580 c m − 1 of CMC were also identified (Bozoğlan et al., 2020).

Figure 4A displays cross-sectional SEM images of both UCs. The 14Fr sample exhibited external and internal diameters of 4.7 mm and 2.7 mm, respectively. In the case of the 20Fr sample, its external and internal diameters were 6.7 mm and 4.7 mm, respectively. Both UCs presented a wall thickness of 1 mm. Moreover, the typical inflation cuff orifice of Foley catheters was also observed in both samples. UCs submitted to the layer-by-layer treatment showed a thin coating of 1.3 ± 0.2 μ m for 14Fr (Figure 4B) and 1.6 ± 0.4 μ m for 20Fr (Figure 4C). To confirm that the observed coating belongs to the LbL film, EDX analysis was performed on samples. In the pristine 14Fr catheter, C, O, and Si were detected, in agreement with their PDMS composition (Figure 4D). EDX mapping also presents these elements on the edge of the surface of the sample. However, in the case of the coated sample (14Fr‐LbL; Figure 4E), Ag was also detected at a mass percent of 1.5 ± 0.6%. EDX mapping shows a high signal of Ag in the coating matrix. For 20Fr-LbL, a similar tendency was observed with a mass percent of 1.6 ± 0.2%.

Rutherford Backscattering Spectrometry (RBS) was performed to analyze the diffusion of silver (Ag) in the UCs. Figure 5A shows the RBS spectrum obtained for the 14Fr-LbL sample. Five element-related signals can be observed, corresponding to carbon (C), nitrogen (N), oxygen (O), silicon (Si), and silver (Ag). The Ag signal shows a peak at a backscattering energy of 3,359 keV, whereas, for Ag atoms on the surface, a backscattering energy of 3,447 keV would be expected. This indicates that silver has infiltrated towards the bulk material of the LbL coating.

The in-depth compositional profile of silver was estimated by fitting the RBS spectrum using the Simnra 7.03 software, as shown in Figure 5B. For this analysis, a structure composed of up to 20 layers of material was considered, each with varying thickness and concentrations of C, N, O, Si, and Ag. The results confirm that the concentration of Ag on the coating surface is only 0.4%, while deeper within the material, the weight percentage increases up to 3.2%.

The surface chemical composition of the UCs was studied via X-ray Photoelectron Spectroscopy (XPS). Figure 5C shows the XPS spectrum of the 14Fr-LbL samples. The survey spectrum reveals the presence of carbon (C), nitrogen (N), oxygen (O), silicon (Si), and silver (Ag). A quantitative estimation of the surface composition was performed based on the intensity of the signals associated to each element, using the sensitivity factors compiled by (Wagner et al., 1981). In this way, the surface concentration of silver was estimated to be 0.4%, consistent with the amount of silver observed in the outermost layer as estimated by RBS.

Figure 5D shows the spectral region corresponding to Ag 3d. Two peaks are observed at 368.3 eV and 374.3 eV, corresponding to the photoemission peaks associated with the Ag 3 d 5 / 2 and 3 d 3 / 2 , respectively. The binding energy of these peaks is consistent with Ag 0 (Ferraria et al., 2012; Firet et al., 2019). In addition to the metallic peaks, two additional bands were detected at 369.5 eV and 376.7 eV. Previous studies have attributed these bands to silver clusters smaller than 4 nm (Calderon V. et al., 2013). However, the intensity of these bands is relatively low, accounting for approximately 15% of the total intensity of the peak associated to silver.

Figure 5E shows the spectral region corresponding to Si 2p. The region is accurately fitted with a single Gaussian-Lorentzian peak centered at 102.7 eV, which is consistent with the formation of Si-O bond (Kaur et al., 2016; Kong et al., 2020).

To evaluate the ciprofloxacin (CFX) delivery performance of the coated samples, experiments were conducted in PBS at 37 °C and pH 7.6 under stirring conditions. A pH of 7.6 is close to physiological conditions and falls within the typical range of urine (pH 4.5–8.0), which can vary depending on hydration levels and overall health status. Both control and coated samples were loaded with highly concentrated solutions of the antibiotic for 12 h and then stored for 24 h (resting stage) prior to use in the experiments. The drug release profiles obtained are presented in Figures 6A,B. The results indicate that pristine samples (14Fr and 20Fr) exhibited significant release of CFX, demonstrating the chemical compatibility between the antibiotic’s structure and the silicone substrates (Hui et al., 2008). In that sense, the released amount of CFX for both control samples was similar, approximately 30 μ g/cm². On the other hand, the coated samples demonstrated a substantial capacity to release higher amounts of CFX, despite having only a micron-thick LbL film on both UCs. Both coated samples, 14Fr-LbL and 20Fr-LbL, released around 70 μ g/cm², more than double the amount released by the control samples. This illustrates the high capacity of the LbL matrix to load and release CFX. Although the long-term durability of the LbL coating under storage conditions was not tested in this study, similar polyelectrolyte-based systems have shown stability over months when stored in dry, dark environments. Future work will evaluate performance degradation over time.

Furthermore, Figures 6C,D illustrate the antibacterial efficacy of the samples against E. coli and S. aureus. For the E. coli strain (Figure 6C), the pristine samples (14Fr and 20Fr) exhibited no antibacterial activity. However, the coated samples 14Fr-LbL and 20Fr-LbL displayed inhibition areas of 10 and 13 mm, respectively. These inhibition zones noted for the LbL-only samples might indicate limited diffusion of silver ions caused by surface-anchored Ag 0 because E. coli, and S. aureus, are generally silver-tolerant up to 1.7–54 mg/L (Gugala et al., 2019; Hochvaldová et al., 2024) by involving the aggregation of AgNPs leading to the formation of black precipitates: E. coli triggers flagellin-mediated AgNPs aggregation, whereas S. aureus accelerates biofilm-induced aggregation (Hochvaldová et al., 2024). Our observation of growth inhibition suggests that the local Ag + concentration at the catheter surface reached, at least transiently, in the 0.05–0.20 ppm (5–20 μ g/L) range since these concentrations fall within the bactericidal window reported for silver-polymer coatings (Jung et al., 2008; Aldakheel et al., 2023). On the other hand, the stronger antibacterial effect observed in the E. coli assay compared to S. aureus can be attributed to the greater efficacy of silver ions against Gram-negative bacteria. This is likely due to the thicker peptidoglycan layer in S. aureus, which may hinder the penetration of silver ions through the bacterial cell wall (Jung et al., 2008). Moreover, due to the loading and release capacity of the control samples for CFX, they exhibited inhibition zones of 35 mm for 14Fr/CFX and 32 mm for 20Fr/CFX. In the case of coated samples loaded with CFX, its enhanced antibacterial activity is chiefly attributable to increased CFX loading and diffusion: 14Fr-LbL/CFX and 20Fr-LbL/CFX achieved inhibition zones of 40 and 35 mm, respectively. It is noteworthy that samples loaded with CFX, both uncoated and coated, exhibited similar inhibition diameters because they released similar amounts of CFX. It is worth noting that both LbL and non-LbL CFX-loaded catheters exceeded the CLSI-defined inhibition zone threshold of ≥ 26 mm for E. coli susceptibility to CFX (Humphries et al., 2019) suggesting that both approaches deliver clinically meaningful antibiotic levels.

Also, it is important to note that an increase in the surface roughness and the transition from a hydrophobic to a hydrophilic surface (as shown in Figures 3A,B) after the LBL procedure may enhance bacterial adhesion. However, this effect could be counteracted by the strong antibacterial properties of silver and CFX, which help prevent early-stage colonization. Nonetheless, further studies are needed to evaluate long-term biofilm formation under dynamic conditions in the urinary tract.

In the case of the S. aureus strain (Figure 6D), pristine samples also showed no antibacterial effect, while coated samples only exhibited an inhibition zone corresponding to the diameter of the UCs: 4.7 mm for 14Fr-LbL and 6.7 mm for 20Fr-LbL. This indicates that the LbL coating solely functioned as an antibacterial surface. However, samples loaded with CFX demonstrated increased antibacterial efficacy: 14Fr/CFX and 20Fr/CFX exhibited inhibition zones of 15 and 10 mm, respectively.Notably, the combined antibacterial activity resulting from the combination of the LbL coating and the CFX antibiotic was significantly enhanced for this strain. The 14Fr-LbL/CFX and 20Fr-LbL/CFX samples demonstrated inhibition zones of 27 mm and 23 mm, respectively. Both values exceed the CLSI threshold of ≥ 21 mm for S. aureus susceptibility to CFX (Din et al., 2022), supporting the potential clinical utility of these coatings, even though S. aureus is not typically a predominant uropathogen. While these results against E. coli and S. aureus are promising, it is important to note that in clinical settings, catheter surfaces are often subject to biofilm formation and encrustation—conditions that can impair antibiotic diffusion and silver efficacy. Biofilm formation and catheter encrustation, often exacerbated by urease-producing bacteria and alkaline urine, can form diffusion barriers that reduce antibiotic release and surface contact activity. Although our dual-action coating may delay biofilm development, future studies should assess release profiles under encrusted or biofilm-laden conditions to more accurately evaluate its performance and long-term clinical relevance.

To gain deeper insights into the mechanisms governing the release of CFX from both uncoated and coated samples, we developed a mathematical model accounting for the drug loading step and the resting stage (Section 2.6). The kinetic parameters obtained from this model are presented in Table 1. Given that the wall thickness of the UCs is 1 mm and the LbL coating is a micron-thick film, we assumed that the diffusivity ( D ) remains constant for both uncoated and coated samples at 1.744 × 1 0 − 5 cm²/h. However, the partition rates ( γ ) listed in Table 1, representing the drug solid–liquid partition coefficient associated with equilibrium, varied depending on the type of solution used: highly concentrated CFX for loading ( γ L ) or the PBS solution used for drug delivery experiments ( γ D ), and the type of UC sample (uncoated or coated). Notably, the values for coated UCs were higher than those for the uncoated controls, indicating a greater affinity of CFX for the LbL film than for pristine samples.

Using the kinetic parameters from Table 1, several simulations were conducted. The Supplementary Figure S1A illustrates the CFX concentration inside the UCs immediately after the 12-h loading period. In all cases, both pristine (14Fr and 20Fr) and coated samples (14Fr-LbL and 20Fr-LbL) exhibited CFX accumulation at the border of the external wall of the UCs. However, after the 24-h rest stage preceding the drug delivery experiments (Supplementary Figure S1B), simulations showed how CFX continues to diffuse through the samples. With these insights, finally, we were able to simulate the experimental data of CFX delivery over the 15-day release period (Figure 7A). Additionally, owing to the characteristics of this mathematical model, we can predict how drug release profiles will differ based on varying drug loading and rest stage times (Supplementary Figure S1C; Figure 7B). Simulations indicated that extending the loading time may not necessarily result in higher CFX loading, as equilibrium is reached at the border of the external UCs wall. However, longer rest stage times enable more controlled release, as CFX becomes more homogeneously distributed throughout the samples.

Figure 8 presents the molecular dynamics simulations. Complex-1 (Figure 8A) shows a stable capacity for CFX loading within PDMS, as reflected by its consistent negative potential energy, approximately −5 E h (Figure 8C), which remains steady over time. In contrast, Complex-2 (Figure 8B), composed of micron-thick LbL layers, exhibits a greater capacity for CFX loading due to strong interactions between CFX and the CMC/CHI-Ag layers. This is illustrated by the notable contraction of the layers at step 500 (Figure 8B), in comparison to the loading of CFX into PDMS at step 2000 (Figure 8A). Furthermore, the observed increase in potential energy from −2 to 3 E h within the first femtoseconds illustrates the relatively lower stability of complex-2. This instability does not endanger the integrity of the structure, but rather, ensures proper biodegradation of the drug delivery system.

This study successfully developed a dual strategy to create an antibacterial surface on commercial Foley silicone urinary catheters (UCs) by combining the contact-killing effect with the controlled release of antimicrobial compounds. Specifically, UCs were coated with an antibacterial drug delivery system using a Layer-by-Layer (LbL) approach that incorporated carboxymethylcellulose (CMC) and chitosan-silver (CHI-Ag) complexes, loaded with ciprofloxacin (CFX) as the model drug. The very thin LbL coating significantly enhanced drug loading and release, achieving over twice the amount of CFX release compared to uncoated catheters. The antibacterial efficacy was markedly improved in the coated samples, particularly against S. aureus than E. coli. Additionally, a mathematical model based on Fick’s second law was proposed to simulate the experimental CFX release profiles, effectively predicting drug release kinetics by incorporating both drug loading and resting stages. Molecular dynamics simulations further confirmed strong compatibility between CFX and the LbL layers, although with relatively low stability. Despite the promising results, the study has certain limitations that should be addressed in future research. First, the antibacterial performance was evaluated only against planktonic forms of E. coli and S. aureus, without directly assessing biofilm formation or eradication on the catheter surface. Given the clinical importance of biofilms in CAUTIs, future studies should focus on quantifying biofilm inhibition and evaluating antimicrobial durability under physiologically dynamic conditions. Second, the release kinetics of silver from the layer-by-layer (LbL) coating were not investigated. Although silver was reduced to its elemental form ( Ag 0 ) and embedded within the matrix, its potential leaching over time could affect long-term antibacterial activity and should be examined in relation to biofilm suppression. Overall, these findings highlight the potential of this coating strategy to enhance the performance and safety of urinary catheters, particularly in reducing the incidence of catheter-associated urinary tract infections (CAUTIs).