The problem of healthcare-associated UTI appears to be deceptively simple: the major extrinsic risk factor is the use of a device that bypasses host defense mechanisms and allows microorganisms to grow in normally sterile body sites. Yet, the pathogenesis is far more complex than is implied by a purely mechanical model, and the epidemiology of the use and complications of urinary catheters is today understood only in its broad outlines.

The relative neglect of this problem by investigators undoubtedly reflects the low importance assigned to UTIs both by clinicians and by infection control programs. Indeed, in 2001, Jarvis (25) reported no epidemics of UTIs among 114 healthcare-associated outbreaks that the Centers for Disease Control and Prevention (CDC) investigated on-site in the previous decade (25). Although this could suggest that the CDC did not elect to participate for various reasons or that individual hospitals simply did not request the CDC’s help, underreporting or failures of surveillance are also likely explanations.

Healthcare-associated UTIs present unique challenges for epidemiologists. Because endemic UTIs occur throughout the hospital and because healthcare-associated epidemics often involve multiple sites of infection, epidemic rates of catheter-associated UTIs may not be readily apparent unless they involve an unusual microbial species (26). Indwelling urinary catheters are used in nearly all hospital nursing units, unlike ventilators and many other devices. For this reason, healthcare-associated UTIs have complex behavioral and social determinants.

Duration of indwelling catheterization is the most important risk factor for the development of catheter-associated infection. Overall, the mean and median durations of catheterization in acute care hospitals are 2 and 4 days, respectively, and catheters are removed within 7 days in nearly 70% of patients (27). Although the prevalence of infection increases steadily with extended durations of catheterization, the daily incidence of newly acquired infection is relatively constant during closed drainage, at least for the first 10 days, with 2% to 16% of previously uninfected patients acquiring infection each day (28,29). Infection becomes nearly universal by 30 days. Nonetheless, this is a dramatic improvement over open drainage systems, for which universal infection followed just 4 days after insertion (30).

Thus, the principal benefit of closed drainage has been to delay, if not prevent, the onset of infection. True prevention begins by avoiding unnecessary catheter use. Catheters that must be used should be removed at the earliest possible time. Unfortunately, epidemiologists have not fully exploited the potential of these simple principles. Studies in many countries suggest that more restrictive policies for catheter use would be beneficial (31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45). Though patterns of use in other countries may differ from those in the United States, investigators in Denmark and Sweden determined that indwelling catheters were used in 13% of patients and 12% of hospital days, respectively, and there was great variation between hospitals for the same type of nursing service (32,33). Catheter usage is most prevalent in ICUs; data from NHSN that examine utilization of urinary catheters indicate that 54% to 90% of all ICU days in medical/surgical intensive care units at major teaching hospitals involve the use of a urinary catheter, in the lowest 10th percentile to the 90th percentile, respectively, of reporting hospitals (18). A study in Israel suggested that patients with intermediate durations of catheterization (7-30 days) and who are catheterized for the indications of obstruction or incontinence are a high-risk group that may benefit most from intervention (35). Such patients had a higher daily risk (8.6%) of acquiring infection even during the early period of catheterization. In Canada, an investigation of overutilization of indwelling urinary catheters in a large tertiary-care hospital found that 20.3% of patients admitted via the emergency room were catheterized upon admission (36). Furthermore, 50% of catheters were inserted for unjustifiable reasons and 60% of those patients who subsequently developed UTI did not meet the study’s criteria for justifiable catheterization. Another study found that 21% of catheterized medical patients did not have any initial indication for placement of the urinary catheter and that continued catheterization was unjustified in 47% of patient-days studied (37).

Other recent studies have produced similar findings. For example, 10.7% of patients on a medical service had an indwelling urinary catheter inserted within the first 24 hours, with 91% having been placed in the emergency room and 38% deemed inappropriate (38). Another study
of emergency department patients found appropriate indication for catheter placement correlated with documentation of a physician order in the chart; only 52% were deemed appropriate in those without orders versus 73% to 82% with orders (46). In a point prevalence study from Spain, only 22% of patients had a correct indication with adequate drainage systems (39).

Lack of awareness of the presence of an indwelling catheter is a further problem. One group found that physicians were not aware of the catheter status in 28% of their patients and in as many as 41% of those whose catheterization was judged inappropriate (40). Thus, physicians appear to discount the importance of the urinary catheter, leading to overuse and misuse, for example, for inappropriate indications such as nursing staff convenience. In a study of patients with urinary incontinence, 37.5% were catheterized even though 55.5% of these were previously incontinent before admission to hospital and had managed this problem by other noninvasive methods. The decision to catheterize was made by physicians in 31.7% and by nursing staff in 37.3% (41). Recent studies directed at prevention of unnecessary catheter use have shown some success of nursing education as a means to reduce the use of urinary catheters. Reinforcing strict adherence to approved indications and use of alternatives to bladder catheterization in non-ICU settings were associated with decreased overall incidence of healthcare-associated UTI through reduction in catheter use. CA-UTI incidence in those with bladder catheters did not change (44). Less comprehensive approaches to reduce duration of urinary catheter use, such as simple reminder systems, have been shown to reduce the median duration of catheter use in an ICU setting from 5 to 4 days (47), without an effect on infection rates. However, a recent meta-analysis of stop orders and reminders for removal of urinary catheters concluded that both approaches appear to reduce rates of infection (48).

All the above studies emphasize that catheter use is frequently inappropriate; inattention to both the proper indications for catheter use and the catheter status in patients appears to be an important factor. Potential solutions include the implementation of hospital-wide protocols for catheter insertion and continued usage, such as allowing removal of a catheter by a nurse without a physician’s order, and systems for computer-based order entry of indwelling catheters (42,43).

Incidence and Costs From a broad epidemiologic perspective, the problem of catheter-associated infections acquires force from the magnitude of the population affected. Each year, 3 to 6 million of the 33 million patients admitted to acute care hospitals in the United States receive indwelling catheters. It has been estimated that about 15% to 25% of patients in general hospitals have a catheter inserted sometime during their stay (49), and that the prevalence of urinary catheters has increased over recent decades (50). The problem encompasses many different medical specialties, local practice patterns, and geographical differences. For example, in a French urology department, 52.4% of the patients received indwelling catheters and the incidence of catheter-related UTI was 13% (51). By contrast, in a pediatric population, healthcare-associated UTI was the fifth most common healthcare-associated infection and only 50% of patients with healthcare-associated UTIs had urethral instrumentation (52). According to an older estimate performed by the CDC, there were 2.39 healthcareassociated UTIs per 100 hospital admissions in 1975 to 1976 (53). Recent NHSN reports provide data on both catheterassociated UTI and urinary catheter utilization rates for a variety of patient settings. For the category of medical/surgical inpatient wards, the largest single category, the pooled mean data suggest that urinary catheters were utilized in 22% of patient days and infection developed in 5.9 patients per 1,000 catheter days (18). In 1992, the CDC estimated that more than 900,000 healthcare-associated UTIs occurred in the United States, and that the resulting extra charges exceeded $600 million (54). This represented nearly 14% of the total charges for healthcare-associated infections, estimated to be $4.5 billion.

These figures, however, may markedly understate the actual costs of UTIs, since they are based on decades-old estimates of an expected increased length of stay of only 1 day and extra charges of $680 (1992 dollars) for each UTI. Moreover, charges reflect cost shifting, and therefore are an inaccurate measure of true costs. Using attribution methods in a case-referent study of true costs at the Salt Lake City LDS Hospital from 1990 to 1992, the mean attributable difference in length of stay for patients with healthcareassociated UTIs was 3.8 days, and the mean increase in hospital costs was $3,803 (55). If this is representative of all US hospitals, the true national cost of healthcare-associated UTIs is likely to be more than $3 billion.

In the managed care environment, costs are a financial loss to healthcare institutions. Thus, market forces should revive interest in preventing all healthcare-associated infections, including UTIs. Based on a theoretical model, the extra costs were estimated for each symptomatic healthcare-associated UTI to be at least $676 and for each catheter-associated bacteremia to be $2,836 (56). Another study estimated the mean costs of a healthcareassociated UTI to be $589, with the lowest costs associated with infections caused by Escherichia coli and higher costs with infections caused by other gram-negative bacilli and yeasts (57). The substantially lower estimates in this study as compared to other earlier retrospective studies were attributed to cost-containment measures implemented in the era of managed care as well as to the availability of newer oral antimicrobials with activity against gram-negative pathogens. Additional economic incentive for hospitals in the United States to reduce rates of healthcare-associated UTI due to urinary catheter use has recently come in the form of a change in reimbursement policy from the federal government (58). Under revised policy, additional payments for certain complications from medical care that are deemed “reasonably preventable” will be curtailed (59).

Mortality The extent of mortality attributed to catheterassociated UTIs is still uncertain since these infections might be effect modifiers or simply markers of high mortality from other causes. The most generally acknowledged cause of death is related to bacteremia, which occurs in 0.3% to 3.9% of patients with healthcare-associated
catheter-associated UTIs (60, 61, 62). Secondary bacteremia from a urinary source is generally considered unequivocal evidence of an invasive UTI. However, when a blood culture was obtained immediately after urethral catheterization from patients with sterile bladder urine, 6.5% were positive (63). Therefore, transient bacteremia secondary to urinary tract instrumentation can be a source of a remote infection, perhaps at the site of an implanted prosthetic device.

As many as 35,000 cases of bacteremia secondary to healthcare-associated catheter-associated UTIs occur each year in the United States. Even though the crude case-fatality rate perhaps exceeds 30%, the mortality rate attributed specifically to bacteremic UTI in one large retrospective study was 12.7% (61). According to this estimate of the attributable mortality, as many as 4,500 deaths occur in the United States each year from healthcare-associated UTIs, but most of these deaths may occur in patients with serious underlying disease processes.

The true mortality rate from bacteremic UTIs for the United States in recent years is also undoubtedly lower than such extrapolations from studies in large tertiary care hospitals. In 1992, the CDC estimated that UTIs directly caused only 932 of the 19,027 deaths from healthcare-associated infections but contributed to an additional 6,500 of 58,092 deaths associated with healthcare-associated infections in US hospitals (54). To appreciate recent advances, consider that, before closed drainage systems were used, Martin et al. (64) estimated that 31,000 deaths occurred in US hospitals each year because of urinary catheter-related bacteremia. This study serves as the principal evidence that closed drainage markedly lowered the mortality rate and suggests that further reductions will be achieved only with great difficulty.

Additional mortality may nonetheless occur from causes unrelated to bacteremia. One study, using logistic regression analysis in a large hospital population, suggested that the actual mortality rate of healthcare-associated UTIs is significantly higher than estimates based on the incidence of bacteremia (65). Acquisition of catheter-related UTI predicted a nearly threefold increase in mortality that was not completely explained by clinical sepsis, documented bacteremia, or underlying disease. If this study is representative of all US hospitals, the actual excess mortality associated with catheter-related infections could be as high as 56,000 deaths per year in acute care hospitals. Possible support for this conclusion also came from observations of women with long-term catheters in whom the incidence of death during fevers of suspected urinary origin was 60 times the incidence during afebrile periods (66). A more recent retrospective study of over 25,000 patients with indwelling urinary catheters, and at least 4 days of hospitalization, documented a relative risk, by multivariate logistic regression modeling, of 1.37 for death among those patients who developed UTI (67).

Morbidity Indwelling urinary catheters pose a risk for many infective and noninfective complications. Catheterrelated infection can spread to any site in the urinary tract and can predispose patients to perinephric, vesical, and urethral abscesses as well as epididymitis, prostatitis, orchitis, and vesicoureteral reflux. The overall incidence of these complications is unknown, although 20% to 30% of patients with asymptomatic catheter-induced UTIs may develop local or systemic symptoms (56,60).

Infection may also have a role in other complications of catheterization. Bladder and renal stones, hemorrhagic pseudopolyps of the bladder (68,69), and squamous metaplasia and carcinoma of the bladder (70) have all been associated with UTIs in patients with long-term or chronic indwelling catheters. Accidental inflation of the catheter balloon in the posterior urethra has caused minor hematuria and subsequent urethral stricture as well as periurethral abscesses, sepsis, and death (71). Neglect of long-term catheters, usually in patients who are discharged from the hospital with an indwelling catheter, can lead to bladder gangrene, perforation, and peritonitis (72, 73, 74). Among noninfectious complications of indwelling urethral catheterization, some cardiovascular surgery units have reported that urethral ischemia during cardiopulmonary bypass caused urethral strictures that could be prevented by the use of silicone rather than latex catheters (75).

Healthcare-associated UTIs may also be a source for other healthcare-associated infections. In one large study, 40% of UTIs occurred in patients with multiple healthcareassociated infections, but the incidence of autoinfection secondary to the urinary site was not evaluated (76). In a study of patients in a university hospital in Spain, an indwelling urinary catheter used for more than 3 days more than doubled the risk of developing bacteremia (77). A more recent study from Spain investigating bacteremia occurring within 30 days of solid organ transplantation found UTI was the most common identifiable cause, in 27% of cases, with E. coli being the most common bacteria isolated (78). Healthcare-associated UTIs can be the source for 10% to 15% of healthcare-associated bloodstream infections (61,62,79). A recent study identified 350 cases of healthcare-associated UTI-related bloodstream infection over a 9-year period at a large academic medical center, predominantly among patients with immunosuppression, liver, or kidney disease (80). Healthcare-associated Staphyloccocus aureus UTI among residents of a longterm care facility was found to be associated with bacteremia in 13%; among the cohort 82% had recent urinary catheterization (81).

Surgical site infection secondary to a healthcare-associated UTI has been documented as a cause of major morbidity with an attack rate of 2.3 secondary surgical site infections per 100 surgical patients with healthcare-associated UTIs (82). Two reports confirm an increased rate of surgical site infections and allograft dysfunction in renal transplant recipients with healthcare-associated UTIs (83,84), whereas others have demonstrated associations between UTIs and infections of prosthetic heart valves (85,86), total hip replacements (87,88), and central venous catheters (89). Rare complications such as gramnegative endocarditis and septic discitis may also complicate urosepsis of healthcare-associated origin (90,91).

Consequences of Antimicrobial Use The indication for antibiotic therapy of healthcare-associated UTIs in acute care settings is a subject of debate and controversy. Nonetheless, treatment of symptomatic UTIs is virtually universal. In one report, among 1,233 patients with healthcare-associated UTIs, only a single patient
was not treated (92). Yet, routine therapy increases not only drug costs but also adverse drug reactions and the emergence of antibiotic-resistant microorganisms. These adverse consequences have not been fully evaluated in epidemiologic studies, although antibiotic use during catheterization influences the patterns of microbial species causing healthcare-associated UTIs. The changing nature of UTIs at one medical center in the last decade was reflected by significant increases in the proportion of certain uropathogens such as yeasts, Klebsiella pneumoniae, and group B streptococcus (93). Antibiotic use was probably largely responsible for these changes. Other reports, such as one that tied the emergence of multidrug-resistant K. pneumoniae to prophylactic use of trimethoprim-sulfamethoxazole in patients with indwelling catheters (94), serve as further evidence that antibiotic use shapes the character of healthcare-associated UTIs.

In the report from the NNIS system with data from 1992 to 1997 for healthcare-associated infections in medical ICUs, fungi accounted for almost 40% of urinary isolates (34). Candida albicans alone accounted for 21% and was the single most frequent microorganism cultured. This was a marked change from a previous report with results from 1986 to 1989 that included all types of ICUs in which all fungi constituted 22.1% and C. albicans 12.8%. Extensive use of broad-spectrum antibiotics and antifungal drugs may have contributed to this increase, especially for the increasing prevalence of non-albicans Candida.

Epidemics of Healthcare-Associated UTIs Epidemics of healthcare-associated UTIs have garnered national attention when the causative microorganisms displayed unusually high levels of antibiotic resistance. In seven large epidemics investigated by the CDC between 1970 and 1975, asymptomatic catheter-associated UTIs were reservoirs of the epidemic microorganisms (95). The most frequently observed risk factor in these epidemics was prior exposure to broad-spectrum antibiotic therapy.

Only three microorganisms caused the seven outbreaks: K. pneumoniae, Serratia marcescens, and Proteus rettgeri. Gastrointestinal carriage was especially prominent in outbreaks caused by K. pneumoniae, but the epidemic microorganism was thought to be transmitted from patient to patient on the hands of healthcare workers in all seven outbreaks. Healthcare-associated UTIs were also sources for other healthcare-associated infections as the epidemic microorganism was isolated repeatedly from nonurinary sites in five of the outbreaks.

Indwelling urinary catheters and other types of urologic instrumentation have contributed to the emergence of healthcare-associated pathogens highly resistant to antimicrobial agents. The urinary drainage bag is a potential site for extraintestinal transfer of resistance plasmids in Enterobacteriaceae as well as an environmental reservoir for cross-infection (96). For instance, interhospital spread of multiply resistant S. marcescens occurred among patients with indwelling catheters in four geographically separate hospitals in one city (97). Hand carriage by personnel rotating among hospitals was the apparent mode of transmission. Indirect contact transmission of highly resistant P. rettgeri appeared to be important in two reported outbreaks of healthcare-associated UTIs (98,99).

Contaminated equipment and inadequate disinfectants have also been responsible for epidemics of UTIs. An outbreak of gentamicin-resistant P. rettgeri and Providencia stuartii UTIs in patients with chronic indwelling catheters in a rehabilitation unit was caused by contaminated urinary leg bags (100). In another hospital, a contaminated drainage pan in a cystoscopy room caused a common-source outbreak of 105 cases of multiple antibiotic-resistant S. marcescens UTIs following cystoscopy, and cross-infection of 29 patients on nursing units amplified the magnitude of the epidemic (101). At yet another hospital, inadequate disinfection of urologic instruments with reuse of 2% glutaraldehyde led to a 12-month-long epidemic of antibiotic-resistant S. marcescens UTIs after a variety of urologic procedures (102). The use of chlorhexidine for hand washing caused an outbreak due to multiply antibiotic-resistant and chlorhexidine-resistant S. marcescens UTIs that lasted over 19 months (103), and use of hexachlorophene solution in preparing patients and cleaning instruments for cystoscopy and transurethral resection of the prostate was associated with Pseudomonas aeruginosa UTIs (104). Contaminated urine measuring containers and urometers were the reservoir for P. aeruginosa that caused 66 catheter-associated UTIs (105). Clearly, rigorous application of existing infection control principles can prevent such epidemics.

Many of these and other reported epidemics had welldefined sources. Others occurred from previously unsuspected environmental reservoirs. For example, uninfected patients with condom catheters who had contaminated urine drainage bags served as a reservoir for infection of patients with indwelling catheters on the same hospital unit (106). Contaminated drainage bags may also mislead surveillance personnel, as false diagnoses of UTIs made from urine specimens obtained from drainage bags can skew surveillance data. Such errors at one hospital led to a pseudoepidemic of Trichosporon beigelii UTIs that, if not recognized, could have subjected patients to the risks of antifungal treatment (107) (see also Chapter 9).

Microbial colonization of bladder urine precedes most invasive UTIs. Urine is an excellent growth medium for common urinary tract pathogens (130,131). Nonetheless, the urinary tract above the distal urethra is normally free of bacteria, and micturition permits nearly complete cyclic emptying of the bladder, thereby rapidly eliminating the small numbers of microorganisms introduced through minor urethral trauma (132). The indwelling transurethral catheter breaches this normal defense mechanism, distending the urethra, and blocking the ducts of the periurethral glands. The retention balloon prevents the complete emptying of the bladder and creates a small pool of residual urine in which microorganisms can multiply. The resulting increase in susceptibility to infection is shown by observations that low-level bacteriuria progresses very rapidly to levels exceeding 100,000 colony-forming units (CFUs)/mL when any microorganisms appear in the catheterized bladder (28,133).

Since the catheter is a continuously open channel, microorganisms can migrate upstream into the bladder through the lumen of the catheter. As long ago as 1957, Dutton and Ralston (134) showed that nonmotile bacteria could ascend sterile tubing against a flow of sterile urine. In addition, the external surface of the catheter stresses the urethral surface, creating a channel for bacterial colonization and entry outside the catheter (24). It has been reemphasized in recent years that the urethra is not merely a passive conduit but has its own complex defense mechanism (135). Exfoliation of urethral cells with bound uropathogens is one example of an overlooked defense mechanism, and differences in the rates of exfoliation of cells in menstruating women or those on hormone replacement as compared with postmenopausal individuals may account for the different rates of UTIs in these populations. The effect of a foreign body on the rate of exfoliation of urethral cells and its contribution to bacteriuria has not been well defined.

The foreign material of the catheter also may promote infection by a number of other mechanisms. For instance, by blunting the local inflammatory response as shown in other types of implanted foreign bodies (136), the catheter may interfere with the removal of bacteria that gain entry to the bladder. In mouse models, Toll-like receptor 4 (TLR4) on both bladder epithelial cells and leukocytes protects against E. coli infection by recruiting inflammatory cells and upregulating chemokine expression needed for an innate immune response (137,138). Although the effect of the urinary catheter on the local innate immune response has not been well studied, a dysfunctional TLR4 may hinder inflammation and bacterial clearance from the urinary tract (138). The possibility that the innate immune response is blunted in catheter-associated UTIs is suggested by observations that, in humans, healthcare-associated UTIs are seldom symptomatic (139) and the sensitivity of detecting catheter-associated UTIs by screening for pyuria is only 37%
(140). Pyuria is also less frequently associated with yeast and gram-positive microorganisms than with gram-negative microorganisms colonizing the catheterized urinary tract.

Recently, the function of defensins, specifically human β-defensin 1, which is produced by renal epithelial cells, and their role in UTI and pyelonephritis have been studied more closely. β-Defensin 1 has activity against E. coli, although the concentration required is tenfold higher than is present in the urine. However, it is also possible that defensins could exist on epithelial cells to form an antimicrobial barrier, a process that may be affected by or compromised by the presence of a urinary catheter (141). Studies have shown that the presence of a catheter may enhance the adherence of gram-negative bacteria to uroepithelial cells. For unknown reasons, 2 to 4 days before the onset of bacteriuria, epithelial cells harvested from the catheterized bladder show a transient increase in the adherence of gram-negative bacteria (142).

Bacteria may also adhere to and migrate along the extraluminal surface of the catheter itself. The physical and chemical properties of the catheter material, therefore, are posited as important determinants of UTI (143). In consequence, efforts have been made to develop an adhesionresistant or colonization-resistant urinary catheter. An in vitro study found marked differences in the ability of various gram-positive and gram-negative bacteria to attach to red rubber catheters and those coated with either a hydrophilic substance, silicone, or tetrafluoroethylene (Teflon) (144). Most bacteria are hydrophobic, and none of the tested bacteria adhered to the hydrophilic catheter. However, studies of hydrophilic catheters in patients have demonstrated no clinical benefit (145, 146, 147). Regardless of their influence on bacterial adhesion, catheters made of Tefloncoated latex or silicone have been introduced for clinical use with the hope of improved biocompatibility, but in the absence of established infection there is little evidence of less irritation and inflammation in the urethra from these catheters than from those made of latex rubber (148,149).

Bacteria that colonize both the external and the internal surfaces of urinary catheters grow in microcolonies within a biofilm that encases the bacterial cells. When urine cultures reveal a single species, the biofilm often contains a mixed community with up to four species. In recent years, there has been greater understanding of the biology of biofilms, with additional insight into the process by which planktonic or freely suspended microorganisms become surface-associated microorganisms or biofilms (150, 151, 152). A biofilm is loosely defined as a collection of microbial cells that is stably associated with a surface and enclosed in a matrix of primary polysaccharide material. The microbial cells in a biofilm are also different from their freely suspended counterparts with respect to gene transcription and growth. The initial step in the process is the formation of a conditioning film by the urine on the catheter surface followed by attachment of the microbial cell to the surface of the urinary catheter or substratum. The surface properties of the catheter appear to play a role, with microorganisms more rapidly attaching to hydrophobic surfaces like Teflon and plastics than to hydrophilic substances like glass (153). It is thought that hydrophobic interactions occur between the cell and the catheter surface and overcome local repulsive forces, thereby enabling attachment. In vitro studies have also shown that the nutrient content of the aqueous medium, in this case urine, also affects the number of bacterial cells that attach to the surface. In certain bacteria, differences in bacterial surface hydrophobicity and presence of fimbriae (pili) and flagella also influence the rate and amount of attachment.

The next step in the formation of a biofilm is a change in gene expression by attached cells upon initial adherence (154). Most of these changes are needed for adaptation to living in a new environment and the change from a planktonic to a surface-associated form. After attachment, the bacterial cells produce extracellular polymeric substances (EPS), which will account for 50% to 90% of the total mass of the biofilm (155). The EPS is composed mainly of polysaccharides and is highly hydrated. The overall charge and composition of the polysaccharides and the amount of EPS produced can vary between different microorganisms and may contribute to antimicrobial resistance by decreasing diffusion of antibiotics through the EPS (156). Genetic variation in the pathways controlling polysaccharide biosynthesis has been shown to result in structural differences in biofilm produced by P. aeruginosa (157).