- Email: [email protected]
Prevention of urinary tract infections
Prevetltion of Urinary Tract Infections
WALTER E. STAMM, M.D. Seattle, Washington
From the Department of Medicine, Harborview Medical Center, 325 9th Avenue, Seattle, Washington. Requests for reprints should be tiddressed to Dr. Walter E. Stamm, Department of Medicine, Harborview Medical Center, University of Washington School of Medicine, Seattle, Washington 98104.
148
May 15, 1994
Potential candidates for measures designed to prevent urinary tract infections include (1) women (or girls) with frequent exogenous reinfections; (2) pregnant women; (3) hospitalized patients with short-term indwelling catheters; (4) men with chronic bacterial prostatitis; (5) patients with neurogenic bladders being managed with either intermittent catheterization or chronic indwelling catheters; and (6) men undergoing urologic surgery. Patients in the first three categories represent the vast majority of men and women with preventable urinary tract infections. On the basis of our current understanding of the pathogenesis of urinary infections in women, potential preventive measures can be envisioned. To date, the most successful methods in susceptible women include continuous low-dose antimicrobial prophylaxis or postintercourse antibiotic prophylaxis. In controlled clinical trials, the efficacy of low-dose trimethoprim-sulfamethoxazole, trimethoprim, or nitrofurantoin has repeatedly been demonstrated, and the cost effectiveness of this approach has also been established. An alternative strategy, intermittent, self-administered, single-dose antimicrobial therapy, may be useful in selected circumstances. Screening of high-risk patients for asymptomatic infections has largely been abandoned except in pregnant women in whom this practice remains an important preventive measure. Exciting newer approaches to prevention based on studies clarifying the fimbriai structures mediating bacterial adherence to receptors on uroepithelial ceils include the use of receptor analogs and immunoprophyiaxis, but these approaches are not yet of demonstrated clinical efficacy. Since urinary tract infections are among the most common bacterial infections, their prevention assumes considerable importance for both economic and public health reasons. An estimated six million visits annually result from these infections in the outpatient setting [ 11, most being in women between the ages of 15 and 35 years. In addition, 5 percent of pregnant women experience bacterial urinary tract infections [2] and these infections produce both maternal and infant morbidity [3]. Ten to 15 percent of hospitalized patients undergo urinary catheterization and incur an approximate 25 percent risk of urinary tract infection, thus resulting in about 600,000 episodes of catheter-associated bacteriuria annually [4]. These three major risk groups (Table I) account for most urinary tract infections, and therefore these three populations are the most important potential candidates for preventive measures. Other high-risk groups include men with chronic bacterial prostatitis, many of whom have recurrent episodes of urinary tract infection [5], men undergoing urologic surgery [6], and men and women with neurogenic bladders managed either by intermittent cateterization or chronic indwelling catheters [7). The
The American Journal of Medicine
PATIENT AT RISK AND ANTIMICROBIALTHERAPY-~TAMM
latter groups often experience complicated and difficult to treat urinary infections, but the patients involved are far fewer in number than the first three populations mentioned. As with other sites of infection, the goals of preventing infection in the urinary tract include reducing morbidity attributable to acute infectious episodes, preventing serious infectious or noninfectious complications of either single or repeated infections, and curtailing health care costs required for diagnosis and treatment. URINARY TRACT INFECTIONS WOMEN
IN NONPREGNANT
Eighty percent of recurrent urinary tract infections in women result from exogenous infection with a newly acquired strain [8,9]. In most instances, the sequence of events resulting in urinary tract infection begins with transperineal migration of infecting Escherichia coli strains from the rectum to the vagina with establishment of introital and urethral colonization. Eventual inoculation of the bladder urine or the posterior urethra with these colonizing organisms then leads to acute lower urinary tract infection (urethritis or cystitis, or both) [ 10,111. Ascending infection of the kidney (acute pyelonephritis) may subsequently occur. Although sexual intercourse produces transient bacteriuria in most women, and although epidemiologic studies clearly associate sexual activity with an increased risk of urinary tract infection, the role of sexual activity in the pathogenesis of these infections has not been universally accepted [ 12- 141. However, two recent studies provide strong support for the argument that sexual intercourse plays an important permissive role in facilitating inoculation of the bladder urine after development of urethral colonization. Nicolle et al [ 151 found that 80 percent of urinary infections have their onset within 24 hours of intercourse. Pfau et al [ 161 reported similar results and showed that postintercourse antibiotics could markedly reduce the episodes of cystitis in such patients. On the basis of this understanding of the pathogenesis of urinary tract infections, several possible pre-
GutFlora
f I I
Antibiotics
Figure 1.
___
-)
Introit01 Colonization
Urethral Colonization
-
$ I I
L-
ventive measures can be envisioned that might interrupt the chain of events culminating in infection (Figure 1). Antimicrobial agents could be used to reduce gastrointestinal colonization with aerobic gram-negative bacteria, thus eliminating the reservoir for infecting strains. Similarly, either systemic or topical antibiotics could be used to prevent the establishment of introital or urethral colonization with rectal E. coli. Antimicrobial agents might prevent the development of such colonization either through direct antibacterial activity or by inhibition of adherence, a property demonstrated to result from subinhibitory levels of ampicillin and nitrofurantoin [ 17,181. Alternatively, preventive measures could be directed at microorganism-specific factors that may enhance the risk of infection. As outlined in the preceding report, uropathogenic E. coli frequently bear specific surface pili that mediate binding with corresponding receptor sites on host cells [ 191. Thus, one potential intervention aimed at reducing introital and urethral colonization could use receptor analogs to inhibit attachment of these uropathogenic E. coli by competitively saturating receptor binding sites. Purified P fimbriae and glycosphyngolipids block attachment of P fimbria bearing E. coli to uroepithelial cells in in vitro adherence assays, whereas d-mannose blocks attachment attributable to Type I pili both in vitro and animal models [20,21]. Alternatively, antibodies directed against fimbriae that mediate attachment can block adherence in in vitro assays and protect against ascending pyelonephritis in a mouse model [ 191. These
Potential Candidates for Prophylaxis
TABLE I
1. Women (girls) with frequent reinfections 2. Pregnant women 3. Indwelling catheter (short-term) 4. Men with chronic bacterial prostatitis 5. Neurogenic bladder Intermittent catheterization Chronic indwelling catheter 6. Urologic
surgery
Bladder Inoculation
T
4
Cystitis
*
__)
Pyelonephritis
ic Antibiotics Receptor Analogues Vaccine
Potential preventive
- J
intervention
I I
I
I Abstinence Postcoital Voiding Postcoital Antibiotics Contraception Behavioral (?)
!
I I :
I Antibiotics Vaccine
l
in women with recurrent urinary tract infections
May 15, 1984
The American
Journal
of Medicine
149
PATIENT AT RISK AND ANTIMICROBIAL
THERAPY-STAMM
data provide a basis for the belief that immunoprophytaxis may be an exciting possibility in the near future. Additional interventions can be envisioned after urethral colonization has been established but before bladder inoculation has occurred. Because sexual intercourse facilitates the transfer of organisms from the urethra to the bladder in many women, postcoital antimicrobial agents have been studied and found effective in preventing infection [ 16,221. Abstinence could achieve similar results. In addition, postcoital voiding may eliminate organisms after their entry into the bladder urine but before active infection is established. In one case-control study, voluntary deferral of micturition was associated with recurrent urinary tract infections [23]. No specific contraceptive practice has been linked directly to urinary tract infection, although the diaphragm has been associated with increased episodes of cystitis in our patients. Lastly, additional interventions can be envisioned that would exert their effect after organisms are introduced into the bladder urine but before infection is established. These might include (1) low concentrations of antimicrobial agents in the urine, which either would kill bacteria on entry into the bladder or would alter these bacteria such that their attachment to uroepithelial cells would be inhibited, or (2) a vaccine that probably would work through the same antiattachment mechanism just outlined. Although many theoretic preventive strategies can be cited, only two approaches to the prevention of recurrent urinary tract infections in highly susceptible women have been effective in clinical trials: continuous low-dose antimicrobial prophylaxis or postintercourse antibiotic prophylaxis. Controlled clinical trials have evaluated the effectiveness of antimicrobial agents in preventing recurrent urinary tract infection in adult women [24-301. These studies show that low-dose antimicrobial agents given either daily or thrice weekly (Table II) can reduce the rate of infection in susceptible women from two to four per year to nearly zero per
TABLE II
Prophylactic Antibiotic Regimens for Women
Drug TMP-SMX TMP Nitrofurantoin Sulfa Methenamine
Dose 40 mg/200 mg 100 mg 50 mg/lOO mg 500 mg 500 mg four times a day
TMP = trimethoprim;
150
Infections per Patient Year o-o.15 o-o.15 0.1-0.8 1-2 0.8-1.6
FecalPerineal Effect + + -
SMX = sulfamethoxazole.
May 15, 1984
The American Journal of Medicine
year. Trimethoprim-sulfamethoxazole, given as 40 mg/200 mg daily, has been the most effective agent in many studies, usually reducing infection to the range of zero to 0.15 per patient year. Trimethoprim-sulfamethoxazole may achieve this effect either by direct antimicrobial action in urine or through the nearly complete elimination of aerobic gram-negative bacteria from the fecal and perineal reservoir that it produces in most patients. Trimethoprim alone has had similar effects on both the fecal-perineal flora and the rate of recurrent infection when it is used prophylactically. Both agents have been continuously effective and relatively free of side effects when used as long as two to four years. Emergence of antimicrobial resistance to trimethoprim-sulfamethoxazole or trimethoprim in E. coli or Klebsiella has not been noted among either urinary or fecal isolates in prophylaxis trials to date. However, increasing trimethoprim resistance has been noted in urine isolates obtained from outpatients in some areas [ 3 1,321. We encountered no trimethoprim-resistant E. coli infections among 30 women receiving trimethoprim or trimethoprim-sulfamethoxazole in a 1975 to 1977 prophylaxis trial, whereas 3 of 30 women had such infections in a 1981 to 1983 study. Nitrofurantoin or nitrofurantoin macrocrystals have also been used successfully to prevent recurrent urinary tract infections. These agents generally have not reduced the rate of infection as dramatically as has either trimethoprim-sulfamethoxazole or trimethoprim alone. However, they are highly effective as compared with placebo or no antimicrobial agents. Because nitrofurantoin has little effect on either the fecal or perineal aerobic gram-negative flora, its effectiveness presumably results from direct antimicrobial activity in urine. Recent reports of frequent side effects in women receiving low-dose continuous nitrofurantoin, including interstitial pneumonia, allergic pneumonitis, hepatitis, and other complications, necessitate careful monitoring and raise concern about its long-term use [33]. Both sulfonamide and methenamine salts plus ascorbic acid reduce the rate of urinary tract infections in susceptible women, but neither can be widely recommended as an effective prophylactic regimen. Sulfonamides, when used on a long-term basis, rapidly promote the development of resistant organisms in the gastrointestinal flora and subsequent superinfection with these resistant strains. Methenamine mandelate (given with vitamin C to acidify the urine) prevents infection, but not to the extent achieved by the other drugs just discussed. In addition, most women can not comply with a regimen requiring four pills four times daily for long periods of time. Naladixic acid reduced infection rates in some prophylactic trials but was not widely used because of concerns about rapid selection of resistant strains
PATIENT
[ 24-261.
Cinoxacin,
a recently
released
derivative, has been used for prophylaxis studies [ 34-361. A significant reduction
quinolone
in three recent in the infection
rate was found in all three trials, but cinoxacin-resistant strains were found in three of 17 women in one study. This agent has an intermediate effect on fecal and perineal aerobic gram-negative flora, having more activity than nitrofurantoin but much less than that of trimethoprim. An alternative strategy to continuous low-dose antimicrobial prophylaxis is postintercourse antimicrobial prophylaxis. Vosti [22] first demonstrated that nitrofurantoin macrocrystals given after coitus prevented recurrent urinary tract infection. More recently, Pfau et al [ 161 have shown that trimethoprim-sulfamethoxazole, naladixic acid, nitrofurantoin, and sulfonamides were all effective in preventing recurrent urinary tract infections when given to young, sexually active women whose infections occurred postcoitally. This management strategy deserves further study, but appears effective when given to women whose infections are often related to intercourse. Compared with continuous low-dose prophylaxis, this approach may reduce medication costs and side effects and lessen the emergence of resistant strains. Practical clinical considerations regarding the use of antimicrobial prophylaxis include the following: (1) In whom should it be used and (2) for how long? The use of prophylactic antibiotics appears warranted from a cost-effectiveness standpoint when women experience more than two infections per year [ 371. Women with three or more infections per year generally prefer prophylaxis to repeated symptomatic episodes and the inconveniences and costs they necessitate. However, antimicrobial prophylaxis only works as long as active drug is given, and does not appear to influence the natural history of infection in a given patient [25]. Knowledge of a woman’s baseline rate of infection in preceding years serves as the best predictor of the likelihood of infection in subsequent years [38]. However, infections may also cluster, and some women move from high-risk to low-risk periods and vice versa 139). The use of prophylaxis for six months followed by reassessment during no prophylaxis seems a reasonable approach in women not known to have baseline infection rates of three per year or more. Postcoital prophylaxis is a useful approach in women whose episodes often follow intercourse. Although not truly preventative, one additional strategy for management of recurrent urinary tract infections deserves mention. Because single-dose antimicrobial agents cure most urinary tract infections, self-diagnosis of infection followed by self-treatment using single-dose therapy could be used in women who have sporadic infections. In a recent study, we found
AT RISK
that (1) women infections,
AND
could
(2) intermittent
ANTIMICROBiAL
accurately
THERAPY
self-diagnose
self-administered
STAMM
their
medication
cured 85 percent of these recurrent urinary tract infections, and (3) this regimen was cost effective when compared with that of continuous low-dose antimicrobial prophylaxis [ 40 1. Intermittent self-medication may be of particular value in women with sporadic infections, (that is, one or two per year) in whom costs, the amount of medication required, and side effects would be reduced. PREVENTION OF URINARY INFECTIONS DURING PREGNANCY Urinary tract infections, which occur in 5 percent of pregnant women, have special significance in that pyelonephritis more often results because of physiologic changes in the urinary tract during pregnancy (decreased bladder and ureteral tone, decreased ureteral peristalsis, hydroureter, and increased residual bladder urine). In addition, urinary tract infections in pregnancy have been associated with prematurity ( 31. Acute pyelonephritis develops in 20 to 30 percent of women with untreated bacteriuria during pregnancy, and of these women 20 percent have premature infants [ 31. The relation of asymptomatic bacteriuria of pregnancy to prematurity remains controversial, but on balance the evidence suggests that asymptomatic bacteriuria of pregnancy is related to prematurity, especially in women with renal infection [ 4 1 1. Thus, the prevention of urinary tract infections in pregnant women reduces morbidity not only in the woman but also in the fetus. For this reason, all pregnant women should be screened for bacteriuria during the first trimester and once again during the third trimester (Figure 2). First trimester screening and treatment of bacteriuric women will reduce the expected cases of acute pyelonephritis from 30 to three per 1,000 pregnant women, and screening during the third trimester will prevent another three to five cases. Women with asymptomatic bacteriuria of pregnancy should have the diagnosis established by at least two positive urine cultures, after which treatment should be initiated I3,41 1. After successful treatment, posttreatment cultures should be obtained on a monthly basis until delivery. After treatment of a urinary tract infection during pregnancy, continuous low-dose antimicrobial prophylaxis for the remainder of pregnancy has been advocated by some, but this practice did not appear more beneficial than careful follow-up in a recent study [ 42 I. PREVENTION OF INFECTION IN PATIENTS URINARY
WITH
CATHETERS
Catheter-associated urinary tract infection remains the most common hospital-associated infection and is the most frequent source of bacteremia in hospitalized
May 15, 1984
The American
Journal
of Medicine
151
PATIENT AT RISK AND ANTIMICROBIAL
THERAPY-STAMM
50 Bacteriuric
Treated
)
, ApN
1000 Pregnant Woman 950 Nonbacteriuric
-------+
1st Trimester
3-5 APN
2nd/3rd
patients [43]. In addition, recent evidence associates bacteriuria in catheterized patients with an approximately three-fold increase in mortality [44]. However, the mechanism through which bacteriuria is related to mortality remains uncertain. Risk factors associated with catheter-associated urinary tract infection include female sex, advanced age, and increasing degree of underlying illness [45]. Although these factors define high-risk groups, they can not easily be altered. In contrast, a variety of alterable factors have been associated with increased risk of catheter-associated infection including lengthy duration of catheterization, improper nonsterile insertion technique, failure to observe closed sterile drainage, and the type of drainage system used [4]. Thus, current efforts at prevention should focus on these alterable factors. The pathogenesis of catheter-associated urinary tract infections has been less well studied than urinary tract infections in noncatheterized patients. However, points of bacterial entry have been well defined and include (1) introduction of bacteria residing in the urethra into the bladder at the time of catheterization; (2) subsequent entry of bacteria colonizing the urethral meatus along the mucous sheath external to the catheter; and (3) ascent of bacteria within the catheter lumen itself. The relative proportion of infections occurring through bacterial entry at each of these sites has not been clearly defined. The impact of closed sterile drainage on the rate of infection suggests that intraluminal ascent may have once been the most frequent source of bacterial entry. However, in most hospitals careful observation of closed sterile drainage techniques has become common practice. Thus, entry of bacteria at the catheter-meatal junction may have become a relatively more frequent point of bacterial entry than in the past. Two recent studies have demonstrated that meatal colonization with gram-negative rods portends a greater risk of urinary tract infection in catheterized patients [ 46,471. In a recent prospective study, we demonstrated that organisms causing bacteriuria in catheter-associated urinary tract infections could often be identified in the urethral or rectal flora two to four days before the onset of bacteriuria. This finding was particularly common in women (70 percent) as opposed
1.52
May 15, 1984
The American Journal of Medicine
Trimester
Figure 2. Prevention of acute pyelonephritis during pregnancy.
to men (40 percent). At this point in time, little has been done to assess possible immune mechanisms in the host that may affect susceptibility to catheter-associated urinary tract infections. Similarly, the organisms that produce these infections have not been intensively studied in terms of virulence factors that may influence the pathogenesis of nosocomial bacteriuria. In fact, it is not possible at this point to clearly define how many episodes of catheter-associated bacteriuria actually represent infection (as opposed to bladder colonization) and whether these infections, when present, are localized largely to the bladder or involve the upper urinary tract. Until more is known about the pathogenesis of nosocomial bacteriuria, the bulk of preventive efforts will continue to focus on aseptic care of the urinary catheter [48]. These measures are directed primarily at preventing entry of bacteria into the sterile closed drainage system and are effective when catheterization is less than seven to 10 days. Because a previous report in this symposium critically reviews this subject in detail, these preventive measures will only be briefly outlined here. The single most effective measure is to avoid the use of the major risk factor, the indwelling catheter. Aseptic insertion of catheters and sterile technique during catheter care are the major mechanisms for preventing periurethral and intraluminal ascent of bacteria into the bladder. Interestingly, it has not been possible to demonstrate the efficacy of local antimicrobial ointments applied to the meatal junction, despite the apparent association of meatal colonization with subsequent infection [ 49,501. The use of antimicrobial irrigants to eradicate organisms after bacteriuria has developed was not effective in reducing the prevalence of bacteriuria in a recent trial [51]. In addition, this technique predisposed to resistant infections and gave rise to junctional disconnections twice those seen in the nonirrigated group. However, this technique may be of some value in situations in which the sterile closed drainage system can be maintained. Systemic antimicrobial agents reduce the occurrence of bacteriuria, at least for the first few days of catheterization [52]. However, their use can not be widely recommended at this time in that the benefit accrued (reduction of
PATIENT AT RISK AND ANTIMICROBIAL
asymptomatic bacteriuria) may not be worth the cost and attendant risk of development of resistant microorganisms. CONCLUSIONS Effective preventive measures have been developed for many subsets of patients at high risk of urinary tract infection. In women with uncomplicated, frequent, exogenous reinfections, prophylaxis is aimed primarily at preventing the morbidity, inconvenience, and costs of repeated infection; serious renal or other sequela have not been linked to frequent reinfection in these patients. Controlled trials have demonstrated the efficacy and cost effective;ess of continuous low-dose antimicrobial therapy in preventing infection in these patients. However, prophylactic antimicrobial agents do not exert any influence on the natural history of recurrent infection in most women, and therefore they must be given for long periods of time. The emergence of resistant
THERAPY-STAMM
microorganisms and the occurrence of side effects limit their effectiveness in some patients. Clinically, it is difficult to know how long to use prophylactic antibiotic therapy and when to stop them. Recent advances in our understanding of factors influencing the pathogenesis of these infections at a molecular level suggest that newer preventive strategies such as immunoprophylaxis and the possible use of receptor analogs for inhibiting adherence should be pursued. Measures for preventing catheter-associated urinary tract infections primarily involve aseptic insertion and care of the catheter and collecting system. Much remains to be learned about the pathogenesis of these infections so that more specific preventive measures can be used in patients at highest risk of serious sequelae. All pregnant women should be screened for bacteriuria and treated if infected, because treatment reduces the risk of maternal pyelonephritis and associated prematurity.
REFERENCES
6.
7.
8.
9. 10.
11.
12.
13.
National Center for Health Statistics: Ambulatory medical care rendered in physicians’ offices: United States, 1975. Adv Data 1977; 12: 1-12. Norden CW, Kass EH: Bacteriuria of pregnancy: a critical appraisal. Annu Rev Med 1968; 19: 432-470. Andriole VT: Urinary tract infections in pregnancy. Urol Clin North Am 1975; 2: 465598. Turck M, Stamm WE: Nosocomial infection of the urinary tract. Am J Med 1961; 70: 651-654. Freeman RB, Smith WM. Richardson JA, et al: Long-term therapy for chronic bacteriuria in men: U.S. Public Health Service Cooperative Study. Ann Intern Med 1975; 83: 133-147. Chodak GW, Plant ME: Systemic antibiotics for prophylaxis in urologic surgery: a critical review. J Urol 1979; 121: 695-708. Warren JW, Muncie HL, Bergquist EJ, et al: Sequelae and management of urinary infection in the patient requiring chronic catheterization. J,Urol 1981; 125: 1-8. Stamey TA, Timothy M, Millar M, Mihara G: Recurrent urinary infections in adult women: the role of introital enterobacteria. Calif Med 1971; 115: 1-19. Kunin CM: The natural history of recurrent bacteriuria in schoolgirls. N Engl J Med 1970; 282: 1443-1448. O’Grady FW, Richards B, McSherry MA, et al: Introital enterobacteria, urinary infection, and the urethral syndrome. Lancet 1970; II: 1208-1210. Cattell WR, McSherry MA, Northeast A, Powell E, Brooks HJL, O’Grady F: Periurethral enterobacterial carriage in the pathogenesis of recurrent urinary infection. Br Med J 1974; 4: 136-139. Bran JL, Levison ME, Kaye D: Entrance of bacteria into the female urinary bladder. N Engl J Med 1972; 286: 626628. Kunin CM: Sexual intercourse and urinary infections. N Engl
14.
15.
16.
17.
18.
19. 20.
21.
22.
23.
24.
J Med 1978: 298: 336-337. Kunin CM, McCormack RC: An epidemiologic study of bacteriuria and blood among nuns and working women. N Engl J Med 1960; 278: 635-642. Nicolle L, Harding GKM, Preiksaitis J, Ronald AR: The association of urinary tract infection with sexual intercourse. J Infect Dis 1982; 146: 579-583. Pfau A, Sacks T, Englestein D: Recurrent urinary tract infections in premenopausal women. Prophylaxis based on an understanding of the pathogenesis. J Urol 1983; 129: 1152-1157. Sandberg T, Stenqvist K, Svanborg-Eden C: Effects of subminimal inhibitory concentrations of ampicillin, chloramphenicol, and nitrofurantoin on the attachment of E. coli to human uroepithelial cells in vitro. Rev Infect Dis 1979; 5: 838-844. Vosbeck K, Handschen H, Menge EB, Zak 0: Effects of antibiotics at concentrations below the minimal inhibitory concentration on the adhesiveness of E. coli in vitro. Rev Infect Dis 1979; 5: 845-851. Schoolnik G: Host-bacterial determinants of pyelonephritis. Am J Med 1984; in press. Svanborg-Eden C, Freter R, et al: Inhibition of experimental ascending urinary tract infection by an epithelial cell-surface receptor analogue. Nature 1982; 298: 560-562. Korhonen TK, Vaisanen V, Kallio P, et al: The role of pili in the adhesion of E. coli to human urinary tract epithelial cells. Stand J Infect Dis 1982; 33 (suppl): 26-32. Vosti KL: Recurrent urinary tract infections: prevention by prophylactic antibiotics after sexual intercourse. JAMA 1975; 231: 934-40. Adatto K, Doebele KG, Galland L. Granowetter L: Behavioral factors and urinary tract infections. JAMA 1981: 241: 2525-2526. Stamey TA, Condy M, Mihara G: Prophylactic efficacy of ni-
May 15, 1984
The American
Journal of Medicine
153
PATIENT AT RISK AND ANTIMICROBIAL THERAPY-STAMM
25.
26.
27.
28.
29.
30. 31.
32.
33.
34.
35.
36.
37.
38.
154
trofurantoin macrocrystals and trimethoprim-sulfamethoxazole in urinary infections: biologic effects on the vaginal and rectal flora. N Engl J Med 1977; 296: 780-782. Stamm WE, Counts GW, Wagner KR, et al: Antimicrobial prophylaxis of recurrent urinary tract infections: a double-blind, placebo-controlled trial. Ann Intern Med 1980; 92: 770-775. Ronald AR, Harding GKM, Mathias R, Wong CK, Muir P: Prophylaxis of recurring urinary tract infection in females: a comparison of nitrofurantoin with trimethoprim-sulfamethoxazole. Can Med Assoc J 1975; 112: 13%1%. Harding GKM, Buckwold FJ, Marrie TJ, Thompson L, Light RB, Ronald AR: Prophylaxis of recurrent urinary tract infection in female patients: efficacy of low-dose, thrice-weekly therapy with trimethoprim-sulfamethoxazole. JAMA 1979; 242: 1975-1977. Harding GKM, Ronald AR: A controlled study of antimicrobial prophylaxis of recurrent urinary infection in women. N Engl J Med 1974; 291: 597-601. Bailey RR, Roberts AP, Gower PE, De Wardener HE: Prevention of urinary tract infection with low-dose nitrofurantoin. Lancet 1971; II: 1112-1114. Holland NH, West CD: Prevention of recurrent urinary tract infections in girls. Am J Dis Child 1963; 105: 560-567. Pearson NJ, Towner KJ, McSherry AM, Cattell WR, O’Grady F: Emergence of trimethoprim-resistant enterobacteria in patients receiving long-term cotrimoxazole for the control of intractable urinary tract infection. Lancet 1979; I: 12051209. Kasanen A, Junnila SYT, Kaarsalo E, Hajba A, Sundquist H: Secondary prevention of recurrent urinary tract infections. Comparison of the effect of placebo, methenamine hippurate, nitrofurantoin, and trimethoprim alone. Stand J Infect Dis 1982; 14: 293-296. Holmberg L, Boman G, Bottiger LE, Eriksson B, Spross R, Wessling A: Adverse reactions to nitrofurantoin: analysis of 921 reports. Am J Med 1980; 69: 733-788. Scheckler WE, Burt RAP, Paulson DF: Comparison of low-dose cinoxacin and placebo in the prevention of recurrent urinary tract infections. J Fam Pratt 1982; 15: 901-904. Landes RR: Long term low dose cinoxacin therapy for the prevention of recurrent urinary tract infections. J Urol 1980; 123: 47-49. Schaeffer AJ, Jones JM, Flynn SS: Prophylactic efficacy of cinoxacin in recurrent urinary tract infection: biological effects of the vaginal and fecal flora. 1982; 127: 11281131. Stamm WE, McKevitt M, Counts GW, Wagner FK, Turck M, Holmes KK: Is antimicrobial prophylaxis of urinary tract infections cost effective? Ann intern Med 1981; 94: 251-255. Stamm WE, Counts GW, McKevitt M, Turck M, Holmes KK: Urinary prophylaxis with trimethoprim and trimethoprim-
May 15, 1984
The American Journal of Medicine
39.
40.
41. 42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
sulfamethoxazole efficacy, influence on the natural history of recurrent bacteriuria, and cost control. Rev Infect Dis 1982; 4: 450-455. Kraft JK, Stamey TA: The natural history of symptomatic recurrent bacteriuria in women. Medicine (Baltimore) 1977; 56: 55-60. Wong ES, McKevitt M, Runnig K, Stamm WE: Comparison of antimicrobial prophylaxis with intermittent self-administered single dose antibiotics for management of recurrent urinary tract infections. Twenty-third Interscience Conference on Antimicrobial Agents and Chemotherapy, American Society of Microbiology, 1983. Sweet RL: Bacteriuria and pyelonephritis during pregnancy. Semin Perinatol 1977; 1: 25-40. Lenke RR, Van Dorsten JP, Schifrin BS: Pyelonephritis in pregnancy: a prospective randomized trial to prevent recurrent disease evaluating suppressive,therapy with nitrofurantoin and close surveillance. Am J Obstet Gynecol 1983; 146: 953-957. Kreger DE, Craven DE, Carling PC, McCabe WA: Gram negative bacteremia. III. Reassessment of etiology, epidemiology and ecology in 612 patients. Am J Med 1980; 68: 332-343. Platt R, Polk BF, Murdock B, Rosner B: Mortality associated with nosocomial urinary tract infection. N Engl J Med 1982; 307: 637-642. Garibaldi RA, Burke JP, Dickman ML, Smith CB: Factors predisposing to bacteriuria during indwelling urethral catheterization. N Engl J Med 1974; 291: 215-220. Garibaldi RA. Burke JP, Britt MR, Miller WA, Smith CB: Meatal colonization and catheter-associated bacteriuria. N Engl J Med 1980; 303: 316-318. Daifuku R, Stamm WE: Association of rectal and periurethral colonization with urinary tract infection in patients with indwelling catheters. Twenty-third Interscience Conference on Antimicrobial Agents and Chemotherapy, American Society of Microbiology, 1983. Wong ES, Hooton TM: Guidelines for prevention of catheterassociated urinary tract infection. Infect Control 1980; 2: 125-130. Burke JP, Jacobson JA, Garbaldi RA, Conti MT, Ailing DW: Evaluation of daily meatal care with poly-antibiotic ointment in prevention of urinary catheter-associated bacteriuria. J Urol 1983; 129: 331-334. Burke JP, Garibaldi RA, Britt MR, Jacobsen JA, Conti M, Ailing DW: Prevention of catheter-associated urinary tract infections. Am J Med 1981; 70: 655-660. Warren JW. Platt R, Thomas RJ, Rosner B, Kass EH: Antibiotic irrigation and catheter-associated urinary tract infections. N Engl J Med 1978; 299: 570-574. Britt MR, Garibaldi RA, Miller WA, Helertson RM, Burke JP: Antimicrobial prophylaxis for catheter-associated bacteriuria. Antimicrob Agent Chemother 1977; 11: 240-243.
WALTER E. STAMM, M.D. Seattle, Washington
From the Department of Medicine, Harborview Medical Center, 325 9th Avenue, Seattle, Washington. Requests for reprints should be tiddressed to Dr. Walter E. Stamm, Department of Medicine, Harborview Medical Center, University of Washington School of Medicine, Seattle, Washington 98104.
148
May 15, 1994
Potential candidates for measures designed to prevent urinary tract infections include (1) women (or girls) with frequent exogenous reinfections; (2) pregnant women; (3) hospitalized patients with short-term indwelling catheters; (4) men with chronic bacterial prostatitis; (5) patients with neurogenic bladders being managed with either intermittent catheterization or chronic indwelling catheters; and (6) men undergoing urologic surgery. Patients in the first three categories represent the vast majority of men and women with preventable urinary tract infections. On the basis of our current understanding of the pathogenesis of urinary infections in women, potential preventive measures can be envisioned. To date, the most successful methods in susceptible women include continuous low-dose antimicrobial prophylaxis or postintercourse antibiotic prophylaxis. In controlled clinical trials, the efficacy of low-dose trimethoprim-sulfamethoxazole, trimethoprim, or nitrofurantoin has repeatedly been demonstrated, and the cost effectiveness of this approach has also been established. An alternative strategy, intermittent, self-administered, single-dose antimicrobial therapy, may be useful in selected circumstances. Screening of high-risk patients for asymptomatic infections has largely been abandoned except in pregnant women in whom this practice remains an important preventive measure. Exciting newer approaches to prevention based on studies clarifying the fimbriai structures mediating bacterial adherence to receptors on uroepithelial ceils include the use of receptor analogs and immunoprophyiaxis, but these approaches are not yet of demonstrated clinical efficacy. Since urinary tract infections are among the most common bacterial infections, their prevention assumes considerable importance for both economic and public health reasons. An estimated six million visits annually result from these infections in the outpatient setting [ 11, most being in women between the ages of 15 and 35 years. In addition, 5 percent of pregnant women experience bacterial urinary tract infections [2] and these infections produce both maternal and infant morbidity [3]. Ten to 15 percent of hospitalized patients undergo urinary catheterization and incur an approximate 25 percent risk of urinary tract infection, thus resulting in about 600,000 episodes of catheter-associated bacteriuria annually [4]. These three major risk groups (Table I) account for most urinary tract infections, and therefore these three populations are the most important potential candidates for preventive measures. Other high-risk groups include men with chronic bacterial prostatitis, many of whom have recurrent episodes of urinary tract infection [5], men undergoing urologic surgery [6], and men and women with neurogenic bladders managed either by intermittent cateterization or chronic indwelling catheters [7). The
The American Journal of Medicine
PATIENT AT RISK AND ANTIMICROBIALTHERAPY-~TAMM
latter groups often experience complicated and difficult to treat urinary infections, but the patients involved are far fewer in number than the first three populations mentioned. As with other sites of infection, the goals of preventing infection in the urinary tract include reducing morbidity attributable to acute infectious episodes, preventing serious infectious or noninfectious complications of either single or repeated infections, and curtailing health care costs required for diagnosis and treatment. URINARY TRACT INFECTIONS WOMEN
IN NONPREGNANT
Eighty percent of recurrent urinary tract infections in women result from exogenous infection with a newly acquired strain [8,9]. In most instances, the sequence of events resulting in urinary tract infection begins with transperineal migration of infecting Escherichia coli strains from the rectum to the vagina with establishment of introital and urethral colonization. Eventual inoculation of the bladder urine or the posterior urethra with these colonizing organisms then leads to acute lower urinary tract infection (urethritis or cystitis, or both) [ 10,111. Ascending infection of the kidney (acute pyelonephritis) may subsequently occur. Although sexual intercourse produces transient bacteriuria in most women, and although epidemiologic studies clearly associate sexual activity with an increased risk of urinary tract infection, the role of sexual activity in the pathogenesis of these infections has not been universally accepted [ 12- 141. However, two recent studies provide strong support for the argument that sexual intercourse plays an important permissive role in facilitating inoculation of the bladder urine after development of urethral colonization. Nicolle et al [ 151 found that 80 percent of urinary infections have their onset within 24 hours of intercourse. Pfau et al [ 161 reported similar results and showed that postintercourse antibiotics could markedly reduce the episodes of cystitis in such patients. On the basis of this understanding of the pathogenesis of urinary tract infections, several possible pre-
GutFlora
f I I
Antibiotics
Figure 1.
___
-)
Introit01 Colonization
Urethral Colonization
-
$ I I
L-
ventive measures can be envisioned that might interrupt the chain of events culminating in infection (Figure 1). Antimicrobial agents could be used to reduce gastrointestinal colonization with aerobic gram-negative bacteria, thus eliminating the reservoir for infecting strains. Similarly, either systemic or topical antibiotics could be used to prevent the establishment of introital or urethral colonization with rectal E. coli. Antimicrobial agents might prevent the development of such colonization either through direct antibacterial activity or by inhibition of adherence, a property demonstrated to result from subinhibitory levels of ampicillin and nitrofurantoin [ 17,181. Alternatively, preventive measures could be directed at microorganism-specific factors that may enhance the risk of infection. As outlined in the preceding report, uropathogenic E. coli frequently bear specific surface pili that mediate binding with corresponding receptor sites on host cells [ 191. Thus, one potential intervention aimed at reducing introital and urethral colonization could use receptor analogs to inhibit attachment of these uropathogenic E. coli by competitively saturating receptor binding sites. Purified P fimbriae and glycosphyngolipids block attachment of P fimbria bearing E. coli to uroepithelial cells in in vitro adherence assays, whereas d-mannose blocks attachment attributable to Type I pili both in vitro and animal models [20,21]. Alternatively, antibodies directed against fimbriae that mediate attachment can block adherence in in vitro assays and protect against ascending pyelonephritis in a mouse model [ 191. These
Potential Candidates for Prophylaxis
TABLE I
1. Women (girls) with frequent reinfections 2. Pregnant women 3. Indwelling catheter (short-term) 4. Men with chronic bacterial prostatitis 5. Neurogenic bladder Intermittent catheterization Chronic indwelling catheter 6. Urologic
surgery
Bladder Inoculation
T
4
Cystitis
*
__)
Pyelonephritis
ic Antibiotics Receptor Analogues Vaccine
Potential preventive
- J
intervention
I I
I
I Abstinence Postcoital Voiding Postcoital Antibiotics Contraception Behavioral (?)
!
I I :
I Antibiotics Vaccine
l
in women with recurrent urinary tract infections
May 15, 1984
The American
Journal
of Medicine
149
PATIENT AT RISK AND ANTIMICROBIAL
THERAPY-STAMM
data provide a basis for the belief that immunoprophytaxis may be an exciting possibility in the near future. Additional interventions can be envisioned after urethral colonization has been established but before bladder inoculation has occurred. Because sexual intercourse facilitates the transfer of organisms from the urethra to the bladder in many women, postcoital antimicrobial agents have been studied and found effective in preventing infection [ 16,221. Abstinence could achieve similar results. In addition, postcoital voiding may eliminate organisms after their entry into the bladder urine but before active infection is established. In one case-control study, voluntary deferral of micturition was associated with recurrent urinary tract infections [23]. No specific contraceptive practice has been linked directly to urinary tract infection, although the diaphragm has been associated with increased episodes of cystitis in our patients. Lastly, additional interventions can be envisioned that would exert their effect after organisms are introduced into the bladder urine but before infection is established. These might include (1) low concentrations of antimicrobial agents in the urine, which either would kill bacteria on entry into the bladder or would alter these bacteria such that their attachment to uroepithelial cells would be inhibited, or (2) a vaccine that probably would work through the same antiattachment mechanism just outlined. Although many theoretic preventive strategies can be cited, only two approaches to the prevention of recurrent urinary tract infections in highly susceptible women have been effective in clinical trials: continuous low-dose antimicrobial prophylaxis or postintercourse antibiotic prophylaxis. Controlled clinical trials have evaluated the effectiveness of antimicrobial agents in preventing recurrent urinary tract infection in adult women [24-301. These studies show that low-dose antimicrobial agents given either daily or thrice weekly (Table II) can reduce the rate of infection in susceptible women from two to four per year to nearly zero per
TABLE II
Prophylactic Antibiotic Regimens for Women
Drug TMP-SMX TMP Nitrofurantoin Sulfa Methenamine
Dose 40 mg/200 mg 100 mg 50 mg/lOO mg 500 mg 500 mg four times a day
TMP = trimethoprim;
150
Infections per Patient Year o-o.15 o-o.15 0.1-0.8 1-2 0.8-1.6
FecalPerineal Effect + + -
SMX = sulfamethoxazole.
May 15, 1984
The American Journal of Medicine
year. Trimethoprim-sulfamethoxazole, given as 40 mg/200 mg daily, has been the most effective agent in many studies, usually reducing infection to the range of zero to 0.15 per patient year. Trimethoprim-sulfamethoxazole may achieve this effect either by direct antimicrobial action in urine or through the nearly complete elimination of aerobic gram-negative bacteria from the fecal and perineal reservoir that it produces in most patients. Trimethoprim alone has had similar effects on both the fecal-perineal flora and the rate of recurrent infection when it is used prophylactically. Both agents have been continuously effective and relatively free of side effects when used as long as two to four years. Emergence of antimicrobial resistance to trimethoprim-sulfamethoxazole or trimethoprim in E. coli or Klebsiella has not been noted among either urinary or fecal isolates in prophylaxis trials to date. However, increasing trimethoprim resistance has been noted in urine isolates obtained from outpatients in some areas [ 3 1,321. We encountered no trimethoprim-resistant E. coli infections among 30 women receiving trimethoprim or trimethoprim-sulfamethoxazole in a 1975 to 1977 prophylaxis trial, whereas 3 of 30 women had such infections in a 1981 to 1983 study. Nitrofurantoin or nitrofurantoin macrocrystals have also been used successfully to prevent recurrent urinary tract infections. These agents generally have not reduced the rate of infection as dramatically as has either trimethoprim-sulfamethoxazole or trimethoprim alone. However, they are highly effective as compared with placebo or no antimicrobial agents. Because nitrofurantoin has little effect on either the fecal or perineal aerobic gram-negative flora, its effectiveness presumably results from direct antimicrobial activity in urine. Recent reports of frequent side effects in women receiving low-dose continuous nitrofurantoin, including interstitial pneumonia, allergic pneumonitis, hepatitis, and other complications, necessitate careful monitoring and raise concern about its long-term use [33]. Both sulfonamide and methenamine salts plus ascorbic acid reduce the rate of urinary tract infections in susceptible women, but neither can be widely recommended as an effective prophylactic regimen. Sulfonamides, when used on a long-term basis, rapidly promote the development of resistant organisms in the gastrointestinal flora and subsequent superinfection with these resistant strains. Methenamine mandelate (given with vitamin C to acidify the urine) prevents infection, but not to the extent achieved by the other drugs just discussed. In addition, most women can not comply with a regimen requiring four pills four times daily for long periods of time. Naladixic acid reduced infection rates in some prophylactic trials but was not widely used because of concerns about rapid selection of resistant strains
PATIENT
[ 24-261.
Cinoxacin,
a recently
released
derivative, has been used for prophylaxis studies [ 34-361. A significant reduction
quinolone
in three recent in the infection
rate was found in all three trials, but cinoxacin-resistant strains were found in three of 17 women in one study. This agent has an intermediate effect on fecal and perineal aerobic gram-negative flora, having more activity than nitrofurantoin but much less than that of trimethoprim. An alternative strategy to continuous low-dose antimicrobial prophylaxis is postintercourse antimicrobial prophylaxis. Vosti [22] first demonstrated that nitrofurantoin macrocrystals given after coitus prevented recurrent urinary tract infection. More recently, Pfau et al [ 161 have shown that trimethoprim-sulfamethoxazole, naladixic acid, nitrofurantoin, and sulfonamides were all effective in preventing recurrent urinary tract infections when given to young, sexually active women whose infections occurred postcoitally. This management strategy deserves further study, but appears effective when given to women whose infections are often related to intercourse. Compared with continuous low-dose prophylaxis, this approach may reduce medication costs and side effects and lessen the emergence of resistant strains. Practical clinical considerations regarding the use of antimicrobial prophylaxis include the following: (1) In whom should it be used and (2) for how long? The use of prophylactic antibiotics appears warranted from a cost-effectiveness standpoint when women experience more than two infections per year [ 371. Women with three or more infections per year generally prefer prophylaxis to repeated symptomatic episodes and the inconveniences and costs they necessitate. However, antimicrobial prophylaxis only works as long as active drug is given, and does not appear to influence the natural history of infection in a given patient [25]. Knowledge of a woman’s baseline rate of infection in preceding years serves as the best predictor of the likelihood of infection in subsequent years [38]. However, infections may also cluster, and some women move from high-risk to low-risk periods and vice versa 139). The use of prophylaxis for six months followed by reassessment during no prophylaxis seems a reasonable approach in women not known to have baseline infection rates of three per year or more. Postcoital prophylaxis is a useful approach in women whose episodes often follow intercourse. Although not truly preventative, one additional strategy for management of recurrent urinary tract infections deserves mention. Because single-dose antimicrobial agents cure most urinary tract infections, self-diagnosis of infection followed by self-treatment using single-dose therapy could be used in women who have sporadic infections. In a recent study, we found
AT RISK
that (1) women infections,
AND
could
(2) intermittent
ANTIMICROBiAL
accurately
THERAPY
self-diagnose
self-administered
STAMM
their
medication
cured 85 percent of these recurrent urinary tract infections, and (3) this regimen was cost effective when compared with that of continuous low-dose antimicrobial prophylaxis [ 40 1. Intermittent self-medication may be of particular value in women with sporadic infections, (that is, one or two per year) in whom costs, the amount of medication required, and side effects would be reduced. PREVENTION OF URINARY INFECTIONS DURING PREGNANCY Urinary tract infections, which occur in 5 percent of pregnant women, have special significance in that pyelonephritis more often results because of physiologic changes in the urinary tract during pregnancy (decreased bladder and ureteral tone, decreased ureteral peristalsis, hydroureter, and increased residual bladder urine). In addition, urinary tract infections in pregnancy have been associated with prematurity ( 31. Acute pyelonephritis develops in 20 to 30 percent of women with untreated bacteriuria during pregnancy, and of these women 20 percent have premature infants [ 31. The relation of asymptomatic bacteriuria of pregnancy to prematurity remains controversial, but on balance the evidence suggests that asymptomatic bacteriuria of pregnancy is related to prematurity, especially in women with renal infection [ 4 1 1. Thus, the prevention of urinary tract infections in pregnant women reduces morbidity not only in the woman but also in the fetus. For this reason, all pregnant women should be screened for bacteriuria during the first trimester and once again during the third trimester (Figure 2). First trimester screening and treatment of bacteriuric women will reduce the expected cases of acute pyelonephritis from 30 to three per 1,000 pregnant women, and screening during the third trimester will prevent another three to five cases. Women with asymptomatic bacteriuria of pregnancy should have the diagnosis established by at least two positive urine cultures, after which treatment should be initiated I3,41 1. After successful treatment, posttreatment cultures should be obtained on a monthly basis until delivery. After treatment of a urinary tract infection during pregnancy, continuous low-dose antimicrobial prophylaxis for the remainder of pregnancy has been advocated by some, but this practice did not appear more beneficial than careful follow-up in a recent study [ 42 I. PREVENTION OF INFECTION IN PATIENTS URINARY
WITH
CATHETERS
Catheter-associated urinary tract infection remains the most common hospital-associated infection and is the most frequent source of bacteremia in hospitalized
May 15, 1984
The American
Journal
of Medicine
151
PATIENT AT RISK AND ANTIMICROBIAL
THERAPY-STAMM
50 Bacteriuric
Treated
)
, ApN
1000 Pregnant Woman 950 Nonbacteriuric
-------+
1st Trimester
3-5 APN
2nd/3rd
patients [43]. In addition, recent evidence associates bacteriuria in catheterized patients with an approximately three-fold increase in mortality [44]. However, the mechanism through which bacteriuria is related to mortality remains uncertain. Risk factors associated with catheter-associated urinary tract infection include female sex, advanced age, and increasing degree of underlying illness [45]. Although these factors define high-risk groups, they can not easily be altered. In contrast, a variety of alterable factors have been associated with increased risk of catheter-associated infection including lengthy duration of catheterization, improper nonsterile insertion technique, failure to observe closed sterile drainage, and the type of drainage system used [4]. Thus, current efforts at prevention should focus on these alterable factors. The pathogenesis of catheter-associated urinary tract infections has been less well studied than urinary tract infections in noncatheterized patients. However, points of bacterial entry have been well defined and include (1) introduction of bacteria residing in the urethra into the bladder at the time of catheterization; (2) subsequent entry of bacteria colonizing the urethral meatus along the mucous sheath external to the catheter; and (3) ascent of bacteria within the catheter lumen itself. The relative proportion of infections occurring through bacterial entry at each of these sites has not been clearly defined. The impact of closed sterile drainage on the rate of infection suggests that intraluminal ascent may have once been the most frequent source of bacterial entry. However, in most hospitals careful observation of closed sterile drainage techniques has become common practice. Thus, entry of bacteria at the catheter-meatal junction may have become a relatively more frequent point of bacterial entry than in the past. Two recent studies have demonstrated that meatal colonization with gram-negative rods portends a greater risk of urinary tract infection in catheterized patients [ 46,471. In a recent prospective study, we demonstrated that organisms causing bacteriuria in catheter-associated urinary tract infections could often be identified in the urethral or rectal flora two to four days before the onset of bacteriuria. This finding was particularly common in women (70 percent) as opposed
1.52
May 15, 1984
The American Journal of Medicine
Trimester
Figure 2. Prevention of acute pyelonephritis during pregnancy.
to men (40 percent). At this point in time, little has been done to assess possible immune mechanisms in the host that may affect susceptibility to catheter-associated urinary tract infections. Similarly, the organisms that produce these infections have not been intensively studied in terms of virulence factors that may influence the pathogenesis of nosocomial bacteriuria. In fact, it is not possible at this point to clearly define how many episodes of catheter-associated bacteriuria actually represent infection (as opposed to bladder colonization) and whether these infections, when present, are localized largely to the bladder or involve the upper urinary tract. Until more is known about the pathogenesis of nosocomial bacteriuria, the bulk of preventive efforts will continue to focus on aseptic care of the urinary catheter [48]. These measures are directed primarily at preventing entry of bacteria into the sterile closed drainage system and are effective when catheterization is less than seven to 10 days. Because a previous report in this symposium critically reviews this subject in detail, these preventive measures will only be briefly outlined here. The single most effective measure is to avoid the use of the major risk factor, the indwelling catheter. Aseptic insertion of catheters and sterile technique during catheter care are the major mechanisms for preventing periurethral and intraluminal ascent of bacteria into the bladder. Interestingly, it has not been possible to demonstrate the efficacy of local antimicrobial ointments applied to the meatal junction, despite the apparent association of meatal colonization with subsequent infection [ 49,501. The use of antimicrobial irrigants to eradicate organisms after bacteriuria has developed was not effective in reducing the prevalence of bacteriuria in a recent trial [51]. In addition, this technique predisposed to resistant infections and gave rise to junctional disconnections twice those seen in the nonirrigated group. However, this technique may be of some value in situations in which the sterile closed drainage system can be maintained. Systemic antimicrobial agents reduce the occurrence of bacteriuria, at least for the first few days of catheterization [52]. However, their use can not be widely recommended at this time in that the benefit accrued (reduction of
PATIENT AT RISK AND ANTIMICROBIAL
asymptomatic bacteriuria) may not be worth the cost and attendant risk of development of resistant microorganisms. CONCLUSIONS Effective preventive measures have been developed for many subsets of patients at high risk of urinary tract infection. In women with uncomplicated, frequent, exogenous reinfections, prophylaxis is aimed primarily at preventing the morbidity, inconvenience, and costs of repeated infection; serious renal or other sequela have not been linked to frequent reinfection in these patients. Controlled trials have demonstrated the efficacy and cost effective;ess of continuous low-dose antimicrobial therapy in preventing infection in these patients. However, prophylactic antimicrobial agents do not exert any influence on the natural history of recurrent infection in most women, and therefore they must be given for long periods of time. The emergence of resistant
THERAPY-STAMM
microorganisms and the occurrence of side effects limit their effectiveness in some patients. Clinically, it is difficult to know how long to use prophylactic antibiotic therapy and when to stop them. Recent advances in our understanding of factors influencing the pathogenesis of these infections at a molecular level suggest that newer preventive strategies such as immunoprophylaxis and the possible use of receptor analogs for inhibiting adherence should be pursued. Measures for preventing catheter-associated urinary tract infections primarily involve aseptic insertion and care of the catheter and collecting system. Much remains to be learned about the pathogenesis of these infections so that more specific preventive measures can be used in patients at highest risk of serious sequelae. All pregnant women should be screened for bacteriuria and treated if infected, because treatment reduces the risk of maternal pyelonephritis and associated prematurity.
REFERENCES
6.
7.
8.
9. 10.
11.
12.
13.
National Center for Health Statistics: Ambulatory medical care rendered in physicians’ offices: United States, 1975. Adv Data 1977; 12: 1-12. Norden CW, Kass EH: Bacteriuria of pregnancy: a critical appraisal. Annu Rev Med 1968; 19: 432-470. Andriole VT: Urinary tract infections in pregnancy. Urol Clin North Am 1975; 2: 465598. Turck M, Stamm WE: Nosocomial infection of the urinary tract. Am J Med 1961; 70: 651-654. Freeman RB, Smith WM. Richardson JA, et al: Long-term therapy for chronic bacteriuria in men: U.S. Public Health Service Cooperative Study. Ann Intern Med 1975; 83: 133-147. Chodak GW, Plant ME: Systemic antibiotics for prophylaxis in urologic surgery: a critical review. J Urol 1979; 121: 695-708. Warren JW, Muncie HL, Bergquist EJ, et al: Sequelae and management of urinary infection in the patient requiring chronic catheterization. J,Urol 1981; 125: 1-8. Stamey TA, Timothy M, Millar M, Mihara G: Recurrent urinary infections in adult women: the role of introital enterobacteria. Calif Med 1971; 115: 1-19. Kunin CM: The natural history of recurrent bacteriuria in schoolgirls. N Engl J Med 1970; 282: 1443-1448. O’Grady FW, Richards B, McSherry MA, et al: Introital enterobacteria, urinary infection, and the urethral syndrome. Lancet 1970; II: 1208-1210. Cattell WR, McSherry MA, Northeast A, Powell E, Brooks HJL, O’Grady F: Periurethral enterobacterial carriage in the pathogenesis of recurrent urinary infection. Br Med J 1974; 4: 136-139. Bran JL, Levison ME, Kaye D: Entrance of bacteria into the female urinary bladder. N Engl J Med 1972; 286: 626628. Kunin CM: Sexual intercourse and urinary infections. N Engl
14.
15.
16.
17.
18.
19. 20.
21.
22.
23.
24.
J Med 1978: 298: 336-337. Kunin CM, McCormack RC: An epidemiologic study of bacteriuria and blood among nuns and working women. N Engl J Med 1960; 278: 635-642. Nicolle L, Harding GKM, Preiksaitis J, Ronald AR: The association of urinary tract infection with sexual intercourse. J Infect Dis 1982; 146: 579-583. Pfau A, Sacks T, Englestein D: Recurrent urinary tract infections in premenopausal women. Prophylaxis based on an understanding of the pathogenesis. J Urol 1983; 129: 1152-1157. Sandberg T, Stenqvist K, Svanborg-Eden C: Effects of subminimal inhibitory concentrations of ampicillin, chloramphenicol, and nitrofurantoin on the attachment of E. coli to human uroepithelial cells in vitro. Rev Infect Dis 1979; 5: 838-844. Vosbeck K, Handschen H, Menge EB, Zak 0: Effects of antibiotics at concentrations below the minimal inhibitory concentration on the adhesiveness of E. coli in vitro. Rev Infect Dis 1979; 5: 845-851. Schoolnik G: Host-bacterial determinants of pyelonephritis. Am J Med 1984; in press. Svanborg-Eden C, Freter R, et al: Inhibition of experimental ascending urinary tract infection by an epithelial cell-surface receptor analogue. Nature 1982; 298: 560-562. Korhonen TK, Vaisanen V, Kallio P, et al: The role of pili in the adhesion of E. coli to human urinary tract epithelial cells. Stand J Infect Dis 1982; 33 (suppl): 26-32. Vosti KL: Recurrent urinary tract infections: prevention by prophylactic antibiotics after sexual intercourse. JAMA 1975; 231: 934-40. Adatto K, Doebele KG, Galland L. Granowetter L: Behavioral factors and urinary tract infections. JAMA 1981: 241: 2525-2526. Stamey TA, Condy M, Mihara G: Prophylactic efficacy of ni-
May 15, 1984
The American
Journal of Medicine
153
PATIENT AT RISK AND ANTIMICROBIAL THERAPY-STAMM
25.
26.
27.
28.
29.
30. 31.
32.
33.
34.
35.
36.
37.
38.
154
trofurantoin macrocrystals and trimethoprim-sulfamethoxazole in urinary infections: biologic effects on the vaginal and rectal flora. N Engl J Med 1977; 296: 780-782. Stamm WE, Counts GW, Wagner KR, et al: Antimicrobial prophylaxis of recurrent urinary tract infections: a double-blind, placebo-controlled trial. Ann Intern Med 1980; 92: 770-775. Ronald AR, Harding GKM, Mathias R, Wong CK, Muir P: Prophylaxis of recurring urinary tract infection in females: a comparison of nitrofurantoin with trimethoprim-sulfamethoxazole. Can Med Assoc J 1975; 112: 13%1%. Harding GKM, Buckwold FJ, Marrie TJ, Thompson L, Light RB, Ronald AR: Prophylaxis of recurrent urinary tract infection in female patients: efficacy of low-dose, thrice-weekly therapy with trimethoprim-sulfamethoxazole. JAMA 1979; 242: 1975-1977. Harding GKM, Ronald AR: A controlled study of antimicrobial prophylaxis of recurrent urinary infection in women. N Engl J Med 1974; 291: 597-601. Bailey RR, Roberts AP, Gower PE, De Wardener HE: Prevention of urinary tract infection with low-dose nitrofurantoin. Lancet 1971; II: 1112-1114. Holland NH, West CD: Prevention of recurrent urinary tract infections in girls. Am J Dis Child 1963; 105: 560-567. Pearson NJ, Towner KJ, McSherry AM, Cattell WR, O’Grady F: Emergence of trimethoprim-resistant enterobacteria in patients receiving long-term cotrimoxazole for the control of intractable urinary tract infection. Lancet 1979; I: 12051209. Kasanen A, Junnila SYT, Kaarsalo E, Hajba A, Sundquist H: Secondary prevention of recurrent urinary tract infections. Comparison of the effect of placebo, methenamine hippurate, nitrofurantoin, and trimethoprim alone. Stand J Infect Dis 1982; 14: 293-296. Holmberg L, Boman G, Bottiger LE, Eriksson B, Spross R, Wessling A: Adverse reactions to nitrofurantoin: analysis of 921 reports. Am J Med 1980; 69: 733-788. Scheckler WE, Burt RAP, Paulson DF: Comparison of low-dose cinoxacin and placebo in the prevention of recurrent urinary tract infections. J Fam Pratt 1982; 15: 901-904. Landes RR: Long term low dose cinoxacin therapy for the prevention of recurrent urinary tract infections. J Urol 1980; 123: 47-49. Schaeffer AJ, Jones JM, Flynn SS: Prophylactic efficacy of cinoxacin in recurrent urinary tract infection: biological effects of the vaginal and fecal flora. 1982; 127: 11281131. Stamm WE, McKevitt M, Counts GW, Wagner FK, Turck M, Holmes KK: Is antimicrobial prophylaxis of urinary tract infections cost effective? Ann intern Med 1981; 94: 251-255. Stamm WE, Counts GW, McKevitt M, Turck M, Holmes KK: Urinary prophylaxis with trimethoprim and trimethoprim-
May 15, 1984
The American Journal of Medicine
39.
40.
41. 42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
sulfamethoxazole efficacy, influence on the natural history of recurrent bacteriuria, and cost control. Rev Infect Dis 1982; 4: 450-455. Kraft JK, Stamey TA: The natural history of symptomatic recurrent bacteriuria in women. Medicine (Baltimore) 1977; 56: 55-60. Wong ES, McKevitt M, Runnig K, Stamm WE: Comparison of antimicrobial prophylaxis with intermittent self-administered single dose antibiotics for management of recurrent urinary tract infections. Twenty-third Interscience Conference on Antimicrobial Agents and Chemotherapy, American Society of Microbiology, 1983. Sweet RL: Bacteriuria and pyelonephritis during pregnancy. Semin Perinatol 1977; 1: 25-40. Lenke RR, Van Dorsten JP, Schifrin BS: Pyelonephritis in pregnancy: a prospective randomized trial to prevent recurrent disease evaluating suppressive,therapy with nitrofurantoin and close surveillance. Am J Obstet Gynecol 1983; 146: 953-957. Kreger DE, Craven DE, Carling PC, McCabe WA: Gram negative bacteremia. III. Reassessment of etiology, epidemiology and ecology in 612 patients. Am J Med 1980; 68: 332-343. Platt R, Polk BF, Murdock B, Rosner B: Mortality associated with nosocomial urinary tract infection. N Engl J Med 1982; 307: 637-642. Garibaldi RA, Burke JP, Dickman ML, Smith CB: Factors predisposing to bacteriuria during indwelling urethral catheterization. N Engl J Med 1974; 291: 215-220. Garibaldi RA. Burke JP, Britt MR, Miller WA, Smith CB: Meatal colonization and catheter-associated bacteriuria. N Engl J Med 1980; 303: 316-318. Daifuku R, Stamm WE: Association of rectal and periurethral colonization with urinary tract infection in patients with indwelling catheters. Twenty-third Interscience Conference on Antimicrobial Agents and Chemotherapy, American Society of Microbiology, 1983. Wong ES, Hooton TM: Guidelines for prevention of catheterassociated urinary tract infection. Infect Control 1980; 2: 125-130. Burke JP, Jacobson JA, Garbaldi RA, Conti MT, Ailing DW: Evaluation of daily meatal care with poly-antibiotic ointment in prevention of urinary catheter-associated bacteriuria. J Urol 1983; 129: 331-334. Burke JP, Garibaldi RA, Britt MR, Jacobsen JA, Conti M, Ailing DW: Prevention of catheter-associated urinary tract infections. Am J Med 1981; 70: 655-660. Warren JW. Platt R, Thomas RJ, Rosner B, Kass EH: Antibiotic irrigation and catheter-associated urinary tract infections. N Engl J Med 1978; 299: 570-574. Britt MR, Garibaldi RA, Miller WA, Helertson RM, Burke JP: Antimicrobial prophylaxis for catheter-associated bacteriuria. Antimicrob Agent Chemother 1977; 11: 240-243.