- Email: [email protected]
Duration of treatment of urinary tract infections
REVIEWS
Duration of Treatment of Urinary Tract Infections
CALVIN M. KUNIN, M.D. Colurrtbus, Ohio
From the Department of Medicine, Ohio State University College of Medicine, Columbus, Ohio. Reprint requests should be addressed to Dr. Calvin M. Kunin, Department of Medicine, Ohio State University College of Medicine, 410 West 10th Avenue. Columbus. Ohio 43210. Manuscriot accepted June 16, 1961.
This review was stimulated by the current interest in use of singledose therapy for uncomplicated “lower tract” infection in females and the potential benefit of long-term prophylaxis for patients with recurrent infections. Duration of therapy is only a tactic. It is dependent on understanding the natural history of urinary tract infections in relation to risk factors and the predictable response to treatment. Based on the pertinent literature, a series of elements are presented that define the current consensus concerning the definition, natural history and risk of renal damage from urinary tract infection. These are then considered in relation to the current diagnostic measures and procedures to localize infection. Single-dose therapy combined with bacteriologic monitoring appears to be a useful method to localize infection. Although it defines individuals who may require more prolonged treatment, it has not yet been shown to predict risk of renal damage or identify a subpopulation requiring further study. The major predictors of renal injury are anatomic and neurologic lesions that alter urine flow and host factors that decrease resistance to infection. These are currently betfer defined by individual patient characteristics and clinical observation than by localization studies. Long-term low-dose prophylaxis has been shown to be an effective means of management of highly recurrent episodes of infection. It does not, however, appear to prevent recurrences, afler therapy has been discontinued, even afler periods of prophylaxis as long as six months. Treatment should be based on reasonable expectancy of reduction in morbidity and/or renal damage. During the past decade, two diverse, but effective, tactics for antimicrobial therapy of female patients with urinary tract infections have emerged. At first glance, these methods appear to be poles apart, but each has the special merit of decreasing morbidity while minimizing cost and side effects due to antibiotics. One tactic, used in uncomplicated infections, is to administer a single dose (or at most, just a few doses over just a few days) of an effective agent combined with follow-up cultures to prove efficacy [l-6]. The other, for managing highly recurrent infections with closely spaced episodes, involves small doses of an effective agent for six months or more [7]. Several intermediate-term courses of therapy are still valuable. These are rouahlv _ - 10 days of treatment for oatients likely to have tissue invasion and four- to six-week courses for patients with persistent foci of infection [8,9]. Each of these tactics has its own special merit and ardent advocates. As with any advance in therapeutics, the wide variety of choices may bewilder the nonspecialist, who must manage most
November 1981
The American Journal of Medicine
Volume 71
849
TREATMENT OF URINARY TRACT INFECTIONS-KUNIN
cases. This brief review was stimulated by publication in this issue (Greenberg RN et al.; 841-845) of a report demonstrating limited efficacy of a new oral cephalosporin used for single-dose therapy. I will attempt to clarify some of the issues and provide some guidelines for management. The various therapeutic and diagnostic tactics must be in accord with the current consensus regarding the natural history of urinary tract infections and their response to therapy [lo]. As the “givens” change, so will the tactics. The current consensus may be summarized as follows: The Current Consensus (1) Intelligent management of urinary tract infections requires diagnosis based on microscopic examination at least, preferably with quantitative culture of the urine. Symptoms of frequency, urgency and dysuria are insufficient for diagnosis; about half of the female patients with these complaints do not have significant bacteriuria, but rather the “acute urethral syndrome” [ 1 l]. Bacterial or chlamydial infection as a cause of the “urethral syndrome” may be suspected in the absence of significant bacteriuria when pyuria (more than 10 white blood cells per high-power field) is present. Suspicions may be confirmed by repeated study or suprapubic bladder aspiration. Overdlagnosis may lead to unnecessary treatment and unwarranted tests. (2) Although urinary tract infections are common in the female, they rarely lead to renal damage sufficient to produce end-stage renal disease. Uncomplicated infections, with low risk of producing renal damage, are defined as those not associated with structural or neurologic abnormalities, foreign bodies, or abnormal host defenses (such as in the presence of diabetes mellitus or after renal transplantation). (3) The goal of management of female patients with uncomplicated urinary tract infections is to decrease morbidity from recurrent infections. Therefore, treatment of prophylaxis should be directed against symptomatic episodes. Asymptomatic bacteriuria need not be treated unless it has been shown to predict occurrence of symptomatic episodes and treatment has been shown to be effective in preventing morbidity (such as in pregnancy, in women with highly recurrent infections or in men with persistent infection). (4) The goal of management of female patients with complicated urinary tract infections (as defined above) is to eradicate the focus of infection and provide prophylaxis, when needed, to prevent frequent recurrence. (5) Most recurrent infections in women (about 80 percent) are due to reinfection with a new serotype of Escherichia coli or a new bacterial species derived from the gut and transferred to the periurethral region. These infections respond well to short courses of therapy; frequent recurrences may be prevented by bedtime chemoprophylaxis. A smaller proportion of recurrences (about 20 percent) is due to failure of antimicrobial therapy to eradicate the organism from the urinary tract.
850
November 1981
The American Journal of Medicine
(6) Male or female patients with complicated infections or in whom recurrence is due to persistence of an organism in the urinary tract require prolonged therapy (four to 12 weeks) with an agent to which the bacteria are susceptible so that the focus of infection may be eradicated. (7) The female patients who will require more prolonged therapy can be predicted with reasonable accuracy by various “localization” studies, such as detection of antibody-coated bacteria (ACB), elevated C-reactive protein, bladder washout methods, ureteral catheterization, urinary alkaline phosphatase isoenzymes and beta-2-microglobulin excretion. The differentiation between relapse and reinfection may be even more clearly made by speciation, serologic typing, biotyping and antibiogram studies of the infecting bacteria. All of these tests are expensive and time-consuming and add to medical costs. Clinical guides are more practical and about equally useful. These include: reappearance of bacteriuria within a few days of single-dose therapy, delay in seeking medical attention until six or more days after onset of symptoms, findings consistent with “complicated” infection and fever, loin pain or other clinical signs or symptoms of upper tract infection. (6) In female patients, treatment of each episode of infection, even with eradication of a tissue focus, only modestly reduces the probability of recurrence. When recurrences are frequent and closely spaced, they may be effectively prevented by small bedtime doses of agents such as trimethoprim, trimethoprim/sulfamethoxazole, nitrofurantoin and several other agents. It makes no difference whether the agents eradicate periurethral bacteria or simply provide antibacterial activity in the bladder urine. Recurrent infections occur after cessation of prophylaxis even when given as long as six months [7]. (9) The most effective way to prevent further morbidity is to allow the natural history of the patient’s problem to be displayed, over time. This is readily achieved by inexpensive culture methods (such as the dip-slide) and by monitoring recurrent symptoms. For patients treated only once without adequate follow-up, such as in emergency rooms, therapy should not be limited to a single dose, but should be more prolonged. In some studies, three days of treatment for recurrent, uncomplicated infection appears to be as effective as treatment given for 10 days or more [ 12,131. For patients for whom good follow-up is available and who meet the criteria of uncomplicated infection, single-dose treatment with selected antimicrobials (see Table I, Group I) is effective, is inexpensive and may be associated with fewer side effects and fewer alterations in the bowel flora. Early recurrence within a few days of treatment strongly suggests tissue infection that requires more prolonged therapy. (10) Urologic studies, including intravenous pyelography, are not cost-effective for adult female patients with “uncomplicated” infections [ 14-161. Such studies are of considerable value in persons at high risk of having complicated infections (those with overt pyelonephritis, stones, diabetes mellitus or neurogenic bladders, or those with repeated lack of response to chemotherapy with an effective agent).
Volume 71
TREATMENT OF URINARY TRACT INFECTIONS-KUNIN
TABLE I
Guidelines to Management Probability Easeof
Group Management
Type of Patient
Clinical Characteristics
of Tissue Invasion
Organism
Therapy
I
Excellent
Female, child or adult
Few previous episodes: Usually E. coli sensitive to most agents reliable, with good follow-up available; less than 2 days between onset of symptoms and treatment
Low
One dose amoxicillin. sulfonamide, TMP/SMZ. kanamycin
II
Good
Female, child or adult
Few previous episodes: follow-up poor
Usually E. coli sensitive to most agents
High or low
3-10 days Prophylaxis for closely spaced recurrences
Ill
Fair
Female, child or adult
Many previous episodes; history of early recurrence, or diabetic, or renal transplantation
Variable, tends to have more resistant bacteria, susceptibility tests essential
High
4-6 weeks Prophylaxis for closely spaced recurrences
IV
Fair
Male, adult
Recurrent infections, some underlying anatomic abnormality
Variable, susceptibility tests needed
High, often prostatic colonization
4-12 weeks Prophylaxis for closely spaced recurrences
V
Poor
Male or female
Neurogenic bladder, large volume residual urine
Variable, susceptibility tests needed
High
Intermittent catheterization (treatment for sympiomaiic infections only)
VI
Very poor
Male or female
Continuous drainage required
Variable, susceptibility tests needed
Very high
Indwelling catheter closed drainage (treatment for sepsis only)
Note: TMP/SMZ = irimethoprim
with sulfamethoxazole.
(11) Men tend to have “complicated” infections. Recurrence is usually due to failure to eradicate the organism from a tissue site. The prostate is a common reservoir of infection. Therapy should be prolonged (four to six weeks) in most cases. Prophylaxis may be useful in reducing morbidity with recurrent infection arising from a focus that cannot be effectively eradicated [a]. (12) Patients with foreign bodies in the urinary tract (stones, catheters) will not have response to therapy; rather, a bacterial flora that is resistant to drugs will develop. Therapy is reserved for episodes of sepsis. (13) Relief of obstruction, when present, is the single most effective method of managing complicated urinary tract infections. Intermittent catheterization achieves this goal without the necessity (and inherent risks) of the continued presence of a foreign body. These 13 “given& are the basis for the recommendations concerning diagnosis and follow-up shown in Table II and the management guide listed in Table I. These tables are only rough outlines and will require continued modification and refinement, but they reflect fairly well the current consensus. The major unanswered question is whether localization studies provide additional helpful information to the clinician. Theoretically, those infections that are associated with tissue invasion (“renal bacteriuria”) should be more difficult to manage and might be expected to indicate a greater potential for renal damage than “bladder bacteriuria.” This concept
may be illusory. For example, many of the localization methods are inexact and their predictive value has been repeatedly challenged [ 17-191. In the report by Greenberg et al., in this issue, slight changes in the definition of what constitutes a positive ACB result led to different assessments of response to therapy. A positive C-reactive protein, which had previously been reported as predictive of upper tract infection in children [ 181, was not predictive in this adult population. Ii is also too much to expect that a test done only at one time can accurately assess clinical events that evolve over the natural course of an illness. Many female patients with uncomplicated urinary tract infections (as many as 50 percent) will be found by various localization procedures to have evidence of upper tract infection. Yet extensive urologic studies in similar populations rarely lead to detection of significant structural abnormalities [ 14- 161. These tests, therefore, cannot be expected to detect so-called high-risk populations with great accuracy. It is much too premature to use localization studies to identify persons who require further evaluation, although some studies suggest this possibility [2]. The most we can now say is that high-risk patients respond poorly to single-dose therapy.
SHORT-COURSE
CHEMOTHERAPY
The tactic of using short courses of chemotherapy for treatment of uncomplicated urinary tract infections is now abundantly supported. Such treatment consists of administering a single dose or three days’ dosage of an antimicrobial agent. Although several recent studies
November 1981
The American Journal of Medicine
Volume 71
851
TREATMENT OF URINARY TRACT INFECTIONS-KUNIN
TABLE
II
Diagnostic and Follow-up Urinary Tract Infections
Methods for
Microscopic examination of urinary sediment or Gram stain of uncentrifuged urine Inexpensive and highly reliable; should be routine. Quantitative urine culture Dip-slide tests recommended for routine ambulatory care: inexpensive highly reliable and allow self-testing by patient. Pyuria Helpful in distinguishing between “low bacterial count” urinary tract infections and other causes of the “acute urethral syndrome.” Susceptibility tests Not needed for management of acute uncomplicated infection particularly when combined with follow-up dip-slide cultures. Essential for management of recurrent or complicated infection. Antibody-coated bacteria, C-reactive protein or other “locaiization” tests Useful mainly for research studies. Sensitivity and specificity, particularly in children, have been questioned. Increases costs of management. One can predict “tissue invasion” about as well by “clinical findings.” Tissue invasion is likely in male patients, in female patients with urinary symptoms of six or more days’ duration, in early recurrence after single-dose therapy, in low socio-economic groups, in renal transplantation patients, and in patients with diabetes mellitus, with structural or neurologic abnormalities or foreign bodies, or with fever, ieukocytosis or classic findings of acute pyelonephritis. Serologic typing and antibiograms to distinguish between reinfection and relapse in recurrent infections Highly desirable for assessment of drug efficacy in clinical trials; less practical for routine clinical care. Alteration of species or antibiograms suggests reinfection; relapse with same organism is usual in men and in patients with complicated infections: reinfection with a new organism accounts for about 80 percent of recurrences in uncomplicated infections in female patients. Natural history of each patient Urinary tract infections vary greatly from patient to patient. Some patients may have only one infection, or only a few scattered over time. Others may have frequent occurrences. No single test is 100 percent accurate in defining whether there is only bladder colonization or tissue invasion. Therefore, management must be tailored to the individual patient.
used amoxicillin in relatively large single doses, many other agents, including sulfonamides, nitrofurantoin, trimethoprim/sulfamethoxazole and tetracycline derivatives, have also been shown to be effective. The results of the study in which cefaclor was used, reported in this issue, are disappointing. Even though a single dose was effective in persons who were ACB-negative, the overall efficacy was inferior to 10 days of treatment because many persons in the population were ACBpositive. This latter factor is the basic problem with this approach to management. Nevertheless the single-dose treatment offers several tactical advantages over more prolonged therapy. Drug costs and side effects are considerably reduced, and
852
November
1981
The American Journal of Medicine
Volume 71
the microbiologic flora of the rectum, vagina and periurethral zone are less likely to be altered. In addition, the concept is of great interest because failure to achieve an immediate bacteriologic response to short-course chemotherapy, judged by failure to eradicate bacteriuria within 2 to 3 days after treatment, is a useful clinical guide to identifying patients with more deeply seated tissue infection [l-5]. This approach is offered as a much less expensive and time-consuming method to localize infection than that of laboratory tests (the seventh consensus). As with many advances in medicine, the issues are often not as straightforward as they may appear. Misinterpretation or inappropriate application of otherwise excellent data may lead to poor practice. Consider, for example, the tendency of some physicians to diagnose and manage urinary tract infections over the telephone or the transient and episodic nature of management offered in emergency rooms or “emergent care centers.” For these reasons, the constraints of this approach must be carefully defined, as summarized below: l The physician is still obligated to obtain objective information to prove that the patient truly has a urinary tract infection (the first consensus). Follow-up cultures are necessary to determine whether bacteriuria has been eliminated. Failure to respond may be due to one of four reasons: (1) the patient did not take the drug, (2) the organism was initially resistant to the agent selected, (3) the focus of infection was not eradicated, or (4) reinfection occurred with a new organism. In general, reinfection usually occurs somewhat later than relapse, but there is considerable overlap. This fact may cause confusion if several weeks are allowed to elapse before the follow-up culture is done. Patients who are feeling well may not return for follow-up unless some simple, inexpensive method such as the self-administered, mailed dip-slide is used.
l Even with more prolonged conventional therapy or long-term prophylaxis, about 80 percent of female patients will have recurrent infection within a year. The same appears to be true after short courses. Some may argue that asymptomatic recurrences of bacteriuria are of no importance for most women: however, early detection of recurrence may be of great importance for persons at high risk of symptomatic illness or with complications such as diabetes mellitus, pregnahdg, neurologic problems or structural abnormalities. Thus, the critical part of management is not necessarily the duration of treatment but careful follow-up (ninth consensus point).
l Women with urinary tract infections are not a homogeneous group. Several important clinical predictors
TREATMENT OF URINARY TRACT INFECTIONS-KUNIN
identify
those
who are unlikely
to respond
to a short
course of treatment. In a large multicenter trial, reported by Rubin et al. [5], those women who were ACB-positive (32.1 percent, indicating probable tissue invasion) more often had symptoms for more than six days before treatment. Short-course treatment was less likely to be effective in this group of patients, who had less immediate access to medical care. The short-term “cure rate” of single-dose treatment with amoxicillin of women who were likely to have tissue invasion was poor (33 percent) compared with the cure rate of a lo-day course of ampicillin or trimethoprim/sulfamethoxazole (84.6 and 93.3 percent, respectively). In women with ACB-negative infections, the cure rates of single-dose therapy and lo-day therapy were equally good. These results are similar to those of Greenberg et al., published in this issue. Only 43 percent of the overall population had response to a single dose of cefaclor compared with 95 percent treated for 10 days, even though the segment of the population who were ACB-negative had good response to a single dose. Thus, immediate institution of longer courses of therapy would be preferred for patients whose management has been delayed, those whose follow-up might be poor, those who may have diabetes mellitus and those with structural or neurologic problems. In some cases, even longer courses of four to 12 weeks may be needed to effectively eradicate the infection. Thus it seems reasonable to treat “transient patients” with at least three to 10 days of therapy, even initially. These comments are not designed to disparage the remarkable efficacy of single-dose chemotherapy for selected patients, but rather to emphasize the responsibility of the physician to assess the probability of success both on clinical grounds and on his estimate of the probability of a patient’s compliance with careful follow-up.
For patients with frequent, closely spaced recurrent infection, a regimen of small doses of an effective antimicrobial agent at bedtime offers considerable advantage in management. The concept is that if antibacterial activity is maintained in the bladder urine or the periurethral zone (the exact location does not appear to matter), then reinfection may be blocked. This tactic works well and has been shown by several groups to be remarkably effective in decreasing morbidity among female patients [7]. A similar approach has been reported to decrease the frequency of acute episodes in male patients as well [8]. Certain cautions must be exercised when this approach is used.
is resistant
l There is a temptation to use long-term prophylaxis when actually the patient has a tissue focus that needs to be eradicated by a shorter course of more intensive therapy. Thus, a patient with a relapsing focus must be differentiated from one who is becoming reinfected. Relapse and reinfection can be differentiated in part by examining the species, biotype and antibiogram of the organism without using expensive serologic methods, or simply on clinical grounds. Patients with complicated infection, for example, are more likely to suffer from relapse than from reinfection and will not have response to a single dose of an agent to which the organism is susceptible.
l A surprising, but well documented, series of observations in females, both children and adults, indicates that regardless of the duration of prophylaxis (at least up to six months), recurrent infection, probably due to reinfection, will occur at about the same rate as after a short course of treatment [7]. This fact implies that the biologic determinants of reinfection are not altered by this duration of prophylaxis. Thus, follow-up is still needed after a prophylactic course is completed. This should not be discouraging, however, provided the patients and physician are aware that the prophylactic tactic may have to be resumed or that short courses of therapy may be used for less closely spaced episodes.
COURSES OF INTERMEDIATE
LONG-TERM PROPHYLAXIS
l Periodic urine cultures are needed to be certain that prophylaxis continues to be effective. Continuing
therapy with a drug to which the organism would be wasteful.
DURATION
Although some studies compare single-dose therapy with 10 days of treatment [l-5] and others compare therapies of three days with those of 10 or more days [ 12,131, I am not aware of studies comparing all of these variables. Furthermore, the single-dose studies were usually conducted in combination with some assessment of “localization,” whereas the three-versus-l O-or-more-day studies were more concerned with general populations. Examination of this “mixed bag” of reports leads to the tentative conclusion that three days of treatment may be about as effective as 10 days in uncomplicated infections in female patients. We might extrapolate then (with some trepidation) that for general populations in which localization studies are not desired or needed, three-day treatment may be more cost-effective. Obviously, some patients will have relapses, but this can be detected by follow-up culture or recurrence of symptoms. Further comparative efficacy studies in representative populations are needed before therapy of three rather than 10 days’ duration becomes standard management for uncomplicated infections in female patients. The major reason for this conservative approach is because of the heterogeneity of the pop-
November 1981
The American Journal of Medicine
Volume 71
853
TREATMENT OF URINARY TRACT INFECTIONS-KUNIN
already have evidence of “tissue invasion.” Three days of treatment appears as effective as 10 days for many persons with uncomplicated infections. Follow-up culture or emergence of recurrent symptomatic illness will define those who require more prolonged treatment or further diagnostic studies. Closely spaced recurrent episodes are readily managed by prolonged bedtime prophylaxis, but this treatment does not eliminate future reinfection. Infection in men and relapsing infection in women may often require prolonged treatment to eradicate a tissue focus. Most important, however, is the need to emphasize that the frequency of recurrence, the magnitude of symptoms, the localization of infection to the “upper tract” or tissue invasion do not appear to be the most important determinants of long-term outcome. Far more important predictors of renal injury are anatomic and neurologic lesions that alter urine flow and host factors, such as diabetes mellitus or polycystic kidney disease or renal transplantation, which decrease resistance to infection. There are the “complicated” infections. Finally, the major objective of therapy in “uncomplicated infections” is relief of symptoms. Treatment of asymptomatic infections can be justified only when there is reasonable certainty that it can prevent future morbidity.
ulations that individual investigators define and include in their studies as having “uncomplicated” urinary tract infections. More prolonged treatment, broadly defined as extending over four to six weeks, appears to be useful for management of patients who have relapses with the same organism soon after short courses (three to 10 days) of treatment. Longer periods up to 12 weeks may be required in some men with more complex problems [20-211. Obviously, no magic number of days fits all cases. It seems to me, however, that the clinician desires some recommendation of a treatment that will most likely cure infection in his patient without the burden of excessive testing. Perhaps the broad recommendations in Table I will meet these needs. CONCLUDING
STATEMENT
The duration of therapy in management of urinary-tract infection is no more than an empiric tactic found to be most useful to deal with specific clinical problems. Single-dose treatment for female patients with uncomplicated infections can be an effective means of localizing infection, but requires closely timed follow-up studies. It is not as effective as 10 days of therapy for most populations, which includes many persons who
REFERENCES
4.
5.
6. 7.
8.
9.
10.
854
Ronald HR, Boutros P, Mourtada H: Bacteriuria localization and response to single-dose therapy in women. JAMA 1976; 235: 1854-1856. Bailey RR, Abbott GD: Treatment of urinary tract infection with a single dose of amoxicillin. Nephron 1977; 18: 316320. Fang LST. Tolkoff-Tubin NE, Rubin RH: Efficacy of single-dose and conventional amoxicillin therapy in urinary-tract infection localized by the antibody-coated bacteria technique. N Engl J Med 1978; 298: 413-416. Anderson JD, Aird MY, Johnson AM: The use of a single large dose of amoxicillin for the treatment of acute urinary tract infections. J Antimicrob Agents Chemother 1978; 5: 481-483. Rubin RH, Fang LST, Jones Sr, et al.: Single-dose amoxicillin therapy for urinary tract infection: multicenter trial using antibody-covered bacteria localization technique. JAMA 1980; 244: 561-564. Stamm WE: Single-dose treatment of cystitis. JAMA 1980; 244: 591-592. Stamm WE, Counts GW, Wagner KF, et al.: Antimicrobial prophylaxis of recurrent urinary tract infections; a double-blind, placebo-controlled trial. Ann Intern Med 1980; 92: 770-775. Freeman RB, Smith WM, Richardson JA, et al.: Long-term therapy for chronic bacteriuria in men; U.S. Public Health Service Corporation. Ann Intern Med 1975; 83: 133147. Turck M, Petersdorf RG: Editorial: optimal duration of treatment of chronic urinary tract infection. Ann Intern Med 1968; 69: 837-839. Kunin CM: Detection, prevention and management of urinary tract infections. Philadelphia: Lea & Febiger, 1979.
November
1981
The American
Journal of Medicine
11.
12.
13. 14.
15.
16.
17.
18.
19.
20.
21.
Volume 71
Stamm WE, Wagner KF, Amsel R, et al.: Causes of the acute urethral syndrome in women. N Engl J Med 1980; 303: 409-415. Charlton CAC, Crowther A, Davies JG, et al.: Three-day and tenday chemotherapy for urinary tract infections in general practice. Br Med J 1976; 1: 124-126. Fair WR. Crane DB, Peterson LJ, et al.: Three-day treatment of urinary tract infections. J Urol 1980; 123: 717-721. Fair WR, McClennan BL. Jost RG: Are excretory urograms necessary in evaluating women with urinary tract infection? J Urol 1979; 121: 313-315. Engel G, Schaeffer AJ, Grayhack JT, Wendel EF: The role of excretory urography and cystoscopy in the evaluation and management of women with recurrent urinary tract infection J Urol 1980; 123: 190-191. Fowler JE, Pulaski ET: Excretory urography, cystography, and cystoscopy in the evaluation of women with urinary-tract infection. N Engl J Med 1981; 304: 462-465. Rumans LW, Vosti KL: The relationship of antibody-coated bacteria to clinical syndromes as found in unselected populations with bacteriuria. Arch Intern Med 1978; 138: 1077-1081. Wientzen RL, McCracken GH Jr, Petruska ML, et al.: Localization and therapy of urinary tract infections of childhood. Pediatrics 1979; 63: 467-474. Mundt KA, Polk BF: Identification of site of urinary-tract infections by antibody-coated bacteria assay. Lancet 1979; II: 1172-1176. Gleckman R, Crowley M, Natsios GA: Therapy of recurrent invasive urinary-tract infections of men. N Engl J Med 1979; 30 1: 878-880. Smith JW, Jones SR, Reed WP, et al.: Recurrent urinary tract infections in men. Ann Intern Med 1979; 91: 544-548.
Duration of Treatment of Urinary Tract Infections
CALVIN M. KUNIN, M.D. Colurrtbus, Ohio
From the Department of Medicine, Ohio State University College of Medicine, Columbus, Ohio. Reprint requests should be addressed to Dr. Calvin M. Kunin, Department of Medicine, Ohio State University College of Medicine, 410 West 10th Avenue. Columbus. Ohio 43210. Manuscriot accepted June 16, 1961.
This review was stimulated by the current interest in use of singledose therapy for uncomplicated “lower tract” infection in females and the potential benefit of long-term prophylaxis for patients with recurrent infections. Duration of therapy is only a tactic. It is dependent on understanding the natural history of urinary tract infections in relation to risk factors and the predictable response to treatment. Based on the pertinent literature, a series of elements are presented that define the current consensus concerning the definition, natural history and risk of renal damage from urinary tract infection. These are then considered in relation to the current diagnostic measures and procedures to localize infection. Single-dose therapy combined with bacteriologic monitoring appears to be a useful method to localize infection. Although it defines individuals who may require more prolonged treatment, it has not yet been shown to predict risk of renal damage or identify a subpopulation requiring further study. The major predictors of renal injury are anatomic and neurologic lesions that alter urine flow and host factors that decrease resistance to infection. These are currently betfer defined by individual patient characteristics and clinical observation than by localization studies. Long-term low-dose prophylaxis has been shown to be an effective means of management of highly recurrent episodes of infection. It does not, however, appear to prevent recurrences, afler therapy has been discontinued, even afler periods of prophylaxis as long as six months. Treatment should be based on reasonable expectancy of reduction in morbidity and/or renal damage. During the past decade, two diverse, but effective, tactics for antimicrobial therapy of female patients with urinary tract infections have emerged. At first glance, these methods appear to be poles apart, but each has the special merit of decreasing morbidity while minimizing cost and side effects due to antibiotics. One tactic, used in uncomplicated infections, is to administer a single dose (or at most, just a few doses over just a few days) of an effective agent combined with follow-up cultures to prove efficacy [l-6]. The other, for managing highly recurrent infections with closely spaced episodes, involves small doses of an effective agent for six months or more [7]. Several intermediate-term courses of therapy are still valuable. These are rouahlv _ - 10 days of treatment for oatients likely to have tissue invasion and four- to six-week courses for patients with persistent foci of infection [8,9]. Each of these tactics has its own special merit and ardent advocates. As with any advance in therapeutics, the wide variety of choices may bewilder the nonspecialist, who must manage most
November 1981
The American Journal of Medicine
Volume 71
849
TREATMENT OF URINARY TRACT INFECTIONS-KUNIN
cases. This brief review was stimulated by publication in this issue (Greenberg RN et al.; 841-845) of a report demonstrating limited efficacy of a new oral cephalosporin used for single-dose therapy. I will attempt to clarify some of the issues and provide some guidelines for management. The various therapeutic and diagnostic tactics must be in accord with the current consensus regarding the natural history of urinary tract infections and their response to therapy [lo]. As the “givens” change, so will the tactics. The current consensus may be summarized as follows: The Current Consensus (1) Intelligent management of urinary tract infections requires diagnosis based on microscopic examination at least, preferably with quantitative culture of the urine. Symptoms of frequency, urgency and dysuria are insufficient for diagnosis; about half of the female patients with these complaints do not have significant bacteriuria, but rather the “acute urethral syndrome” [ 1 l]. Bacterial or chlamydial infection as a cause of the “urethral syndrome” may be suspected in the absence of significant bacteriuria when pyuria (more than 10 white blood cells per high-power field) is present. Suspicions may be confirmed by repeated study or suprapubic bladder aspiration. Overdlagnosis may lead to unnecessary treatment and unwarranted tests. (2) Although urinary tract infections are common in the female, they rarely lead to renal damage sufficient to produce end-stage renal disease. Uncomplicated infections, with low risk of producing renal damage, are defined as those not associated with structural or neurologic abnormalities, foreign bodies, or abnormal host defenses (such as in the presence of diabetes mellitus or after renal transplantation). (3) The goal of management of female patients with uncomplicated urinary tract infections is to decrease morbidity from recurrent infections. Therefore, treatment of prophylaxis should be directed against symptomatic episodes. Asymptomatic bacteriuria need not be treated unless it has been shown to predict occurrence of symptomatic episodes and treatment has been shown to be effective in preventing morbidity (such as in pregnancy, in women with highly recurrent infections or in men with persistent infection). (4) The goal of management of female patients with complicated urinary tract infections (as defined above) is to eradicate the focus of infection and provide prophylaxis, when needed, to prevent frequent recurrence. (5) Most recurrent infections in women (about 80 percent) are due to reinfection with a new serotype of Escherichia coli or a new bacterial species derived from the gut and transferred to the periurethral region. These infections respond well to short courses of therapy; frequent recurrences may be prevented by bedtime chemoprophylaxis. A smaller proportion of recurrences (about 20 percent) is due to failure of antimicrobial therapy to eradicate the organism from the urinary tract.
850
November 1981
The American Journal of Medicine
(6) Male or female patients with complicated infections or in whom recurrence is due to persistence of an organism in the urinary tract require prolonged therapy (four to 12 weeks) with an agent to which the bacteria are susceptible so that the focus of infection may be eradicated. (7) The female patients who will require more prolonged therapy can be predicted with reasonable accuracy by various “localization” studies, such as detection of antibody-coated bacteria (ACB), elevated C-reactive protein, bladder washout methods, ureteral catheterization, urinary alkaline phosphatase isoenzymes and beta-2-microglobulin excretion. The differentiation between relapse and reinfection may be even more clearly made by speciation, serologic typing, biotyping and antibiogram studies of the infecting bacteria. All of these tests are expensive and time-consuming and add to medical costs. Clinical guides are more practical and about equally useful. These include: reappearance of bacteriuria within a few days of single-dose therapy, delay in seeking medical attention until six or more days after onset of symptoms, findings consistent with “complicated” infection and fever, loin pain or other clinical signs or symptoms of upper tract infection. (6) In female patients, treatment of each episode of infection, even with eradication of a tissue focus, only modestly reduces the probability of recurrence. When recurrences are frequent and closely spaced, they may be effectively prevented by small bedtime doses of agents such as trimethoprim, trimethoprim/sulfamethoxazole, nitrofurantoin and several other agents. It makes no difference whether the agents eradicate periurethral bacteria or simply provide antibacterial activity in the bladder urine. Recurrent infections occur after cessation of prophylaxis even when given as long as six months [7]. (9) The most effective way to prevent further morbidity is to allow the natural history of the patient’s problem to be displayed, over time. This is readily achieved by inexpensive culture methods (such as the dip-slide) and by monitoring recurrent symptoms. For patients treated only once without adequate follow-up, such as in emergency rooms, therapy should not be limited to a single dose, but should be more prolonged. In some studies, three days of treatment for recurrent, uncomplicated infection appears to be as effective as treatment given for 10 days or more [ 12,131. For patients for whom good follow-up is available and who meet the criteria of uncomplicated infection, single-dose treatment with selected antimicrobials (see Table I, Group I) is effective, is inexpensive and may be associated with fewer side effects and fewer alterations in the bowel flora. Early recurrence within a few days of treatment strongly suggests tissue infection that requires more prolonged therapy. (10) Urologic studies, including intravenous pyelography, are not cost-effective for adult female patients with “uncomplicated” infections [ 14-161. Such studies are of considerable value in persons at high risk of having complicated infections (those with overt pyelonephritis, stones, diabetes mellitus or neurogenic bladders, or those with repeated lack of response to chemotherapy with an effective agent).
Volume 71
TREATMENT OF URINARY TRACT INFECTIONS-KUNIN
TABLE I
Guidelines to Management Probability Easeof
Group Management
Type of Patient
Clinical Characteristics
of Tissue Invasion
Organism
Therapy
I
Excellent
Female, child or adult
Few previous episodes: Usually E. coli sensitive to most agents reliable, with good follow-up available; less than 2 days between onset of symptoms and treatment
Low
One dose amoxicillin. sulfonamide, TMP/SMZ. kanamycin
II
Good
Female, child or adult
Few previous episodes: follow-up poor
Usually E. coli sensitive to most agents
High or low
3-10 days Prophylaxis for closely spaced recurrences
Ill
Fair
Female, child or adult
Many previous episodes; history of early recurrence, or diabetic, or renal transplantation
Variable, tends to have more resistant bacteria, susceptibility tests essential
High
4-6 weeks Prophylaxis for closely spaced recurrences
IV
Fair
Male, adult
Recurrent infections, some underlying anatomic abnormality
Variable, susceptibility tests needed
High, often prostatic colonization
4-12 weeks Prophylaxis for closely spaced recurrences
V
Poor
Male or female
Neurogenic bladder, large volume residual urine
Variable, susceptibility tests needed
High
Intermittent catheterization (treatment for sympiomaiic infections only)
VI
Very poor
Male or female
Continuous drainage required
Variable, susceptibility tests needed
Very high
Indwelling catheter closed drainage (treatment for sepsis only)
Note: TMP/SMZ = irimethoprim
with sulfamethoxazole.
(11) Men tend to have “complicated” infections. Recurrence is usually due to failure to eradicate the organism from a tissue site. The prostate is a common reservoir of infection. Therapy should be prolonged (four to six weeks) in most cases. Prophylaxis may be useful in reducing morbidity with recurrent infection arising from a focus that cannot be effectively eradicated [a]. (12) Patients with foreign bodies in the urinary tract (stones, catheters) will not have response to therapy; rather, a bacterial flora that is resistant to drugs will develop. Therapy is reserved for episodes of sepsis. (13) Relief of obstruction, when present, is the single most effective method of managing complicated urinary tract infections. Intermittent catheterization achieves this goal without the necessity (and inherent risks) of the continued presence of a foreign body. These 13 “given& are the basis for the recommendations concerning diagnosis and follow-up shown in Table II and the management guide listed in Table I. These tables are only rough outlines and will require continued modification and refinement, but they reflect fairly well the current consensus. The major unanswered question is whether localization studies provide additional helpful information to the clinician. Theoretically, those infections that are associated with tissue invasion (“renal bacteriuria”) should be more difficult to manage and might be expected to indicate a greater potential for renal damage than “bladder bacteriuria.” This concept
may be illusory. For example, many of the localization methods are inexact and their predictive value has been repeatedly challenged [ 17-191. In the report by Greenberg et al., in this issue, slight changes in the definition of what constitutes a positive ACB result led to different assessments of response to therapy. A positive C-reactive protein, which had previously been reported as predictive of upper tract infection in children [ 181, was not predictive in this adult population. Ii is also too much to expect that a test done only at one time can accurately assess clinical events that evolve over the natural course of an illness. Many female patients with uncomplicated urinary tract infections (as many as 50 percent) will be found by various localization procedures to have evidence of upper tract infection. Yet extensive urologic studies in similar populations rarely lead to detection of significant structural abnormalities [ 14- 161. These tests, therefore, cannot be expected to detect so-called high-risk populations with great accuracy. It is much too premature to use localization studies to identify persons who require further evaluation, although some studies suggest this possibility [2]. The most we can now say is that high-risk patients respond poorly to single-dose therapy.
SHORT-COURSE
CHEMOTHERAPY
The tactic of using short courses of chemotherapy for treatment of uncomplicated urinary tract infections is now abundantly supported. Such treatment consists of administering a single dose or three days’ dosage of an antimicrobial agent. Although several recent studies
November 1981
The American Journal of Medicine
Volume 71
851
TREATMENT OF URINARY TRACT INFECTIONS-KUNIN
TABLE
II
Diagnostic and Follow-up Urinary Tract Infections
Methods for
Microscopic examination of urinary sediment or Gram stain of uncentrifuged urine Inexpensive and highly reliable; should be routine. Quantitative urine culture Dip-slide tests recommended for routine ambulatory care: inexpensive highly reliable and allow self-testing by patient. Pyuria Helpful in distinguishing between “low bacterial count” urinary tract infections and other causes of the “acute urethral syndrome.” Susceptibility tests Not needed for management of acute uncomplicated infection particularly when combined with follow-up dip-slide cultures. Essential for management of recurrent or complicated infection. Antibody-coated bacteria, C-reactive protein or other “locaiization” tests Useful mainly for research studies. Sensitivity and specificity, particularly in children, have been questioned. Increases costs of management. One can predict “tissue invasion” about as well by “clinical findings.” Tissue invasion is likely in male patients, in female patients with urinary symptoms of six or more days’ duration, in early recurrence after single-dose therapy, in low socio-economic groups, in renal transplantation patients, and in patients with diabetes mellitus, with structural or neurologic abnormalities or foreign bodies, or with fever, ieukocytosis or classic findings of acute pyelonephritis. Serologic typing and antibiograms to distinguish between reinfection and relapse in recurrent infections Highly desirable for assessment of drug efficacy in clinical trials; less practical for routine clinical care. Alteration of species or antibiograms suggests reinfection; relapse with same organism is usual in men and in patients with complicated infections: reinfection with a new organism accounts for about 80 percent of recurrences in uncomplicated infections in female patients. Natural history of each patient Urinary tract infections vary greatly from patient to patient. Some patients may have only one infection, or only a few scattered over time. Others may have frequent occurrences. No single test is 100 percent accurate in defining whether there is only bladder colonization or tissue invasion. Therefore, management must be tailored to the individual patient.
used amoxicillin in relatively large single doses, many other agents, including sulfonamides, nitrofurantoin, trimethoprim/sulfamethoxazole and tetracycline derivatives, have also been shown to be effective. The results of the study in which cefaclor was used, reported in this issue, are disappointing. Even though a single dose was effective in persons who were ACB-negative, the overall efficacy was inferior to 10 days of treatment because many persons in the population were ACBpositive. This latter factor is the basic problem with this approach to management. Nevertheless the single-dose treatment offers several tactical advantages over more prolonged therapy. Drug costs and side effects are considerably reduced, and
852
November
1981
The American Journal of Medicine
Volume 71
the microbiologic flora of the rectum, vagina and periurethral zone are less likely to be altered. In addition, the concept is of great interest because failure to achieve an immediate bacteriologic response to short-course chemotherapy, judged by failure to eradicate bacteriuria within 2 to 3 days after treatment, is a useful clinical guide to identifying patients with more deeply seated tissue infection [l-5]. This approach is offered as a much less expensive and time-consuming method to localize infection than that of laboratory tests (the seventh consensus). As with many advances in medicine, the issues are often not as straightforward as they may appear. Misinterpretation or inappropriate application of otherwise excellent data may lead to poor practice. Consider, for example, the tendency of some physicians to diagnose and manage urinary tract infections over the telephone or the transient and episodic nature of management offered in emergency rooms or “emergent care centers.” For these reasons, the constraints of this approach must be carefully defined, as summarized below: l The physician is still obligated to obtain objective information to prove that the patient truly has a urinary tract infection (the first consensus). Follow-up cultures are necessary to determine whether bacteriuria has been eliminated. Failure to respond may be due to one of four reasons: (1) the patient did not take the drug, (2) the organism was initially resistant to the agent selected, (3) the focus of infection was not eradicated, or (4) reinfection occurred with a new organism. In general, reinfection usually occurs somewhat later than relapse, but there is considerable overlap. This fact may cause confusion if several weeks are allowed to elapse before the follow-up culture is done. Patients who are feeling well may not return for follow-up unless some simple, inexpensive method such as the self-administered, mailed dip-slide is used.
l Even with more prolonged conventional therapy or long-term prophylaxis, about 80 percent of female patients will have recurrent infection within a year. The same appears to be true after short courses. Some may argue that asymptomatic recurrences of bacteriuria are of no importance for most women: however, early detection of recurrence may be of great importance for persons at high risk of symptomatic illness or with complications such as diabetes mellitus, pregnahdg, neurologic problems or structural abnormalities. Thus, the critical part of management is not necessarily the duration of treatment but careful follow-up (ninth consensus point).
l Women with urinary tract infections are not a homogeneous group. Several important clinical predictors
TREATMENT OF URINARY TRACT INFECTIONS-KUNIN
identify
those
who are unlikely
to respond
to a short
course of treatment. In a large multicenter trial, reported by Rubin et al. [5], those women who were ACB-positive (32.1 percent, indicating probable tissue invasion) more often had symptoms for more than six days before treatment. Short-course treatment was less likely to be effective in this group of patients, who had less immediate access to medical care. The short-term “cure rate” of single-dose treatment with amoxicillin of women who were likely to have tissue invasion was poor (33 percent) compared with the cure rate of a lo-day course of ampicillin or trimethoprim/sulfamethoxazole (84.6 and 93.3 percent, respectively). In women with ACB-negative infections, the cure rates of single-dose therapy and lo-day therapy were equally good. These results are similar to those of Greenberg et al., published in this issue. Only 43 percent of the overall population had response to a single dose of cefaclor compared with 95 percent treated for 10 days, even though the segment of the population who were ACB-negative had good response to a single dose. Thus, immediate institution of longer courses of therapy would be preferred for patients whose management has been delayed, those whose follow-up might be poor, those who may have diabetes mellitus and those with structural or neurologic problems. In some cases, even longer courses of four to 12 weeks may be needed to effectively eradicate the infection. Thus it seems reasonable to treat “transient patients” with at least three to 10 days of therapy, even initially. These comments are not designed to disparage the remarkable efficacy of single-dose chemotherapy for selected patients, but rather to emphasize the responsibility of the physician to assess the probability of success both on clinical grounds and on his estimate of the probability of a patient’s compliance with careful follow-up.
For patients with frequent, closely spaced recurrent infection, a regimen of small doses of an effective antimicrobial agent at bedtime offers considerable advantage in management. The concept is that if antibacterial activity is maintained in the bladder urine or the periurethral zone (the exact location does not appear to matter), then reinfection may be blocked. This tactic works well and has been shown by several groups to be remarkably effective in decreasing morbidity among female patients [7]. A similar approach has been reported to decrease the frequency of acute episodes in male patients as well [8]. Certain cautions must be exercised when this approach is used.
is resistant
l There is a temptation to use long-term prophylaxis when actually the patient has a tissue focus that needs to be eradicated by a shorter course of more intensive therapy. Thus, a patient with a relapsing focus must be differentiated from one who is becoming reinfected. Relapse and reinfection can be differentiated in part by examining the species, biotype and antibiogram of the organism without using expensive serologic methods, or simply on clinical grounds. Patients with complicated infection, for example, are more likely to suffer from relapse than from reinfection and will not have response to a single dose of an agent to which the organism is susceptible.
l A surprising, but well documented, series of observations in females, both children and adults, indicates that regardless of the duration of prophylaxis (at least up to six months), recurrent infection, probably due to reinfection, will occur at about the same rate as after a short course of treatment [7]. This fact implies that the biologic determinants of reinfection are not altered by this duration of prophylaxis. Thus, follow-up is still needed after a prophylactic course is completed. This should not be discouraging, however, provided the patients and physician are aware that the prophylactic tactic may have to be resumed or that short courses of therapy may be used for less closely spaced episodes.
COURSES OF INTERMEDIATE
LONG-TERM PROPHYLAXIS
l Periodic urine cultures are needed to be certain that prophylaxis continues to be effective. Continuing
therapy with a drug to which the organism would be wasteful.
DURATION
Although some studies compare single-dose therapy with 10 days of treatment [l-5] and others compare therapies of three days with those of 10 or more days [ 12,131, I am not aware of studies comparing all of these variables. Furthermore, the single-dose studies were usually conducted in combination with some assessment of “localization,” whereas the three-versus-l O-or-more-day studies were more concerned with general populations. Examination of this “mixed bag” of reports leads to the tentative conclusion that three days of treatment may be about as effective as 10 days in uncomplicated infections in female patients. We might extrapolate then (with some trepidation) that for general populations in which localization studies are not desired or needed, three-day treatment may be more cost-effective. Obviously, some patients will have relapses, but this can be detected by follow-up culture or recurrence of symptoms. Further comparative efficacy studies in representative populations are needed before therapy of three rather than 10 days’ duration becomes standard management for uncomplicated infections in female patients. The major reason for this conservative approach is because of the heterogeneity of the pop-
November 1981
The American Journal of Medicine
Volume 71
853
TREATMENT OF URINARY TRACT INFECTIONS-KUNIN
already have evidence of “tissue invasion.” Three days of treatment appears as effective as 10 days for many persons with uncomplicated infections. Follow-up culture or emergence of recurrent symptomatic illness will define those who require more prolonged treatment or further diagnostic studies. Closely spaced recurrent episodes are readily managed by prolonged bedtime prophylaxis, but this treatment does not eliminate future reinfection. Infection in men and relapsing infection in women may often require prolonged treatment to eradicate a tissue focus. Most important, however, is the need to emphasize that the frequency of recurrence, the magnitude of symptoms, the localization of infection to the “upper tract” or tissue invasion do not appear to be the most important determinants of long-term outcome. Far more important predictors of renal injury are anatomic and neurologic lesions that alter urine flow and host factors, such as diabetes mellitus or polycystic kidney disease or renal transplantation, which decrease resistance to infection. There are the “complicated” infections. Finally, the major objective of therapy in “uncomplicated infections” is relief of symptoms. Treatment of asymptomatic infections can be justified only when there is reasonable certainty that it can prevent future morbidity.
ulations that individual investigators define and include in their studies as having “uncomplicated” urinary tract infections. More prolonged treatment, broadly defined as extending over four to six weeks, appears to be useful for management of patients who have relapses with the same organism soon after short courses (three to 10 days) of treatment. Longer periods up to 12 weeks may be required in some men with more complex problems [20-211. Obviously, no magic number of days fits all cases. It seems to me, however, that the clinician desires some recommendation of a treatment that will most likely cure infection in his patient without the burden of excessive testing. Perhaps the broad recommendations in Table I will meet these needs. CONCLUDING
STATEMENT
The duration of therapy in management of urinary-tract infection is no more than an empiric tactic found to be most useful to deal with specific clinical problems. Single-dose treatment for female patients with uncomplicated infections can be an effective means of localizing infection, but requires closely timed follow-up studies. It is not as effective as 10 days of therapy for most populations, which includes many persons who
REFERENCES
4.
5.
6. 7.
8.
9.
10.
854
Ronald HR, Boutros P, Mourtada H: Bacteriuria localization and response to single-dose therapy in women. JAMA 1976; 235: 1854-1856. Bailey RR, Abbott GD: Treatment of urinary tract infection with a single dose of amoxicillin. Nephron 1977; 18: 316320. Fang LST. Tolkoff-Tubin NE, Rubin RH: Efficacy of single-dose and conventional amoxicillin therapy in urinary-tract infection localized by the antibody-coated bacteria technique. N Engl J Med 1978; 298: 413-416. Anderson JD, Aird MY, Johnson AM: The use of a single large dose of amoxicillin for the treatment of acute urinary tract infections. J Antimicrob Agents Chemother 1978; 5: 481-483. Rubin RH, Fang LST, Jones Sr, et al.: Single-dose amoxicillin therapy for urinary tract infection: multicenter trial using antibody-covered bacteria localization technique. JAMA 1980; 244: 561-564. Stamm WE: Single-dose treatment of cystitis. JAMA 1980; 244: 591-592. Stamm WE, Counts GW, Wagner KF, et al.: Antimicrobial prophylaxis of recurrent urinary tract infections; a double-blind, placebo-controlled trial. Ann Intern Med 1980; 92: 770-775. Freeman RB, Smith WM, Richardson JA, et al.: Long-term therapy for chronic bacteriuria in men; U.S. Public Health Service Corporation. Ann Intern Med 1975; 83: 133147. Turck M, Petersdorf RG: Editorial: optimal duration of treatment of chronic urinary tract infection. Ann Intern Med 1968; 69: 837-839. Kunin CM: Detection, prevention and management of urinary tract infections. Philadelphia: Lea & Febiger, 1979.
November
1981
The American
Journal of Medicine
11.
12.
13. 14.
15.
16.
17.
18.
19.
20.
21.
Volume 71
Stamm WE, Wagner KF, Amsel R, et al.: Causes of the acute urethral syndrome in women. N Engl J Med 1980; 303: 409-415. Charlton CAC, Crowther A, Davies JG, et al.: Three-day and tenday chemotherapy for urinary tract infections in general practice. Br Med J 1976; 1: 124-126. Fair WR. Crane DB, Peterson LJ, et al.: Three-day treatment of urinary tract infections. J Urol 1980; 123: 717-721. Fair WR, McClennan BL. Jost RG: Are excretory urograms necessary in evaluating women with urinary tract infection? J Urol 1979; 121: 313-315. Engel G, Schaeffer AJ, Grayhack JT, Wendel EF: The role of excretory urography and cystoscopy in the evaluation and management of women with recurrent urinary tract infection J Urol 1980; 123: 190-191. Fowler JE, Pulaski ET: Excretory urography, cystography, and cystoscopy in the evaluation of women with urinary-tract infection. N Engl J Med 1981; 304: 462-465. Rumans LW, Vosti KL: The relationship of antibody-coated bacteria to clinical syndromes as found in unselected populations with bacteriuria. Arch Intern Med 1978; 138: 1077-1081. Wientzen RL, McCracken GH Jr, Petruska ML, et al.: Localization and therapy of urinary tract infections of childhood. Pediatrics 1979; 63: 467-474. Mundt KA, Polk BF: Identification of site of urinary-tract infections by antibody-coated bacteria assay. Lancet 1979; II: 1172-1176. Gleckman R, Crowley M, Natsios GA: Therapy of recurrent invasive urinary-tract infections of men. N Engl J Med 1979; 30 1: 878-880. Smith JW, Jones SR, Reed WP, et al.: Recurrent urinary tract infections in men. Ann Intern Med 1979; 91: 544-548.