Urinary Tract Infections

0025-7125/91 $0.00

+ .20

Laboratory in the Diagnosis and Management of Urinary Tract Infections Peter C. Pappas, MD*

The diagnosis of urinary tract infections (UTIs) is usually confirmed on the basis of results from one or more commonly available laboratory diagnostic tests. The importance of laboratory confirmation of a UTI is underscored by the fact that the clinical diagnosis is frequently inaccurate, particularly in women with frequency and dysuria, in whom one must consider other causes of these symptoms, including vaginitis, urethritis due to Chlamydia trachomatis or other urethral pathogens, urethritis due to physical or chemical agents, urethral trauma, and symptoms without any recognized cause. 20 • 21 Thus, along with a careful history and physical examination, urine laboratory studies become essential tools in the evaluation of patients with symptoms suggestive of a UTI. Of additional importance is the use of urine laboratory studies in screening asymptomatic individuals, such as pregnant women, who are at high risk for developing significant complications of a UTI. The purpose of this article is to describe the important laboratory tests for the evaluation of patients with UTIs, including the routine assessment of pyuria and bacteriuria, rapid assays of pyuria and bacteriuria, and quantitative urine cultures, and discuss the utility of these studies in the diagnosis and management of UTIs. Other applications of the laboratory, such as localization techniques to distinguish between upper and lower UTI, and practical office procedures to assist the clinician in the outpatient diagnosis and management of patients with suspected UTIs are included in this article. URINALYSIS A carefully performed urinalysis, including a microscopic examination of the urine, cannot be overemphasized in the assessment of a patient with *Assistant Professor of Medicine, Division of Infectious Diseases, University of Alabama School of Medicine, Birmingham, Alabama

Medical Clinics of North America-Vo!. 75, No. 2, March 1991

313

314

PETEH

C.

PAPPAS

possible UTI. The standard urinalysis should include a description of the color, measurement of specific gravity and pH, and a determination of the concentration of glucose, protein, ketones, blood, and bilirubin. These de terminations are usually made by use of a dipstick, which allows rapid interpretation of these results, usually within 1 minute. Measurement of pyuria and bacteriuria, two key indicators of UTI, can be made by direct microscopic examination of urine (centrifuged or uncentrifuged) or by newer, more rapid assays using the dipstick method. The importance of these determinations and their correlation with culture-proven UTIs are discussed later. Pyuria

One of the most important and readily available laboratory tests in patients with suspected UTIs is the detection of pyuria. 50 Pyuria is present in almost all symptomatic UTIs, and its absence should strongly suggest another diagnosis. The most accurate method of measuring pyuria is to measure urinary leukocyte excretion rate. 2 An excretion rate of 400,000 leukocytes/hour or greater correlates with symptomatic UTI, whereas excretion rates of less than 400,000 leukocytes/hour are generally not seen in symptomatic bacteriuric patients. 28.50 Unfortunately, determining urinary leukocyte excretion is so cumbersome and impractical that other methods are necessary to measure pyuria. A hemocytometer can accurately measure pyuria and is easier to perform than microscopic urinary sediment examination. 12, 50 Studies in the 1960s correlated ~1O leukocytes/mm3 of urine with an hourly excretion rate of 400,000 or more leukocytes,2 and in a number of studies, virtually all symptomatic patients with bacterial urine cultures revealing ~ 105 CFU Iml of urine had ~ 10 leukocytes/mm3 of urine or greater as measured by the hemocytometer. 2 • 28 However, the hemocytometer method of determining pyuria is infrequently utilized despite its relative simplicity and accuracy because of the unavailability of a hemocytometer or unfamiliarity with its use for this purpose, Thus, the quantification of pyuria is usually made on the basis of direct microscopic examination of urinary sediment from a centrifuged specimen, Unfortunately, this is an inaccurate method of quantifying pyuria for several reasons: (1) variable or unmeasured initial urine volume, (2) variable or imprecise speed and time of centrifugation, (3) inconsistent resuspension volume after centrifugation, (4) inconsistent or unmeasured amount of urine placed on the slide for microscopic examination, (5) counting inaccuracy due to no grid lines for reference, and (6) observer bias in favor of areas on the slide where leukocytes are more easily seen. 3 , 50 If these inconsistencies and inaccuracies are minimized by accurate measurement of initial and res uspension urine volume, speed and time of centrifugation, and consistent counting technique, this method correlates more closely with hemocytometer measurements. l Less is understood about the correlation of pyuria and bacteriuria when bacterial counts are less than 105 CFU/ml of urine. Stamm and colleagues54 found significant pyuria (~8 leukocyteslmm 3 urine) in 26 of 27 female patients with documented "low level" bacteriuria «10'5 CFU/ml) determined by suprapubic aspiration or urethral catheterization. In those

LABORATORY IN THE DIAGNOSIS AND MANAGEMENT OF

UTls

315

individuals with significant pyuria but without a bacterial isolate from the urine, Chlamydia trachomatis was recovered from 10 of 16. 54 Based on these and other data, a symptomatic patient without significant pyuria should cause the clinician to question the diagnosis of a UTI. The measurement of pyuria using a rapid assay to determine the presence of leukocyte esterase in the urine obviates many of the problems that may be encountered during the routine examination of urinary sediment for the presence and number of leukocytes. Leukocyte esterase, an enzyme found in primary neutrophil granules, reacts with a reagent impregnated into the dipstick pad, producing a blue col or at room temperature within 1 to 2 minutes. 34, 42 In addition to being less time consuming than standard methods of measuring pyuria, the leukocyte esterase test is also relatively inexpensive. The reliability of the test is based on comparisons with standard methods of defining UTIs (isolation of ;::: 105 CFUiml of pathogenic bacteria) or significant pyuria (;:::10 white blood cells/mm 3 urine). The sensitivity of this test is between 75% and 96%, and the specificity varies between 94% and 98%.21,37 The test has a positive predictive value of only 50%, but it has a negative predictive value of 92%Y Thus, the leukocyte esterase assay is a reasonable method of quickly determining the presence of significant pyuria, and it may be used most appropriately as a screening test to determine the need for urine cultures in symptomatic patients for whom routine microscopy is either unavailable or impractical. Bacteriuria

Most UTIs are defined by the presence of significant numbers of pathogenic bacteria in urine that has been appropriately collected. The isolation of significant quantities of pathogenic bacteria from sterilely collected urine is the standard with which all other methods of detection of bacteriuria are compared. Other methods, including microscopic examination for bacteriuria and rapid tests such as the nitrite test, are inexpensive and simple to perform. These tests also can be performed and interpreted quickly, thus avoiding the inevitable delay of routine urine cultures. The various methods of determining bacteriuria are described later. Direct microscopy for the detection of bacteriuria is a readily available but highly variable method of determining bacteriuria. The methodology for performing microscopic examination of the urine varies greatly depending on the investigator; thus, the data that support its utility as a diagnostic test for UTIs are difficult to interpret because there is no consistent standard method of specimen preparation and interpretation. Most studies have used the isolation of ;:::105 CFU/ml of urine as the standard for comparison. In a review of urine microscopy for bacteriuria, Jenkins et aPl reviewed several studies of UTIs and analyzed the various methods of detecting bacteriuria microscopically, including the examination of uncentrifuged, unstained urine; centrifuged, unstained urine; stained, uncentrifuged urine; and stained, centrifuged urine. The sensitivity and specificity of these methods were determined using culture of 105 CFU/ml or greater of urine as the "gold standard" for comparison. l l These authors determined that uncentrifuged Gram-stained urine that revealed at least one organism per oil immersion field correlated with ;:::105 CFU/ml of urine with a

316

PETER

C.

PAP PAS

sensitivity and specificity of almost 90%. Additionally, finding five or more organisms per oil immersion field increased the specificity to 99%.1l Thus, the most convenient and reasonably reliable method of microscopically assessing bacteriuria is the examination of Gram-stained uncentrifuged urine. Kunin 23 suggested the use of unstained, centrifuged urine as a convenient and reliable method of determining significant bacteriuria, but the method was most reliable only when 106 CFU/ml or greater were isolated by culture and was less sensitive below this level of bacteriuria. A rapid diagnostic test for the detection of bacteriuria, the nitrite test, is both widely available and easily performed. The test is performed by the dipstick method, which utilizes an amine-impregnated pad to detect the presence of urinary nitrite. Nitrite in the urine is produced by the action of bacteria on dietary nitrate through nitrate reductase, a bacterial enzyme. The presence of urinary nitrite is indicated by the development of a pink color on the pad within 60 seconds. False-negative assays are common and may be the result of the lack of dietary nitrate or insufficient urinary nitrate levels due to diuretics. 3. 40 False-negative assays may also result when infection is due to an organism that is unable to produce nitrite in the urine through a lack of nitrate reductase. Organisms in this category include Staphylococcus sp., Enterococcus sp., and Pseudomonas Sp.3. 40, 46 Given the convenience of this test, it is easily used in an office setting and has also been successfully used by female patients at home with frequently recurring UTls.24 Sensitivity of the test ranges between 35% and 85%, and the specificity ranges between 92% and 100%.34,40 Several assays are presently available that combine the methodology for rapid detection of pyuria and bacteriuria, The Bac-T-Screen and Chemstrip LN are two examples of such tests that are more sensitive and specific than either the leukocyte esterase or nitrite test alone, 18, ,33, 36, 38 Both tests have sensitivities of between 88% and 92% and specificities of between 66% and 76% when using 105 CFU/ml or greater as a comparative standard. At lower levels of bacteriuria (:::::103 CFU/ml), the sensitivity range is considerably less at 75%.14 Notably, a negative predictive value of 96% to 97% at a level of 105 CFU/ml or greater supports the potential use of these assays as screening tests prior to urine culture when 105 CFU/ml is used as a cutoff value for significance. 38 Another test, the bioluminescence assay, utilizes the firefly luciferin-Iuciferase reaction to detect bacterial ATP in urine specimens. The test has been used successfully as a screening test for significant bacteriuria and has a sensitivity of94% to 97% and a specificity of 70% to 75% in detecting :::::105 CFU/ml of urine. 19, 29

URINE CULTURE The detection of significant numbers of pathogenic bacteria from culture of the urine has remained the gold standard for the diagnosis of UTI since Kass defined :::::105 CFU/ml of a single pathogenic bacterium isolated from urine culture as being significant in women with pyelonephritis or asymptomatic bacteriuria. 16,17 This finding was supported soon afterwards by Monzon et al,31 who demonstrated the correlation between cultures of

LABORATOlW IN THE DIAGNOSIS AND MA:-.!AGEMENT OF

UTls

317

sterilely obtained bladder aspirates, catheterized urine specimens, and voided urine obtained from women with laboratory evidence of a UTI. These culture methods correlated best when bacterial isolation exceeded 105 CFU/ml of urine. Lower bacterial counts from voided urine correlated poorly with either bladder aspiration or catheterization, and the data supported the use of 2:105 CFU/ml as a meaningful definition of "significant" bacteriuria. 31 For many years, clinicians applied these data to the diagnosis of all UTIs, including acute uncomplicated cystitis, a disorder that was poorly studied. More recently, Stamm et aP3. 54 have attempted to redefine the level of significant bacteriuria as it relates to acutely dysuric women in the absence of clinical pyelonephritis. Comparing cultures of bladder urine aspirates or catheter urine specimens to midstream specimens, as few as 102 coliform bacteria were found to be significantly related to acute urinary symptoms. By using 102 coliforms/ml from midstream voided urine as a cutoff, a sensitivity and specificity of 9.5% and 85%, respectively, were achieved when compared to bladder aspirate or catheter specimen cultures. 53 Importantly, in this study, the criterion of 2:105 CFU/ml would have failed to identify almost 50% of truly infected symptomatic women. Still, the controversy regarding the appropriate level of significant bacteriuria continues, 51 particularly because the practicality of screening urine cultures for "low colony" bacteriuria not only depends on the prevalence of UTIs within a population but also on the availability of laboratory technology and expertise. 21, 25, .39 The standard urine culture in most laboratories is performed on midstream clean-catch specimens collected into sterile containers. In women, the external genitalia are washed two to three times with a cleansing agent and water prior to collection of the specimen. This process reduces but does not eliminate urethral contamination. The routine cleansing of the urethral meatus in men prior to a midstream clean-catch specimen to reduce urethral contamination is of questionable value,27 but this technique is practiced widely nonetheless. Once the urine specimen is collected, it is ideally transported to the laboratory, where the specimen is inoculated by calibrated platinum wire loops onto blood agar and a selective gramnegative medium such as MacConkey or eosin-methylene blue (EMB) media. A urine specimen that is allowed to sit at room temperature for several hours may yield falsely elevated colony counts; thus, it is important to refrigerate urine if the specimen cannot be inoculated onto media expeditiously. Urine colony counts are relatively stable for up to 24 hours at refrigerator temperature (4°C). Following inoculation onto appropriate media, approximately 24 hours of incubation are required to obtain an accurate colony count, and an additional 12 to 24 hours are required for organism identification and antibiotic susceptibility testing. This method of urine culture is practiced widely among US hospitals and is expensive, time-consuming, labor-intensive, and usually fails to give the clinician specific information prior to initiation of treatment. Thus, more efficient and rapid urine culture techniques are an attractive alternative to the conventional culture techniques. A number of automated cultures systems are currently available. 35 These methods require between 5 hours and 13 hours to detect significant

318

PETER

C.

PAPPAS

bacteriuria and correlate well with isolation of 105 CFU!ml or greater using standard culture techniques. Most of these systems have sensitivities of 95% to 98% when 105 CFU!ml or greater is used as the standard of comparison. Some methods will allow detection of 104 CFU!ml or greater, but these methods are less sensitive in detecting "low colony" infections of less than 104 CFUlml and are of limited value in this setting. Antimicrobial susceptibilities may also be determined on bacteria that are isolated by these rapid methods and correlate well with standard susceptibility testing. 35 The automated methods of bacterial isolation can best be applied to the ambulatory population, in whom the Enterobacteriaceae, including Escherichia coli, Klebsiella sp., Proteus sp., and gram-positive organisms, including Enterococcus sp. and Staphylococcus sp. (especially S. saprophyticus), remain the most common causative agents of community-acquired, noncatheter-related UTIs.48 It is unfortunate that this same population is most likely to have "low colony" UTIs; thus, these methods may be somewhat insensitive in this population. Nosocomial and catheter-related UTIs are more commonly polymicrobial and are frequently due to organisms such as Pseudomonas aeruginosa, Acinetobacter sp., Citrobacter sp., or Candida Sp.48 These organisms may require substantially longer periods to achieve significant growth in the automated systems; therefore, these systems may not be ideal for the nosocomial environment. This is a particularly important point because nosocomial UTIs may be substantially more difficult to treat because of antimicrobial resistance or the presence of an indwelling urinary catheter. 57 Thus, the rapid isolation and identification techniques are imperfect in both the outpatient and nosocomial settings owing to poor sensitivity, but they offer advantages to laboratories that are processing large numbers of urine cultures because they increase efficiency and decrease cost. NONINVASIVE METHODS OF URINARY TRACT INFECTION LOCALIZATION The ability to localize the source of bacteriuria to the upper or lower urinary tract can be important in the evaluation and management of patients with UTIs. Infections of the upper urinary tract (acute or subclinical pyelonephritis) can be significantly more complicated than most lower UTIs (cystitis) and generally require more extensive antimicrobial therapy.48 In contrast, most cases of acute cystitis in nonpregnant women have fewer significant complications and can be treated with 3 days or less of an appropriate antimicrobial agent. Thus, it can be important to distinguish between upper and lower UTI even when this distinction is not apparent on clinical examination, because of the frequency of complications associated with upper UTI and the impact on length of therapy. Several laboratory methods are available to help distinguish between upper and lower UTIs. These methods vary with respect to their reliability and are generally compared to selective urethral catheterization (the Stamey test)49 or the bladder washout technique (the Fairley test)6 as reference methods. Details of selective ureteral catheterization and the bladder

LABORATORY IN THE DIAGNOSIS AND MANAGE~1ENT OF

UTIs

319

washout technique are not included in this discussion, but the reader is referred to selected references on these topics. 4. 6, 47, 49 Antibody-Coated Bacteria

The antibody-coated bacteria (ACB) technique, initially described in 1974,15,56 is based on the premise that "invasive" UTIs (e.g., pyelonephritis) lead to a specific host immune response to the invading bacteria and that mucosal infections such as uncomplicated cystitis do not elicit the same host response. 55 Specimens considered to be positive for ACB would thus indicate tissue invasion, such as renal (upper tract) infection or "invasive" cystitis in women. Unfortunately, the test is not as useful in men, in whom a positive test cannot distinguish between prostate and renal infection. 55 The technique of performing ACB in urine specimens is described in detail elsewhere, 15. 55, 56 and the reader is referred to these references for a complete description of the methodology. Several authors have examined the utility of the ACB assay for localization of UTIs. The sensitivity and specificity of the assay vary somewhat from study to study depending on the patient population examined, the criteria for a positive assay, the length of time from onset of infection to the time of study (theoretically, earlier diagnosis may not allow enough time for an appropriate host antibody response), and the reference method used for comparison. As an example, Thomas et al 56 examined 55 patients, 35 of whom had clinical evidence of acute or chronic pyelonephritis; the remaining 20 had uncomplicated cystitis. The diagnosis of pyelonephritis was made on the basis of the clinical findings of acute infection with fever and costovertebral angle tenderness or characteristic radiographic changes and impairment of renal function. Only a small portion of these patients underwent ureteral catheterization or bladder washout to document the source of bacteriuria. Thirty-four (97%) of 35 females with pyelonephritis had positive ACB tests compared to only 1 (5%) of 20 patients with uncomplicated cystitis. 56 Jones et aP5 similarly studied a group of 29 men and women with bacteriuria using bladder washout as the direct localization procedure. Only 1 (5.5%) of 18 with upper UTI and none (0%) of 8 patients with lower UTI revealed discordant results; one patient with early acute pyelonephritis with a confirmatory positive bladder washout procedure had a negative test for ACB. 15 A number of other studies subsequent to these two initial studies provide less supportive evidence of the utility of ACB as a localization technique. These studies suggest caution when interpreting the results in children,8 patients with spinal cord injury, 9, 22 and chronically catheterized patients. lO, 30 In addition, substantial false-positive and false-negative rates have been described, particularly in women with asymptomatic bacteriuria.7. 32, 41 In summary, the ACB test could be a useful adjunct to the clinician in determining the site of UTI, but the test is not widely available to clinicians; thus, it remains a research tool. It correlates best with acute or chronic renal infection as determined by clinical, radiographic, and direct localization procedures, but it may be falsely negative in patients with early upper UTI.

320

PETER

C.

PAPPAS

Urinary Beta-2 Microglobulin Another indirect localization method is the beta-2 microglobulin urinary assay. Beta-2 microglobulin is a small protein synthesized by most nucleated cells and is closely associated with immunoglobulin and histocompatibility antigens on the cell surface. Daily production of the protein is fairly constant. It is filtered by the glomeruli and almost completely reabsorbed by the renal tubules in the undamaged kidney. In the presence of tubular damage (as is seen in pyelonephritis), the urinary concentration of beta-2 microglobulin increases. Schardijn et al 44 studied 24 patients with UTIs, 10 of whom had upper tract involvement, to determine the levels of beta-2 microglobulin in the urine of these individuals. All 10 (100%) of the patients with upper UTI as opposed to none (0%) of the 14 patients with lower UTI had markedly elevated urinary beta-2 microglobulin assays.44 In a similar study,43 Schardijn et al described 34 patients with UTIs, 19 of whom had clinical or radiographic evidence of acute pyelonephritis. All 19 (100%) patients with pyelonephritis but none (0%) of 15 patients with cystitis and none (0%) of 44 controls had elevated levels of beta-2 microglobulin in urine. 43 Sheldon and Gonzalez47 more recently reviewed the utility of this assay and found it to be a promising test, but they concluded that it has technical problems that presently preclude its widespread use. 47 Maximal Urinary Concentration Ability Maximal urinary concentrating ability was one of the first methods utilized to determine the presence of upper UTI. Theoretically, patients with ongoing or recent upper UTI lack the ability to concentrate urine greater than 700 mOsm/L following a 16-hour water-deprivation test. More recently, intranasal DDAVP has been used to measure urinary concentrating ability in children in whom the water-deprivation test could be dangerous. Despite its theoretical advantages, the measurement of maximal urinary concentrating ability remains an insensitive indicator of upper UTI47 and is thus used infrequently as a localization technique. Other Indirect Methods of Localization A number of other methods to identifY upper UTI are available, but these tests are infrequently utilized in clinical practice. Some of these methods include the measurement of antibodies to Tamm-Horsfall protein, urinary lactic dehydrogenase (LDH), serum C-reactive protein, urinary beta-glucuronidase, and serum antibody to 0 antigen of aerobic gramnegative rods. These assays are generally insensitive and nonspecific, and they do not substantially add to the careful clinical evaluation of patients and a careful examination of the urine. 4. 47 OFFICE DIAGNOSIS OF URINARY TRACT INFECTIONS The office diagnosis of UTls employs many of the same techniques and principles described earlier in this article, particularly as this relates to the performance of microscopy and dipstick analysis of urine specimens. Cost,

LABORATORY IN THE DIAGNOSIS AND YiANAGEMEKT OF

UTIs

321

efficiency, and practicality are of paramount importance in the office setting; thus, the following discussion will focus on the diagnostic approaches that meet these criteria and that are most helpful to the clinician faced with the diagnosis and management of outpatients with suspected UTIs. Urine microscopy is an important aid to the diagnosis of UTI, and most offices have the capability to perform reliable urine microscopy if proper attention is given to technique and the laboratory personnel develop a consistent approach to the procedure. This will eliminate much of the variability that accompanies microscopic assessment of urine. The presence or absence of pyuria should be determined on every urine specimen obtained from patients with proven or suspected UTI because of its importance as a marker of bacterial infection. 50 Most offices do not have a hemocytometer at their disposal, and a manual count may be the only method of obtaining an estimate of leukocytes per cubic millimeter of urine. As previously described, urine leukocytes of 8 to 10/mm 3 or greater are considered abnormal and correlate well with the presence of UTI. 2, 50 With proper attention to technique, the measurement of pyuria from centrifuged urine can be reproducible and have reasonable correlation with hemocytometer values. Specific ways in which a laboratory can minimize the variables of urine microscopy include standardizing initial urine volume (10 ml), resuspension volume (0.2 ml), centrifuge time (5 minutes), and speed (2000 rpm), and the use of slides with grid lines. Urine microscopy for bacteriuria can also be accomplished in the office setting, and, as discussed earlier, the presence of one or more bacteria per oil immersion field in an uncentrifuged Gram-stained specimen correlates highly with 105 CFU/ml or greater of urine. 11 "Lower count" UTIs are much more difficult to assess by this method; thus, urine microscopy for bacteriuria has significant limitations. 11, 21 Rapid detection assays for bacteriuria and pyuria offer a reasonable alternative to urine microscopy in the office setting. The leukocyte esterase assay for determining pyuria is described earlier in this article, This dipstick method is easily applied to the office setting and has the advantage of speed, low cost, easy interpretability, and reasonable accuracy. Sensitivity for the test ranges between 75% and 94%, with specificity ranging between 94% and 98%,21.45 making it a reasonable rapid screening tool for significant pyuria and, by inference, bacteriuria. Many commercially available dipsticks offer assays for both urinary nitrite and leukocyte esterase on the same test strip (e.g., Chemstrip LN) and can be interpreted visually without the aid of special instrumentation. Other combination assays, such as Bac-T-Screen or the bioluminescence assay, offer similar accuracy and speed but require special equipment for interpretation and are therefore less attractive to many office practitioners. The rapid assays are probably best utilized as screening tests for asymptomatic individuals, with urine microscopy being reserved for symptomatic individuals in whom UTI seems likely. Urine cultures and sensitivities done in the routine manner are expensive, cumbersome, and not well suited to the office environment. Cohen and Kass 5 described the first slide culture technique that was easily applicable to the ambulatory setting. Since the initial description of this technique, a number of culture methodologies applicable to the outpatient

322

PETER

C.

PAPPAS

setting have been described, all of which are reasonably accurate provided that the lower limit of significant bacteriuria is 105 CFU/ml. Unfortunately, as described by Stamm et al,53 about 50% of acutely symptomatic women with UTIs will have fewer than 105 CFU/ml; thus, the sensitivities of most of these less well standardized culture methods are insufficient to identify a large proportion of "low colony" UTIs. Some have questioned the value of urine cultures in acutely symptomatic, nonpregnant woman without clinical evidence of upper UTI. 12, 13 It is argued that urine cultures done on ambulatory patients are unnecessary and should be omitted from the initial assessment unless other complicating features are evident. The reasons given for this approach to acutely dysuric women include (1) the cost and delay of routine urine culture, (2) the frequency of "low level" bacteriuria «105 CFU/ml) causing symptomatic infection that might be overlooked or considered insignificant in the microbiology laboratory, (3) the availability of rapid, less expensive "presumptive" tests that provide evidence of UTI, and (4) the observation that the result of a urine culture usually does not alter the treating physician's therapy. 12 Johnson and Stamm 12 have recommended that urine cultures be done in women with persistent symptoms following therapy for presumed cystitis, in women with frequent recent infections, and in all patients with symptoms of cystitis who lack pyuria. Further, the routine use of posttreatment cultures for uncomplicated cystitis in nonpregnant women is not recommended unless symptoms of cystitis persist. 58 Thus, the office diagnosis of UTIs is dependent on the ability to perform reliable urine microscopy for pyuria and bacteriuria or, in the absence of microscopy, the ability to perform rapid assays of pyuria and bacteriuria using one of the many available commercial assays. Office urine cultures are relatively unimportant in the evaluation of healthy nonpregnant women with acute symptoms of cystitis. Cultures should probably be reserved for patients with a higher risk of complications or treatment failure, such as pregnant women, patients with symptoms lasting more than 7 days or those with signs and symptoms suggestive of upper UTI,52 and in all men with a suspected UTI. 26 SUMMARY

The laboratory is essential in the diagnosis and management of UTIs. The presence of pyuria and bacteriuria, the two most important indicators of UTIs, are most accurately determined by standard techniques. In quantitating pyuria, the finding of 2:10 leukocytes/mm 3 of urine by either hemocytometry or direct microscopy correlates highly with symptomatic, culture-proven UTIs. The determination of bacteriuria by direct microscopy is inaccurate, particularly at lower levels of bacteriuria; thus, quantitative urine cultures remain the most accurate measure of bacteriuria. Significant bacteriuria, previously defined as 2: 105 CFU Iml of urine, has been redefined with the observation that as few as 102 CFU/ml can be associated with significant pyuria and symptoms suggestive of cystitis. The need for routine and posttreatment urine cultures in nonpregnant women with acute dysuria

LABORATORY IN THE DIAGNOSIS AND MANAGEMENT OF

UTls

323

remains controversial, but current data suggest that they are usually unneccesary. Rapid diagnostic tests for detection of pyuria and bacteriuria are designed to increase efficiency and decrease cost in the diagnosis of UTI. Unfortunately, none of these techniques can quantitate pyuria or bacteriuria as accurately as the standard methods, but the level of accuracy offered by the standard methods is not always necessary in the care of patients with uncomplicated UTIs. These tests are particularly well suited for screening asymptomatic high-risk populations. Noninvasive localization techniques continue to be explored as possible alternatives to invasive localization procedures, but they remain largely research tools that are not readily available to the practicing clinician. Understanding the applicability and appropriate use of newer technologies in the evaluation of patients with UTIs and how these technologies complement the standard diagnostic techniques will lead to better, more efficient, and less costly patient care.

REFERENCES 1. Alwall N: Pyuria: Deposit in high-power microscopic field WBC/hpf versus WBC/mm 3 in counting chamber. Acta Med Scand 194:537, 1973 2. Brumfitt W: Urinary cell counts and their value. J Clin PathoI18:550, 1965 3. Cobbs CC: Presumptive tests for urinary tract infection. In Kaye D (ed): Urinary Tract Infection and Its Management. St. Louis, CV Mosby, 1972, p 43 4. Cobbs CC: Localization of urinary tract infection. In Kaye D (ed): Urinary Tract Infection and Its Management. St. Louis, CV Mosby, 1972, p 52 5. Cohen SN, Kass EH: A simple method for quantitative urine culture. N Engl J Med 277:176,1967 6. Fairley KF, Bond AC. Brown RB, et al: Simple test to determine the site of urinary tract infection. Lancet ii:7513, 1967 7. Harding CKM, Marrie TJ, Ronald AR, et al: Urinary tract infection localization in women. JAM A 240:1147, 1978 8. Hellerstein S, Kennedy E, Nussbaum L, et al: Localization of the site of urinary tract infections by means of antibody-coated bacteria in the urinary sediments. J Pediatr 92:188, 1978 9. Hooton TM, O'Shaughnessy EJ, Clowers D, et al: Localization of urinary tract infection in patients with spinal cord injury. J Infect Dis 150:85, 1984 10. Hulter HN, Borchardt KA, Mahood JA, et al: Localization of catheter-induced urinary tract infections: Interpretation of bladder washout and antibody-coated bacteria tests. Nephron 38:48, 1984 11. Jenkins RD, Fenn JP, Matsen JM: Review of urine microscopy for bacteriuria. JAM A 255:3397, 1986 12. Johnson JR, Stamm WE: Diagnosis and treatment of acute urinary tract infections. Infect Dis Clin North Am 1:773, 1987 13. Johnson JR, Stamm WE: Urinary tract infections in women: Diagnosis and treatment. Ann Intern Med 111:906, 1989 14. Jones C, MacPherson DW, Stevens DL: Inability of the Chemstrip LN compared with quantitative urine culture to predict significant bacteriuria. J Clin Microbiol 23:160, 1986 15. Jones SR, Smith JW, Sanford JP: Localization of urinary-tract infections by detection of antibody-coated bacteria in urine sediment. N Engl J Med 290:591, 1974 16. Kass EH: Asymptomatic infections of the urinary tract. TraIlS Assoc Am Physicians 69:56, 1956 17. Kass EH: Bacteriuria and the diagnosis of infections of the urinary tract. Arch Intern Med 100:709, 1957

324

PETER

C.

PAPPAS

18. Kellogg JA, Manzella JP, Shaffer SN, et al: Clinical relevance of culture versus screens for the detection of microbial pathogens in urine specimens. Am J Med 83:739, 1987 19. Kolbeck JC, Padgett RA, Estevez EG, et al: Bioluminescence screening for bacteriuria. J Clin Microbiol 21:527, 1985 20. Komaroff AL: Acute dysuria in women. N Engl J Med 310:368, 1984 21. Komaroff AL: Urinalysis and urine culture in women with dysuria. Ann Intern Med 104:212, 1986 22. Kuhlemeier KV, Lloyd LK, Stover SL: Failure of antibody-coated bacteria and bladder washout tests to localize infection in spinal cord injury patients. J UroI130:729, 1983 23. Kunin CM: The quantitative significance of bacteria visualized in the unstained urinary sediment. N Engl J Med 265:589, 1961 24. Kunin CM, DeGroot JE: Self-screening for significant bacteriuria: Evaluation of dips trip combination nitrite/culture test. JAMA 231:1349, 1975 25. Latham RH, Wong ES, Larson A, et al: Laboratory diagnosis of urinary tract infection in ambulatory women. JAMA 254:3333, 1985 26. Lipsky BA: Urinary tract infections in men: Epidemiology pathophysiology, diagnosis and treatment. Ann Intern Med 110:138, 1989 27. Lipsky BA, Inui TS, Plorde JJ, et al: Is clean-catch midstream void procedure necessary for obtaining urine culture specimens from men? Am J Med 76:257, 1984 28. Little PJ: A comparison of urinary white cell concentration with the white cell excretion rate. Br J Urol 36:360, 1964 29. Males BM, Bartholomew WR, Amsterdam D: Leukocyte esterase-nitrate and bioluminescence assays as urine screens. J Clin Microbiol 22:531, 1985 30. Montplaiser S, Courteau C, Martineau B, et al: Limitations of the direct immunofluorescence test for antibody-coated bacteria in determining the site of urinary tract infections in children. Can Med Assoc J 125:993, 1981 31. Monzon OT, Ory EM, Dobson HL, et al: A comparison of bacterial counts of the urine obtained by needle aspiration of the bladder, catheterization and midstream-voided methods. N Engl J Med 259:764, 1958 32. Mundt KA, Polk BF: Identification of site of urinary-tract infections by antibody-coated bacteria assay. Lancet ii:1l72, 1979 33. Murray PR, Smith TB, McKinney TC Jr: Clinical evaluation of three urine screening tests. J Clin Microbiol 25:467, 1987 34. Pels RJ, Bor DH, Woolhandler S, et al: Dipstick urinalysis screening of asymptomatic adults for urinary tract disorders. JAMA 262:1221, 1989 35. Pezzlo MT: Automated methods for detection of bacteriuria. Am J Med 75 (suppl 1):71, 1983 36. Pezzlo MT, Wetkowski MA, Peterson EM, et al: Detection of bacteriuria and pyuria within two minutes. J Clin Microbiol 21:578, 1985 37. Pfaller MA, Koontz FP: Laboratory evaluation of leukocyte esterase and nitrite tests for the detection of bacteriuria. J Clin Microbiol 21:840, 1985 38. Pfaller MA, Koontz FP: Use of rapid screening tests in processing urine specimens by conventional culture and the AutoMicrobic system. J Clin Microbiol 21:783, 1985 39. Platt R: Quantitative definition of bacteriuria. Am J Med 75 (sup pI 1):44, 1983 40. Pollock HM: Laboratory techniques for detection of urinary tract infection and assessment of value. Am J Med 75 (suppl 1):79, 1983 41. Rumans LW, Vosti KL: The relationship of antibody-coated bacteria to clinical syndromes. Arch Intern Med 138:1077, 1978 42. Sawyer KP, Stone LL: Evaluation of a leukocyte dip-stick test used for screening urine cultures. J Clin Microbiol 20:820, 1984 43. Schardijn GHC, Statius van Eps LW, Pauw W, et al: Comparison of reliability of tests to distinguish upper from lower urinary tract infection. Br Med J 289:284, 1984 44. Schardijn G, Statius van Eps LW, Swaak AJG: Urinary beta-2 microglobulin in upper and lower urinary-tract infections. Lancet i:805, 1979 45. Scheer WD: The detection of leukocyte esterase activity in urine with a new reagent strip. Am J Clin Pathol 87:86, 1987 46. Scheifele DW, Smith AL: Home-testing for recurrent bacteriuria, using nitrite strips. Am J Dis Child 132:46, 1978 47. Sheldon CA, Gonzalez R: Differentiation of upper and lower urinary tract infections: How and when? Med Clin North Am 68:321, 1984

LABORATORY IN TilE DIAGNOSIS AND MA1'iAGEMENT OF

UTIs

325

48. Sobel JD, Kaye D: Urinary tract infections. In Mandell GL, Douglas HG, Bennett JE (eds): Principles and Practice of Infectious Diseases, ed 3. New York, Churchill Livingstone, 1990, p 582 49. Stamey TA, Govan DE, Palmer JM: The localization and treatment of urinary tract infections: The role of bactericidal urine levels as opposed to serum levels. Medicine 44:1, 1965 50. Stamm WE: Measurement of pyuria and its relation to bacteriuria. Am J Med 75(suppl):53, 1983 51. Stamm WE: Quantitative urine cultures revisited. Eur J Clin Microbiol 3:279, 1984 52. Stamm WE: When should we use urine cultures? Infect Control 7:431, 1986 53. Stamm WE, Counts GW, Running R, et al: Diagnosis of coliform infection in acutely dysuric women. N Engl J Med 307:463, 1982 54. Stamm WE, Wagner KF, Amsel R, et al: Causes of the acute urethral syndrome in women. N Engl J Med 303:409, 1980 55. Thomas VL: Antibody-coated bacteria in urinary tract infections. Kidney Int 21:1, 1982 56. Thomas V, Shelokov A, Forland M: Antibody-coated bacteria in the urine and the site of urinary-tract infection. N Engl J Med 290:588, 1974 57. Warren JW: Catheter-associated urinary tract infections. Infect Dis Clin North Am 1:823, 1987 58. Winickoff RN, Wilner SI, Gall G, et al: Urine culture after treatment of uncomplicated cystitis in women. South Med J 74:165, 1981

Address reprint requests to Peter G. Pappas, MD University of Alabama School of Medicine 229 Tinsley Harrison Tower University Station Birmingham, AL 35294