ilure of the urinalysis and ntitative urine culture in

iagnosing symptomatic urinary tra ctions in patients with long&r ary catheters

Dallas, Texas

Since urine culture and urinalysis are both of uncertain reliability in diagnosing symptomatic urinary tract infections in patients with chronic urinary catheters, we performed sequential quantitative cultures and urinalyses on 177 urine specimens from 14 patients with long-term urinary catheters during a 12-month period. We found high concentrations of 32 species of aerobic bacteria or fungi in most specimens examined. Pyuria was common even during asymptomatic periods; hematuria was less common. During symptomatic urinary tract infections, neither urinalyses nor quantitative urine cultures exhibited changes specific for such infections. Thus neither urinalysis nor urine culture appears to be a reliabie test for symptomatic urinary tract infections in patients with chronic urinary catheters. (AM J INFECT CONTROL 13:154-160, 1985.)

The indwelling urinary catheter serves a number of essential functions in modern med- ical practice. It is used to relieve urinary ob- struction, to facilitate surgical repair of the urethra, to maintain a dry environment for the incontinent patient, and to ensure accurate measurement of urinary output in severely ill

From the Infection Control Program, Laboratory Service, and Medical Service, Veterans Administration Medical Center, and the Department of Medicine, University of Texas Southwestern Medical School.

Supported by the Veterans Administration.

Presented at the Tenth Annual Educational Conference of the Association for Practitioners in Infection Control, San Diego, June 1963.

Reprint requests: Philip A, Mackowiak, M.D., Medical Service, Veterans Administration Medical Center, 4500 South Lancaster Rd., Dallas, TX 75216.

patients.’ Unfortunately, because it provides a portal of entry for microorganisms into the lower urinary tract, the urinary catheter is also singularly important in predisposing patients to nosocomial urinary tract infections?, 3 The risk of such infections is all the more pernicious because the infections are difficult to diagnose. The difficulty arises because bacteriuria, an es- sential diagnostic criterion of acute urinary in- fections, is a constant finding in patients with long-term urinary catheters.4-7 Pyuria, another cardinal feature of acute urinary infections, is of uncertain value in establishing the existence of such infections, because characteristics of the baseline urinalysis in patients with chronic uri- nary catheters have not previously been de- scribed.

Therefore, we prospectively performed uri- nalysis on 14 patients with long-term urinary

Valume 13 Ntimber 4

August. 1985 Urinary catheters 1%

catheters. In addition, because previous micro- biologic surveys of such patients have failed to examine the incidence of bacteriuria in a pro- spective, systematic mannei? ‘ml2 or have fo- cused primarily on female patients,7 we also performed sequential quantitative urine cul- tures on our patients. In so doing, we attempted not only to characterize the baseline urinary flora of catheterized patients, but also to deter- mine whether specific changes in either the uri- nalysis or the urine culture accompany symp- tomatic urinary tract infections and to see whether a relationship exists between the com- position of the urinary flora and the incidence of such infections.

OD

Patients followed up in a hospital-based home care (HBHC) program in Dallas who had indwelling urinary catheters for a period of more than 14 days were considered for entry into the study. Fourteen patients with urinary catheters in place for a mean duration of 6.3 years agreed to participate in the study. They ranged in age from 38 to 93 years (mean = 63 years). One subject was a woman and the re- mainder were men. Ten patients had Foley catheters, two had coude catheters, and two had suprapubic catheters. All had underlying neurologic deficits and a variety of other med- ical problems. Although most patients had for- merly received some form of antimicrobial therapy in an attempt to suppress urinary bac- teria, none were receiving suppressive agents at the time of entry into the study. Nurses working

HC program visited each patient weekly to monitor patients’ conditions, to pro- vide catheter care, and to obtain specimens re- quired for analysis.

On the average, urinary catheters were changed monthly. Five subjects (36%) required intermittent bladder irrigations to control ex- cessive urinary sediment. Periurethral cleans- ing was performed daily with either mild soap and water or an iodophor by the patient or an attendant. Urine specimens were obtained by aspiration from the catheter specimen port af- ter it had been cleansed with an iodophor or alcohol pad. The urine was placed in a pre- packaged sterile container, then into an airtight

polyester pouch, and finally into a container of ice for transport to the Dallas Veterans Admin- istration Medical Center bacteriology labora- tory. All specimens were processed within 6 hours of collection. Serial tenfold dilutions of each urine specimen were subcultured aerobi- cally on blood agar, MacConkey agar, and Co- lumbia colistin-nalidixic acid agar by using a 0.001 ml calibrated platinum loop. These were incubated for 18 to 24 hours at 35” C. Gram- negative bacilli isolated from the specimens were identified by using the API 20E profile rec- ognition system (Analytab Products, Inc., Plain- view, N.Y .). Gram-positive cocci were identified according to criteria outlined in the American Society for Clinical Microbiology Mantrai of Clinical MicrobioZogy.13 Antibiotic susceptibility determinations were performed by using the standardized disc diffusion method of B et a1.14 Weekly cultures were obtained from e patient for 5 to 6 weeks and at monthly intervals thereafter. When appropriate, additional urine specimens were examined immediately before, during, and after antibiotic therapy.

Sequential urinalyses were performed on these same specimens. After centrifuging 12 ml of urine at 1200 rpm for 5 minutes, approxi- mately 11.5 ml of supernatant was decanted and discarded. The sediment contained within the remaining 0.5 ml was resuspended by gentle agitation and examined. The pH of the suspen- sions was analyzed with N-Multistix-C reagent strips (Miles Laboratories, Inc., Elkhart, Ind.) and the number of white blood cells (W red blood cells (RBCs), and casts per dry held (x400) was quantitated by light microscopy.

RESULTS

All urine specimens except those obtained during or immediately after antibiotic therapy contained at least one microbial species. Se- quential quantitative cultures showed 3 10’ cfui ml of two to four different microbial species in 111 of 177 specimens (63%) examined (Table 1). The total number of different species isolated from each subject varied fr four to eleven species. Some species appear in the urine of patients only transiently, whereas other (pre- dominant) microorganisms exhibited remark-

American Journal of

INFECTION CONTROL

Results of quantitative cultures of 177 urine specimens obtained from 14 patients with iong- term urinary catheters during a 12-month period of observation

Patient No.

NO. of specimens examined

No. of different species in all

specimens

Species per specimen

(mean +‘ SD) Total cfuiml

(mean i: SD [loglO])

It 17 5 2.2 2 0.7 6.2 i: 1.4 Pseudomonas aeruginosa (100) 8 8 4 1.8 + 0.7 5.9 1 1.6 Pseudomonas aeruginosa (87) 3 13 6 2.5 -c 1.2 5.4 i: 1.5 Pseudomonas aeruginosa (77) 4 14 5 1 9 -c 0.7 6.4 -I- 1.6 Pseudomonas aeruginosa (64) 3 9 7 2.7 i 0.9 6.3 + 0.5 Pseudomonas aeruginosa (78) 6 10 5 2.4 i 1 1 6.9 r 1 1 Enterococcus (70)

7 13 6 1.9 i 0.7 6.3 c 1.2 Enterococcus (69) Wit: 22 11 2.5 + 1.0 7.0 f 1.1 Enterococcus (45) 9 10 6 2.5 ” 0.7 7.0 _t 1.3 Morganeila morganii (100)

10 11 8 2.6 -t 0.7 6.4 i: 1.7 Morgane!la morganii (82) 11* 12 9 3.6 i 2.8 6.9 i: 1.2 Klebsiella pneumoniae (92) 12 12 11 4.0 k 0.9 6.0 +- 1.2 Escherichia co/i (83) I3$$j 14 9 21 2 1.1 5.8 i 1.8 Proteus mirabilis (67)

.14 12 7 2.2 -t- 1.1 7.2 f 0.9 Citrobacter diversus (42)

*Numbers in parentheses indicate percent of specimens containing the predominant species. tPatients receiving antibiotics for infections other than urinary tract rnfectrons durrng ihe study. $Patient receiving antibiotics for symptomatic urinary tract infections during the study SFemale patient.

able capacities for persisting from one sampling period to the next. Pseudomonas aeruginosa was the most common predominant microorganism identified (five patients), followed by entero- cocci (three patients) and Morganella morganii (two patients).

Urinalyses performed on these same speci- mens showed pyuria to be common in patients with chronic urinary catheters (Fig. 1). Of the specimens examined, 66% contained numbers of WBCs exceeding the upper limit of normal for our laboratory (35 WBC/hpf); 28% of urine specimens examined contained >20 WBC/hpf. These values are to be contrasted with those for red cells. Only 16% of the specimens examined contained >5 I?BC/hpf (the upper limit of nor- mal for our laboratory). The mean pH of spec- imens containing urease-producing bacteria was 7.3 compared with 6.7 for urine specimens lacking such microorganisms.

Three patients experienced febrile illnesses during the investigation, which were diagnosed by their attending physicians as acute urinary tract infections, Comparison of the results of urinalysis performed just before four such epi- sodes with those performed during the infec- tions or just after completion of antibiotic ther- apy showed little difference in concentration of

BCs or pH values (data not shown). In addi-

tion, there was little difference between the WBC concentrations in specimens obtained be- fore infections and those obtained during the infections. However, WBC concentrations ex- hibited a marked decrease during the conva- lescent period compared with specimens ob- tained before and at the time of symptomatic urinary tract infections (Fig. 2).

A careful examination of the results of se- quential urine cultures obtained from these same patients revealed that in three of four in- stances it was possible to sterilize the urine dur- ing therapy with broad-spectrum antibiotics (Fig. 3). In each case, however, microorganisms returned to their original concentration 2 to 12 days after antibiotics were discontinued. In three cases, repopulation of the urine was by one or more of the organisms present before antibiotic therapy (Fig. 3), whereas in another case a sensitive microorganism was replaced by an antibiotic-resistant one (Fig. 4).

Comparison of the urinary flora of patients with histories of symptomatic urinary tract in- fections to those having no such infections showed only one potentially important differ- ence between the two groups (Table 2). Patients with histories of symptomatic urinary tract in- fections were less likely to have enterococci as the predominant urinary colonist. However, the

VoiLime 13 Wmber 4 august, 1985

Fig. 1. Results of urinalyses performed on 177 se- quential specimens from 14 patients. Bar graphs depict the percent of specimens having the characteristics in- dicated along the ordinate. WBCs per high-power field are depicted in the top panel, RBCs per high-power field in the middle panel, and pH in the lower panel.

percentage of urine specimens containing en- Like numerous other investigators we have terococci (including patients with enterococci found the urine of patients with long-term uri- present as nonpredominant microorganisms) nary catheters to be continuously colonized by was similar for the two groups. The mean con- large numbers of aerobic bacteria. Indeed, most centration of microorganisms and the average of the specimens we examined contained sub- number of microorganisms isolated from urine stantially higher concentrations of bacteria

50

LL. 4o %I 30 3; ..A 20 ii 0 10

0

Fig. 2. WBCs per high-power field detected in urinal- yses performed before, during, and just after treatment of four episodes of symptomatic urinary infections in three study subjects”

Table 2. Comparison of results of quantitative urine cultures in subjects with and without histories of symptomatic urinary tract infections

tiistwy of $yffl~t~~~ti~ urinary tract

infection

Factors

Number of subjects Mean cfuiml (loglO) per specimen Mean microorganisms per specimen Subjects in whom enterococci

predominated

Yes NO

6 8 6.6 6.6 2.5 2.4 0 3

Percent of specimens containing enterococci

44 36

specimens was also similar for patients with and without histories of symptomatic urinary tract infections.

DISCUSSION

American Joxnal oi

INFECTION CONTROL

m GENTAMICIN I 1

GDS-Group D Streptococcus

Fig. 3. Results of quantitative cultures of urine specimens obtained from patient No. 13 before, during, and after treatment with gentamicin. Treatment succeeded in temporarily sterilizing this patient’s urine. However, KlebsieNa pneumoniae isolated just prior to therapy was again present shortly after cessation of antibiotic therapy. Both isolates of K. pneu- moniae had identical API biotype numbers and antibiotic susceptibility patterns.

SULFA /TRIMETHOPRIM 10 GENTAM ICIN

73 9 I 1

a KETOCONAZOLE

g- 8 pBS,Kp,Pr,E

Pa GBS=Group B Streptococcus

Kp+.pneumoniae Pr =P.rettcrerl Cs=Candlda sp. Pa=P.aeruainora

E~Enterococcus

Fig. 4. Results of quantitative cultures of urine specimens obtained from patient No. 11 before, during, and after treatment with sulfaitrimethoprim, gentamicin, and ketoconazole. The antimicrobial combination did not succeed in sterilizing this patient’s urine, but did result in replacement of susceptible microorganisms by progressively more resistant species.

than the I x lo5 cfu/ml generally regarded as indicative of infection for “clean-catch” speci- mens. Others have observed that such bacteria fall into two general categories, those that are constantly found and those that are transient colonists6. 7 Our findings support the existence of this dichotomy. Unlike previous investi- gators, however, who have emphasized the importance of gram-negative bacilli in the ‘“constantly found” category, we identified en- terococci as the predominant urinary colonist

in 3 of our 14 patients. As such, enterococci were second in importance only to P. aeruginosa as predominant microorganisms in OUT palient population. The frequency with which P. aeru- ginosa established itself as the predominant colonist among our patients also differed from the findings of other investigators who have reported Pvovidencia stuarti? and Proteus mirabW to be the most prevalent and persis- tent urinary colonists in catheterized patients.

To our knowledge no previous prospective

Voiume 13 Number 4

August, 1985 Uvinavy catheters “1

survey of urinalyses among patients with chronic urinary catheters has been reported. In conducting such a survey, we have shown py- uria to be a common finding among patients with chronic urinary catheters. Hematuria is less common. Although the urinary pH of such patients varies, urine specimens exhibit a ten- dency toward alkalinity, particularly when such specimens are colonized by urease-pro- ducing bacteria.

Three of our patient s experienced clinical signs and symptoms of acute urinary tract in- fections during the course of our survey. These episodes provided us with an opportunity to compare results of urine cultures and urinal- yses performed during asymptomatic and con- valescent periods with those performed on spec- imens obtained during the acute infections. This comparison failed to identify specific changes in the microorganisms present in the urine, the level of pyuria or hematuria, or the urinary pH that could be considered as predic- tive of the presence of a symptomatic urinary tract infection. Like Warren et al.7 we observed a decrease in the mean number of WBCs per high-power field in urine specimens obtained after treatment with broad-spectrum antibiot- ics. These findings suggest that the chronic py- uria of patients with long-term urinary cathe- ters is due to bacterial invasion of genitourinary tissues rather than the simple mechanical ir- ritation of the urethra and bladder by the uri- nary catheters. Unfortunately, the level of py- uria generally does not increase when such low-grade infections progress to symptomatic (febrile) urinary tract infections. Our failure to identify specific changes in either the quanti- tative urine culture or the urinalysis during symptomatic urinary tract infections suggests that such infections are, by necessity, diagnoses of exclusion.

Musher et alI5 have provided data suggesting that urinary WBC counts are more reliable when performed on unspun urine with a hemo- cytometer than by the semiquantitative meth- od used in this investigation and by most clin- ical laboratories. Future studies will have to determine if more quantitative methods for de- termining urinary WBC concentrations can re- liably detect symptomatic urinary tract infec-

tions in patients with long-term urinary catb- eters. Although tests for antibody-coated bacteria have been used with some success in distinguishing upper from lower urinary tract infections in noncatheterized patients,16 they have generally not been reliable in making such a distinction in patients with urinary catheters.”

There is increasing evidence that the indig- enous microorganisms of various body surfaces constitute a limited barrier to colonization and infection by pathogenic microorganismsi wished to determine whether the composition of the urinary flora of patients with long-term urinary catheters might be similarly important in dictating the likelihood of symptomatic in- fections of the genitourinary tract. To do so we compared patients with histories of one or more symptomatic urinary tract infections with those having no such infections. We found a difference in predominant urinary microorgan- isms in these two groups in that patients free from symptomatic urinary tract infections more commonly had enterococci as their pre- dominant urinary colonist. However, low con- centrations of enterococci were no more com- mon among these subjects than among those with histories of symptomatic urinary tract in- fections. Consequently, a cause-and-effect re- lationship between colonization of the urinary tract by enterococci and freedom from symp- tomatic urinary tract infections is doubtful. Furthermore, symptomatic urinary tract infec- tions in these patients were diagnoses of exclu- sion-made by the patients’ attending physi- cians in response to febrile illnesses whe source other than the urinary tract coul identified as a cause for the fever. As such, these diagnoses may not always have been accurate. Thus our data neither establish nor refute the existence of the capacity of indigenous micro- organisms of the urinary tract to act as a barrier to infection in patients with long-term urinary catheters.

It is clear from our data and the work of others2, 3 that urinary tract infections in pa- tients with long-term urinary catheters are dif- ficult problems, because they are frequent and also hard to diagnose. For these reasons the search for effective prophylactic regimens

60 Steward et al. Ame:ican iournal oi

INFECTION CONTROL

against such infections has been intensive. Un- fortunately, prophylactic use of antibiotics in chronically catheterized patients has been shown repeatedly to be ineffective in suppress- ing symptomatic urinary tract infections.” Al- though methenamine hippurate has been re- ported by Scandinavian investigators to de- crease the incidence of such infections (presum- ably because of a physiochemical action on salt formation rather than a direct antibacterial ef- fect),*O investigators working in this country have not succeeded in substantiating these re- ports. 21 Additional research is needed to deter- mine whether nonpathogenic inhibitors (e.g., enterococci) do exist in the urinary flora of cath- eterized patients and whether these inhibitory microorganisms can be manipulated in ways to increase the resistance of such patients to symp- tomatic urinary tract infections.

The authors thank Margaret Aull, R.N., Vera Baily, L.V.N., Annie Finnerty, R.N., Diane Ragsdale, R.N., Susan Thorn- bury, R.N., Ramona Young, R.N., and Joseph W. McKeever, M.D., for closely monitoring the patients and collecting urine specimens; Ann Wolf, MS., for assisting with the planning of the study; and Mr. Walter Cregg for graphic illustrations.

fere

1. Kunin CM, McCormack RC: Prevention of catheter- induced urinary-tract infections by sterile drainage. N Engl J Med 274:1155-1161, 1966.

2. Scheckler WE, Garner JJ, Kaiser AB, et al: Prevalence of infections and antibiotic usage in eight community hospitals. In Brachman PS, Eichoff TC, editors: Pro- ceedings of the International Conference on Nosoco- mial Infections, Center for Disease Control, August 3- 6, 1970. Chicago, 197 1, American Hospital Association, pp 299-305.

3. Garibaldi RA, Burke JP, Dickman ML, et al: Factors predisposing to bacteriuria during indwelling urethral catheterization. N Engl J Med 291:215-219, 1974.

4. Ailing B, Brandberg A, Seeberg S, et al: Aerobic and anaerobic microbial flora in the urinary tract of geriatric patients during long-term care. J Infect Dis 127:34-39, 1973.

5. Newman E, Price M: Bacteriuria in patients with spinal cord lesions: Its relationship to urinary drainage ap- pliance. Arch Phys Med Rehabil 58:427-430, 1977.

6. Bergqvist D, Bronnestam R, Hedelin H, et al: Changes in the aerobic bacterial Flora in the urinary tract of pa- tients with long-term indwelling Foley catheters. Ural Res 8~43-47, 1980.

7. Warren JW, Tenney JH, Hoopes JM, et al: A prospective microbiological study of bacteriuria in patients with chronic indwelling urethral catheters. J Infect Dis

146:719-723, 1982. 8. Morals P, Tsau AY: Quantitative bacteriological study

of urinary infections among paraplegics. J Ural 87:191- 198, 1962.

9. Hunt TE, Hader WJ: Intensive investigation and treat- ment of urinary tract infections in paraplegics. Med Serv J Canada 22:548-558, 1966.

10. Sapico FL, Wideman PA, Finegold SM: Aerobic and anaerobic flora in bladder urine of patients with in- dwelling urethral catheters. Urology 3:382-384, 1978.

11. McLeod JW, Glasg MB, Mason MJ, et ai: Survey of the different urinary infections which develop in the para- plegic and their relative significance. Microbiology 3:124-143, 1965.

12. Sherman FT, Tucci B, Libow LS, et al: Nosocomial urinary-tract infections in a skilled nursing facility. J Am Geriatr Sot 23:456-461, 1980.

13. Lennett EH, Spaulding EH, Truant JP, editors: Manual of clinical microbiology, ed 2. Washington, DC, 1974, American Society for Clinical Microbiology.

14. Bauer AW, Kirby WM, Sherris JC, et al: Antibiotic susceptibility testing by a standardized single disc method. Am J Clin Path01 45:493-496, 1966.

15. Musher DM, Thoesteinsson SB, Airola VM II: Quanti- tative urinalysis: Diagnosing urinary tract infections in men. JAMA 236:2069-2072, 1976.

16. Jones SR, Smith JW, Sanford JP: Localization of uri- nary tract infections by detection of antibody-coated bacteria in urine sediment. N Engl J Med 290:591-593, 1974.

17. Thorley JD, Babin GK, Reinarz JA: The prevalence of antibody-coated bacteria in urine. Am J Med Sci 27§:75-80, 1978.

18. Mackowiak PA: The normal microbial flora. N Engl J Med 307:83-93, 1982.

19. Liedberg CF: Nosocomial urinary tract infections, with special reference to the role of the indwelling catheter. Acta Chir Stand 207:469-473, 1980.

20. Wibell L, Scheynius A, Norrman K: Methenamine- hippurate and bacteriuria in the geriatric patient with a catheter. Acta Med Stand 207~469-473, 1980.

21. Vainrub B, Musher DM: Lack of effect of methenamine in suppression of, or prophylaxis against, chronic uri- nary infection. Antimicrob Agents Chemother 12:625- 629, 1977.