ORIGINAL ARTICLE Evaluation of the leukocyte esterase test

SF Yuen  FN Ng  Evaluation of the accuracy of LY So  leukocyte esterase testing to detect pyuria in young febrile children: prospective study 

○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○○ !  !"!#$%&'()*+,-./01 Objective. To study the accuracy and clinical application of the dipstick leukocyte esterase test in the detection of pyuria in young febrile children suspected to have urinary tract . Design. Prospective study. Setting. Regional hospital, Hong Kong. Patients. Urine samples were taken from 215 children younger than 2 years who were suspected to have (fever without an obvious focus of infection). Main outcome measures. The accuracy of the dipstick leukocyte esterase test in detecting significant pyuria defined as a leukocyte count ≥10 mm3 (≥0.01 x 109 /L). Results. Two hundred and fifty-four urine samples collected by bag, mid- stream clean-catch, suprapubic bladder aspiration, or urethral catheteriza- tion were examined. Using urine microscopy results as a reference, the sensitivity and specificity of the leukocyte esterase test in detecting signifi- cant pyuria were found to be 72.0% and 85.8%, respectively; the positive and negative predictive values were 55.4% and 92.6%, respectively; and the positive and negative likelihood ratios were 5.1 and 0.3, respectively. Key words: Conclusions. The dipstick leukocyte esterase test cannot accurately detect Child; Pyuria/diagnosis; pyuria in young febrile children. It is also not appropriate as a screening test Reagent strips; to exclude pyuria, reduce the need for the microscopic examination of urine, Sensitivity and specificity; or indicate when a hospital admission for probable urinary tract infection is Urinalysis; needed. Urinary tract   !"#$%&'()*+,-./0123456+789  !  !"#$%&'(    !"  L    !"#$%  !  !"#215 !"2 !"#$%&'()*+(  !"#$%   !"#$%&)  ! ≥ 3 ≥  !  !"#$ !"#$%&'()*+!"#, 10 mm ( 0.01 x 109 /L) !"#$%&'( HKMJ 2001;7:5-8   !"#$%&'()*+,%-./01234%5678  !"#254 !"#$%&'()*+,-./01234 Department of Paediatrics, Pamela Youde Nethersole Eastern Hospital, 3 Lok Man  !"#$%&'()*+,-./012372.0%85.8% Road, Chai Wan, Hong Kong  !"#$%55.4%92.6% !" #$ %&'(5.10.3 SF Yuen, LMCHK, MRCP, FHKAM (Paediatrics) FN Ng, MRCP, FHKAM (Paediatrics)   !"#$%&'()*+,-.-/012345678( LY So, FRCP (Edin), FHKAM (Paediatrics)  !"#$%&'()*+,-./0#*1234567 !89 Correspondence to: Dr SF Yuen  !"#$%&'()*+

HKMJ Vol 7 No 1 March 2001 5 Yuen et al

Introduction care: by bag collection, midstream clean-catch, supra- pubic bladder aspiration, or urethral catheterization. Urinary tract infection (UTI) is the most common ser- For most patients, urine was first collected non-invasively. ious bacterial infection found in febrile young children, Urine cultures obtained by suprapubic aspiration or with a reported prevalence of 4.1% to 7.5%.1-4 A pre- catheterization were performed for those patients with a sumptive diagnosis of UTI is often made on the basis higher likelihood of having a UTI (positive for pyuria of the clinical presentation and the presence of pyuria. by microscopic examination of urine) or for those under- To diagnose UTI in young children, a quantitative going a full work-up. urine culture of a sample obtained by suprapubic tap or catheterization should be completed.5 The detection Examination of urine of pyuria can assist in selecting appropriate patients Urine specimens were examined within 1 hour of col- for proper urine culture, thus reducing unnecessary lection by dipstick LE testing and by light microscopy. invasive procedures. Dipstick analysis Microscopic examination of urine is the standard Dipstick analysis is a semi-quantitative test that method used to detect pyuria. However, the dipstick test detects -specific esterase activity by the to measure urinary leukocyte esterase (LE) activity is conversion of indoxyl carboxylic acid ester in the re- quick, inexpensive, and does not require technical ex- agent strip into an indoxyl moiety, which then reacts pertise. This test is commonly used to identify pyuria in with a diazonium salt to produce a violet pigment. The accident and emergency departments and in out-patient Ecur-Test reagent strip (Boehringer Mannheim, UK clinics in which a urine microscopy service is not avail- Limited) was used in this study.4 Nurses performed able. Studies have investigated the efficacy of the the test according to the manufacturer’s instructions, dipstick LE test in detecting pyuria in adults.6-10 The without knowing the result of the microscopic exam- sensitivity of the test ranges from 78.0% to 99.3% whereas ination. A leukocyte count of 10-25 mm3 (0.01 to 0.025 the specificity of the test ranges from 69.0% to 99.3%.11 x 109 /L) or higher was considered a positive result. Studies of children have suggested that dipstick tests (the LE test with or without the nitrite test) are as accu- Microscopic examination rate as microscopic examination in predicting bacteri- Using the Fuchs-Rosenthal cell counting chamber, uria.12-15 The accuracy of using the dipstick LE test to uncentrifuged urine from the same urine sample was detect pyuria in children, however, remains uncertain. examined under the light microscope by the doctor on call, who was not aware of the dipstick LE test The dipstick LE test is seldom used in the Depart- result. A leukocyte count of ≥10 mm3 (≥0.01 x 109 /L) ment of Paediatrics at the Pamela Youde Nethersole was taken to be a positive result. Eastern Hospital. Referrals of children with suspected UTI are often received from accident and emergency Statistical analysis departments based on a positive result from the dip- Using results from the microscopic examination as a stick LE test alone. Pyuria or UTI is subsequently reference, the sensitivity and specificity of the dipstick diagnosed in only a few such patients. This pro- LE test, positive and negative predictive values, and spective study aimed to determine the accuracy of the likelihood ratios were calculated. dipstick LE test and its usefulness as a screening test for pyuria. The study was confined to children younger Results than 2 years—the age-group at the greatest risk of renal scarring resulting from UTI. During the study period, 296 urine specimens were examined microscopically. All but 42 specimens were Methods tested with the dipstick LE test. The results of micro- scopic examination and the dipstick LE test were Specimen collection compared in 254 urine specimens from 215 patients. Urine specimens were obtained from children younger The urine collection methods were as follows: 204 than 2 years who presented to the paediatric department were obtained by bag collection, 12 by midstream of the Pamela Youde Nethersole Eastern Hospital clean-catch, 17 by suprapubic bladder aspiration, and from July 1998 through October 1998 for in-patient 21 by urethral catheterization. care because of suspected UTI (fever without an obvious focus of infection). The method of urine collection was The dipstick LE test and microscopic examination decided by the paediatrician managing the patient’s results are shown in the Table. The dipstick LE test

6 HKMJ Vol 7 No 1 March 2001 Evaluation of the leukocyte esterase test

Table. Urine test results* selecting urine samples for microscopy. The high nega- Leukocyte Microscopic Total tive predictive value (92.6%) appears favourable in this esterase test examination regard. Nevertheless, six (23%) of the 26 confirmed Positive Negative cases of UTI tested negative according to the dipstick Positive 36 29 65 LE test. If microscopic examination had not been Negative 14 175 189 performed, these six patients would not have been con- Total 50 204 254 firmed to have UTI. Hence, the dipstick LE test cannot * Sensitivity = 36/50 (72.0%), specificity = 175/204 (85.8%); positive be used reliably as a screening test in selecting urine predictive value = 36/65 (55.4%); negative predictive value = 175/189 (92.6%); positive likelihood ratio = 36/50:29/204 (5.1); negative samples for microscopic examination or bacterial culture. likelihood ratio = 14/50:175/204 (0.3) Although this study was performed in children had a sensitivity and specificity of detecting clinically younger than 2 years who were admitted to hospital, significant pyuria of 72.0% and 85.8%, respectively. it is reasonable to believe that findings will be similar The positive and negative predictive values were 55.4% for paediatric patients in emergency or out-patient and 92.6%, respectively, and the positive and negative settings. If the dipstick LE test is used widely in likelihood ratios were 5.1 and 0.3, respectively. febrile children and a positive test is taken to indicate pyuria and probable UTI, many children would be Twenty-six patients were subsequently shown to admitted to hospital unnecessarily. have a UTI based on a positive culture of urine that had been sampled by suprapubic tap or catheterization. Six Conclusion (23%) of the 26 patients tested had negative results according to the dipstick LE test. The dipstick LE test is not an accurate method of detecting pyuria in young febrile children. Further- Discussion more, it cannot be used as a screening test to exclude pyuria because, although it would reduce the number The risk of renal damage from UTI is greatest in chil- of microscopic examinations needed, this reduction dren younger than 2 years; thus, early diagnosis and would be at the expense of failing to diagnose UTI in prompt treatment are important. Studies of the dip- some patients. The widespread use of dipstick LE stick LE test in adults6-10 have shown that the test is testing for the screening or diagnosis of suspected both sensitive and specific in detecting pyuria micro- UTI in young children is not supported by the find- scopically.11 Only a limited number of studies in chil- ings of this study. Using a positive dipstick LE test dren have so far been reported.16-18 Hoberman and Wald18 result as an indication of pyuria and probable UTI in have demonstrated that the dipstick LE test has a low accident and emergency departments appears to result sensitivity (52.9%) in detecting pyuria (as defined by in unnecessary hospital admissions. a leukocyte count of ≥10 mm3 [≥1.0 x 107 /L]) in febrile children younger than 2 years. The difference References between studies of adults and children might relate to the degree of pyuria, the enzyme content of immature 1. Hoberman A, Chao HP, Keller DM, Hickey R, Davis HW, leukocytes, or both. Hoberman and Wald18 calculated Ellis D. Prevalence of urinary tract infection in febrile infants. a positive predictive value for the dipstick LE test of J Pediatr 1993;123:17-23. 82.1%, but the specificity was not reported in their 2. Roberts KB, Charney E, Sweren RJ, et al. Urinary tract infec- tion in infants with unexplained fever: a collaborative study. study. J Pediatr 1983;6:239-42. 3. Bauchner H, Philipp B, Dashefsky B, Klein JO. Prevalence In this study, the dipstick LE test was found to have of in febrile children. Pediatr Infect Dis J 1987;6: a sensitivity of 72.0% and a specificity of 85.8% (Table). 239-42. The positive and negative predictive values (55.4% 4. Crain EF, Gershel JC. Urinary tract infections in febrile in- fants younger than 8 weeks of age. Pediatrics 1990;86:363-7. and 92.6%, respectively) imply that using the dipstick 5. American Academy of Pediatrics. Practice parameter: the LE test alone to detect pyuria would result in a large diagnosis, treatment, and evaluation of the initial urinary tract number of false-positive and some false-negative infection in febrile infants and young children. Pediatrics 1999; results. More invasive urine sampling procedures would 103:843-52. consequently be performed unnecessarily, whereas 6. Gillenwater JY. Detection of urinary leukocytes by chemstrip- 1. J Urol 1981;25:383-4. some patients with pyuria and possible UTI would be 7. Kusumi RK, Grover PJ, Kunin CM. Rapid detection of pyuria overlooked. A further question that arises is whether by leukocyte esterase activity. JAMA 1981;245:1653-5. the dipstick LE test could act as a screening test for 8. Herlihy RE, Wilkerson R, Roy JB. New and rapid method for

HKMJ Vol 7 No 1 March 2001 7 Yuen et al

detection of pyuria by leukocyte esterase reaction. Urology 14. Weinberg AG, Gan VN. Urine screen for bacteriuria in symp- 1984;23:148-9. tomatic pediatric outpatients. Pediatr Infect Dis J 1991;10: 9. Mariani AJ, Luangphinith S, Loo S, et al. Dipstick chemical 651-4. urinalysis: an accurate cost-effective screening test. J Urol 15. Lohr JA, Portilla MG, Geuder TG. Making a presumptive 1984;132:64-6. diagnosis of urinary tract infection by using a urinalysis 10. Shaw ST, Poon SY, Wong ET. “Routine urinalysis”: is the performed in an on-site laboratory. J Pediatr 1993;122:22-5. dipstick enough? JAMA 1984;253:1596-600. 16. Kaltenis P, Baciulis V, Liubsis A. Reliability of detection of 11. Kiel DP, Moskowitz MA. The urinalysis: a critical appraisal. leucocyturia by means of the dip-stick Cytur-Test. Int Urol Med Clin North Am 1987;71:607-24. Nephrol 1984;16:233-5. 12. Goldsmith BM, Campos JM. Comparison of urine dipstick, 17. Hiraoka M, Hida Y, Hori C, Tuchida S, Kuroda M, Sudo M. microscopy, and culture for the detection of bacteriuria in Rapid dipstick test for diagnosis of urinary tract infection. Acta children. Clin Pediatr 1990;29:214-8. Paediatr Jpn 1994;36:379-82. 13. Shaw KN, Hexter D, McGowan KL, Schwatz JS. Clinical 18. Hoberman A, Wald ER. Pyuria and bacteriuria in urine evaluation of a rapid screening test for urinary tract infection specimens obtained by catheter from young children with fever. in children. J Pediatr 1991;118:733-6. J Pediatr 1994;124:513-9.

8 HKMJ Vol 7 No 1 March 2001