Raheela Hussain  ( Department of Pathology, Postgraduate Medical Institute, 6, Birdwood Road, Lahore. )
Naseer Ahmed Chaudhry  ( Department of Pathology, Postgraduate Medical Institute, 6, Birdwood Road, Lahore. )
Muhammad Saeed Anwar  ( Department of Pathology, Postgraduate Medical Institute, 6, Birdwood Road, Lahore. )
Saeed Akhtar Khan  ( Department of Pathology, Postgraduate Medical Institute, 6, Birdwood Road, Lahore. )
Muhammad Munir  ( Department of Pathology, Postgraduate Medical Institute, 6, Birdwood Road, Lahore. )
Muhammad Tayyab  ( Department of Pathology, Postgraduate Medical Institute, 6, Birdwood Road, Lahore. )

Two hundred and fifty cases of clinically suspected urinary tract infection were analysed for the detection of bacteriuria, Parameters studied included direct Gram staining, pyuria on microscopic examination of uncentrifuged urine and d. ‘strip method for the detection of blood, protein, nitrite and leucocyte esterase. Significant bacteriuria (colony count 10 5 per ml) was found in 112 cases with a positivity ranging from 65 to 83% for the presence of food, protein, nitrite and leucocyte esterase. Highest positive predictive values were obtained with the presence of nitrite and leucocyte esterase (98%), blood, protein and nitrite (94%) as well as with blood, protein, nitrite and leucocyte esterase (98%). Both pyuria and direct Gram staining were positive in 85% cases. The combined presence of both these parameters gave 100% positive predictive value. Gram staining combined with pyuria was more effective and economical as compared to the dipstrips for the detection of bacteriuria (JPMA 46: 38, 1996).

Urinary tract infection (UTI) cannot be diagnosed on clinical grounds alone¹. Confirmation of UTI requires docu­mentalion of bacteria in bladder urine². However, the culture in clinical practice is usually done on voided urine which in women is easily contaminated with perineal bacteria. Quanti­tative culture and specific identification of the organisms in urine are used to distinguish contamination from true urinary tract infection^2,3. Bacterial counts by conventional methods are difficult to adopt in an ordinary diagnostic laboratory and are expensive intime and material^4-6. The simple methods like calibrated ioops, blotting papers, strips and dipslides also require overnight incubation^4-8. As an alternative to the conventional methods, several presumptive tests are available which either quantitate bacteria by non-cultural methods or suggest infection by the demonstration of an inflammatory reaction. Such tests are microscopy for pyuria and bacteriuria, chemical tests for detection of pus cells and bacteria in urine and automated methods for bacteriuria detection^9-11.
Pyuria is a term used to describe the appearance of increased number of polymorphonuclear leucocytes in the urine¹². It is an evidence of acute inflammation¹³. However, pyuriais not always associatedwithbacteriuria and its absence does not exclude infection^14,15. Leucocytes are normally absent in midstream urine from healthy children and male subjects, The urine of adult women frequently contains leucocytes and a small number should be regarded as normal⁹. Ten cells per cubic mm in uncentrifuged urine is the upper limit of normal leucocyte excretion¹². The presence of 100 pus cells per cubic mm in uncentrifuged urine specimens coffe­lates well with infection^9,12,16.
Microscopy of urine for bacteriuria is a rapid screening technique which has been used for years as a standard laboratory practice¹¹. Urine can be examined stained or unstained and centrifuged or uncentrifuged. The most widely used method is the Gram stained smear of fresh uncentrifuged specimen of unne examined under oil immersion objective¹¹. Visualizing any bacteria by this technique is reported positive, with the sensitivity being 69 to 96%^17-20.
As Enterobacteriaceae, the most prevalent orpnisms found in infected urine, reduce nitrate to nitrite^21,22. , so the nitrate reduction test is a useful tool and nitrite detection is an indicator of UTI²³. Presence of blood or protein in the urine detected by multipatch strips also indicates urinary tract infection^24,25. Sensitivity of more than 85% and negative predictive value of more than 86% has been reported^24,26. Pyuria can also be assessed with distrips. The estcrases present in neutrophils are detected with the help of a dipstrip 27. This parameter with the presence of blood, protein and nitrite in urine gives a sensitivity from 91% to 94% and negative predictive value from 89% to 95%^24,25.
The present study compares the screening of urine by Gram staining and pyuria assessment in uncentrifuged urine with the dipstrip tests for blood protein mtnte and leucocyte esterase.

Two hundred and fifty specimens of urine were obtained from both patients and in-patients of Sir Ganga Ram Hospital and Mayo Hospital, Lahore from patients with suspected UTI irrespective of age and sex. Brief history of patients was obtained and patients on antibiotics were excluded.
Urine samples from each patient were labelled and immediately transported to the laboratory where they were gently shaken and divided into two aliquotes. In case of delay, samples were kept at 4°C and analysed within two hours of collection. One aliquot was used for screening tests and the other for inoculating culture media (blood agar and Mac­Conkey agar).
The dipstnp examination of the urine samples was carried out with Multistix 8 (Ames, Miles Inc. Diagnostic Division, Elkhart, In 46515, USA) urine strips. The strip was dipped in fresh uncentrifuged specimen of urine and read according to the manufacturers instructions.
Method of Collee et al28 was used to determine the presence of pus cells in uncentrifuged urine.
Gram stained smears from uncentrifuged urine were made by spreading one p.! (0,001 ml) of urine on a 3x1 inch clean, grease free glass slide wit]ha calibrated loop overan area of 3-5 mm diameter. Eight similar smears were made per slide and examined under oil immersion objective. Presence of Gram negative rods in any high power field constituted a positive smear for infection.

Among the 250 cases, there were 112 clinically signifi­cant cultures, the chemistry of which is given in Table 1.

The comparison of clinically significant culture with blood, protein, nitrite and leucocyte esterase in urine, singly as well as in combination is shown in Table II.

The presence of blood, protein, nitrite and leucocyte esterase were found to be significant individually and in combination, in detecting infected urine samples. They also provided a high sensitivity and specificity as well as positive and negative predictive values..
The comparison of clinically significant culture with Gram staining of uncentrifuged urine and with pyuria, was also found to be significant in detecting infected urine samples (Tables III and IV).


Sensitivity of more than 84% was found with both these parameters while positive predictive value with Gram staining was 100% and with pyuria assessment 97%.

The present study supports the view that presence of blood, protein and nitrite and leucocyte esterase in urine are good indicators of urinary tract infection^24-26. High positive predictive values obtained in the present study could be due to collection of samples from symptomatic patients not taking any antibiotics. The study samples were either processed promptly or were kept at 4°C and processed within two hours of collection. Other workers examined random urine samples including catheter, midstream, clean catch and bag specimens, irrespective of time of collection, storage or transportation conditions¹².
In this study the nitrite test was positive in significant numbers of culture positive cases. Vaiying results have been reportedby otherworkers. Low sensitivity of nitrite test(39%) was reported by Moffat et al⁵, whereas Sinaniotis et al²⁹ found nitrite test unable to detect 40% of cases with proven bacteriuria. Craig et al⁸, observed 100% sensitivity and specificity ma field study and 42%positivity inhospital cases with significant bacteriuria.
Most common organisms responsible for urinary tract infection reduce nitrate to nitrite²¹. Consequently the presence of non- nitrate reducing organisms in urine can give false-negative results. In the present study, non-nitrate reducing organisms were present in 12% of the cases. Craig et al⁸ found that positive nitrite test strongly suggests UTI when done on first morning sample which allows a long period for nitrite to accumulate in the bladder. The test was not found useful and gave false negative results on randomly collected samples, or for patients receiving parenteral nutrition without a nitrogen source when concentration of bacteria was less than 10⁵ per ml or when urine was diluted or slow growing organisms were present^8,30,31. James et al²³ showed urobilinogen to interfere with nitrite analysis and the test was not reliable in the presence of elevated levels of urobilinogen. Factors such as phenzopyridine, blood, ascofoic acid and bilirubin also influence the sensitivity of the test²².
The sensitivity of leucocyte esterase in detecting in­fected urine samples reported by Marquette et a1³² was 83% and by Roberston and Duff³¹ 77%. However, the specificity was 70% and 96% respectively. This test is based on esterases present in polymorphonuclear cells which convert indoxyl cathomc ester, the substrate of dipstrip, to indoxyl and this in the presence of atmospheric oxygen becomes indigo blue³³. The esterase activity is not influenced by bacteria, commonly used drugs or variable composition of urine and is not dependant on intact cells³⁴, Trichomonas can give a positive leucocyte esterase reaction³⁵. Two samples of urine obtained from female patients in the present study gave a positive leucocyte esterase test in absence of pyuria due to trichomonas infection.
The combined strip tests when compared with culture positive cases gave comparable results. However, lower specificity and positive predictive value were reported as compared to the present study²⁴.
The presence of pus cells in urine is an evidence of acute inflammation¹³. However, pyuria without bacteriuria can occur in tuberculosis, urinary calculi, toxic nephropathy, bladder tumors, resolving bacterial infections, bacterial infec­tions with antimicrobial agents in urine, infection adjacent to the urinary tract and abacterial infections³⁶. In this study pus cells were found in 85% of culture positive and 2% of culture negative cases. While 15% of infected cases were negative for pus cells. The overall sensitivity of this indicator for infection was 85% and specificity was 98%. The predictive value for positive result was 97% and for negative result 89%. Results reported by various workers^37-39. vary depending upon the patient population and the presence of complicating factors such as diabetes. calculus or analgesic nephropathy.
The Gram staining of uncentrifuged urine was found to be significant (P<0.001) in detecting culture positive cases in the present study. Sensitivity was 75% when compared with total culture positive cases and 85% withclinically significant cultures. The results are comparable to those reportedby other workers^17,19,20. where sensitivity varied from 89% to 93%, whereas Jorgenson and Jones¹⁸ detected 69% of all urine samples by Gram staining and 74% of those containing Gram negative bacilli. The chemical tests by dipstrips were intro­duced as cost effective measures^20,33, but in our country the strips are costly and uneconomical. As technical staff is easily available and the volume of work is large, it is more feasible to examine the uncentrifuged urine for pus cells. This alongwith Gram staining are strong indicators of infection and have high positive predictive values. Both techniques have proved to be effective and inexpensive for diagnosing urinary tract infec­tion.

The help of Dr. Taj Bano, Dr. Zaibunmsa and others at Mayo and Sir Ganga Ram Hospital in conducting this study and Mr. Abdul Ghaffar Naeem for typing the manuscript is gratefully acknowledged.

1. Maskell, R. A controlled trial of the use of dip slides in general practice for the diagnosis of urinary infection. j. Clin. Pathol., 1973;26:18 1-83.
2. Johnson, JR. and Stamm, W.E. Urinary tract infections in women: Diagnosis and treatment. Ann. Intern. Med., 1989;111 :906-17.
3. Smellie, S. and Normand, C. Urinary tract infection and reflux nephropathy in children. Med. Inteme, 1991 ;87:361 0-3617.
4. Gould, J.C. Quantity and quality in the diagnosis of urinary tract infections. Br. J. Urol,, 1965;37:7-12.
5. Moffat, C.M., Britt, M.R. and Burke, JR Evaluation of miniature test for bacteriuria using dehydrated media and nitrite pads. Appl. Microbiol. 1974;28:95-99.
6. Dureden, B.1. and Moyes, A. Corpparison oflaboratorymethods in the diagnosis of urinary tract infection. J. Clin. Pathol., 1976;29:286:291.
7. Goodwin, CS. and Stewart, A.E. Detection of bacteriuria in women attending local authority antenatal clinics using the dip- in oculum transport medium spoon method. S. Clin. Pathol., 1968;21 :445-48.
8. Craig, WA., Kunin, C.M. and Degroot, J. Evaluation of new urinary tract infection screening devices. App!. Microbiol., 1973;26: 196-201.
9. Coltman, K.D. Urinary screening in general practice. Practitioner, 1981 ;225 :668-672.
10. PeSo, MT. Automated methods for detection of bacteriuria. Ant J. Med., 1983;(suppl IB)75:71-78,
11. Jenkins, R.D., Fenn, JR and Matsen, J.M. Review of urine microscopy for bacteriuria. JAMA, 1986;255:3397-3403.
12. Topley, N. and Williams, J.D. Urinary tract infection in adults. Med. Inteme., 1991;87:3604-3609.
13. Brumfitt, W. Urology and thefamily doctor. Practitioner, 1984;228:97-98.
14. Moore, T., Parker, C. and Edwards, E.C. Sterile non-tuberculous pyuris. Br. J. Urol., 1971;43:47-51.
15. Cattell, W.R. Urinary infections in adults - 1985. Postgrad. Med. J., 1985;61:907-913.
16. Jones, K.V Urinary tract infections in childhood. Practitioner, 1991 ;255:1 35-143.
17. Heinze, PA., Lauri, M.S., Thrupp, D. et al. A rapid (4-6 hours) urine culture system for direct identification and direct antimicrobial susceptibility testing. Am. J. Clin. Pathol., 1979;71 :177-183.
18. Jorgenson, J.H. and Jones, PM. Comparative evaluation ofthe limulus assay and the direct gram stain for detection ofaignificant bacteriuria. Am. J. Clin. Pathol., 1975;63:142-148.
19. Tilton, RE. and Tilton, R.C. Automated direct antimicrobial susceptibility testing of microscopically screened urine cultures. 3. Clin. Microbiol.,1980;11:157-161.
20. Sewell, D.L., Burt, SF, Gabbert, N.J. et al. Evaluation of the Chemstrip 9TM as a screening test for urinalysis and urine culture in men. Am. J. Clin. Pathol., 1985;83:740-43.
21. Lie, J.T. Evaluation of a nitrite test kit (Stat test) for the detection ofsignificant bacteriuria. J. Clin. Pathol., 1968;21 :443-444.
22. Rabi, T. Evaluation of a new sensitive nitrite test as a reliable screening tool for bacteriuria. J. Clin. Pathol., 1981;34:723- 729.
23. James, G.P,Paul, K.L. and Fuller, J.B. Urinary nitriteandurinary tract infection. Am.J. Clin. Pathol., 1978;70:671-78.
24. Lowe, PA. Chemical screening and prediction of bacteriuria a new approach. Med. Lab. Sci., 1985;42:28-32.
25. Doran, H.M. and Kensit, J.G. Screening forbacteriuria with clinitec-200, J. Clin. Pathol., 1988;41:1126-1129.
26. Boreland, PC. and Stoker, M. Dipstick analysis forscreening ofpaediatric urine. 3. din. Pathol., 1986;39:1360-1362,
27. Komaroff, A.L. Urinalysis and urine culture inwomenwith dysuria. Ann. Intern. Med., 1986;104:212-218.
28. Collee, J.G., Duguid, J.P, Fraser, A.G. Practical Medical Microbiology. 12th S. Edinburgh, Churchill Livingatone, 1989.
29. Sinaniotis, C.A., Haratsaris, M.N. and Papatatos, C.J. Nitrite Indicater Strip Test for Bacteriuria. Lancet, 1978;i:776-777.
30. Wilkins, E.G.L., Ratcliffe, J.C. and Roberts, C. Leucocyte estrase nitrite screening method for pyuria and bacteriuria. 3. Clin. Pathol., 1985;38:1342­1345.
31. Robertson, A.W. and Duff, P. The nitrite and leukocyte estrase tests for the evaluation of asymptomatic bacteriuria in obstetric patients. Obstet. Gynecol., 1988;71:878-881.
32. Marquete, G.P, Dillard, T, Bietla, S. et al. The validity ofthe leucocyte estrase reagent teststrip in detecting significant leucocyteuria. Am. J. Obstet. Gynecol., 1985;153:888-889.
33. Leighton, P.M and Little, J.A. Leucocyte estrase determination as a secondary procedure for urine screening. J. Clin. Pathol., 1985;3 8:229-232.
34. Kusumi, R.K., Grover, P.J. and Kunin, C.M. Rapid detection of pyuria by leucocyte estrase activity. JAMA., 1981 ;245:1653-1655.
35. Mariani, A.J., Lunangphinith, S., Loo, S. etal. Dipstick chemical urinalysis; An accurate cost effective screening test. J. Urol., 1984;132:6466.
36. Moore, t, Parker, C. and Edwards, Ed. Sterile non-tuberculous pyuria. Br. 3. Urol., 1971 ;43:47-51.
37. Robins, D.C., White, P.H.R,, Rogers, K.B. et al. Urine microscopy as an aid to detection ofbscteriuria. Lancet, 1975;i:476-478.
38. Tapsall, J.W., Taylor, PC., Bell, S.M. et al. Relevance of significant bacteriuria to aetiology and diagnosis ofurmnary tract infection.Lancet, 1975;i:637-639.
39. Latham, RH., Wong, E.S . ,Larson, A. et al. Laboratory diagnosis of urinary tract infection in ambulatory women. JAMA, 1985;254:3333-3336.