1179
of the request forms enabled the donor to be identified. For most incidents donor-recipient blood pairs were not received. 23 viral infections were identified in the serum donors, the most common being HCV. Of the 14HCV ELISA-reactive sera, 12 were confirmed by recombinant immunoblot assay (RIBA 2, Chiron); the other 2 reacted with c22c (p22) alone. 1 needlestick accident donor infected with HCV was also HIV infected. 5 donors were HBV infected but all were low-infectivity carriers whose serum contained anti-HBe. 3 HIV infections and 1 HTLV-I infection were also identified. This retrospective analysis reveals that the information and samples received by laboratories that provide microbiological follow-up of victims in needlestick accidents are insufficient. This should be rectified, especially since we have specific interventions for HBV and HIV infection. This study shows also that HCV infection is the most common parenterally transmitted pathogen to which needlestick recipients are exposed. The donors infected with HIV and HTLV were known before the time of the accident but none of those infected with either HBV or HCV were considered as hepatitis risks at the time. The estimate by Kiyosawa and colleagues3 of a post-needlestick transmission rate of HCV of 2-7% (95% CI 0’6%-8%) suggests that HCV infection poses a substantial risk to staff, in London anyway. Blood from any patient should be regarded as a source of infection when there is an accidental needlestick injury, and when such accidents do happen records must be kept at the time. Division of Virology, Department of Medical
Time
(minutes)
Microbiology,
University College and Middlesex School of Medicine, London W1P 6DB. UK
A. J. CUMMINS R. S. TEDDER
1. Klein RS, Freeman K, Taylor PE, et al.
Occupational risk for hepatitis C virus among New York City dentists. Lancet 1991; 338: 1539-42. 2. Tedder RS, Gilson RJC, Briggs M, et al. Hepatitis C virus: evidence for sexual transmission. Br Med J 1991; 302: 1299-302. 3. Kiyosawa K, Sodeyama T, Tanaka E, et al. Hepatitis C in hospital employees with needlestick injuries. Ann Intern Med 1991; 115: 367-69.
Time
(minutes)
Amylin concentrations and glucose control SIR,-Amylin is a 37 aminoacid protein isolated from pancreatic amyloid ; it is synthesised in and secreted from pancreatic &bgr; cells, along with insulin.! Amylin potently stimulates glycogenolysis2 and reduces insulin-stimulated glucose incorporation into glycogen in isolated skeletal muscle; in animals it elicits a rapid increase in plasma lactate followed by a sustained increase in plasma glucose.3 In hyperinsulinaemic glucose clamp experiments in rats and dogs, amylin caused "insulin resistance", as judged from the reduced rate of glucose infusion needed to maintain euglycaemia.4,5 There have been several reports of measurement of plasma amylin by radioimmunoassay.6-8 We have established a competitive radioimmunoassay for amylin-like immunoreactivity in human plasma using rabbit antisera (Peninsula Laboratories) and a robotic procedure for extracting and concentrating amylin. This assay gives coefficients of variation of 9-13% inter-assay and 2-10% intra-assay with spikes of synthetic human amylin calculated to increase amylin from 8 to 30 pmol/l. Recovery of unlabelled peptide from plasma at these concentrations of amylin is about 35%. We have measured amylin in people fasted overnight and after a 75 g oral glucose load (figure). 5 healthy volunteers (aged 40-61) had good glucose control in that the glucose concentration at 2 h had returned to fasting concentrations. Amylin rose to a peak averaging 9-3 pmol/1 above the basal concentration, and returned to baseline by 2 h. In 5 patients with adolescent type I diabetes aged 13-18, no significant change in amylin over the 2 h test was seen. This absence of amylin response to glucose in type I diabetes (where &bgr; cells are destroyed by autoimmune attack) is consistent with work implicatingP cells as the primary source of amylin secretion and supports the proposal that in type I diabetes there is a deficiency in not one but two islet-cell hormones. In 6 apparently healthy controls aged 53-71 grouped on the basis of compromised glucose control, the 2 h glucose concentration (mean 145 mg/dl) exceeded the fasting value by at least 35 % (average + 60%), placing these subjects in the upper
Time
(minutes)
Changes in amylin after oral glucose challenge in patients with type I diabetes (&Dgr;) and in healthy individuals with good (8)or compromised (D) glucose control. Results as mean (SE). Fasting amylin concentrations: type 13.1 (0-6), good glucose control 87 (46), and compromised glucose control 44 (1 5) pmot/L Also shown are incremental changes in insulin (• = non-diabetic, good glucose control, fasting 49.4[15’6] pmoljl; 0 = compromised control, fasting 75-7 7 [23.0] pmoljl) and in glucose (&Dgr; =type I, fasting 143-6 [37’6] mg/dl 8 = non-diabetic, good control, fasting 904 [25] mg/dl; [] non-diabetic, compromised control, fasting 88.55 [4.3] mg/dl)=
50% of published norms for age-matched controls.9 (By contrast, the 2 h glucose for those with good glucose control fell in the lower 50%.) Amylin concentrations in those with compromised glucose control remained high, averaging 142 pmol/1 above basal at the 2 h point and significantly above the concentration seen at 2 h in those with good glucose control (p < 0-0005). Thus, in this small group of
1180
otherwise normal individuals, compromised glucose control was associated with a sustained amylin increment after a glucose challenge. For the whole group of 11apparently healthy adults, the increment in amylin concentrations at 2 h was correlated with the increase in glucose (r 0-84, p < 0-01); the 2 h increment in glucose was less well correlated with the increment in insulin (r=0-65,
COMPARISON OF VIABLE AND CALCULATED MICROSCOPIC COUNTS WITH CORRECTION FACTORS
=
p > 005). These data indicate that in type I diabetes there is no secretion of amylin in response to a glucose challenge; individuals with good glucose control show transiently increased amylin concentrations while those with compromised control maintain high amylin values for at least 2 h after ingestion of glucose. Amylin Pharmaceuticals, Inc,
JOY E. KODA
9373 Towne Centre Drive, San Diego, California 92121, USA
MARK FINEMAN TIMOTHY J. RINK
Scripps Clinic and Research Foundation, La Jolla, California
GEORGE E. DAILEY DOUGLAS B. MUCHMORE
San
LOUIE G. LINARELLI
Diego, California
Cooper GJS, Day AJ, Willis AC, Roberts AN, Reid KBM, Leighton B. Amylin and the amylin gene: structure, function and relationship to islet amyloid and to diabetes mellitus. Biochim Biophys Acta 1989; 1014: 247-58. 2. Young AA, Mott DM, Stone K, Cooper GJS. Amylin activates glycogen phosphorylase in the isolated soleus muscle of the rat. FEBS Lett 1991; 281: 1.
149-51.
Young AA, Wang M-W, Cooper GJS. Amylin injection causes elevated plasma lactate and glucose in the rat. FEBS Lett 1991; 291: 101-04. 4. Frontoni S, Choi SB, Banduch D, Rossetti L. In vivo insulin resistance induced by amylin primarily through inhibition of insulin-stimulated glycogen synthesis in 3.
5.
skeletal muscle. Diabetes 1991; 40: 568-73. Koopsmans SJ, vanMansfeld ADM, Jansz JS, et al. Amylin-induced in vivo insulin resistance in conscious rats: the liver is more sensitive to amylin than peripheral
tissues. Diabetologia 1991; 34: 218-24. 6. Butler PC, Chou J, Carter WB, et al. Effects of meal ingestion on plasma amylin concentration in NIDDM and nondiabetic humans. Diabetes 1990; 39: 752-56. 7. Hartter E, Svoboda T, Ludvik B, et al. Basal and stimulated plasma levels of pancreatic amylin indicate its co-secretion with insulin in humans. Diabetologia 1991; 34: 52-54. 8. Sanke T, Hanabusa T, Nakano Y, et al. Plasma islet amyloid polypeptide (amylin) levels and their responses to oral glucose in type 2 (non-insulin-dependent) diabetic patients. Diabetologia 1991; 34: 129-32. 9. Sayetta RB, Murphy RS. Summary of current diabetes-related data from the National Center for Health Statistics. Diabetes Care 1979; 2: 105-19.
microscopy and clinically significant bacteriuria
Quantitative
SIR,-Vickers et all concluded that quantitative microscopy was preferable to culture as a cheap, rapid, reliable, and easily taught method for screening paediatric urine samples for bacteriuria. They maintained that counts of 107 organisms/ml on microscopy equated to viable counts on culture of 105 colony-forming units (cfu)/ml, and that microscopic enumeration is accurately achieved by applying a fixed correction factor of 10 to bacteria counted in a single plane of focus. Since they did not provide a reference for the technique and we were unaware of any publication, we decided to evaluate the method before considering its introduction. On three occasions, nutrient-broth cultures of five clinical isolates representing common urinary pathogens (Escherichia coli, Staphylococcus saprophyticus, Proteus mirabilis, Enterococcus faecalis, and Pseudomonas aeruginosa) were examined by preparing serial 10-fold dilutions in broth. For each dilution the viable bacterial count was measured2 and quantitative microscopy done1 (table). More than 100 organisms per grid were recorded as greater than 107 organisms/ml. Cocci and bacilli were assessed separately since they can be differentiated on microscopy. Microscopy reliably detected cultures of greater than 106 cfu/ml of bacilli and greater than 105 cfu/ml of cocci. Below these values it remained more sensitive for cocci but identified only 78% (7 of 9) of bacillary cultures containing greater than 105 cfu/ml and 54% (7 of 13) of coccal cultures containing 103 -105 cfu/ml. with those of Vickers et al, in which These results are 29% (13 of 45) urines with pure growths of greater than 105 cfu/ml were negative on microscopy-a finding curiously attributed to contamination. Vickers et al also acknowledge that "microscopy could quite easily miss a UTI [urinary tract infection] if the count is only just over lOS/ml" and found that "true" UTIs, in which, by
orgamsms/ml on microscopy, factor= Calculatedcfu/ml viable count
*CorrecUOn factor Correction NA= not
applicable
’
All
microscopic counts
>
107
organisms/ml
their definition, microscopy and culture both had to be positive, were usually associated with counts much higher than 105 cfu/ml. We found microscopy difficult to do accurately because the method calls for the entire Neubauer marked grid to be examined, with organisms tending to move in and out of the single plane of focus. Furthermore, at the lower limit of observable counts, very small numbers (occasionally only single organisms) were being recorded. We believe that to differentiate between mixed and pure growths would therefore be impossible. Our results show that whereas positive microscopy in a single plane of focus correlates with high bacterial counts, there is no constant relation between the two, and the method will only reliably detect counts equal to or greater than 106 cfu/ml for bacilli and 105 cfu/ml for cocci. In our laboratory, 90% of UTIs are due to bacilli, and 4-3% and 28-9% are positive at 104-5 and 105-6 cfu/ml, respectively. Although counts as low as 102 cfu/ml in midstream urine are important,3 present methods impose a limit of culture in most laboratories of 1 if cfuml. However, since midstream urinary bacterial counts of 1 if or greater cfu/ml are associated with UTI,3 we conclude that quantitative microscopy cannot be used to screen urine samples from culture, which we believe remains essential. Central Microbiological Laboratories, Western General Hospital, Edinburgh EH4 2XU, UK
P. A. CRAWFORD R. G. MASTERTON
1. Vickers D, Ahmad T, Coulthard MG. Diagnosis of urinary tract infection in children: fresh urine microscopy or culture? Lancet 1991; 338: 767-70. 2. Miles AA, Misra SS, Irwin JO. The estimation of the bactericidal power of the blood. J Hyg Camb 1938; 38: 732-49. 3. Kellogg JA, Manzella JP, Shaffer SN, Schwartz BB. Clinical relevance of culture versus screens for the detection of microbial pathogens in urine specimens. Am J Med 1987; 83: 739-45.
SIR,-We have been using "bacterioscopy" for many years as a and reliable technique to distinguish noninfected from infected urine samples. Our method was validated in a prospective study on 933 samples from infants and children attending our outpatient paediatric nephrology clinic. Fresh, cleanly voided, unspun urine specimens were examined by phase-contrast microscopy at x 40 in a Fuchs-Rosenthal counting chamber and bacteria were classified as: rare (no or only a few bacteria encountered in the whole square of 1 mm3), 1 + (a few bacteria in every square), 2 + (many, but readily countable number, of bacteria per square), and 3 + (uncountable bacteria). Comparison with quantitative bacteriological culture yielded the following results:
rapid, inexpensive,
Culture
compatible
*Colony-formmg units (cfu)/ml
A finding of "rare" bacteria on bacterioscopy virtually excluded significant bacteriuria (6 [1’ 3 %] false negatives and in half of them negativity was confirmed by dip-slide culture) while uncountable
of the request forms enabled the donor to be identified. For most incidents donor-recipient blood pairs were not received. 23 viral infections were identified in the serum donors, the most common being HCV. Of the 14HCV ELISA-reactive sera, 12 were confirmed by recombinant immunoblot assay (RIBA 2, Chiron); the other 2 reacted with c22c (p22) alone. 1 needlestick accident donor infected with HCV was also HIV infected. 5 donors were HBV infected but all were low-infectivity carriers whose serum contained anti-HBe. 3 HIV infections and 1 HTLV-I infection were also identified. This retrospective analysis reveals that the information and samples received by laboratories that provide microbiological follow-up of victims in needlestick accidents are insufficient. This should be rectified, especially since we have specific interventions for HBV and HIV infection. This study shows also that HCV infection is the most common parenterally transmitted pathogen to which needlestick recipients are exposed. The donors infected with HIV and HTLV were known before the time of the accident but none of those infected with either HBV or HCV were considered as hepatitis risks at the time. The estimate by Kiyosawa and colleagues3 of a post-needlestick transmission rate of HCV of 2-7% (95% CI 0’6%-8%) suggests that HCV infection poses a substantial risk to staff, in London anyway. Blood from any patient should be regarded as a source of infection when there is an accidental needlestick injury, and when such accidents do happen records must be kept at the time. Division of Virology, Department of Medical
Time
(minutes)
Microbiology,
University College and Middlesex School of Medicine, London W1P 6DB. UK
A. J. CUMMINS R. S. TEDDER
1. Klein RS, Freeman K, Taylor PE, et al.
Occupational risk for hepatitis C virus among New York City dentists. Lancet 1991; 338: 1539-42. 2. Tedder RS, Gilson RJC, Briggs M, et al. Hepatitis C virus: evidence for sexual transmission. Br Med J 1991; 302: 1299-302. 3. Kiyosawa K, Sodeyama T, Tanaka E, et al. Hepatitis C in hospital employees with needlestick injuries. Ann Intern Med 1991; 115: 367-69.
Time
(minutes)
Amylin concentrations and glucose control SIR,-Amylin is a 37 aminoacid protein isolated from pancreatic amyloid ; it is synthesised in and secreted from pancreatic &bgr; cells, along with insulin.! Amylin potently stimulates glycogenolysis2 and reduces insulin-stimulated glucose incorporation into glycogen in isolated skeletal muscle; in animals it elicits a rapid increase in plasma lactate followed by a sustained increase in plasma glucose.3 In hyperinsulinaemic glucose clamp experiments in rats and dogs, amylin caused "insulin resistance", as judged from the reduced rate of glucose infusion needed to maintain euglycaemia.4,5 There have been several reports of measurement of plasma amylin by radioimmunoassay.6-8 We have established a competitive radioimmunoassay for amylin-like immunoreactivity in human plasma using rabbit antisera (Peninsula Laboratories) and a robotic procedure for extracting and concentrating amylin. This assay gives coefficients of variation of 9-13% inter-assay and 2-10% intra-assay with spikes of synthetic human amylin calculated to increase amylin from 8 to 30 pmol/l. Recovery of unlabelled peptide from plasma at these concentrations of amylin is about 35%. We have measured amylin in people fasted overnight and after a 75 g oral glucose load (figure). 5 healthy volunteers (aged 40-61) had good glucose control in that the glucose concentration at 2 h had returned to fasting concentrations. Amylin rose to a peak averaging 9-3 pmol/1 above the basal concentration, and returned to baseline by 2 h. In 5 patients with adolescent type I diabetes aged 13-18, no significant change in amylin over the 2 h test was seen. This absence of amylin response to glucose in type I diabetes (where &bgr; cells are destroyed by autoimmune attack) is consistent with work implicatingP cells as the primary source of amylin secretion and supports the proposal that in type I diabetes there is a deficiency in not one but two islet-cell hormones. In 6 apparently healthy controls aged 53-71 grouped on the basis of compromised glucose control, the 2 h glucose concentration (mean 145 mg/dl) exceeded the fasting value by at least 35 % (average + 60%), placing these subjects in the upper
Time
(minutes)
Changes in amylin after oral glucose challenge in patients with type I diabetes (&Dgr;) and in healthy individuals with good (8)or compromised (D) glucose control. Results as mean (SE). Fasting amylin concentrations: type 13.1 (0-6), good glucose control 87 (46), and compromised glucose control 44 (1 5) pmot/L Also shown are incremental changes in insulin (• = non-diabetic, good glucose control, fasting 49.4[15’6] pmoljl; 0 = compromised control, fasting 75-7 7 [23.0] pmoljl) and in glucose (&Dgr; =type I, fasting 143-6 [37’6] mg/dl 8 = non-diabetic, good control, fasting 904 [25] mg/dl; [] non-diabetic, compromised control, fasting 88.55 [4.3] mg/dl)=
50% of published norms for age-matched controls.9 (By contrast, the 2 h glucose for those with good glucose control fell in the lower 50%.) Amylin concentrations in those with compromised glucose control remained high, averaging 142 pmol/1 above basal at the 2 h point and significantly above the concentration seen at 2 h in those with good glucose control (p < 0-0005). Thus, in this small group of
1180
otherwise normal individuals, compromised glucose control was associated with a sustained amylin increment after a glucose challenge. For the whole group of 11apparently healthy adults, the increment in amylin concentrations at 2 h was correlated with the increase in glucose (r 0-84, p < 0-01); the 2 h increment in glucose was less well correlated with the increment in insulin (r=0-65,
COMPARISON OF VIABLE AND CALCULATED MICROSCOPIC COUNTS WITH CORRECTION FACTORS
=
p > 005). These data indicate that in type I diabetes there is no secretion of amylin in response to a glucose challenge; individuals with good glucose control show transiently increased amylin concentrations while those with compromised control maintain high amylin values for at least 2 h after ingestion of glucose. Amylin Pharmaceuticals, Inc,
JOY E. KODA
9373 Towne Centre Drive, San Diego, California 92121, USA
MARK FINEMAN TIMOTHY J. RINK
Scripps Clinic and Research Foundation, La Jolla, California
GEORGE E. DAILEY DOUGLAS B. MUCHMORE
San
LOUIE G. LINARELLI
Diego, California
Cooper GJS, Day AJ, Willis AC, Roberts AN, Reid KBM, Leighton B. Amylin and the amylin gene: structure, function and relationship to islet amyloid and to diabetes mellitus. Biochim Biophys Acta 1989; 1014: 247-58. 2. Young AA, Mott DM, Stone K, Cooper GJS. Amylin activates glycogen phosphorylase in the isolated soleus muscle of the rat. FEBS Lett 1991; 281: 1.
149-51.
Young AA, Wang M-W, Cooper GJS. Amylin injection causes elevated plasma lactate and glucose in the rat. FEBS Lett 1991; 291: 101-04. 4. Frontoni S, Choi SB, Banduch D, Rossetti L. In vivo insulin resistance induced by amylin primarily through inhibition of insulin-stimulated glycogen synthesis in 3.
5.
skeletal muscle. Diabetes 1991; 40: 568-73. Koopsmans SJ, vanMansfeld ADM, Jansz JS, et al. Amylin-induced in vivo insulin resistance in conscious rats: the liver is more sensitive to amylin than peripheral
tissues. Diabetologia 1991; 34: 218-24. 6. Butler PC, Chou J, Carter WB, et al. Effects of meal ingestion on plasma amylin concentration in NIDDM and nondiabetic humans. Diabetes 1990; 39: 752-56. 7. Hartter E, Svoboda T, Ludvik B, et al. Basal and stimulated plasma levels of pancreatic amylin indicate its co-secretion with insulin in humans. Diabetologia 1991; 34: 52-54. 8. Sanke T, Hanabusa T, Nakano Y, et al. Plasma islet amyloid polypeptide (amylin) levels and their responses to oral glucose in type 2 (non-insulin-dependent) diabetic patients. Diabetologia 1991; 34: 129-32. 9. Sayetta RB, Murphy RS. Summary of current diabetes-related data from the National Center for Health Statistics. Diabetes Care 1979; 2: 105-19.
microscopy and clinically significant bacteriuria
Quantitative
SIR,-Vickers et all concluded that quantitative microscopy was preferable to culture as a cheap, rapid, reliable, and easily taught method for screening paediatric urine samples for bacteriuria. They maintained that counts of 107 organisms/ml on microscopy equated to viable counts on culture of 105 colony-forming units (cfu)/ml, and that microscopic enumeration is accurately achieved by applying a fixed correction factor of 10 to bacteria counted in a single plane of focus. Since they did not provide a reference for the technique and we were unaware of any publication, we decided to evaluate the method before considering its introduction. On three occasions, nutrient-broth cultures of five clinical isolates representing common urinary pathogens (Escherichia coli, Staphylococcus saprophyticus, Proteus mirabilis, Enterococcus faecalis, and Pseudomonas aeruginosa) were examined by preparing serial 10-fold dilutions in broth. For each dilution the viable bacterial count was measured2 and quantitative microscopy done1 (table). More than 100 organisms per grid were recorded as greater than 107 organisms/ml. Cocci and bacilli were assessed separately since they can be differentiated on microscopy. Microscopy reliably detected cultures of greater than 106 cfu/ml of bacilli and greater than 105 cfu/ml of cocci. Below these values it remained more sensitive for cocci but identified only 78% (7 of 9) of bacillary cultures containing greater than 105 cfu/ml and 54% (7 of 13) of coccal cultures containing 103 -105 cfu/ml. with those of Vickers et al, in which These results are 29% (13 of 45) urines with pure growths of greater than 105 cfu/ml were negative on microscopy-a finding curiously attributed to contamination. Vickers et al also acknowledge that "microscopy could quite easily miss a UTI [urinary tract infection] if the count is only just over lOS/ml" and found that "true" UTIs, in which, by
orgamsms/ml on microscopy, factor= Calculatedcfu/ml viable count
*CorrecUOn factor Correction NA= not
applicable
’
All
microscopic counts
>
107
organisms/ml
their definition, microscopy and culture both had to be positive, were usually associated with counts much higher than 105 cfu/ml. We found microscopy difficult to do accurately because the method calls for the entire Neubauer marked grid to be examined, with organisms tending to move in and out of the single plane of focus. Furthermore, at the lower limit of observable counts, very small numbers (occasionally only single organisms) were being recorded. We believe that to differentiate between mixed and pure growths would therefore be impossible. Our results show that whereas positive microscopy in a single plane of focus correlates with high bacterial counts, there is no constant relation between the two, and the method will only reliably detect counts equal to or greater than 106 cfu/ml for bacilli and 105 cfu/ml for cocci. In our laboratory, 90% of UTIs are due to bacilli, and 4-3% and 28-9% are positive at 104-5 and 105-6 cfu/ml, respectively. Although counts as low as 102 cfu/ml in midstream urine are important,3 present methods impose a limit of culture in most laboratories of 1 if cfuml. However, since midstream urinary bacterial counts of 1 if or greater cfu/ml are associated with UTI,3 we conclude that quantitative microscopy cannot be used to screen urine samples from culture, which we believe remains essential. Central Microbiological Laboratories, Western General Hospital, Edinburgh EH4 2XU, UK
P. A. CRAWFORD R. G. MASTERTON
1. Vickers D, Ahmad T, Coulthard MG. Diagnosis of urinary tract infection in children: fresh urine microscopy or culture? Lancet 1991; 338: 767-70. 2. Miles AA, Misra SS, Irwin JO. The estimation of the bactericidal power of the blood. J Hyg Camb 1938; 38: 732-49. 3. Kellogg JA, Manzella JP, Shaffer SN, Schwartz BB. Clinical relevance of culture versus screens for the detection of microbial pathogens in urine specimens. Am J Med 1987; 83: 739-45.
SIR,-We have been using "bacterioscopy" for many years as a and reliable technique to distinguish noninfected from infected urine samples. Our method was validated in a prospective study on 933 samples from infants and children attending our outpatient paediatric nephrology clinic. Fresh, cleanly voided, unspun urine specimens were examined by phase-contrast microscopy at x 40 in a Fuchs-Rosenthal counting chamber and bacteria were classified as: rare (no or only a few bacteria encountered in the whole square of 1 mm3), 1 + (a few bacteria in every square), 2 + (many, but readily countable number, of bacteria per square), and 3 + (uncountable bacteria). Comparison with quantitative bacteriological culture yielded the following results:
rapid, inexpensive,
Culture
compatible
*Colony-formmg units (cfu)/ml
A finding of "rare" bacteria on bacterioscopy virtually excluded significant bacteriuria (6 [1’ 3 %] false negatives and in half of them negativity was confirmed by dip-slide culture) while uncountable
