Kidney International, Vol. 7 (1975), p. 63—67
EDITORIAL
Urinary tract infection: Value of early diagnosis The value of presymptomatic diagnosis may be judged by the criteria of Wilson and Jungner [1].
The only means of obtaining accurate data would be to
collect the urine by suprapubic aspiration of the bladder, but this technique is not acceptable as a
These are 1) the condition sought must be an important health hazard; 2) a latent phase of the condition must
routine screening procedure. Among 16,000 school entrants, Kunin et al [6, 7] observed significant bacteriuria in 1.2% of the girls and 0.03% of the boys. Studies in the United Kingdom [8,
be recognizable by a simple, acceptable test; 3) the natural history of the condition must be understood and the beneficial effect of treatment upon the natural history must have been established; and 4) the cost of
9] have confirmed this. Follow-up of bacteriuric
case finding and treatment should be economically
schoolgirls over a period of seven years revealed that the prevalence of bacteriuria among schoolgirls rises steadily with age with an annual acquisition rate of 0.32% [101. This is an important observation since it establishes that if bacteriuria screening among schoolchildren is of preventive value, it must be carried out repeatedly rather than be confined to school entrants. This would add considerably to its cost and staff requirements.
balanced against the expenditure on medical care as a whole. Screening procedures
Screening for urinary tract infection (UTI) became possible as a result of the pioneering work of Kass [2]. When bacteriuria is due to infection, large numbers of bacteria are found in the urine, usually in excess of
The prevalence of bacteriuria in adults has been studied among defined populations in South Wales, Jamaica and Japan [11, 12]. About 4% of females between the ages of 16 and 65 yr show significant
100,000 organisms/mI. This state is referred to as significant bacteriuria (SB). In contrast, contaminating organisms which are added to the urine during or after voiding are found in small numbers provided multi-
bacteriuria as compared with 0.5% of males. Spontaneous remissions and new infections occur at the rate of
plication of organisms in the urine after it has been voided is prevented. Quantitative urine culture, therefore, allows the distinction between contamination and infection without the need for catheterization or suprapubic aspiration of the bladder. Urine collection procedures and methods of transporting the urine, as well as the various bacteriological and chemical methods for detecting SB, have been reviewed extensively [3, 4] and will not be reiterated. Suffice it to say that the best available semiquantitative method of screening for SB
about 1% of the total female population per annum; consequently, at any given time the prevalence remains
at about 4%. The incidence of bacteriuria, therefore, far exceeds the prevalence. The prevalence of bacteriuria rises with age and parity but age is the more important factor; it seems to be less dependent on race and has not, so far, been shown to have a particular geographical distribution [13]. In pregnant women, bacteriuria does not appear to be more common than in the nonpregnant population but bacteriuria in preg-
is the "dip-slide" technique [5]. This technique is acceptable to public health authorities, bacteriologists and the population alike.
nancy rarely remits spontaneously, whereas in the nonpregnant adult, as well as in schoolgirls, spontaneous cure is a frequent occurrence [14, 15].
Prevalence of SB
In neonates and children below school age, there are no reliable data on the prevalence of SB because of the difficulties in obtaining uncontaminated urine samples.
Clinical aspects of SB
SB is an abnormal laboratory finding. Evaluation of
its clinical significance requires description of the clinical, laboratory and radiological findings in bacteriuric and matched control populations, together with a study of their subsequent fate. Bacteriuric
Received for publication September 6, 1974; and in revised form October 21, 1974. © 1975, by the International Society of Nephrology.
63
64
Asscher
populations differ in respect to their past medical history and blood pressure from non-infected control populations. Documentation of the clinical histories obtained from bacteriuric subjects has been somewhat
neglected, since field studies of the prevalence of bacteriuria do not lend themselves to good historytaking. Kunin, Zacha and Paquin [7] noted that a previous history of UTI can be obtained from 27.9% of bacteriuric schoolchildren as compared with 9.7%
of noninfected controls. Among adult bacteriuric women, 91% give a past history of UT! as compared with 62.5% of the matched controls [16]. Tn 70% of the
bacteriuric subjects, symptoms occur during the year immediately preceding the detection of bacteriuria, whereas such a recent history of urinary infection is obtained from only 17% of controls [16]. These observations cast doubt on the use of the term "asymptomatic bacteriuria" and some authors favor the term
"covert bacteriuria" [8]. More importantly, these findings show that in adults, and even among school entrants, screening for significant bacteriuria often fails to detect urinary infection at an early stage. Measurement of blood pressure in adult bacteriuric and matched control populations reveals that bacteriuric women have a somewhat higher blood pressure than do the controls [11, 12, 16, 17]. In pregnancy, an association between toxemia and bacteriuria has been demonstrated by some workers [18—20] but denied by
Our own unpublished data on kidney function in children with SB show that blood urea and serum creatinine concentrations are also within the normal range. The radiological findings in schoolgirls and adults with bacteriuria have been well documented. In a series of 89 white schoolgirls, Kunin and Paquin [29] found abnormalities on excretory urograms in 22°/s. Micturating cystograms revealed abnormalities in 43%, including vesico-ureteric reflux in 23%. Recent studies in the
United Kingdom have confirmed these findings [8] (McLachlan MSF, et al, unpublished data). KincaidSmith and Bullen [20] performed excretory urograms six weeks after delivery in 148 women found to have bacteriuria during pregnancy. Abnormalities were demonstrated in 51%, and acquired abnormalities (scars, stones, caliectasis, papillary necrosis and differences in kidney size) were found in 45% of this series. Similar studies carried out by Whalley, Martin and Peters [30], Gower eta! [31], and Leigh, GrUneberg and Brumfitt
[32] have yielded a somewhat lower prevalence of acquired renal disease. Although none of the foregoing studies was controlled, it seems likely that significant bacteriuria delineates a population with a high prevalence of structural abnormalities of the urinary tract. This has been confirmed in a controlled study in which the prevalence of acquired renal abnormalities
demonstrated by excretory urography was l8/ in
others [21—231. The mechanism of the associatioii
adult nonpregnant bacteriuric women as compared
between hypertension and urinary infection is obscure. Shapiro [24] suggested that hypertension sensitizes the
with 6% among matched noninfected controls [16].
kidney to infection. Eradication of bacteriuria does not lower blood pressure. This would suggest that the renal damage so frequently seen in bacteriuric subjects (see following), rather than the bacteriuria itself, is responsible for the elevation of blood pressure. It has been recognized for some years that patients with acute pyelonephritis fail to concentrate their urine maximally [25] and that subjects with so-called chronic pyelonephritis may show a disproportionate decrease in concentrating power as compared with glomerular filtration rate [26]. Kaitz [27] demonstrated that in 9 of 20 pregnant women with asymptomatic bacteriuria,
the urine could not be concentrated to more than 700 mOsmjkg. Norden and Tuttle [28] showed that this defect of renal concentrating power could be reversed almost completely by eradicating the bacteriuria. Blood urea concentrations are marginally higher in bacteriuric women than among matched controls [16,
Natural history of SB Assessment of the value of screening for SB must be
based on a thorough understanding of its natural history and its response to treatment. At present our knowledge is incomplete. Existing studies have made the unjustifiable assumption that SB defines an homogenous population. Moreover, the majority of studies of the natural history of SB have been concerned with the fate of pregnant women with bacteriuria who represent but a small part of the whole problem and, lastly, those studies in which the fate of subjects with untreated bacteriuria have been followed for prolonged periods have concerned relatively small populations and their
follow-up has been of short duration. Despite these difficulties, certain important facts have emerged. In pregnant women it has been shown that SB predisposes to the development of acute pyelonephritis [33]. Some 20% of pregnant bacteriuric women develop this com-
20]. None of the bacteriuric women were found to have blood urea concentrations in excess of 55 mg/lOO ml.
plication and the risk can be largely avoided if the bacteriuria is eradicated or suppressed by suitable
This seems surprising if bacteriuria in the adult is, in fact, an important cause of progressive kidney failure.
treatment [33]. The prevention of maternal acute pyelonephritis is one of the strongest arguments in favor
Urinary tract infrction
of bacteriuria screening in pregnancy. Whereas the de-
velopment of acute pyelonephritis in pregnant bacteriuric women is beyond question, claims [33] that pregnant bacteriuric women run a higher than expected risk of pre-eclamptic toxemia, premature birth and fetal loss are more controversial. Even less certain is the suggestion that eradication of the bacteriuria can reduce all these risks. The natural history of significant bacteriuria in nonpregnant women differs in two important respects from
that of pregnant women. First, there is a high spon-
65
There is now evidence to suggest that the urinary pathogens in children with covert bacteriuria differ from those in children with symptomatic infection. In children with covert bacteriuria, the frequency of the common urinary 0-serotypes of E. coli (0k, 02, 04, 06, 07, 0 and 075) is similar to that in the stools of healthy children, whereas in children with symptomatic infection these common strains are found twice as often in the urine as in the stools of the healthy children drawn from the same population [38]. Also, urinary strains of E. coli from children with covert bacteriuria are more
taneous cure rate. Nearly 4Ø0/ of untreated bacteriuric
sensitive to the cidal effect of serum than those ob-
women lose their bacteriuria over a period of one year [14]. The second difference is that in the non-
tained from children with symptomatic infection [38]. These findings may indicate differences in host-parasite relationships in covert as compared with overt infections and such differences may well influence the longterm natural history of these infections. Studies of the natural history of covert infections of the urinary tract
pregnant woman acute pyelonephritis develops much
less frequently than in the pregnant woman with bacteriuria. In a one-year follow-up of 92 nonpregnant bacteriuric women, only 2 developed loin pain with fever, whereas symptoms of lower UTI (frequency and dysuria) developed in 30 subjects. What is more, the
development of symptomatic infection was not prevented by a short course of treatment. In fact, such treatment encouraged the development of symptoms, since the re-infections (i.e., infections with different organisms from those originally isolated) which followed initially successful treatment were more commonly associated with the development of symptoms than the persistent infections among untreated bacteriuric controls [14]. Similar experiences have been reported following treatment of schoolgirls with SB [34]. These findings would suggest that tolerance to the
infecting strain of Escherichia coli exists in subjects with covert bacteriuria and that treatment may disturb this symbiosis. The long-term effects of SB in women are still largely unknown. Freedman [36] followed up 250 women with UTI for up to 12 yr. One-half these patients had been
followed for five or more years. No evidence of deterioration of kidney function or elevation of blood pressure has so far been observed. Our own experience [36], which concerns 107 women with bacteriuria and 88 matched controls, extends over three to five years. It has yielded no evidence to suggest that untreated SB in the adult produces progressive kidney damage, pro-
vided the bacteriuria is not associated with raised blood pressure or obstructive uropathy.
The sequelae of bacteriuria in childhood may be more serious. Smellie and Normand [37] showed that impairment of kidney growth and kidney scarring occur in those children in whom symptomatic infection is associated with vesico-ureteric reflux. This finding is
not necessarily applicable to children with covert bacteriuria even though infection in them is accompanied by vesico-ureteric reflux in 23% of cases [29].
in childhood are now in progress in a number of centers in the United Kingdom. Although it is too early to draw conclusions from them, two significant facts have emerged. First, in 36% of bacteriuric schoolgirls aged 4 to 11 yr, bacteriuria clears spontaneously
during the first year of follow-up [15]. Second, persistence of bacteriuria is more common in children with vesico-ureteric reflux [39]. It has yet to be settled whether children with such persistent bacteriuria and
vesico-ureteric reflux show a reduction of kidney growth or develop kidney scarring since repeat radiographs in our untreated children with bacteriuria have not been performed as yet. It is important that studies
such as ours are allowed to proceed. Unfounded claims to the effect that treatment of SB in children is beneficial could force health authorities into the introduction of screening programs before their value is
proven. In our study in the cities of Cardiff and Oxford, no benefit from treating SB in schoolgirls has so far been observed.
Studies of the natural history of untreated UTI should also attempt to identify high risk groups, such as girls with infection associated 'with vesico-ureteric reflux or girls in whom infection involves the renal parenchyma. Renal parenchymal infection can now be detected by non-invasive means. Thomas, Shelokov and Forland [40] have shown that organisms which derive from the kidney are coated with antibody whereas those which are confined to the bladder are not. Anti-
body coating of urinary pathogens can readily be identified with the use of fluorescent-labelled anti-IgG. Jones, Smith and Sanford [41] confirmed the value of this procedure, and it will now be possible to separate the natural history of covert bacteriuria with renal involvement from that of subjects in whom the infection is confined to the bladder.
66
Asscher
The role of vesico-ureteric reflux in the pathogenesis
UTI. Mortality statistics reveal an increase in deaths
of kidney scarring has recently been clarified. In a group of 175 infants presenting with UTI, Rolleston
from infections of the kidney which may not be entirely
[42] showed that gross reflux may lead to progressive kidney damage even in the absence of continuing infection. Hodson [43] showed that in young pigs with vesico-ureteric reflux produced by ureterotomy and urethral narrowing, focal kidney scars developed in the absence of infection; these scars were situated in areas of the kidney in which intrarenal refiux due to pyelo-
tubular back flow had previously been shown on micturating cystography. In a retrospective study, Rolleston, Maling and Hodson [44] have recently shown that kidney scars in children occur particularly in areas of the kidney in which intrarenal reflux is demonstrable. It was also shown that such intrarenal reflux was confined to children below the age of 4 yr. Morgan and Asscher (unpublished data) showed that in the rat, vesico-ureteric reflux, produced by squeezing the urinary bladder (filled with normal saline) while the external urethral orifice is occluded, regularly produces slightly depressed and/or discolored areas in the perihilar region of the kidneys. Kidneys bearing such
lesions were found to be sterile on culture and the
lesions persisted throughout a two-month observation period. These lesions, though similar in location and microscopic appearance to scars produced by refluxing
E. coil from the bladder to the kidney [45], were
due to a change in diagnostic habit and acuity [46]. The diagnosis of UTI in a latent phase of the disease is now possible, by methods acceptable to the population, at reasonable cost. Brumfitt and Reeves [47] esti-
mated that the cost of screening for bacteriuria amounted to less than 2Sp a subject. Facilities for screening would be relatively simple to provide in ante-
natal clinics, but if bacteriuria screening were to be extended to schoolgirls and nonpregnant women, the provision of facilities and staff would involve considerable expenditure. The value of screening for significant bacteriuria in pregnancy is well established, since treatment of the bacteriuria largely prevents maternal pyelonephritis. This conclusion does not apply to the vast majority of bacteriuric women, i.e., those who are not pregnant. Treatment of bacteriuria by means suitable for largescale use confers no lasting benefit on them. Screening of nonpregnant women could be justified only if longterm prospective studies show that bacteriuria produces progressive kidney damage leading to kidney failure which can be prevented by treatment. So far, the results of studies do not show such a trend. Finally, it is important that the controlled treatment trials of covert bacteriuria in childhood which are now in progress are completed and that medical opinion on
smaller and much less-depressed than the scars produced by reflux and infection. It was concluded that reflux alone produces a "locum resistantiae minor" where organisms settle and produce the markedly depressed infected scars. The method whereby intrarenal reflux produces kidney damage is obscure. Leakage of
the value of screening for UTI in schoolgirls is not
urine into the renal interstitium and/or release of
Reprint requests to Dr. A. W. Asscher, K.R.U.F. Institute of Renal Disease, Welsh National School of Medicine, Royal In-
tubular lysosomal enzymes are just two of the possibilities which are being examined in the experimental
allowed to crystallize before the results of these studies are to hand. A. W. Asscher Cardiff, United Kingdom
firmary, Cardiff, United Kingdom.
model.
References
It follows from these observations that vesicoureteric reflux is an important determinant of the kidney damage associated with infection. At present, the best guide to the detection of vesico-ureteric reflux is the discovery of SB because persistent bacteriuria and vesico-ureteric reflux are associated (see foregoing). It would be preferable, if a noninvasive method of screening healthy populations for vesico-ureteric reflux were available since screening for SB will miss a proportion of subjects with vesico-ureteric reflux. Rationale of screening for bacteriuria
There is no doubt that UTI is an important public health problem. In general practice, 12 per 1000 consultations are the result of symptoms which suggest
G: Principles and practice of screening for disease. W.H.O. Geneva, 1968
1. WILSON JMG, JUNGER
2. KASS EH: Asymptomatic infections of the urinary tract. Trans Assoc Am Physicians 69:56—64, 1956 3. NORDEN CW,
KASS EH: Bacteriuria of pregnancy: A
critical appraisal. Ann Rev Med 19:431—470, 1968 4. ASSCHER AW: The early diagnosis of urinary tract infection:
Early diagnosis. Paper No. 8. London, Office of Health Economics, 1970 5. GUTTMANN DE, NAYLOR GRE: Dip-slide: An aid to quantitative urine culture in general practice. Br MedJ 3:343—345, 1967 6. KUNIN CM, SOUTFIALL I, PAQUIN AJ: Epidemiology of
urinary tract infections: A pilot study of 3057 schoolchildren. N Engl J Med 263:817—823, 1960 7. KUNIN CM, ZACHA E, PAQUIN AJ: Urinary tract infection in
schoolchildren: Prevalence of bacteriuria and associated urological findings. N Engi J Med 266:1287-1296, 1962
67
Urinary tract inJection 8. SAVAGE DCL, WILSON MI: Covert bacteriuria in childhood, in Urinary Tract Infection, edited by BRUMFITT W, ASSCHER AW, Oxford, Oxford University Press, 1973, pp. 39—47
uria, in Progress
in Pyelonephritis, edited by KASS EH, Philadelphia, F. A. Davis, 1965, pp. 73—80 29. KUNIN CH, PAQUIN AJ: Frequency and natural history of
urinary tract infection in schoolchildren, in Progress in
9. ASSCHER AW, MCLACHLAN MSF, JONES RV, MELLER S, SUSSMAN M, HARRISON S. JOHNSTON HH, SLEIGHT
Pyelonephritis, edited by KAss EH, Philadelphia, F. A. Davis,
G, FLETCHER EW: Screening for asymptomatic urinary
30. WHALLEY PJ, MARTIN FG, PETERS PC: Significance of asymptomatic bacteriuria detected during pregnancy. JAMA
tract infection in schoolgirls. Lancet 2:1—4, 1973
10. KUNIN CM: Emergence of bacteriuria, proteinuria and symptomatic urinary tract infections amongst a population of schoolgirls followed for 7 years. Paediatrics 41 968—976, 1968 11. KASS EH, SAVAGE W, SANTAMARINA BAG: The significance
of bacteriuria in preventive medicine, in Progress in Pyelonephritis, edited by KAss El-I, Philadelphia, F. A. Davis, 1965, pp. 3—10 12. FREEDMAN LR, PHAIR JP, SOKI M, HAMILTON HB, NEFZGER
MD, HIRATA M: The epidemiology of urinary tract infections in Hiroshima. Yale J Biol Med 37:262—282, 1965
13. KASS EH: Geographic pathology of bacteriuria, in The Kidney. International Academy of Pathology monograph No. 6, Baltimore, Williams & Wilkins, 1966, pp. 469—475 14. ASSCHER AW, SUSSMAN M, WATERS WE, EVANS JAS, CAMPBELL H, EVANS KT, WILLIAMS JE: Asymptomatic
significant bacteriuria in the nonpregnant woman: II. Response to treatment and follow-up. Br Med J 1:804—806, 1969
1965, pp. 33—44
193:879—881, 1965 31. GOWER PE, HASWELL B, SIDAWAY ME, DE WARDENER HE:
Follow-up of 164 patients with bacteriuria of pregnancy. Lancet 1:990—994, 1968 32. LEIGH DA, GRONEBERG RN, BRUMFITT W: Long-term
follow-up of bacteriuria in pregnancy. Lancet
1:603—605,
1968
33. Kss EH: Bacteriuria and pyelonephritis of pregnancy. Arch Intern Med 105:194, 1960 34. LINDBERG U, CLAESSON 1, HANSON LA, JODAL U: Asym-
ptomatic bacteriuria in schoolchildren: A clinical and laboratory study. Acta Paediat, in press 35. FREEDMAN LR: Long-term follow-up of urinary infections,
in Abstracts
Vth International
Congress of Nephrology.
Mexico City, 1972, p. 25 36. ASSCHER AW, CHICK S, RADFORD N, WATERS WE, SLJSSMAN
M, EVANS JAS, MCLACHLAN MSF, WILLIAMS JE: Natural
history of asymptomatic bacteriuria (A.S.B.) in nonpregnant women, in Urinary Tract Infection, edited by
15. MELLER 5, JONES RV, MCLACIILAN M, HOLED, SUSSMAN M,
BRUMFITT W, ASSCHER AW, Oxford, Oxford University
ASSCHER AW: The significance of bacteriuria in Cardiff schoolgirls (abstract). Kidney mt 6:61, 1974
37. SMELLIE JM, NORMAND ISS: Experience of follow-up of
16. SUSSMAN M, ASSCHER AW, WATERS WE, EVANS JAS, CAMPBELL H, EVANS KT, WILLIAMS JE: Asymptomatic
significant bacteriuria in the nonpregnant woman: I. Description of population. Br MedJ 1:799—803, 1969
Press, 1973, pp. 55—61
children with urinary tract infection, in Urinary
Tract In-
by O'GRADY F, BRUMFITT W, Oxford, Oxford University Press, pp. 123—135 38. OLLING 5, HANSON LA, HOLMGREN J, JODAL U, LINCOLN K, fection, edited
17. MIALL WH, KASS EH, LING J, STUART KL: Factors influenc-
LINDBERG U: The bactericidal effect of normal human
ing arterial pressure in the general population in Jamaica.
serum on E. coli strains from normal and from patients with urinary tract infections. Infection 1:24—28, 1973
Br MedJ 2:497—506, 1962 18. NORDEN CW, KILPATRICK WH: Bacteriuria of pregnancy,
in Progress in Pyelonephritis, edited by KASS EH, Philadelphia, F. A. Davis, 1965, pp. 64-72 19. STUART KL, CUMMINS GTM, CHIN WA: Bacteriuria, pre-
maturity and the hypertensive disorders of pregnancy. Br MedJ 1:554—556, 1965 20. KINCAID-SMITH P, BULLEN M: Bacteriuria in pregnancy. Lancet 1:395—399, 1965
21. LJTTLE P: The incidence of urinary tract infection in 5,000 pregnant women. Lancet 2:925—928, 1966 22. BRYANT RE, WJNDOM RE, VINEYARD JP JR, SANFORD JP:
Asymptomatic bacteriuria in pregnancy and its association with prematurity. JLab C/in Med 63:224—231, 1964 23. MONZON OT, ARMSTRONG D, PION RJ, DEIGH R, HEWITT
WL: Bacteriuria during pregnancy. Am J Obstet Gyncol 85: 511—518, 1963
24. SHA?IR0AP: Experimental pyelonephritis and hypertension:
Implications for the clinical problem. Ann Intern Med 59: 37—52, 1963
25. WINBERG J: Renal concentrating capacity during acute non-
obstructive urinary tract infections in infancy and early childhood. Acta Paediatr 47:635—645, 1958 26. BROD J: Chronic pyelonephritis. Lancet 2:973—981, 1956
27. KAITZ AL: Urinary concentrating ability in pregnant women with asymptomatic bacteriuria. J C/in Invest 40: 1131—1338, 1961
28. NORDEN CW, TUTTLE EP: Impairment of urinary concentrat-
ing ability in pregnant women with asymptomatic bacteri-
39. JONES RV, MELLER S, MCLACHLAN MSF, SLJSSMAN M, ASSCHER AW, WAYON-WHITE R, LEDINGHAM JGG, SMITH
JC, FLETCHER EWL, JOHNSTON HH: Treatment of covert bacteriuria in schoolgirls (abstract). Kidney mt 6:61, 1974 40. 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—590, 1974
41. JONES SR, SMITH JW, SANFORD JP: Localization of urinary
tract infections by detection of antibody-coated bacteria in urine sediment. NEng1J Med 290:591—593, 1974 42. ROLLESTON GL: The significance and management of Vesico-ureteric reflux in infancy: II. Radiological aspects, in Renal Infrction and Renal Scarring, edited by KINCAIDSMITH P, FAIRLEY KF, Melbourne, Mercedes Publishing Services, 1970, pp. 246—256
43. Hoeso JC: Vesico-ureteric reflux and renal scarring with and without infection (abstract). Kidney mt 5:308, 1974 44. ROLLESTON GL, MALING TMJ, HODSON CJ: Intrarenal reflux and the scarred kidney. Arch Dis Child 49:531—539,1974
45. ASSCHER AW, CHICK 5: Increased susceptibility of the kidney to ascending Escherichia co/i infection following unilateral nephrectomy. Br J Uro/44:202—206, 1972 46. KESSNER DM, FLOREY C DU V: Mortality trends for acute and chronic nephritis and infections of the kidney. Lancet 2: 979—982, 1967
47. BRUMFITT W, REEVES DS: Screening procedures for urinary infection, in Presymptomatic and Early Diagnosis, edited by
SHARP CLEH, KEEN H, London, Pitman Medical, 1968, pp. 179—201