Postgrad Med J: first published as 10.1136/pgmj.41.478.485 on 1 August 1965. Downloaded from POSTGRAD. MED. J. (1965), 41, 485 Clinical Review PYELONEPHRITIS IN CHILDREN An Interim Review of Recent Literature MALCOLM MACGREGOR, M.D., F.R.C.P. From the South Warwickshire Hospital Group PYELONEPHRITIS is now a fast-changing subject, But in infancy the diagnosis is often missed: with which the general reader may keep abreast in one autopsy series it had been missed clinic- only if frequent attempts are made to bring ally in 8.3% (Pryle and Neumann, 1962). T-he together published work from different sources. initial urinary infection often occurs in the This survey aims to provide a balanced account newborn period; in one series 0.3%/, of hospital of recent concepts, but is in no sense exhaustive. births (Smellie and others, 1964) and in another 1.5% (James, 1959) were considered to be in- Incidence of the Disease fected. In fact, the incidence among the new- Among childhood infections those of the born may be higher than in other age groups; urinary tract are second in frequency only to congenital defects in the kidney may pre- respiratory infections, and are the commonest dispose (Porter and Giles, 1956). Postmortem bacterial infections under two years of age studies suggest that the prevalence of urinary in- Protected by copyright. (Pryles, 1960; Deluca, Fisher and Swenson, fection is still underestimated (Kleeman, Hewitt 1963). The incidence of overt urinary infections and Gaze, 1960); about 2%/, of routine autopsies in the general population is estimated at 8 per on American children disclose evidence of 1,000 per annum (Percival, Brumfitt and pyelonephritis (Pryles and Neumann, 1962; Louvois, 1964), and in American schoolgirls Spark, Travis, Dodge, Dalschmer and Hopps, at 1.4% of the school population per annum 1962; Macaulay, 1964), but the difficulties in (Kunin, Deutscher and Paquin, 1964). Between diagnosis of this infection from postmortem I and 4%/, of hospital admissions of children material must be borne in mind. An increased are for this disease (Stansfeld, 1954); 3 to 4%/ incidence of pyelonephritis in siblings has been of recent admissions to an acute p,ediatric ward noted (Kunin, Deutscher and Paquin, 1964). in London were for urinary infections (Burke, 1961). On the other hand, in the Newcastle The Course and Prognosis of Childhood survey of 1,000 families, only 3 out of 847 Urinary Infections children were noted to have pyelonephritis in A tendency for the infection to relapse or http://pmj.bmj.com/ five years; all relapsed. The incidence of "signi- to become persistent is an alarming feature of ficant bacteriuria", which is often symptomless, oyelonephritis. Some 80-95%/, of treated primary (for definition see below) is higher, especially in infections are "cured" (when defined as free girls; 1% of schoolgirls in a recent large from bacteriuria six weeks after cessation of American survey were found to have this treatment), but about half of these have further (Kunin, Southall and Paquin, 1962; Kunin and periods of infection (Macaulay and Sutton, seem more others, 1964). Forty per cent of young women 1957; Lancet, 1963). Recurrences on September 26, 2021 by guest. with symptomless bacteriuria in early pregnancy common in older children than in infants later developed pyelitis of pregnancy (Kincaid- (Burke, 1961) and especially so in girls of the Smith, 1964; Percival, Brumfitt and Louvois, six to ten age group (Williams and Sturdy, 1964). The highest incidence of bacteriuria 1961). Persistence of infection is common, was found in girls aged between 15 and 19 mainly in girls (Dunn, Hine and MacGregor, (Kunin, Deutscher and Paquin, 1964). Most 1964; Woodruff and Everett. 1954), and one clinicians find that the onset of infection is third to one half of continuing infections are most frequent under one year old (Stansfeld, asymptomatic (Steele, Leadbetter and Crawford, 1954: Smellie, Hodson, Edwards and Normand, 1963: Dunn and others, 1964). Admittedly 1964), though some assert that the peak in- the significance of "significant bacteriuria" as cidence is among girls aged three to five the only finding is still debatable, but there (Deluca, Fisher and Swenson; Gross, Randolph seems to be little tendency for this to die out and Wise, 1963). spontaneously (Kass, 1956; Kleeman and others, 486 POSTGRADUATE MEDICAL JOURNAL August, 1965 Postgrad Med J: first published as 10.1136/pgmj.41.478.485 on 1 August 1965. Downloaded from 1960; Spence, Murphy, McGovern, Herdion years still show a 7% relapse rate afterwards and Pryles, 1964). Persistent infection has been (Campanacci, Bonomini and Zuchelli, 1963). found with different frequency in different sur- In chronic or recurrent infections in adults veys, in 10% (Burke, 1961), in 25% (Dunn there is only a 10% (Lancet, 1963) or 20% and others, 1964), in 39% (Woodruff and (Turck and others, 1962) cure rate, in spite of Everett, 1954). Of the chronic and recurrent really prolonged treatment (up to two years) groups the outcome varies, but the prognosis (Williams, 1963). In children, after conventional is believed to be worst when the onset is under treatment of acute infections there may be a two years of age. Some clear up, spontaneously 250/% persistence rate (Dunn and others, 1964). (Kunin, Deutsches and Paquin, 1964) or with There is not much evidence that prolonged treatment (Turck, Browder, Lindmeyer, Brown, drug treatment will produce better figures Anderson and Petersdorf, 1962; Macaulay, (Lancet, 1963), unless perhaps in infants (Stans- 1964) or at adolescence (Turner-Warwick, 1962; feld and Webb, 1954), though assertions are Williams, 1964). Others progress to renal failure made to the contrary (Smallpeice, 1958). In (Rosenheim, 1963). The follow-up in adults adults, non-antibiotic urinary disinfectants such after 20 years can be appalling. In a survey as methenamine mandelate (Holland and West, relating to the pre-antibiotic era, 43%/0 of young 1963) are capable of giving as good results as women who had been treated for acute pye- the powerful modem antibiotics tetracycline, lonephritis had serious urological disease after chloramphenicol and kanamycin (Turck and 16 years (Hanley, 1964). There is not yet much others, 1962), especially in older age groups information about the long-term prognosis for (Lindmeyer, Turck and Petersdorf, 1962). Be- children. The incidence of toxemia of pregnancy cause so many infections are clinically silent. essential and pyelitis of pregnancy is known to be un- prolonged follow-up of treated cases is Protected by copyright. usually high when there is a history of urinary (Lancet, 1963). Such follow-up studies in infection in childhood (Steele and others, 1963), children are in progress in a number of centres although a history of childhood infection was and should soon provide facts about the true found in only five out of 75 pregnant women prognosis in childhood. Early treatment of with pyelitis in one enquiry (Woodruff and primary infections is important (Deluca and Everett, 1954). There seems to be little correla- others, 1963), for delay in diagnosis increases tion between the number of recurrences and the the relapse rate (Stansfeld and Webb, 1954). effect after some years on the kidneys (Steele Once pyelonephritic scarring of a kidney has and others, 1963). The equal sex incidence occurred, infection is very difficult to eradicate of chronic pyelonephritis at autopsy in adults (British Medical Journal, 1964). Delay in diag- does not seem consistent with the much higher nosis is common, and averaged 18 months in incidence of symptomatic infections in girls, one recent survey (Deluca and others, 1963). suggesting that the one is not the direct ante- In another (Spence and others, 1964), 47% cedent of the other (Macaulay, 1964). However, of children sent to hospital had had symptoms http://pmj.bmj.com/ some girls with recurrent infections have been for one year or more, and in a third series, observed to develop progressive loss of kidney only 30%/O of children had been sent to hospital parenchyma with a fatal issue (Williams, 1965). with a suggestion of the correct diagnosis Although pvelonephritis is the second common- (Smellie and others, 1964). est cause of hypertension in adults, in two-thirds of such cases no history of recurrent urinary Clinical Features could be obtained (Kincaid-Smith. of vomiting infection In infancy, anorexia, loss weight, on September 26, 2021 by guest. McMichael and Murphy, 1958). Indeed, it is and failure to thrive are the main symptoms; said that a history of acute pyelonephritis is in older children, fever or abdominal pain rare in advanced chronic nvelonephritis (Kim- (Burke, 1961: Smellie and others, 1964). Only melstiel, Kim, Beres and Wellman. 1961). On 25%/, have micturition symptoms (Burke, 1961). the other hand, in a recent series of 200 children Hematuria occurs in 10% of acute attacks seen for urinary infection at a hospital, 13%/, (Kleeman and others, 1960; Burke, 1961). Re- had X-ray evidence of chronic pyelonephritis current febrile urinary infections are known to (Smellie and others, 1964). occur in older girls (Williams and Sturdy, 1961). Whatever its ultimate results, chronic urinary and the symptoms tend to be the same each infection is difficult to eradicate. With chemo- time (Burke, 1961). In older children chronic therapy the prognosis is still depressing (Mac- infection may be quite silent, but close question- aulay and Sutton, 1957). Uncomplicated primary ing will often uncover some disregarded urinary infections treated continuously for up to two symptom, such as enuresis, frequency or August, 1965 MAcGREGOR: Pyelonephritis in Children 487 Postgrad Med J: first published as 10.1136/pgmj.41.478.485 on 1 August 1965. Downloaded from dysuria (Dunn and others, 1964). Ureteric re- berg, 1964). The plastic bags, much used for flux occasionally gives rise to pain in the loin obtaining specimens from young children, are on micturition (Forsythe and Wallace, 1958). liable to introduce contamination into bacterial Episodes of partial or complete urinary reten- counts, but have less influence on cell counts tion may be described (Gross and others, 1963). (Houston, 1964). Impor,tant clues may be revealed if questions Proteinuria is of little aid in detecting urinary about the rate and force of the urine stream, infection (Burke, 1961; Lancet, 1962a and b; and whether straining is present, are a part of Steele and others, 1963; Dunn and others, 1964; the routine enquiry (New England J. Med., Smellie and others, 1964); albumin was found 1963; Gross and others, 1963). However, ana- as frequently in the urines of unaffected school- lysis of symptoms does not allow a judgement girls, as in those with significant bacteriuria to be made of which cases have urological (Kunin and others, 1964). The lack of a simple abnormalities, and which have not. (Forsythe but reliable screening test for infection is still and Wallace, 1958; Kunin and others, 1964). lamented (Lancet, 1962). Controversy has Neither is the severity of symptoms proportional focussed upon the relative accuracy of methods to the extent of existing, or of future, renal which estimate bacteria, and of those which damage (Steele and others, 1963). The growth count cells, in the urine. The correlations bet- of patients with X-ray changes of pyelonephritis ween urinary cell concentration, timed cell ex- is sometimes a little reduced. There is occasion- cretion rates, and bacterial colony counts are ally hypertension, or another congenital mal- not close, and misleading results can come from formation (Smellie and others, 1964). Polyuria relying upon any one of them (Osborn and and nocturia are early signs of serious renal Smith, 1963). In America colony counts are impairment (Kleeman and others, 1960). preferred (Brumfitt and Percival, 1964; Spence and others, 1964). For children in hospital pus Protected by copyright. Diagnosis of Urinary Infection cell counts have been found as useful as The danger of catheterisation as a means of bacterial counts in identifying infections; with introducing infection is increasingly recognized, either method there were 13-25% doubtful especially in obstetric practice Davies results; when the methods were used in con- and Rosser, 1961). The risk(Brumfit't,of infection in junction uncertainty was reduced to 6% children is probably less (tho-ugh it was 70/o (Houston, 1964). in one series (Vertanen, Oksanen and Pettonen, Direct examination of freshly voided urine 1962); catheterisation should not be withheld may show organisms under the microscope; if needed for urgent diagnosis, retention, non- bacteriuria of this magnitude is never due to cooperation or uncertainty (Pryles, 1960). But contamination, so this is a valuable screening in children the bacteriological correlation bet- technique (Gardborg, 1959; Dunn and others, ween catheter and clean-voided specimens can 1964). Again, if the uncentrifuged sediment of

be 96% (Pryles, 1960) and at all ages mid- a urine shows organisms after Gram staining, http://pmj.bmj.com/ stream specimens without bacterial contamina- the bacterial count is nearly always above tion can be obtained from a high proportion 100,000/ml. (Pryles, 1960; Kunin and others, of both sexes (Clarke, 1960; Pryles, 1960; 1964). Laboratory techniques for demonstrating Vertanen and others, 1962; Cattell and Lefford, bacteriuria employ quantitative, semiquantita- 1963). The most precise results are obtained tive, or chemical screening methods. Urine is from suprapubic aspiration of the bladder an excellent culture medium, so within two (Monzon, Ory, Dobson, Carter and Yow, 1958); hours at room temperature a large increase in this method gives only a 93%/0 bacterial cor- bacterial numbers may occur (Pryles, 1960). on September 26, 2021 by guest. relation with catheter urine, but is higher if the This can be prevented by refrigeration of fresh first few millilitres of catheter urine are dis- specimens at 40C, which preserves the bacterial carded,(Pryles, 1960). Many different,techniques content unaltered for 2-7 days (Pryles, 1960; of perineal cleansing are employed before taking Houston, 1964). Bacterial colony counts on clean specimens (Brumfitt and others, 1961; large numbers of fresh urines show a bimodal Cattell and Lefford, 1963; Houston, 1964); distribution, with peaks at less than 10,000 and benzylkonium chloride 1 in 1000 is much used at more than 100,000 organisms/ml. There is (Little, 1962), polybactrin spray has advocates much evidence that organisms derived from (McLeod, Mason and Pilley, 1963), but soap- the urethra or vulva never exceed 104 per ml., and-water seems satisfactory (Clarke, 1960). whereas bacteria rapidly multiply in the bladder Different methods of cleansing do not seem to to reach numbers in excess of 105 per ml., influence urine cell counts (Lincoln and Win- the level of "significant bacteriuria" (Kass, 488 POSTGRADUATE MEDICAL JOURNAL August, 1965 Postgrad Med J: first published as 10.1136/pgmj.41.478.485 on 1 August 1965. Downloaded from 1956, 1960). Significant levels may have to be and others, 1963). But this standard is no more slightly higher in the case of young children acceptable than are the still much employed (Houston, 1964). Quantitative bacterial counts descriptive reports ("moderate numbers" or on urine are made by the surface viable, or "scanty pus cells") which for CSF, with a similar pour plate techniques (Brumfitt and Percival, range of cell numbers, are never countenanced 1964). These methods are too time-consuming (Stansfeld, 1962; Little, 1964). Of course, for most routine laboratories, so semi- quantita- chronic pyelonephritis with a cell-free urine is tive modifications are often used. Using fresh well recognized (Clarke, 1960; Deluca and uncentrifuged urine and a standard loop, colony others, 1963; Leather, Wills and Gault, 1963), counts of more than 200 equate witih more and both pyuria and bacteriuria may be inter- than 105 organisms per ml., and counts of less mittent (Stansfeld, 1954; Dunn and others, than 90 with less than 104 organism per ml. 1964); but bacilluria without pyuria, using a This is sufficiently accurate for clinical practice sensitive criterion, occurred only three times in (O'Sullivan, Fitzgerald, Meynell and Malins, 250 specimens from children (Stansfeld, 1962). 1960). Another technique employs a blotting Timed white-cell excretion rates, which do not paper strip dipped into urine and pressed on to seem to be much influenced by the rate of the culture medium; results have been stand- urine flow, are available for adults, but not yet ardised against pour plate counts (Leigh and for children (Houston, 1964). An upper limit Williams, 1964). Conventional methods of des- of excretion of 400,000 cells per hour is not criptive reporting on the bacteriology of urine exceeded in normal men or women (Kass, have been compared with the results of bacterial 1956; Little, 1962; Gadeholt, 1964); this is counting; it is claimed that if standards are probably the best measurement of pyuria

rigorous, closely similar conclusions will be (Gadeholt, 1964). Centrifugation in any cellProtected by copyright. drawn from each method, and that experimental counting technique introduces important errors, results are not necessarily invalidated by failure as it leads to considerable and unpredictable to carry out bacterial counts (Clark, 1960; loss of cells (Gadeholt, 1964; Little, 1964). Cell Guttman and Stokes, 1963). The T.T.C. test counts in uncentrifuged specimens have been is a chemical screening method based upon found in adul,ts to give a good correlation with naked-eye observation of colour change in the timed cell excretion rates; more than 10 cells/ urine after incubation of urine with test material cu. mm. means more than 400,000 excreted per in a water bath (Simmons and Williams, 1962). hour, while less than three cells per cu. mm. It is claimed to give a positive reading with means less than 400,000 per hour with fair bacterial counts of gram negative organisms certainty; there is no sex difference (Little, over 100,000/mI., and has proved successful 1964). Quantitative methods of leucocyte count- in detecting 86% of antenatal women with ing in a counting chamber are increasingly used asymptomatic bacteriuria of pregnancy (Kin- for children (Stansfeld, 1954), and have less caid-Smith, Bullen, Mills, Fussell, Huston, variation than methods based on high power http://pmj.bmj.com/ and Goon, 1964), and even better results are field counts. It is however important to ensure reported. The level of certain enzymes is also a urinary pH of between six and eight (Stansfeld, reported to be raised in infected urines (Bren- 1962). ner and Gilbert, 1963). Semiquantitative wet film screening techniques Pyuria may result from other causes than of cell counting have proved useful in some urinary infection, from vaginitis, trauma, de- hands (Dunn and others, 1964), though criticized as well as at the end by others in respect of variability (McGrackie hydration and nephritis, on September 26, 2021 by guest. of a urinary infection when organisms have and Kennedy, 1963). After infancy, 92% of gone (Pryles, 1960). Until recently the criteria urines from normal children contain less than accepted by many authors as "significant 10 leucocytes/cu. mm. (Stansfeld, 1962). Older pyuria" were not uniform (Lancet, 1962). A girls give more misleading cell counts (Houston, commonly employed criterion has been "more 1964). One centre accepts less than 10/cu. mm. than five cells -per high power field" (represent- as normal, more than 100 as infected, and inter- ing, if centrifuged, approximately 25 cells/ mediate values as dubious (Houston, 1964). A cu. mm., if uncentrifuged, approximately 250 level of 50/cu. mm. or more is regarded by cells/cu. mm. (Stansfeld, 1962, 1963). On this others as indicating infection (Stansfeld, 1954). basis, only 35% of children with significant In the newborn, counts greater than 25/cu. mm. baoteriuria had also significant pyuria (Kunin in boys, or 50/cu. mm. in girls, are abnormal and others, 1962), and, in another series, only (James, 1959; Lincoln and Winberg, 1964). half the infected children had pyuria (Steele Neonates, especially males, may excrete epithe- August, 1965 MAcGREGOR: Pyelonephritis in Children 489 Postgrad Med J: first published as 10.1136/pgmj.41.478.485 on 1 August 1965. Downloaded from lial cells in considerable number and cause child in whom serious obstruction had been confusion (Lincoln and Winberg, 1964); differ- overlooked (Williams and Sturdy, 1961). But ential staining may then make counting easier about 10%/, of affected girls show recurrence or (Sternheimer and Malbin, 1957; Prescott and persistence of infection, and a large proportion Brodie, 1964; Houston, 1964). Taking a level (variously estimated at 25% (Dunn and others, of 10 leucocytes/cu. mm. as "significant 1964), 50% (Spence and others, 1964) or 75% pyuria", some 14% of admissions to a children's (Macaulay, 1964) have structurally normal hospital were found to have it, though counts urinary tracts. It is plain that damage from below 50/cu. mm. seldom turned out to be one attack confers in some way a susceptibility abnormal (Stansfeld, 1954, 1962). In general, to further attacks (Lancet, 1963), and in ex- those who have studied the subject over a perimental pyelonephritis intra-renal scarring long period now accept that the diagnostic permits of infection more readily, as also do value of leucocyte counts is comparable to areas of renal dysplasia (Kleeman and others, that of bacterial counts (Osborn and Smith, 1960). Non-bacterial agencies, such as an under- 1963). Wlhen chronic pyelonephritis is suspected, lying virus infection, or auto-immune reactions, attempts to stimulate leucocyte excretion can or progressive renal vascular damage have been be made, using intravenous bacterial pyrogen considered, without good evidence to favour (Leather and others, 1963), or, probably with them (Brumfitt and Percival, 1964). Bacteria more reliability ('Montgomerie and North, are in fact probably responsible; it has been 1963), with oral steroid drugs (Huth, Chalmers, suggested that -they could persist within the Macdonald and de Wardener, 1961; Little and kidney as spheroplasts or protoplasts (Brum- de Wardener, 1962). There are no reports of fitt, Percival and Williams, 1964), but the the use of these methods for children. There is tendency for one organism to be replaced by need for a test to distinguish reliably between another after treatment argues to the contrary Protected by copyright. infection involving the kidney, and that confined (Gard'borg, 1959; Turck and others, 1962). to the lower urinary tract (Brumfltt and Perci- There may be "pyelopathic" strains of E.Coli val, 1964). The presence of leucocyte casts in (Brumfitt and Percival, 1964; Sweet and Wolin- urine means renal infection, and by using a sky, 1964). Again, anti-bacterial substances are special stain, morphological differences are known to exist in normal urine, and may be apparent between leucocytes of renal origin and reduced in diseased states (Kleeman and others, those from more distal sources (Sternheimer 1960). But of late years increasing attention is and Malibin, 1957; Poirier and fackson, 1957), being given to vesico-ureter reflux as a factor which can be used as a diagnostic aid. When promoting persistence of infection. the serum from a patient with an acute urinary infection is used for direct agglutination against Vesico-ureteric Reflux organisms obtained from the urine, a high Retrograde movement of urine from the and rising titre seems to correlate with involve- bladder up the ureters rarely if ever occurs in http://pmj.bmj.com/ ment of renal tissue. A titre of more than 1: 160 normal children (Hodson and Edwards, 1960; suggests such infection: it falls rapidly when Williams, 1964; Spence and others, 1964), infection is controlled (Percival and others, although it is a normal finding in rats, and can 1964). However, in another large group of be borne in dogs under experimental conditions bacteriuric girls no such relationship was found for long periods without ill-effects except some (Kunin and others, 1964). progressive depression in ureteric peristalsis (Scott, 1962). Reflux occurs in man in congenital Reasons for Persistence of Infection megaureter, in bladder wall abnormalities, with on September 26, 2021 by guest. The insidious advance towards chronic neurogenic lesions, with lower urinary obstruc- pyelonephritis, which is so resistant to treat- tion, and with chronic pyelonephritis (Turner- ment, has prompted much speculation about Warwick, 1962; Gross and others, 1963; New predisposing factors. Major structural abnormal- England J. Med, 1963). It is also observed ities of the renal tract are readily demonstrated, frequently in association with trigonitis and and do not contribute to the problem urethritis in young women as a transient pheno- under discussion. The megacystis-megaureter menon, and in this way probably disposes to syndrome, for example, is found in 6.5%/, of ascending infection of the kidneys (Hanley, persistent urinary infections (Kunin and others, 1964). Its occurrence in obstructive uropathy 1964). A retrospective survey of a group of is generally accounted for by the protrusion of girls with recurrent urinary infections, pre- a mucosal saccule at the ureteric orifice, while viously investigated, failed to demonstrate any in non-obstructive urinary infections reflux is 490 POSTGRADUATE MEDICAL JOURNAL August, 1965 Postgrad Med J: first published as 10.1136/pgmj.41.478.485 on 1 August 1965. Downloaded from ascribed either to a congenital deficiency of be a necessary condition for damage to occur ureteric muscle, to a shortened intramural (Allen and Burrows, 1964). The ill-effect of re- course from a contracted bladder, or to rigidity flux is well demonstrated in patients with con- from cedema of the bladder wall (Williams, genital doulble ureter, the upper calyx of whose 1964). In differen.t patients, reflux may occur kidneys is always drained by a ureter opening at different levels of intra-vesical pressure, and low in the bladder, after an intramural course of bladder filling (Melick, Brodeur and Karel- of exceptional length. The other ureter has a los, 1962b). Ureteral competence is achieved short intramural course, and therefore is apt to -by a longitudinal crumpling of intra-vesical become incompetent, so allowing reflux to occur. mucosa (a "whistle-valve" effect), and if this In these patients it is always the lower calyces fails reflux may occur without any rise of in- which become damaged and infected (Williams, travesical pressure (Melick, 1962a; Stephens and 1962). In adults, where reflux and chronic Lenaghan, 1962). Reflux may remain unilateral, infection are associated, the untreated prog- as free reflux quickly reduces pressure within nosis before renal failure is two or three decades the bladder. It may disappear spontaneously at most (Hodson and Edwards, 1960). If there (Turner-Warwick, 1962), and because i.t is not is an element of obstruction, the renal failure present at one examination, it is no proof that may be of a kind with a normal blood urea it was never present (Johnston, 1963); it is level, but with excessive water and electrolyte occasionally seen with a sterile urine and no losses (Lancet, 1962; Jones and Mills, 1964). evidence of previous infection. It may occur There may be renal osteodystrophy (Hodson for the first time during a urinary infection and Edwards, 1964). (Hodson and Edwards, 1960), and is recorded in 18% of all infections in girls (Kunin and Lower Urinary Obstruction others, 1964). In recurrent urinary infections re- There is no general agreement upon howProtected by copyright. flux is commoner in children than in adults, and often persistent infection is associated with is reported in 15-30% of non-obstructive minor degrees of obstruction to the lower pyelonephritis (Palken and Kennelly, 1960; urinary tract, though this is a well-recognized Deluca and others, 1963). Radiographic evid- association with ureteric reflux (Hodson, 1959; ence of chronic pyelonephritis without reflux Hodson and Edwards, 1960; Williams and at any time has seldom been found (Williams, Sturdy, 1961; Turner-Warwick, 1962; Deluca 1965), and the triad of chronic pyuria, vesico- and others, 1963; New England J. Med., 1963; ureteric reflux and pyelonephritic scarring of Gross and others, 1963). In a long follow-up the kidney is a familiar one (Johnston, 1963). averaging sixteen years it was found that all Reflux has been reported in 70-80% of children children with advancing renal disease had wi.th chronic pyelonephritis (Allen and Burrows, obstruction to urinary outflow (Steele and others, 1964; Hodson, 1964). There is frequent associa- 1963). Organic abnormalities of the urinary tion with hydronephrosis, with renal scarring tract, including reflux, are found in 50-75%/ of http://pmj.bmj.com/ and with duplex kidneys (Smellie and others, children with persistent infections (Allen and 1964). There is good evidence that the effects Burrows, 1964; Dunn and others, 1964). Obs- of reflux back pressure on the kidneys may tructive uropathy has been reported in 16% be serious (New England J. Med., 1963), and (Deluca and others, 1963) and in 35%/, (Steele that pyelographic abnormalities may be caused, and others, 1963) of large surveys of such presumably via effects upon the calyceal blood children. Obstruction is most often atthebladder vessels (Williams, 1964). These effects are neck in girls, and the lack of clear-cut objective calyceal blunting, general diminution in size of criteria for this diagnosis, has made the entity on September 26, 2021 by guest. the kidney, and failure of the kidney to grow an unsatisfactory one. Bladder neck obstruc- (Hodson, 1959; Hodson and Edwards, 1960; tion cannot be diagnosed endoscopically, and Hodson, Drewe, Karn and King, 1962). In diagnosis depends upon the observation of non- most cases a combination of backpressure and progressive bladder muscle hypertrophy, with infection is present, but the only certain radio- overgrowth of tissue at the bladder outlet, in graphic indication of infection (whose other -the absence of more distal obstruction (Williams effects are indistinguishable from those of re- and Sturdy, 1961). In practical terms this means flux) is coarse localised scarring of the renal the presence of residual urine (not due to nerv- substance (Hodson, 1959), which takes two to ousness or to reflux), and, on cystoscopy, the seven years to develop (Hodson, 1964). Reflux presence of trabeculation of the bladder wall. is present so often when there is X-ray evidence Suggestive additional features are raised intra- of renal damage in pyelonephritis, that it may vesical voiding pressure, ureteric reflux and August, 1965 MAcGREGOR: Pyelonephritis in Children 491 Postgrad Med J: first published as 10.1136/pgmj.41.478.485 on 1 August 1965. Downloaded from sacculation of the bladder (Williams and Sturdy, dilatation of the calyceal system in albout 13% 1961; Gross and others, 1963). Except for of persisting childhood urinary infections bladder neck hypertrophy itself, all these (Kunin and others, 1964). Blunting or shorten- features can be due to other causes, and even ing of the calyces, the size of the kidneys, and now bladder neck obstruction is not accepted by the presence of coarse localised depressions all surgeons (Johnston, 1963), but most consider upon them, may also be revealed. For small it real and important (Williams and Sturdy, children tomography (one or two cuts) may 1961; Gross and others, 1963; Williams, 1965), help to evaluate the pyelogram. Concentration with an onset usually under three years old, and of dye may be remarkably good even with present, according to one estimate (Spence and severely damaged kidneys (Hodson, 1959). others, 1964), in 25% of recurrent infections, Weakening of the picture in later films may though this may be too high a figure (Williams, indicate dilution from ureteric reflux (;Hodson, 1965). The "spinning-top appearance" of the 1964). Tonelessness of the lower ureters (a urethra during a voiding cystogram, previously "flalbby dilatation"), or a ureter visible in its thought to indicate bladder neck obstruction, is entire length may be pyelographic pointers to probably a normal appearance. Likewise the the presence of reflux (Marshall, 1962; Turner- "wide bladder neck deformity" in girls, re- Warwick, 1962; Johnston, 1963). Kinking may garded as an abnormality predisposing to in- be visible at the upper end of such a ureter fection by some workers (Forsythe and Wallace, (Williams, 1964). The "ureteric spurt reaction". 1958), may be a normal variant (Burrows and a jet of urine from the ureteric orifice across Allen, 1964). the bladder, may be noted in the I.V.P. of children with urinary infections (Nevin, Cline, Surgical Treatment of Ureteric Reflux and and Haug, 1952). A nephrogram may be ob- Bladder Neck Obstruction tained by the rapid intravenous injection of Protected by copyright. About ureteric reflux there is not yet sufficient 50 ml. of 70% organic iodine solution .(Hodson, evidence to justify invariable corrective surgery 1959). A pyelogram is advocated in all first (Scott, 1962; TurnerAWarwick, 1962; Macaulay, a,ttacks of urinary infections, as soon as the 1964), nor has time enough elapsed to assess infection is controlled (British Medical Journal, fully the results of corrective procedures 1964; Spence and others, 1964). (Johnston, 1963). Most surgeons advocate first Cysto-urethrography. A voiding or micturat- a period of conservative management, with anti- ing cystogram is necessary for the diagnosis of biotics and the use of "triple micturition" (Gross ureteric reflux. If X-ray screening is possible and others, 1963). For bladder neck obstruction, during the procedure, more useful results will graduated urethral dilatations are urged by be obtained (Allen and Burrows, 1964). The some surgeons (Palken and Kennelly, 1960; procedure is indicated in recurrent infections, Knappenberger, 1963). The presence of dilata- or if reflux is suspected on the pyelogram, or tion of the upper urinary tract is a widely if renal scarring is visible (British Medical http://pmj.bmj.com/ accepted indication for operation (Turner- Journal, 1964). The procedure is often more Warwick, 1962; Gross and others, 1963), most useful than a pyelogram as a first investigation often a reimplantation of the ureters into the in neonates (Smellie and others, 1964). bladder wall, as well as a plastic procedure Reflux may be difficult to demonstrate. The upon the bladder neck. Postoperative follow- earliest indication may be fleeting filling of up for several years in one series showed that the lower ureter (best seen in an oblique-view) infection was unlikely to return if there had during micturition. More obvious reflux pro- been no pre-operative dilatation of the upper duces distension of the pelvis of the kidney on September 26, 2021 by guest. urinary tract (Gross and others, 1963). Another during voiding (Johnston, 1963). At all ages, experienced view is thaat successful reimplanta- a cystogram is more often revealing in urinary tion of the ureters will lead in most instances to infections than a pyelogram. Thus, in one series control of infection without chemotherapy of 500 children submitted to both procedures, (Williams, 1964), and that results justify blad- the cystogram was abnormal on 237 occasions der neck surgery alone in a relatively small and the pyelogram on only 27 (Allen and number of girls. Unless reflux coexists, pye- Burrows, 1964). Another large series reported lonephritis is not usual with bladder neck an albnormal IVP in 22% and an abnormal obstruction (Williams, 1965). cystogram in 44% of cases (Lancet, 1962). Using the two procedures together, the need Techniques of Investigation for retrograde pyelography is virtually Excretion pyelography. This will display eliminated (Forsythe and Wallace, 1958 492 POSTGRADUATE MEDICAL JOURNAL August, 1965 Postgrad Med J: first published as 10.1136/pgmj.41.478.485 on 1 August 1965. Downloaded from Hodson, 1959; Burrows and Allen, 1964). and others, 1964). Pseudomonas and proteus Cineradiography. This is a refinement of the are both related to instrumentation or to sur- voiding cystogram, but provides little informa- gery (Kleeman and others, 1960), or to previous tion that cannot be obtained by other techniques drug treatment (Turck and others, 1963). (Clarke, 1960; New England J. Med., 1963; Staphylococcus aureus has been recorded as the Smellie and others, 1964). second commonest organism in pure growth Cystoscopy. An essential investigation in the (Pryles, 1960). In recurrences, a new strain of diagnosis of bladder neck obstruction. organism is commonly found; estimates for Estimation of residual urine. Aid may be this even vary from 30-87% (Kunin and others, obtained by introducing floating lipiodol into 1964; Spence and others, 1964). Treatment the bladder at the time of the cysitogram. If frequently leads to replacement by a different there is no residue, all should be excreted with- organism in the urine (Gardborg, 1959; Turck in 24 hours. With reflux, radio opaque material and others, 1962). Renal biopsy, presumably may ascend to the pelvis of the kidney (New because of the patchy nature of the inflam- England J. Med., 1963). matory process, has not yielded useful Estimation of voiding pressures within the bacteriological information; results have been bladder. This is useful for assessing the results described as "mystifying and sparse" (Kleeman of operation; measurement can be made through and others, 1960). a small urethral polythene tube (New England J. Med., 1963; Gross and others, 1963). Normal Pathology intravesical pressures are 0-18 mm. Hg. resting, The route of infection of the urinary trace in rising to 70-100 mm. Hg. voiding (Melick, most cases is regarded as ascending. Haema- 1962a). Other authors record lower voiding togenous infection which certainly occurs in pressures of 15 to 20 mm. Hg., and regarded staphylococcal cases, must be accepted as aProtected by copyright. pressures of 40-50 mm. Hg. or more as in- rare possibility (Kleeman and others, 1960). dicative of obstruction (Gross and others, 1963; Experimentally, haematogenous infection by Spence and others, 1963). Reflux may reduce gram-negative organisms will onlly occur if this. abnormal renal conditions, such as ureteric Detection of reflux by a radioactive isotope. ligature, are present (Brumfitt and Percival, By using intravenous 1311-hippuran, a nephro- 1964). Organisms can ascend a stagnant film gram can be obtained with a scintillation of fluid, and they can be found, according to counter during the passage of ;the dye through one paper, at 5 cms. depth in the male urethra the kidney, and again during micturition if (Spence and others, 1964), so ascent to the there is reflux (Dodge, 1963). bladder would not be difficult. Urethritis is Aortogram. This investigation is of special common in many adult women (Brumfitt and value if hypertension is present, or in other Percival, 1964), and cystourethritis precedes

cases where chronic disease may be unilateral. acute pyelonephritis in many young women; http://pmj.bmj.com/ In one centre an aortogram is regarded as the transient reflux up the ureter has been observed next step if excretion pyelography and a voiding many times in the former condition (Hanley, cystogram prove insufficient (Hodson, 1959). 1964). The concepit in some cases of lower Renal biopsy. The results of this are urinary infection without pyelonephritis is gain- generally disappointing in pyelonephritis (see ing ground (Mackinnon, 1964). To allow of below). multiplication of organisms within the bladder It is generally agreed that incomplete study a disturbance of dynamics is usually necessary, of these cases, which may lead to premature especially deficient emptying (Brumfitt and on September 26, 2021 by guest. or misplaced surgery, is dangerous (New Percival, 1964; Spence and others, 1964). England J. Med., 1963). When inflammation of the renal pelvis is present, renal parenchymal infection is almost Bacteriology invariable (Weiss and Parker, 1939). Acute A pure growth of organisms is obtained pyelonephritis is primarily an inflaimmation of from the urine in some three quarters of child- the interstitial tissues of the kidney. Starting hood urinary infections. Of these E.Coli in the medulla, the septic process soon bursts accounts for the majority, 60-80%. B.Proteus, into the tubule, and leucocytes and organisms often in mixed growth, is found in 12-17%, are released into the urine. Glomeruli are cocci (Staph. aureus and enterococci) in 6-8%, remarkably resistant to the suppurative process, and pseudomonas in 4% (Gardborg, 1959; although they too become involved in an Pryles, 1960; Deluca and others, 1963; Kunin " i n v a s i v e glomerulitis". Haematogenous August, 1965 MAcGREGOR: Pyelonephritis in Children 493 Postgrad Med J: first published as 10.1136/pgmj.41.478.485 on 1 August 1965. Downloaded from staphylococcal pyelonephritis, the so-called crescents which suggest nephritis. Dysplastic carbuncle of the kidney, gives rise to constitu- lesions of the nephron may contribute to the tional symptoms with local tenderness, but incidence at this age (Porter and Giles, 1956). often at first no pyuria. Such infections usually There is no other type of nephropathy in heal completely. which fluctuation in renal function occurs so Features by which to differentiate kidneys frequently; patients may approach and recede damaged by healed or chronic pyelonephritis from renal failure in a most variable manner, 'from other renal disturbances have been much according to the state of activity of the discussed, and absolute criteria are hard to find inflammatory process. The residual renal units, '(Weiss and Parker, 1939; Kimmelstiel and though reduced in number, function normally; others, 1961). In general, such kidneys are the consequence is decreasing regulatory flexi- irregularly contracted, often nodular, with bility with increasing azotaemia. Hyper- adherent thickened capsules and flat U-shaped chloraemic acidosis, or the salt-wasting scars. The pelvis may be thickened and infil- syndrome, are rare biochemical accompani- 'trated. Renal papillae are atrophic, and ments of chronic pyelonephritis. calyces dilated and shortened. Within, there is a characteristic patchiness of involvement, with Treatment hyperplasia of unaffected nephrons. The In the treatment of primary uncomplicated interstitial tissues are much inifilltrated with infections immediate good results can be lymphocytes and plasma cells but the presence obtained from most drugs, if the organism is of many polymorphs is essential to the inflam- sensitive (Pryles, 1964), but there is a 10-20% 'matory diagnosis. Many tubules show dilatation failure rate no matter what chemotherapeutic

and atrophy and contain colloid material, so- is used. Reliable data for judging the claims Protected by copyright. called "thyroid-like" areas, which are typical of of one therapeutic regime over another are still pyelonephritis. The glomeruli are relatively scant. The need to obtain adequate renal tissue well preserved, even in scarred areas; peri- levels of drugs, as well as urinary levels, is fibrosis is common, and when being emphasised lately (Brumfitt, 1964), glomerular though the value of this has been questioned in advanced renal failure existed, focal necrosis the face of considerable successes with of the kind descri'bed as "alterative glomerulitis" mandelamine and furadantin which do not is frequently seen. Hyperplastic arteriosclerosis, achieve tissue levels at all (Kleeman and others, ("onion-peel arteritis") of varying severity is 1960). But for many gram negative organisms, regularly seen, and in many such kidneys there the levels of antibiotic reached in the serum by is an intermediate or "barrier" zone of relative conventional therapeutic dosage barely exceed, preservation between an overlying necrotic and and seldom fully overreach, the minimum scarred cortex, and the contracted medulla inhibitory concentration range (Brumfitt, 1964).

with dilated calyces. Most of these changes The correlation between clinical effects of a http://pmj.bmj.com/ are non-specific and the presence of abundant drug in urinary infections and the results of in nolymorphs and of thyroid-like areas are pro- vitro bacterial sensitivity tests is not close; bably the most reliable registers of clinicians do not always pay regard to them nyelonephritis. "Chronic active pyelonephritis" (Pryles, 1964; Spence and others, 1964). The is a useful concept to denote the scarred disc method is unreliable; the tube dilution pyelonephritic kidney in which some acute technique is better, except for sulphonamides inflammatory change is still evident. A less for whom "the patient himself is tihe best common group of chronic pyeloneohritic sensitivity test." (Pryles, 1964). Aittention must on September 26, 2021 by guest. 'kidneys are smaller and evenly and diffusely be paid to pH adjustment of the urine in order contracted (Kincaid-Smith and others, 1958). to get the best results from drugs (Brumfitt and They may show adenoma-like nodularity (Weiss Percival, 1962). This is especially important for and Parker, 1939). These are seen in younger streDtomycin which needs an alkaline urine, people, often with growth failure and renal and for tetracycline which is more effective with rickets. They probably represent inflammation an acid urine. The naturally occurring anti- that has begun in very early life. ("Congenital bacterial substances in urine are thought to be hyponlastic kidneys," which are bilaterally more active at a low pH (Kleeman and others, small but of normal structure. are rarely if ever 1960). Minimum and optimum lengths of seen (Weiss and Parker, 1939)). In neonatal treatment for acute infections have not been pyelonephritis, histological appearances may be decided upon. In a small group of infants confusing because of the presence of glomerular better results were obtained from a 6 months 494 POSTGRADUATE MEDICAL JOURNAL August, 1965 Postgrad Med J: first published as 10.1136/pgmj.41.478.485 on 1 August 1965. Downloaded from than a 2 weeks course of drugs (Stansfeld and DELUCA, F. G., FISHER, J. H., and SWENSON, 0. (1963): Review of Recurrent Urinary Infection in Webb, 1954), but when acute pyelitis was Childhood. New Engl. J. Med., 268, 75. treated for 14 days in each month for a total DODGE, E. A. (1963): Vesico-ureteric Reflux: Diag- of two years, there was still a 7%/0 recurrence nosis with Iodine 131, Lancet, i, 303. rate afterwards (Campanacci and others, 1963). DUNN, P. M., HINE, L., and MACGREGOR, M. E. (1964): Search by Clinical Methods for Persistent In a controlled study of children with acute I Trinary Infection in Children. Brit. med. J.. i. pyelonephritis, a relapse rate of 60% was found 1081. after 2 weeks treatment, 40% after 4 weeks, EDWARDS, D. (1964): Cysto-urethrography in 'Recent and 25% after 6 weeks (Spence and others, Advances in Radiology', edited T. Lodge, p. 179, 1964). In persistent infection there is proved London: J. & A. Churchill. FORSYTHE, W. I., and WALLACE, I. R. (1958): The effectiveness in reducing recurrences from giv- Investigation and Significance of Persistent and ing continued treatment for 6 months, in pre- Recurrent Urinary Infections in Children. Brit. J. ference to short courses lasting 3 to 30 days Urol., 30, 297. (Deluca and others, 1963). If treatment in GADEHOLT, H. (1964): Quantitative Estimation of Urinary Sediment, with Special Regard to Sources persistent cases is going to work, the urine is of Error. Brit. med. J., i, 1547. always sterile after 6 weeks. (Turck and others, GARDBORG, 0. (1959): Bacteriological Findings and 1962). A sulphonamide drug is still the most Resistance to Antibiotics in Chronic and Recurring widely used for first attacks, and furadantin Urinary Tract Infections in Childhood. Acta Paediat. ,(Uppsala) Suppl., 118, 62. for chronic uncontrolled infections. Ampicillin GROSS, R. E., RANDOLPH, J., and WISE, H. M. Jnr. is receiving much study and is the first choice (1963): Surgical Correction of Bladder Neck for Proteus and Enterococci (Naumann, 1964); Obstruction. New Engl. J. Med., 268, 5. many strains of E.Coli show in vitro resistance GUTTMAN, D. and STOKES, E. J. (1963): Diagnosis of to therapeutic levels of this drug, unless it is Urinary Infection. Brit. med. J., i, 1383. HANLEY, H. G. (1964): Pyelonephritis and AscendingProtected by copyright. given parenterally (Brumfitt, 1964). Infection from the Lower Urinary Tract., J. Urol REFERENC(ES (Baltimore), 91, 1. ALLEN, R. P., and BURROWS, E. H. (1964): Micturat- HODSON, C. J. (1959): Radiological Diagnosis of ing Cysto-urethrography in Investigation of Pyelonephritis. Proc. roy. Soc. Med., 52, 669. Urinary Tract Diseases in Children. Arch. Dis. HODSON, C. J., DREWE, J. A., KARN, M. N., and Childh., 39, 95. KING, A. (1962): Renal Size in Normal Children. BRENNER, B. M., and GILBERT, V. E. (1963): Arch. Dis. Childh., 37, 616. Elevated Levels of Lactic Dehydrogenase Glutamic HODSON, C. J., and EDWARDS, D. (1960): Chronic Oxalacetic Transaminase and Catalase in Infected Pyelonephritis and Vesico-ureteric Reflux. Clin. Urine. Amer. J. med. Sci., 245, 31. Radiol., 11, 219. British Medical Journal. (1964): Pyelonephritis in HODSON, C. J. (1964): The Urinary Tract in 'Recent Childhood, 'ii, 1215. Advances in Radiology', edited T. Lodge, p. 156, BRUMFIrr, W. (1964): An Assessment of Ampicillin London: J. & A. Churchill. in the Treatment of Urinary Tract Infection. HOLLAND, N. H., and WEST, C. D. (1963): Prevention Postgrad. med. J. Suppl., 40, 55. of Recurrent Urinary Tract Infection in Girls. BRUMFITr, W., DAVIES, B. I., and ROSSER, E. Amer. J. Dis. Child., 105, 160. (1961): Urethral Catheter as a Cause of Urinary HOUSTON, I. B. (1964): Pus Cell and Bacterial Counts http://pmj.bmj.com/ Tract Infection. Lancet, ii, 1059. in the Diagnosis of Urinary Tract Infe4tions in BRUMFITr, W., and PERCIVAL, A. (1962): Adjustment Childhood. Arch. Dis. Childh., 38, 600. of Urinary pH in the Chemotherapy of Urinary HuTH, M. S. R., CHALMERS, J. A., MACDONALD, J. S., Tract Infections, Lancet, i, 186. and DE WARDENER, H. E. (1961): Pyelonephritis. BRUMFITr, W., and PERCIVAL, A. (1964): Patho- Lancet, i, 351. genesis and Laboratory Diagnosis of Non-tuber- JAMES, U. (1959): Urinary Infection in the Newborn. culous Urinary Tract Infection. J. Clin. Path., 17, Lancet, ii, 1001. 482. JOHNSTON, J. H. (1963): Vesico-ureteric Reflux in the

BRUMFITrI, W., PERCIVAL, A., and WILLIAMS, J. D. Child. Clin. Paediat, 2, 72. on September 26, 2021 by guest. (1964): Problems in the Treatment of Urinary JONES, N. F., and MILLS, I. H. (1964): Reversible Tract Infection. Brit. J. clin. Pract., 18, 503. Renal Potassium Loss with Uxinary Tract Infec- BURKE, J. B. (1961): Pyelonephritis in Infancy and tion. Amer. J. Med., 37, 305. Childhood. Lancet, ii, 1116. KASS, E. H. (1956): Asymptomatic Infections of the BURROWS, E. H., and ALLEN, R. P. (1964): Urethral 56. Lesions in Infancy and Childhood Studied by Urinary Tract. Trans. Ass. Amer. Phycns., 69, Micturition Cysto-Urethrography. Brit. J. Radiol., KASS, E. H. (1960): Biology of Pyelonephritis, edited 37, 187. Quinn, E. L., and Kass, E. -H. London: J. & A. CAMPONACCI, D., BoNoMINI, V., and ZUCHELLI, P. Churchill. (1963): Chemotherapy in Acute and Chronic KIMMMELSTIEL, P., KIM, 0. J., BERES, J. A., and Pyelonephritis, Lancet ii, 601. WELLMAN, K. (1961): Chronic Pyelonephritis. CATrELL, W. R. and LEFFORD, M. J. (1963): Amer. J. Med., 30, 589. Bacteriological Examination of Urine. Brit. med. KINCAID-SMITH, P. (1964): Ampicillin in Bacteriuria J., i, 97. and Pyelonephritis of Pregnancy. Postgrad. Med. CLARKE, S. H. C. (1960): Investigation into Methods J. Suppl. 40, 74. of Collecting Urine. Brit. med. J., ii, 1491. KINCAID-SMITH, P., MCMICHAEL, J., and MURPHY, August, 1965 MAcGREGOR: Pyelonephritis in Children 495 Postgrad Med J: first published as 10.1136/pgmj.41.478.485 on 1 August 1965. Downloaded from E. A. (1958): The Clinical Course and Pathology Aspiration of the Bladder Catheterisation, and Mid- of Hypertension with Papilloedema. Quart. J. Med., stream-voided Methods. New Engl. J. Med., 259, 27, 117. 764. KINCAID-SMITH, P., BULLEN, M., MILLS, J., FUSSELL, NAUMANN, P. (1964): Treatment of Urinary Tract U., HUSTON, N., and GooN, F. (1964): The Reliabil- Infections with Ampicillin, Postgrad. med. J. ity of Screening Tests for Bacteriuria in Pregnancy. Suppl. 40, 62. Lancet, ii, 61. New England Journal of Medicine (1963): Lower KLEEMAN, C. R., HEWITr, W. L., and GAZE, L. B. Urinary Tract Obstruction in Children., 268, 52. (1960): Pyelonephritis. Medicine (Baltimore), 39, 3. NEVIN, I. M., CLiNE, F. A., and HAUG, T. M. (1962): KNAPPENBERGER, S. T. (1963): Importance of Urethral Forceful Ureteric Spurt: a Common Roentgen Dilatations in Treatment of Chronic Urinary Tract Manifestation of Urinary Tract Infection in Child- Infections in Female Children. J. Urol (Baltimore), ren. Radiology, 79, 933. 89, 95. OSBORN, R. A., and SMITH, A. J. (1963): Comparison KUNIN, C. M., DEUTSCHER, R., and PAQUIN, A. of Quantitative Methods in the Investigation of (1964): Urinary Infection in School-children. Urinary Tract Infection, J. clin. Path., 16, 46. Medicine (Baltimore), 43, 91. O'SULLIVAN, D. J., FITZGERALD, M. G., MEYNELL, KuNIN, C. M., SOUTHALL, I., and PAQUIN, A. J. M. J., and MALINS, J. M. (1960): A Simplified (1962): Urinary Tract Infections in School- Method for the Quantitative Bacterial Culture of children. New Engl. J. Med., 266, 1287. Urine, J. clin. Path., 13, 527. Lancet, (1962a): Obstructive Nephropathy, i, 31. PALKEN, M., and KENNELLY, J. M. (1960): Recurrent Lancet, (1962b): Bacteriuria in School-girls, ii, 598. Urinary Tract Infection in Girls, J. Urol. (Balti- Lancet, (1963): Chemotherapy of Urinary Infections, more) 83, 749. i, 148. PERCIVAL, A., BRUMFITr, W., and DE LouVoIs, J. LEATHER, H. M., WILLS, M. R., and GAULT, H. M. 1964): Serum Antibodies as an Indication of (1963): Bacterial Pyrogen in the Diagnosis of Clinically Inapparent Pyelonephritis. Lancet, ii, Pyelonephritis. Brit. med. J., i, 92. 1027. LEIGH, D. A., and WILLIAMS, J. D. (1964): Method POIRIER, K. P., and JACKSON, G. G. (1957): for Detection of Significant Bacteriuria in Large Characteristics of Leucocytes in the Urine Sedi-

Groups of Patients. J. Clin. Path., 17, 498. ment in Pyelonephritis. Amer. J. Med., 23, 579. Protected by copyright. LINCOLN, K., and WINBERG, J. (1964): Quantitative PORTER, K. A. and GILES, H. MCC. (1956): A Estimation of Cellular Excretion in Unselected Pathological Study of Five Cases of Pyelonephritis Neonates. Acta Paediat. (Uppsala), 53, 447. in the Newborn. Arch. Dis. Childh., 31, 303. LINDMEYER, R. I., TURCK, M., and PETERSDORF, R. G. PRESCOTT, L., and BRODIE, D. E. (1964): A Simple (1963): Factors Determining the Outcome of Differential Stain for Urinarv Sediment. Lancet ii, Chemotherapy in Infections of the Urinary Tract. 940. Ann. intern. Med., 58, 201. PRYLES, C. V. (1960): The Diagnosis of Urinary LITTLE, P. J. (1962): Urinary White Cell Excretion. Tract Infection. Pediatrics 26, 441. Lancet, i, 1149. PRYLES, C. V. (1964): Anti-microbial Therapy of LITTLE, P. J. (1964): A Comparison of Urinary White Urinary Tract Infections in 'Current Pediatric Cell Concentration with the White Cell Excretion Therapy'. Edited Gellis, S. S., and Kagan, B. M. Rate, Brit. J. Urol., 36, 360. Philadelphia: W.B. Saunders. LITTLE, P. J., and DE WARDENER, H. E. (1 962): Use PRYLES, C. V., and NEUMANN, C. G. (1962): of Prednisolone Phosphate in the Diagnosis of Pyelonephritis in Infants and Children. Amer. J. Pyelonephritis in Man. Lancet, L, 1145. Dis. Child., 104, 215. MACAULAY, D., and SUTToN, R. N. P. (1957): The ROSENHEIM, M. L. (1963): Problems of Chronic

Prognosis of Urinary Tract Infection in Child- Pyelonephritis. Brit. med. J., i, 1433. http://pmj.bmj.com/ hood. Lancet, ii, 1318. ScoTT, J. E. (1962): Experimental Asplects of MACAULAY, D. (1964): Recurrent Urinary Infection Vesico-ureteric Reflux, Proc. roy. Soc. Med., 55, in Girls. Lancet, ;;, 1319. 422. MCGRACKIE, J., and KENNEDY, A. C. (1963): Simpli- SENECA, H. (1964): Current Therapy of Infections of fied Quantitative Methods for Bacteriuria and the Renal Excretory System. J. Amer. Geriat. Soc., Pyuria, J. clin. Path., 16, 32. 12, 1100. MACKINNON, K. J. (1964): Ampicillin in Bacteriuria SIMMONS, N. H., and WILLIAMS, J. D. (1962): A and Pyelonephritis; discussion. Postgrad. med. J., Simple Test for Significant Bacteriuria. Lancet, i, Suppl. 40, 79. 1377. MCLEOD, J. W., MASON, J. M., and PILLEY, A. A. SMALLPIECE, V. (1958): Prognosis of Urinary Infec- on September 26, 2021 by guest. (1963): Prophylactic Control of Urinary Infection tions in Childhood. Lancet, i, 103. from Catheterisation. Lancet, i, 292. SMELLIE, J. M., HODSON, C. J., EDWARDS, D., and MARSHALL, F. C. (1962): Excretion Urographic NORMAND, C. S. (1964): Clinical and Radiological Changes which Suggest Reflux, J. Urol. (Baltimore), Features of Urinary Infection in Children. Brit. 87, 681. med. J., ii, 1222. MELICK, W. F., BRODEUR, A. E., and KARELLOS, D. N. SPARK, H., TRAVIS, L. B., DODGE, W. F., DAESCHMER, (1962a): Abnormal Hydrodynamics of Bladder and C. W., and HoPPs, H. C. (1962): Prevalence of Ureter as a Cause of Reflux, J. Urol. (Baltimore), Pyelitis in Children at Autopsy. Pediatrics, 30, 88, 38. 737. MONTGOMERIE, J. Z., and NORTH, J. D. K. (1963): SPENCE, H. M., MURPHY, J. J., McGOVERN, J. H., Evaluation of Pyrogen Test in Chronic Pyelone- HENDRON, W. H., and PRYLES, C. V. (1964): phritis. Lancet, i, 690. Urinary Tract Infections in Infants and Children. MONZON, 0. T., ORY, E. M., DOBSON, H. L., CARTER, J. Urol. (Baltimore) 91, 623. E., and Yow, E. M. (1958): A Comparison of STANSFELD, J. M. (1954): Chronic Pyelonephritis in Bacterial Counts of the Urine Obtained by Needle Children. Proc. roy. Soc. Med., 47, 631. 496 POSTGRADUATE MEDICAL JOURNAL August, 1965 Postgrad Med J: first published as 10.1136/pgmj.41.478.485 on 1 August 1965. Downloaded from STANSFELD, J. M. (1962): The Measurement and lem and Investigation of Ureteric Reflux, Proc. roy. Meaning of Pyuria. Arch. Dis. Childh., 37, 257. Soc. Med., 55, 419. STANSFELD, J. M. (1963): Personal communication. VERTANEN, S., OKSANEN, T., and PErONEN, T. (1962): STANSFELD, J. M., and WEBB, J. H. (1954): A Plea Colony Count and Diagnosis of Urinary Tract for Longer Treatment of Chronic Pyelonphritis in Infection in Infants and Children, Ann. Paediat. Children, Brit. med. J., i, 616. Fenn., 8, 269. STEPHENS, F. D., and LENAGHAN, D. (1962): Anatom- WEISS, S., and PAKER, F. (1939): Pyelonephritis: its ical Basis and Dynamics of Ureteral Reflux, J. Relation to Vascular Lesions and to Arterial Hyper- Urol. i(Baltimore), 87, 669. tension, Medicine (Baltimore), 18, 221. STEELE, R. E., LEADBETTER, G., and CRAwFORD, J. WILLIAMS, D. I. (1962): Reflux in Double Ureters, (1963): Prognosis of Childhood Urinary Tract Proc. roy. Soc. Med., 55, 423. Infection, New Engl. J. Med., 269, 883. WILLIAMS, D. I. (1963): Personal communication. STERNHEIMER, R., and MALBIN, B. (1957): Clinical WILLIAMS, D. I. (1964): Vesico-uretic Reflux in Recognition of Pyelonephritis, with a New Stain 'Recent Advances in Surgery', edited Taylor, S. for Urinary Sediments, Amer. J. Med., 11, 312. P. 110. London: J. & A. Churchill, SWEET, A. Y., and WOLINSKY, E. (1964): An Outbreak WILLIAMS, D. I. (1965): Recurrent Urinary Infection of Urinary Tract and Other Infections due to in Girls, Lancet, i, 109. E. Coli, Pediatrics., 33, 865. WILLIAMS, D. I., and STURDY D. E. 1(1961): Recurrent TURCK, M., BROWDER, A., LINDMEYER, R. I., BROWN, Urinary Infection in Girls, Arch. Dis. Childh., 36, N. K., and PETERSDORF, R. G. (1962): Failure of 130. Prolonged Treatment in Chronic Urinary Infection WOODRUFF, J. D., and EVERETr, H. S. (1954): Prog- with Antibiotics, New Engl. J. Med., 267, 999. nosis in Childhood Urinary Tract Infections in TURNER-WARWICK, R. T. (1962): The Clinical Prob- Girls, Amer. J. Obstet. Gynec., 68, 798. Protected by copyright. http://pmj.bmj.com/ on September 26, 2021 by guest.