Brirish Joririral of Urology (1976), 48, I 11-1 17 0
Influence of Vesico-ureteral Reflux on the Response to Treatment of Urinary Tract Infections in Female Children WILLIAM R. FAIR
and
DUNCAN E. GOVAN
Department of Surgery (Urology), Stanford University School of Medicine, Stanford, California
The purpose of this article is to present data on 3 groups of female children, comprising a total of 169 patients, who were followed through successive episodes of urinary tract infection (UTI) with respect to recurrence of infection. Specifically, the authors will consider: ( a ) the influence of vesico-ureteral reflux on the recurrence rate of infection and (b) whether or not surgical correction of reflux influenced the rate of infection.
Materials and Methods A total of 169 patients followed at Stanford University Medical Centre constitutes the basis of the study. All patients had intravenous pyelograms and voiding cysto-urethrograms performed and were personally followed by one of the authors. Most of the children were white and of middle class socio-economic background. The diagnosis of infection was based, in all instances, on the results of the urine cultures. The criteria for considering a urine culture positive have been outlined in detail previously (Govan and Palmer, 1969). E. coli was the most common urinary tract pathogen recovered and was responsible for approximately 70% of all infections. No serotyping was done, so it was not possible to differentiate between relapse and reinfection in these patients. Group A consisted of 66 patients with no evidence of vesico-ureteral reflux. Group B was composed of 42 patients with demonstrated vesico-ureteral reflux, in whom urinary tract infections were treated with medical therapy alone and in whom the reflux was not corrected. Group C comprised 61 patients, all of whom had successful surgical correction of reflux. Only those children presenting with UTI with or without vesico-ureteral reflux, who were free of other congenital urologic or neurologic abnormalities and who were personally followed by one of the authors for at least I year, were selected for this study. Of the 169 children presented here, 62 (31 in Group A and 31 in Group C) have been reported on in a previous publication from this institution (Govan and Palmer, 1969).
Treatment Modalities. None of the infections were treated with any long-term suppressive medication. However, those patients who had surgical correction of reflux were prescribed an appropriate antimicrobial agent (r.ide infru) postoperatively for a period of 4 to 6 weeks only. In all other cases treatment of a documented infection consisted of 10 days of a specific antibacterial agent following diagnosis by urine culture. A patient was considered free of infection following treatment only if a sterile urine culture was obtained 3-4 days after the cessation of all antibiotics. A variety of antibacterial agents were used; generally the choice was made on the basis of bacterial sensitivity results. The primary antibacterial drugs used for therapy were sulfisoxazole, penicillin-(;, and nitrofurantoin. Of the 66 children without vesico-ureteral reflux, who were followed through at least one and in some instances as many as 1 I episodes of infection, the follow-up time averaged 40 months. 111
119
BRITISH JOURNAL 0 1 URC)I.OGY
Table I Grade of Ureterovesical Reflux Group B (42 patients), medication, no surgery
Group C (61 patients), surgery plus medication ~~
No. of ureters Grade I Grade 11 Grade I I I
66 26 35 5
I05 12 46 47
In the groups of children presenting with reflux, those treated conservatively with medication alone (Group B) and those whose reflux was corrected by surgery (Group C ) , both were followed for a minimum of 12 months. Decision for Surgical Correction. Those children, in whom adequate control of UTl's wab not obtained with medical therapy only, underwent surgical correction of reflux. In general, "adequate control" indicates that the patient had persistently sterile urine or was experiencing infrequent episodes of minimally symptomatic bladder infections or asymptomatic bacteriuria. Grading of Reflux. The degree of reflux was graded as previously described (Filly r t NI., 1974). The severity of reflux varied in the 2 groups of children presenting with vesico-ureteral rellux. As seen in Table I, some children had only unilateral reflux; thus, Group B represents 42 children with a total of 66 refluxing ureters. Group C comprised 61 patients with a total of 105 retluxing ureters. I n general, those children undergoing surgery had more severe degrees of reflux. Originally the Hutch techniques of ureteroneocystostomy were employed (Hutch, 1963; Hutch. Smith and Osborne, 1968). Because of the high incidence of persistent reflux associated with those procedures, we later adopted the Politano-Leadbetter approach as our primary surgical technique. The majority of the patients had a variation of this original technique i n that the ureter was exposed from outside the bladder as well as circling the orifice froiii within. The ureter was then reanastomosed under direct vision, using the tunnelling procedures advocated by the above authors (Politano and Leadbetter, 1958). If the ureter was severely dilated the distal one-third of the ureter was tapered prior to reimplantation into the bladder. All of the children in Group C had successful correction of vesico-ureteral reflux. Results The results will be considered under the following 3 groups: Group A : Those girls who had no evidence of vesico-ureteral reflux: 66 patients. Group B: Girls having vesico-ureteral reflux treated with medical therapy only: 42 children. Group C : Patients having vesico-ureteral reflux corrected by surgery, plus medical therapy : 61 children. Children without Reflux (Group A ) The 66 patients were followed through as many as 1 1 episodes of urinary tract infections, cuch treated with a short course of a specific antibacterial agent (Table 11). The mean time of observation in the entire group was 40 months with all patients followed for at least 12 months. Table I I
113
URINARY TRACT INFECTIONS IN FEMALE CHILDREN
Table I1 Results of Short-term Treatment in Female Children without vesico-ureteral Reflux (Group A)
Treatment n uni ber ~
Patients at start
Patients cured*
-~
~~
1 2 3 4 5 6 7 8 9 10 11
Patients with recurrent infections
66 50 46 31 30 24 18
14 10 9 8
16 4 9
1 6 6 4
~~
50 46 31 30 24
15 70 56 49 39 30 23 16 15 13 8
18
1
14 10 9 8
3
5
4 I
Percentage with recurrences
* No further infections during observation period (mean
= 40
months).
indicates exactly what happened in terms of reinfection when these children were treated for a specific infection and urine cultures confirmed that the treatment was indeed successful. Thus, of the 66 patients (lOOo/o) who were treated for one infection, 50 (75”,) went on to develop a second infection at some time during the follow-up period. After treatment of this second infection 4 additional patients were “cured” and had no further infections during the period of observation, while the majority went on to experience an additional infection. The same tendency occurred following treatment of subsequent infections. Thus, more than one-half of the girls in the study had only 3 or 4 infections and did not have a positive culture during the remainder of the follow-up period. In general, of those patients treated for any given infection, approximately 202; were “cured” each time and did not experience another infection. However, 80‘;; of those given this short-term treatment with a specific antibacterial agent went on to develop another infection. This was true whether it was the patient’s first or subsequent infection. Children with Rejux treated with Medical Therapy only (Group B ) In comparing patients with vesico-ureteral reflux treated with medical and surgical modalities (Group B versus Group C) it is important to consider first the rate of infection prior to referral to Stanford. Both groups of children averaged slightly more than 2 infections per year during this time. Hence, there was no significant difference in the incidence of infection prior to entry into the study. Also, about 50”/: of the infection episodes in each group were clinically diagnosed as pyelonephritis (fever, with or without chills and back or abdominal pain). Those children with recurrent severe episodes of pyelonephritis, particularly if accompanied by pyelographic changes consistent with this diagnosis and who were not kept sterile with antibacterial medication, were operated upon for correction of the reflux and, hence, were included in Group C. Thus, it appeared that the groups were quite comparable in regard to both the incidence and type of infection prior to entering the study, although those in Group C were more likely to have had recurrent pyelonephritis not readily controlled by medication. The Figure indicates the percentage of patients who were “cured” following each treatment episode in these 42 female children (Group B) with demonstrated vesico-ureteral reflux who
1 I4
BRITISH JOURNAL OF UROLOGY
I
I
I
I
I
I
I
I
All patients followed over 1 year Non-reflux 66 patients 0 Reflux- Medical treatment 42 patients o Reflux- Surgical treatment 61 patients
I
-
10 -
0
1
I
1
2
1
3
1
1
I
5 6 Treatments
4
7
8
9
10
Fig. The recurrence rate of new infection for the 3 groupsofchildrenfollowingeachsu~ccssfiiltrcatrnent I S plotted against the theoretical 20 ,, "cure" rate per treatment (black line).
were followed for at least 1 year after the demonstration of reflux and in whom surgery w a s not performed. As with Group A children, 20 to 25 of those treated for a specific infection did not develop another infection but maintained sterile urine throughout the follow-up period. The Figure also indicates that the 2 patient populations showed a similar response to treatment whether it was the first infection or a subsequent infection.
x,
Children with Vesico-Ureteral Refrux treated surgically Group C consisted of 61 patients followed for at least 12 months after the correction of the reflux. As indicated in the Figure these patients had a very similar postoperative infection rate when compared with the patients without reflux (Group A) and those children in whom the reflux was not surgically corrected (Group B). That is, approximately 20:4 of the patients were cured with each course of short-term medical therapy. An important observation in both groups of refluxing children ( B and C ) was that the incidence of clinical pyelonephritis was approximately 50:';; of all infections in both Groups B and C prior to referral. This was reduced to 10";) of the total number of infections in those surgically treated (Group C ) and less than 20n;,of the infections in those medically treated (Group B). The differences are not statistically significant but indicated a marked decrease in symptomatic pyelonephritis in both populations compared with the rate prior to the initiation of close medical and/or surgical supervision. The children in Groups B and C were not arbitrarily assigned to either a medically or surgicallytreated group. In general, the non-surgical group was comprised of children with lesser degrees of reflux than those children who were treated operatively (Table I). Also, those children whose infections could not be adequately controlled with close medical supervision were considered as surgical candidates. The 2 groups, therefore, are not strictly comparable in the severity of the reflux prior to entering into our study.
Comparison of the Results As shown in the Figure, the results of treatment of infection were essentially the same. In the children without reflux, and in those whose reflux was surgically corrected, the percentage
URINARY TRACT INFECTIONS IN FEMALE CHILDREN
115
experiencing recurrent infections following each episode of treatment paralleled the theoretical 20% “cure” rate. Thus, there appears to be no difference in the recurrence rate of infections in those children with reflux corrected surgically and those treated medically. More importantly, the recurrence rate in both of these groups does not differ significantly from the group of children with UTl’s who do not have ureterovesical reflux. Comments There were several interesting observations apparent during the course of these studies. All patients on the study received 10 days of a specific antibacterial agent at the time the infection was diagnosed; no long-term suppressive medication was given to this group. As shown in the Figure our results in the group of children without reflux confirmed the earlier results of Kunin (1970) which indicate, essentially, that each time we treat a group of children presenting with UTI’s, 20% of the patients will have no further infection but that 80% will develop another infection. If the remaining 80 who develop another infection are again treated and sterilised, another 20% of this population will be “cured” but 80% of this group will still go on to develop a further infection. Prior to referral to our institution over 50% of the infections in the patients comprising both Group Band Group C were associated with fever, with or without flank pain, symptoms clinically suggestive of pyelonephritis (Govan et al., 1974). This is quite in contrast to the benign type of clinical infection experienced in the children of Group A (Fair et al., 1974) in whom a diagnosis o f pyelonephritis was suspected in fewer than 10% of the total number of infections. It should be noted that Groups B and C were not strictly comparable, in that there was a marked difference in the degree of reflux found in the patients comprising the 2 groups. However, the overall infection rate was similar in both populations under the same type of medical management. Of particular interest is the observation that the incidence of clinical pyelonephritis was reduced to less than 20% of all infections in both groups. This compares favourably with an incidence of clinical pyelonephritis of less than 10% of all infections in non-refluxing children (Group A ) and is in agreement with other studies in similar patient populations (Winberg, Larson and Bergstrom, 1970). Winberg et d.(1970) found that in children with reflux, whose infections were under bacteriological control, the incidence of clinical pyelonephritis was similar to that occurring in children without reflux. In addition to the similarities between the 2 groups of patients with reflux as to the type and incidence of infection prior to entry into the study, it is also of note to compare the “cure” rate of these children with each successful treatment. As shown in the Figure both groups had about 20% of the children “cured” following each course of treatment, whether the reflux was surgically corrected or not. Quite significantly, the “cure” rate was similar for the 3 groups of children with or without reflux and did not differ in those children with reflux corrected by surgery plus medication and those treated with medication alone. All 3 groups paralleled the theoretical 20 % “cure” rate previously described by Kunin ( 1970). At this point it is perhaps important to realise in discussing a prevention of recurrences, we are, for the most part, talking about preventing reinfection rather than relapse of an improperlytreated infection. Kunin and Halmagyi (1962) and Bergstrom et a/. (1967) have shown that 85 ?,) of urinary tract infections are new infections caused by a different organism each time. Thus it becomes clear that long periods of treatment are simply preventing reinfections and not eradicating relapses due to inadequate treatment. In summary, from these data one may draw the following conclusions: I . Urinary tract infections are not totally prevented by correction of vesico-ureteral reflux; 2. The rate of reinfections in surgically-treated ver.su.r medically-treated children with reflux is similar; 3. The “cure” rate is similar, when one compares the 2 groups of children with reflux (Groups B and C ) , with the children without vesico-ureteral reflux (Group A).
1 I6
BRITISH JOURNAL 0 1 UROLOCiY
From a prospective point of view it becomes apparent that it would be most helpful t o identify those girls who are likely to develop multiple recurrences and treat those children prophylactically with long-term suppressive therapy. The results of this study, which confirms the earlier work of Kunin (1970), show that medication alone will produce a long-term cure i n 20”,, of young white females whether it is the first infection o r a subsequent infection. Use of long-term SLIPpressive medication such as advocated by Normand and Smellie ( 1965) gave excellent results i n a group of children without reflux similar to ours. The fact that this approach to the problem has been shown to be effective underscores the major problem i n treating these children; that i h ? identifying those girls who are most susceptible to recurrent infections and directing long-term therapy to that group. Since more than SOY{, of the children will have no more than 3 or 4 infections, we suggest that in those female children whose reflux is not deemed serious enough to warrant surgery (Govan et d., 1974) o r who continue to have reinfections following surgery. as well as in those children who d o not have vesicoureteral reflux a t all, it would be appropriatc to begin these patients on long-term suppressive medication after the third o r fourth infection i n an effort to select those children who are most likely to benefit from this form of therapy. Our current practice in such instances of recurrent bacteriuria is initially to sterilise the child with appropriate medication for 10 days and then to place the child immediately on it low dose of nitrofurantoin (usually in the form of Macrodantin 25 or 50 nig) given just before bedtime. This medication is prescribed for 6 months, then discontinued. Should the child develop a further infection, then the whole procedure is repeated. This regimen has been highly successful i n lowering the frequency and morbidity of infections in this small, but difficult population.
Summary 3 groups of female children, comprising a total of 169 patients, were followed through succcssive episodes of urinary tract infection with respect to recurrence of infection. Group A consisted o f 6 6 patients with no evidence of vesico-ureteral reflux. Group B comprised 42 patients with demonstrated vesico-ureteral reflux. The infections i n these patients were treiited with medical therapy alone and the reflux was not corrected. Group C consisted o f 61 children who had surgical correction of vesico-ureteral reflux. While Groups B and C were n o t strictly comparable, the data indicate that there was no demonstrable difference in the rate o f urinury tract infection among the 3 groups. Following short-term specific antibacterial therapy of a new infection, about X ” , 01’ ,childrcn in each group were “cured”, i.c., liud t i o ,furthrr i i l f ~ c t i o n .clirritig ~ tlic 12 niont1i.s or t t i o w tliiit ~ l i i , , i > \i’ori’ .$ uhsc~qlic~tit I!, follo KW/ 6). thc mrtllors. Patients with less severe grades of reflux treated by medication alone ( G r o u p B) experienced no greater rate of reinfection than those children presenting with infections i n thc itbscncc of vcsico-ureteral reflux. In those children whose infections could not be adequately controlled by antiniicrobial therapy. and those patients with severe (Grade I l l ) reflux, surgery was performed to eliminote the vesicoureteral reflux. The data also reinforced the concept that correction of reflux does not eliminate the need for constant vigilance and proper treatment of recurrent infections even after successful surgery has been carried out.
References BI i
Influence of Vesico-ureteral Reflux on the Response to Treatment of Urinary Tract Infections in Female Children WILLIAM R. FAIR
and
DUNCAN E. GOVAN
Department of Surgery (Urology), Stanford University School of Medicine, Stanford, California
The purpose of this article is to present data on 3 groups of female children, comprising a total of 169 patients, who were followed through successive episodes of urinary tract infection (UTI) with respect to recurrence of infection. Specifically, the authors will consider: ( a ) the influence of vesico-ureteral reflux on the recurrence rate of infection and (b) whether or not surgical correction of reflux influenced the rate of infection.
Materials and Methods A total of 169 patients followed at Stanford University Medical Centre constitutes the basis of the study. All patients had intravenous pyelograms and voiding cysto-urethrograms performed and were personally followed by one of the authors. Most of the children were white and of middle class socio-economic background. The diagnosis of infection was based, in all instances, on the results of the urine cultures. The criteria for considering a urine culture positive have been outlined in detail previously (Govan and Palmer, 1969). E. coli was the most common urinary tract pathogen recovered and was responsible for approximately 70% of all infections. No serotyping was done, so it was not possible to differentiate between relapse and reinfection in these patients. Group A consisted of 66 patients with no evidence of vesico-ureteral reflux. Group B was composed of 42 patients with demonstrated vesico-ureteral reflux, in whom urinary tract infections were treated with medical therapy alone and in whom the reflux was not corrected. Group C comprised 61 patients, all of whom had successful surgical correction of reflux. Only those children presenting with UTI with or without vesico-ureteral reflux, who were free of other congenital urologic or neurologic abnormalities and who were personally followed by one of the authors for at least I year, were selected for this study. Of the 169 children presented here, 62 (31 in Group A and 31 in Group C) have been reported on in a previous publication from this institution (Govan and Palmer, 1969).
Treatment Modalities. None of the infections were treated with any long-term suppressive medication. However, those patients who had surgical correction of reflux were prescribed an appropriate antimicrobial agent (r.ide infru) postoperatively for a period of 4 to 6 weeks only. In all other cases treatment of a documented infection consisted of 10 days of a specific antibacterial agent following diagnosis by urine culture. A patient was considered free of infection following treatment only if a sterile urine culture was obtained 3-4 days after the cessation of all antibiotics. A variety of antibacterial agents were used; generally the choice was made on the basis of bacterial sensitivity results. The primary antibacterial drugs used for therapy were sulfisoxazole, penicillin-(;, and nitrofurantoin. Of the 66 children without vesico-ureteral reflux, who were followed through at least one and in some instances as many as 1 I episodes of infection, the follow-up time averaged 40 months. 111
119
BRITISH JOURNAL 0 1 URC)I.OGY
Table I Grade of Ureterovesical Reflux Group B (42 patients), medication, no surgery
Group C (61 patients), surgery plus medication ~~
No. of ureters Grade I Grade 11 Grade I I I
66 26 35 5
I05 12 46 47
In the groups of children presenting with reflux, those treated conservatively with medication alone (Group B) and those whose reflux was corrected by surgery (Group C ) , both were followed for a minimum of 12 months. Decision for Surgical Correction. Those children, in whom adequate control of UTl's wab not obtained with medical therapy only, underwent surgical correction of reflux. In general, "adequate control" indicates that the patient had persistently sterile urine or was experiencing infrequent episodes of minimally symptomatic bladder infections or asymptomatic bacteriuria. Grading of Reflux. The degree of reflux was graded as previously described (Filly r t NI., 1974). The severity of reflux varied in the 2 groups of children presenting with vesico-ureteral rellux. As seen in Table I, some children had only unilateral reflux; thus, Group B represents 42 children with a total of 66 refluxing ureters. Group C comprised 61 patients with a total of 105 retluxing ureters. I n general, those children undergoing surgery had more severe degrees of reflux. Originally the Hutch techniques of ureteroneocystostomy were employed (Hutch, 1963; Hutch. Smith and Osborne, 1968). Because of the high incidence of persistent reflux associated with those procedures, we later adopted the Politano-Leadbetter approach as our primary surgical technique. The majority of the patients had a variation of this original technique i n that the ureter was exposed from outside the bladder as well as circling the orifice froiii within. The ureter was then reanastomosed under direct vision, using the tunnelling procedures advocated by the above authors (Politano and Leadbetter, 1958). If the ureter was severely dilated the distal one-third of the ureter was tapered prior to reimplantation into the bladder. All of the children in Group C had successful correction of vesico-ureteral reflux. Results The results will be considered under the following 3 groups: Group A : Those girls who had no evidence of vesico-ureteral reflux: 66 patients. Group B: Girls having vesico-ureteral reflux treated with medical therapy only: 42 children. Group C : Patients having vesico-ureteral reflux corrected by surgery, plus medical therapy : 61 children. Children without Reflux (Group A ) The 66 patients were followed through as many as 1 1 episodes of urinary tract infections, cuch treated with a short course of a specific antibacterial agent (Table 11). The mean time of observation in the entire group was 40 months with all patients followed for at least 12 months. Table I I
113
URINARY TRACT INFECTIONS IN FEMALE CHILDREN
Table I1 Results of Short-term Treatment in Female Children without vesico-ureteral Reflux (Group A)
Treatment n uni ber ~
Patients at start
Patients cured*
-~
~~
1 2 3 4 5 6 7 8 9 10 11
Patients with recurrent infections
66 50 46 31 30 24 18
14 10 9 8
16 4 9
1 6 6 4
~~
50 46 31 30 24
15 70 56 49 39 30 23 16 15 13 8
18
1
14 10 9 8
3
5
4 I
Percentage with recurrences
* No further infections during observation period (mean
= 40
months).
indicates exactly what happened in terms of reinfection when these children were treated for a specific infection and urine cultures confirmed that the treatment was indeed successful. Thus, of the 66 patients (lOOo/o) who were treated for one infection, 50 (75”,) went on to develop a second infection at some time during the follow-up period. After treatment of this second infection 4 additional patients were “cured” and had no further infections during the period of observation, while the majority went on to experience an additional infection. The same tendency occurred following treatment of subsequent infections. Thus, more than one-half of the girls in the study had only 3 or 4 infections and did not have a positive culture during the remainder of the follow-up period. In general, of those patients treated for any given infection, approximately 202; were “cured” each time and did not experience another infection. However, 80‘;; of those given this short-term treatment with a specific antibacterial agent went on to develop another infection. This was true whether it was the patient’s first or subsequent infection. Children with Rejux treated with Medical Therapy only (Group B ) In comparing patients with vesico-ureteral reflux treated with medical and surgical modalities (Group B versus Group C) it is important to consider first the rate of infection prior to referral to Stanford. Both groups of children averaged slightly more than 2 infections per year during this time. Hence, there was no significant difference in the incidence of infection prior to entry into the study. Also, about 50”/: of the infection episodes in each group were clinically diagnosed as pyelonephritis (fever, with or without chills and back or abdominal pain). Those children with recurrent severe episodes of pyelonephritis, particularly if accompanied by pyelographic changes consistent with this diagnosis and who were not kept sterile with antibacterial medication, were operated upon for correction of the reflux and, hence, were included in Group C. Thus, it appeared that the groups were quite comparable in regard to both the incidence and type of infection prior to entering the study, although those in Group C were more likely to have had recurrent pyelonephritis not readily controlled by medication. The Figure indicates the percentage of patients who were “cured” following each treatment episode in these 42 female children (Group B) with demonstrated vesico-ureteral reflux who
1 I4
BRITISH JOURNAL OF UROLOGY
I
I
I
I
I
I
I
I
All patients followed over 1 year Non-reflux 66 patients 0 Reflux- Medical treatment 42 patients o Reflux- Surgical treatment 61 patients
I
-
10 -
0
1
I
1
2
1
3
1
1
I
5 6 Treatments
4
7
8
9
10
Fig. The recurrence rate of new infection for the 3 groupsofchildrenfollowingeachsu~ccssfiiltrcatrnent I S plotted against the theoretical 20 ,, "cure" rate per treatment (black line).
were followed for at least 1 year after the demonstration of reflux and in whom surgery w a s not performed. As with Group A children, 20 to 25 of those treated for a specific infection did not develop another infection but maintained sterile urine throughout the follow-up period. The Figure also indicates that the 2 patient populations showed a similar response to treatment whether it was the first infection or a subsequent infection.
x,
Children with Vesico-Ureteral Refrux treated surgically Group C consisted of 61 patients followed for at least 12 months after the correction of the reflux. As indicated in the Figure these patients had a very similar postoperative infection rate when compared with the patients without reflux (Group A) and those children in whom the reflux was not surgically corrected (Group B). That is, approximately 20:4 of the patients were cured with each course of short-term medical therapy. An important observation in both groups of refluxing children ( B and C ) was that the incidence of clinical pyelonephritis was approximately 50:';; of all infections in both Groups B and C prior to referral. This was reduced to 10";) of the total number of infections in those surgically treated (Group C ) and less than 20n;,of the infections in those medically treated (Group B). The differences are not statistically significant but indicated a marked decrease in symptomatic pyelonephritis in both populations compared with the rate prior to the initiation of close medical and/or surgical supervision. The children in Groups B and C were not arbitrarily assigned to either a medically or surgicallytreated group. In general, the non-surgical group was comprised of children with lesser degrees of reflux than those children who were treated operatively (Table I). Also, those children whose infections could not be adequately controlled with close medical supervision were considered as surgical candidates. The 2 groups, therefore, are not strictly comparable in the severity of the reflux prior to entering into our study.
Comparison of the Results As shown in the Figure, the results of treatment of infection were essentially the same. In the children without reflux, and in those whose reflux was surgically corrected, the percentage
URINARY TRACT INFECTIONS IN FEMALE CHILDREN
115
experiencing recurrent infections following each episode of treatment paralleled the theoretical 20% “cure” rate. Thus, there appears to be no difference in the recurrence rate of infections in those children with reflux corrected surgically and those treated medically. More importantly, the recurrence rate in both of these groups does not differ significantly from the group of children with UTl’s who do not have ureterovesical reflux. Comments There were several interesting observations apparent during the course of these studies. All patients on the study received 10 days of a specific antibacterial agent at the time the infection was diagnosed; no long-term suppressive medication was given to this group. As shown in the Figure our results in the group of children without reflux confirmed the earlier results of Kunin (1970) which indicate, essentially, that each time we treat a group of children presenting with UTI’s, 20% of the patients will have no further infection but that 80% will develop another infection. If the remaining 80 who develop another infection are again treated and sterilised, another 20% of this population will be “cured” but 80% of this group will still go on to develop a further infection. Prior to referral to our institution over 50% of the infections in the patients comprising both Group Band Group C were associated with fever, with or without flank pain, symptoms clinically suggestive of pyelonephritis (Govan et al., 1974). This is quite in contrast to the benign type of clinical infection experienced in the children of Group A (Fair et al., 1974) in whom a diagnosis o f pyelonephritis was suspected in fewer than 10% of the total number of infections. It should be noted that Groups B and C were not strictly comparable, in that there was a marked difference in the degree of reflux found in the patients comprising the 2 groups. However, the overall infection rate was similar in both populations under the same type of medical management. Of particular interest is the observation that the incidence of clinical pyelonephritis was reduced to less than 20% of all infections in both groups. This compares favourably with an incidence of clinical pyelonephritis of less than 10% of all infections in non-refluxing children (Group A ) and is in agreement with other studies in similar patient populations (Winberg, Larson and Bergstrom, 1970). Winberg et d.(1970) found that in children with reflux, whose infections were under bacteriological control, the incidence of clinical pyelonephritis was similar to that occurring in children without reflux. In addition to the similarities between the 2 groups of patients with reflux as to the type and incidence of infection prior to entry into the study, it is also of note to compare the “cure” rate of these children with each successful treatment. As shown in the Figure both groups had about 20% of the children “cured” following each course of treatment, whether the reflux was surgically corrected or not. Quite significantly, the “cure” rate was similar for the 3 groups of children with or without reflux and did not differ in those children with reflux corrected by surgery plus medication and those treated with medication alone. All 3 groups paralleled the theoretical 20 % “cure” rate previously described by Kunin ( 1970). At this point it is perhaps important to realise in discussing a prevention of recurrences, we are, for the most part, talking about preventing reinfection rather than relapse of an improperlytreated infection. Kunin and Halmagyi (1962) and Bergstrom et a/. (1967) have shown that 85 ?,) of urinary tract infections are new infections caused by a different organism each time. Thus it becomes clear that long periods of treatment are simply preventing reinfections and not eradicating relapses due to inadequate treatment. In summary, from these data one may draw the following conclusions: I . Urinary tract infections are not totally prevented by correction of vesico-ureteral reflux; 2. The rate of reinfections in surgically-treated ver.su.r medically-treated children with reflux is similar; 3. The “cure” rate is similar, when one compares the 2 groups of children with reflux (Groups B and C ) , with the children without vesico-ureteral reflux (Group A).
1 I6
BRITISH JOURNAL 0 1 UROLOCiY
From a prospective point of view it becomes apparent that it would be most helpful t o identify those girls who are likely to develop multiple recurrences and treat those children prophylactically with long-term suppressive therapy. The results of this study, which confirms the earlier work of Kunin (1970), show that medication alone will produce a long-term cure i n 20”,, of young white females whether it is the first infection o r a subsequent infection. Use of long-term SLIPpressive medication such as advocated by Normand and Smellie ( 1965) gave excellent results i n a group of children without reflux similar to ours. The fact that this approach to the problem has been shown to be effective underscores the major problem i n treating these children; that i h ? identifying those girls who are most susceptible to recurrent infections and directing long-term therapy to that group. Since more than SOY{, of the children will have no more than 3 or 4 infections, we suggest that in those female children whose reflux is not deemed serious enough to warrant surgery (Govan et d., 1974) o r who continue to have reinfections following surgery. as well as in those children who d o not have vesicoureteral reflux a t all, it would be appropriatc to begin these patients on long-term suppressive medication after the third o r fourth infection i n an effort to select those children who are most likely to benefit from this form of therapy. Our current practice in such instances of recurrent bacteriuria is initially to sterilise the child with appropriate medication for 10 days and then to place the child immediately on it low dose of nitrofurantoin (usually in the form of Macrodantin 25 or 50 nig) given just before bedtime. This medication is prescribed for 6 months, then discontinued. Should the child develop a further infection, then the whole procedure is repeated. This regimen has been highly successful i n lowering the frequency and morbidity of infections in this small, but difficult population.
Summary 3 groups of female children, comprising a total of 169 patients, were followed through succcssive episodes of urinary tract infection with respect to recurrence of infection. Group A consisted o f 6 6 patients with no evidence of vesico-ureteral reflux. Group B comprised 42 patients with demonstrated vesico-ureteral reflux. The infections i n these patients were treiited with medical therapy alone and the reflux was not corrected. Group C consisted o f 61 children who had surgical correction of vesico-ureteral reflux. While Groups B and C were n o t strictly comparable, the data indicate that there was no demonstrable difference in the rate o f urinury tract infection among the 3 groups. Following short-term specific antibacterial therapy of a new infection, about X ” , 01’ ,childrcn in each group were “cured”, i.c., liud t i o ,furthrr i i l f ~ c t i o n .clirritig ~ tlic 12 niont1i.s or t t i o w tliiit ~ l i i , , i > \i’ori’ .$ uhsc~qlic~tit I!, follo KW/ 6). thc mrtllors. Patients with less severe grades of reflux treated by medication alone ( G r o u p B) experienced no greater rate of reinfection than those children presenting with infections i n thc itbscncc of vcsico-ureteral reflux. In those children whose infections could not be adequately controlled by antiniicrobial therapy. and those patients with severe (Grade I l l ) reflux, surgery was performed to eliminote the vesicoureteral reflux. The data also reinforced the concept that correction of reflux does not eliminate the need for constant vigilance and proper treatment of recurrent infections even after successful surgery has been carried out.
References BI i
