Bacteriology of urinary tract infection in patients of renal failure undergoing dialysisSurangi K Jadav1, Suman M Sant2, Vidya N Acharya1
1 Department of Medicine, Seth G. S. Medical College and K. E. M. Hospital, Parel, Bombay-400012, India
2 Department of Pathology, Seth G. S. Medical College and K. E. M. Hospital, Parel, Bombay-400012, India
In a study of 114 patients who underwent 808 dialyses for renal failure, a high incidence of urinary tract infection (U.T.I.) was observed both in acute renal failure (A.R.F.) group (73.0%) and chronic renal failure (C.R.F.) group (57.5;;) . Although E. coli was the predominating urinary pathogen, the present study revealed that other gram negative micro-organisms besides E. coli were also frequently isolated from the U.T.I. of both A.R.F. and C.R.F. cases. On the whole, both gram negative and gram positive microorganisms exhibited high degree of resistance to antimicrobial agents. Antibiotics like Gentamycin Sulphate (98%), Polymyxin B (88%) and Colimycin Sulphate (86%) exhibited better range of total activity as compared to others.
The life of patients with renal failure could be saved in A.R.F. and prolonged in C.R.F. with the advent of haemo or peritoneal dialysis. However, it is widely recognised that infections at different sites are the most serious hazard and account for an increased proportion of the morbidity and mortality in renal failure ,,, The injudicious use of antibiotics and chemotherapeutic agents, their reduced and decreased dosage in renal failure resulting in the emergence of resistant strains have added to the hazard.
In cases of renal failure, there is a change in the composition of urine with oliguria, anuria, albuminuria and haematuria. The resultant changes in pH, osmolality and urinary urea definitely have their own effects in urinary infection .  Added to this is the danger of urethral catheterisation. 
Because of these serious consequences the prevention of such infections became the cornerstone in the management of renal failure. The present study was undertaken (1) to assess the frequency of U.T.I. in patients of renal failure undergoing dialysis; (2) to isolate the infecting agents and (3) to find out their sensitivity to various anti-microbial agents. The bacteriological study of other sites of infections in human body was also carried out to find out whether these infections can lead to the development of U.T.I. in these patients.
A total of 808 dialyses were performed on 114 cases (41 cases of A.R.F. and 73 cases of C.R.F.) to include 586 haemodialyses and 222 peritoneal dialyses. 9 cases of A.R.F. and 20 cases of C.R.F. were on haemodialyses, 22 cases of A.R.F. and 34 cases of C.R.F. were treated with peritoneal dialyses while the remaining 10 cases of A.R.F. and 19 cases of C.R.F. were treated with both dialyses. Haemodialysis was done by cannulating the peripheral blood vessels (artery and vein) by means of plexitron catheters made out of PVC material in cases of A.R.F, and by the use of Scribner's silastic teflon shunts for the cases of C.R.F. The dialysis through the peritoneal route was done by the technique of alternate puncture in the iliac fossa with the intermittent drainage through a single catheter.
On enquiry about the history, 23 cases stated that they had undergone catheterisation prior to the present admission. Out of these cases, 16 (69.5%) belonged to A.R.F. group and 7 (30.4X) belonged to C.R.F. group. Going through the data on administration of antibiotic therapy before being referred for dialysis it was noted that 28 patients (24.4%) were treated with chloramphenicol, 25 patients (21.9%) were treated with ampicillin, 16 patients (14.02%) were treated with penicillin, 10 patients (8.7%) were treated with erythromycin and streptomycin, 8 (7.01%) patients were treated with kanamycin and 5 patients (4.35%) were treated with gentamycin.
[Table 1] shows the biological specimens obtained from these patients for bacteriological studies.
The colony count of urine samples was determined by dip-slide inoculum method  which was modified in our laboratory.  When the mixed growth was observed, colony count of individual microorganism was determined by pour plate method.  In the present study, the diagnosis of U.T.I. was established by the presence of bacteriuria (10 4 orgs./ml.) in at least three-clean-voided urine collections with the same species observed. This was in view of the history of previous antibiotic therapy in many cases and the fact that uraemic state was associated with immunological aberration and low urinary outputs with the possibility of intermittent bacteriuria. 
All these materials were cultured on blood agar and MacConkey's agar. Saboraud's glucose agar was used for the isolation of candida species. All these media were incubated at 37°C for 24-72 hours. Indentification of micro-organisms was done by using standard biochemical tests.  The antibiotic sensitivity was determined in vitro by impregnated disc method.  The following antimicrobial agents were used in the concentration per disc mentioned against them.  [Table 2].
The positive urine culture was obtained from 30 (73%) out of 41 cases of A.R.F. and from 42 (57.5%) out of 73 cases of C.R.F. However, the difference in the incidences of U.T.I. in these two groups was not statistically significant (X 2 = 2.759; df = 1, n.s). It was noted that 28/% of cases with positive urine cultures had undergone catheterisation.
The different types of micro-organisms isolated from urine cultures in A.R.F. and C.R.F. cases are shown in [Table 3]. Gram negative bacteria were more frequently isolated than Gram positive micro-organisms. E. coli was the predominant urinary pathogen in cases of C.R.F. while Klebsiella was the predominant pathogen in A.R.F. cases. The statistical analysis, however, shows that the types of micro-organisms isolated from both the groups are not significantly different from each other (p>0.05). It is also seen that other gram negative micro-organisms besides E. coli were also frequently isolated from both A.R.F. and C.R.F. cases.
The number of mixed isolates from urine of renal failure cases are shown in [Table 4]. Most of the strains of E. coli, Klebsiella and Enterobacter groups were isolated in pure cultures while those of Proteus and Pseudomonas were more frequently found from mixed cultures and also from the cases who were repeatedly catheterised.
Out of 22 cases who had low colony counts (10 4 ), 10 had mixed infections, 5 cases were of chronic pyelonephritis and 7 patients had U.T.I. due to gram positive micro-organisms.
The antibiotic sensitivity of microorganisms isolated from U.T.I. in A.R.F. and C.R.F. is shown in [Table 5]. On comparing the overall sensitivity of all organisms with the individual sensitivity of Klebsiella, E. coli and the group of the remaining organisms there was no statistically significant difference in activity for any antibiotics in two groups of A.R.F. and C.R.F. except for chloramphenicol (p<0.05). The overall sensitivity to chloramphenicol was better for C.R.F. (20.8%) than A.R.F. (5.3% ).
The anti-bacterial agents could be grouped into 4 distinct categories depending on the extent of antibacterial activity pattern. The first group consisted of Sulphonamides, Penicillin, Erythromycin, Chloramphenicol and Ampicillin which had activity below 25%. The second group consisted of Oxytetracyclin, Nitrofurantoin and Carbenicillin having activity range of 25% to 50%. The third group comprised of Trimethoprim + Sulphamethoxazole, Mandelamine, Streptomycin and Cephaloridine having activity range of 51% to 75% while the fourth group of Kanamycin, Colimycin, Polymyxin and Gentamycin had 76% to 100 % activity. The analysis of variance indicates that the different antibiotics which manifested overall activity of more than 25% (Group 2, 3, 4) have different activity pattern (p<0.001). The organisms like E. coli, Klebsiella and a group of other organisms also differ in their sensitivity (p<0.01). The activity of antibiotics to the three groups of micro-organisms differ significantly (p<0.05).
Relationship between Infections at Different Sites with Respect to Micro-Organisms
It was possible to study the interrelationship of infections at different sites in 37 patients (32.5%). In 20 (53.3%) of these, the same micro-organisms were isolated from different infectious sites [Table 6].
It is well known that `conservative' management suffices for the majority of patients with acute urinary suppression, however, the dialysis methods are valuable adjuncts where the suppression is persistent. The overwhelming and fulminent infections developing during the management create a major problem in the treatment of patients of renal failure while undergoing dialysis. The present paper deals with the study of bacteriology of U.T.I. in 114 patients of renal failure undergoing dialysis and also makes observations on the relationship of U.T.I. with infections at different sites. As these patients are more vulnerable to infections, the question of the magnitude of risk of renal failure among patients with bacteriuria is a complicated one. However, Kaye  has shown that this risk exists only in patients who have bilateral renal infection since only one normal kidney is enough to sustain health.
In the present study, 30 out of 41 patients (73%) with A.R.F. and 42 out of 73 patients (57.5%) had urinary infection. The frequency of U.T.I. has been reported to be 33% to 89% among the patients with renal failure. ,,,, In the present study, 28% of the patients with U.T.I. had been subjected to at least one bladder catheterisation procedure during the acute phase of the disease and mere avoidance of this procedure could have reduced the incidence of U.T.I. to a considerable extent.
In the current series, 22 out of 72 cases (30.5%) of positive urine cultures had low colony count (10 4 ). Antibacterial therapy and diuretic drugs, in massive dosages converting oliguric renal failure phase into non-oliguric phase, may be the key factors accounting for the relatively low range of colony count. It is interesting to note that some of low colony counts seen in the present study were in cases of chronic pyelonephritis and also in the specimens where the infections were with gram positive cocci. In such situation low counts need to be considered significant and treated to prevent morbidity due to U.T.I.
The observations on incidence of different micro-organisms and their sensitivity pattern in patients with recurrent U.T.I without renal failure have been reported by many workers. ,,,,,,, However, there is no data available on bacteriuria in a population with renal insufficiency comparable with the present findings. It was observed that the urinary infection with multiple organisms was about twice more frequent in chronic U.T.I. cases (14.2% ' ) as compared to acute ones (7.2%) in patients with no evidence of renal failure.  In contrast to this, in the present study, 30% of the cases with A.R.F. and 24%, of patients with C.R.F. had mixed infection.
It was interesting to observe one patient who had primary U.T.I. due to Klebsiella and Candida albicans. This bacterial and fungal infection disappeared without any specific antifungal treatment when it was replaced by infections with Ps. aeruginosa. The patient also developed overwhelming Pseudomonas sepsis. It has been shown that Candida albicans growth is inhibited by gram negative bacteria. ,
In our experience of 37 cases, 20 cases (53.3%) showed the same microorganisms from infections at different sites like urine, pus from the wounds, sputum, vaginal swab, ear swab, peritoneal fluid, blood and pericardical fluid. This observation was noted by Leigh  in peritoneal sepsis and U.T.I. Montgomerei et al  noted this in cases of septicaemia and U.T.I. where as Balslov and Jorgensen  made similar observations in wound 'infection and U.T.I. Thus it was evident that there may be an inter-relationship between infections at different sites developing during renal failure. However, the cause and effect relationship between them would be difficult to surmise, except in areas of close proximity like vagina and anal canal where infections from these two areas are likely to travel up and lead to U.T.I. On the other hand, peritoneal wound and cavity could get contaminated by the infected urine leading to sepsis due to common organisms producing U.T.I. Wounds at the cannulation sites of blood vessels could lead to U.T.I. by blood stream spread. This brings into focus the need of scrupulous and frequent cleaning of all wounds, while managing patients of renal failure. The low urine output itself precludes frequent bladder emptying in an azotaemic patient making the urinary tract vulnerable to bacterial invasion. Hence frequent examination of urine for bacteriuria would help early detection of U.T.I.-a valuable step in reducing the morbidity by instituting appropriate and very early therapy.
The data of antibiotic sensitivity pattern has revealed a close relationship with their prophylactic usage. The more frequently used antibiotics like Penicillin, Erythromycin, Chloramphenicol, and Ampicillin revealed very low levels of sensitivity (<25%) to all organisms in a group as a whole. The avoidance of prophylactic usage of antibiotics may help surmount this to a certain extent and would also be a logical step to avoid the threat of R.-factor resistance. 
We are most thankful to the nursing staff of Artifical Kidney Department of the K.E.M. Hospital, Bombay-400 012, for their ready assistance in collection of the pathological samples. Our sincere thanks are due to Mrs. K. D. Lotlikar for carrying out the statistical analysis of the present data. For the facilities and also for giving us permission, we thank Dr. C. K. Deshpande, the Dean of the K.E.M. Hospital.
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5], [Table 6]