- Email: [email protected]
Urinary Tract Infection in Percutaneous Surgery for Renal Calculi
0022-t,347/86/1351-0015$02.0D/O 'THE JOURNAL OF URCLOG'Y
Copyright© 1986 by The 'lVilliaz:Gs 8r 'llilkins Co,
URINARY R VEILLON
M. CHARTON, G.
AND
J.M. BRISSET
From the Department of Uro-Nephrology, C, M, C, Porte de Choisy, Paris, France
ABSTRACT
Percutaneous extraction of renal stones is associated with a risk of infection, which sometimes can be severe as a result of the intraoperative introduction of a ureteral catheter, the itself and the fact that a nephrostomy tube sometimes is left in place. It generally is accepted that patients with a preoperative urinary tract infection should be covered during the operation by an appropriate antibiotic. However, the need for routine prophylactic antibiotic treatment in ,,~,u,v,,.vu with sterile urine preoperatively still is a subject of debate. We report the bacteriological results of 126 cases of percutaneous extraction of renal stones. Of the patients 107 had sterile urine preoperatively and deliberately did not receive prophylactic antibiotics so that the mechanisms of urinary tract infection after percutaneous nephrolithotomy could be studied. Of these patients 37 (35 per cent) suffered a postoperative urinary tract infection, usually owing to Escherichia coli, streptococcus or staphylococcus. The responsible organism was isolated in the bladder urine only in 22 cases, in the nephrostomy tube in 2 and in both sites in 13. Eleven patients (10 per cent) presented with a fever of or more. All of the infected patients received appropriate antibiotic therapy and there were 2 ,v,,vh'""'' failures on long-term followup (5 per cent). A total of 19 patients had a tract infection appropriate antibiotic therapy at I.east 24 hours for a minimum of 3 weeks. Five patients (26 per cent) presented a fever but there were no serious septic complications. All of the patients were discharged from the hospital with sterile urine and there was only 1 long-term bacteriological failure (5 per cent). Both with Pseudomonas infection were cured. The risk of clinical infection following 'nn~m,t-,'!.~,,u~,u the fact that 35 per cent of the patients have bacteriuria postoperatively, rrnrmorl<>rl a bacteriological examination is performed preoperatively and the patients with urinary tract infection are treated appropriately_ These results are in favor of short-term antibiotics adapted to the bacterial ecology. ?YOLlrnC.<'.>'C'>-,
Among the new techniques available to the urologist, percutaneous nephrolithotomy is a method in full expansion. 1- 6 Although the various surgical complications currently are well recognized, the infectious risks of an operation that involves the percutaneous insertion of material (nephroscopy and sometimes a nephrostomy often associated with manipulation Of the lOVVef tract C'mff(),Qf'rffm 1nt1•,>,C>n,P1"SiJ-1,7p H1Sertion of a retrograde ureteral <-a,,11tct,
an,,c,cnc•i-,
of residual stone following the first of 126 operations was performed. "'"'''""''"'"ci patients 65 per cent were men and 35 cent were women, with a mean age of 45 years. A ~"'~""=" or nephrological disease was noted in 78 µm,,,.,,,010, per cent of the men and 66 per cent of had tract infection in the past. A total of 39 ne, 1ha,nto underthe calculi. upon ,,v,w_c"""'U of a 1-stage dilation and extraction with the anesthesia, 6 A ureteral catheter was inserted in ""';c110,,m, at the u~,,,rn,ni,,i; operation. After tube was left in place in 48 patients urinary tract infection) depending on the local The nephrostomy tube was left in for an average of 3 days (range 1 to 7 The mean dimensions of the were 15 x 13 mm., and 53 per cent were located in the renal pelvis, 15 per cent in a renal calix, 7 per cent in the ureter and 25 per cent in multiple sites. Bacteriology. All of the patients had a preoperative cytobacteriological examination of the urine. A further examination was performed 2 days postoperatively on the bladder and nephrostomy urine, when a nephrostomy tube was present. Additional urine cultures were performed depending on the duration of the hospital stay. A bacteriological examination was
results of 126 cases of and attempt to pathway and implications of bacterial contamination of the urine in 107 patients who had sterile urine risks Although this type of operation has a inadequate asepsis, we deliberately chose not to administer any antibiotic prophylaxis to these patients to study accurately the bacterial ecology of the urine following percutaneous nephrolithotomy. We adopted this approach to define the most appropriate antibiotic prophylaxis for the prevention of infections occurring after this type of operation. .v,v~·~~·
PATIENTS AND METHODS
Between May 1983 and May 1984, 121 patients underwent percutaneous extraction of renal calculi. Of these patients 5 were operated on twice: 2 had a bilateral operation for calculi and 3 underwent a second operation on the same side because Accepted for publication August 23, 1985. 15
16
CHARTON AND ASSOCIATES
requested 1 month after the patient was discharged from the hospital for noninfected patients and 1 month after the end of treatment for those with urinary tract infection. We considered the bacteriuria to be significant when the bacterial count was 105 or more organisms per ml. urine. Blood cultures were performed when indicated by clinical context (fever and rigors). A total of 107 patients (75 men and 32 women) had sterile urine preoperatively and did not receive any prophylactic antibiotic therapy, since the aim of this study was to define the modalities of urinary tract infection after percutaneous nephrolithotomy. The patients who did not suffer an infection postoperatively did not receive any antibiotic treatment after they were discharged from the hospital. Patients with an infection were treated according to the urine culture and antibiotic sensitivity results for 20 days. Of the patients 7 men and 12 women had infected urine preoperatively, and were treated preoperatively and postoperatively according to the results of the antibiotic sensitivity tests. RESULTS
Patients with sterile urine preoperatiuely. There were 107 operations performed in patients with sterile urine preoperatively and 105 cases were interpretable bacteriologically. The patients with a positive cytobacteriological examination during the hospital stay were considered to have urinary tract infection. In 87 per cent of the patients the infection was diagnosed 2 days postoperatively and in 13 per cent there was a delay in the diagnosis. Four patients required secondary insertion of a ureteral catheter because of urine leak: 3 were already infected following the percutaneous extraction and 1 was infected following insertion of the catheter. Of 64 patients who did not have a nephrostomy tube 14 (22 per cent) had a postoperative urinary tract infection. In 41 patients the nephrostomy tube was maintained for 1 to 7 days and 23 (56 per cent) suffered infected bladder or nephrostomy urine. Over-all, 37 patients (35 per cent) suffered postoperative urinary tract infection. The infection often was detected in the bladder urine, and the predominant organisms were Escherichia coli and group D streptococci (table 1). Bladder origin of the postoperative infection also was observed in the patients with a nephrostomy tube, with an equal incidence regardless of the duration of the nephrostomy. The presence of a nephrostomy tube was associated with an increased incidence of urinary tract infection (23 of 41 patients or 56 per cent, versus 14 of 64 or 22 per cent, which is statistically significant at p <0.001). However, these results should be analyzed in detail: 8 of the 23 patients had infection of the bladder urine only and 2 had infection of the nephrostomy urine only. Of the 13 patients with bipolar infection 8 were infected with identical organisms (impossible to define the primary site of infection) and 5 had different organisms. Therefore, the nephrostomy was definitely responsible for only 7 infections among the 41 patients with a nephrostomy tube in situ (17 per cent). TABLE
1. Organisms detected after percutaneous nephrolithotomy in
patients with sterile urine preoperatively Site of Pos. Culture Bladder Nephrostomy Po st0P· No. Pts. No. Pts. Ureteral Catheter No. Pts. E.coli Streptococcus D Staphylococcus Streptococcus Klebsiella Acinetobacter Pseudomonas aeruginosa plus Streptococcus D E. coli plus Streptococcus D Staphylococcus plus Streptococcus D Totals
10 10
2 1
8
7
3
1
1 1
2
2
35
15
1
TABLE 2.
Postoperative infection in patients with sterile urine before nephrolithotomy and the course under treatment Bacteriology Sterile preop. Infected postop. Antibiotics effective Antibiotics failed Lost to followup
TABLE 3.
No. Pts. (%) 105 37 27 2 8
(35) (73) (5) (22)
Infected urine preoperatively
Bacteria E.coli Proteus Streptococcus D Providencia Citrobacter Pseudomonas aeruginosa Pseudomonas plus streptococcus plus staphylococcus
No. Pts. 9
4 2
1 1
1 1
With the same criteria of selection for the 105 patients who had intraoperative insertion of a retrograde ureteral catheter, we found that only 2 had infected nephrostomy urine and in 35 cases the bacteria were isolated in the bladder (22 patients had infected bladder urine only and 13 had bipolar infection, with an identical organism in 8 and different organisms in 5). The insertion of the retrograde ureteral catheter can be considered to be definitely responsible for the postoperative infection in 27 of 105 cases (26 per cent). Finally, in the 41 patients with a nephrostomy the urine obtained from this route was infected in 15 (36 per cent), while among the entire series the bladder urine was infected in 35 (33 per cent). We also demonstrated a statistically significant difference in the type of bacteria isolated after nephrolithotomy in men and women: there were more E. coli infections in women (8 of 15 versus 2 of 30, p <0.05, Yates' correction), and more streptococcal and staphylococcal infections in men (16 of 20 versus 6 of 15, p <0.02) with a higher incidence of infection in women (47 per cent) than in men (30 per cent) (pis not significant). Clinically, 11 patients (10 per cent) had a fever of more than 38.5C postoperatively. There were no cases of septic shock and all of the blood cultures performed remained sterile. Of the 37 patients with urinary tract infection 8 (22 per cent) were lost to followup after 1 month. Among the remaining patients we obtained 27 bacteriological cures (73 per cent) and 2 failures (5 per cent): 1 patient with E. coli infection and a residual stone, and 1 patient with bipolar Staphylococcal infection (bladder plus nephrostomy tube) (table 2). Of the 68 patients who were discharged from the hospital with sterile urine only 1 had a urinary tract infection at followup 1 month later. Patients with urinary tract infection preoperatiuely. There were 19 patients infected preoperatively, 14 of whom had definite chronic infection that had been detected on several occasions, suggesting an infected stone. Table 3 shows the bacteriological results of the preoperative urine specimens. In all of the patients appropriate antibiotic therapy was commenced at least 24 hours preoperatively and was continued for a minimum of 3 weeks postoperatively. Of the patients 5 suffered fever postoperatively. Only 1 of these patients had a nephrostomy tube. However, 3 of the patients had a residual calculus. There were no severe septic complications and the blood cultures of the 5 febrile patients remained sterile. All 4 patients (21 per cent) who were lost to followup at 1 month were discharged from the hospital with sterile urine under treatment. Among the remaining patients 14 (74 per cent) were cured of the infection, including 2 with Pseudomonas
URINARY TRACT INFECTION IN PERCUTANEOUS NEPHROLITHOTOMY
aeruginosa infection, and there was only 1 failure (5 per cent) (E. coli infection and residual stone). Of the total of 56 infections detected there were only 2 bacteriological failures of treatment (4 per cent) and no signs of clinical severity of these infections, whether they were acquired as a result of surgery or of operating on a primarily infected patient. DISCUSSION
The extraction of renal calculi raises 2 bacteriological issues: 1) what is the risk of infection with this type of operation in patients without preoperative urinary tract infection and 2) is it possible to operate safely on patients with renal calculi associated with urinary tract infection? Of the 105 patients with sterile urine preoperatively and in the absence of any prophylactic antibiotic therapy we detected 37 cases of urinary tract infection (35 per cent). Patients with a nephrostomy tube suffered an infection more often than those without a nephrostomy tube (p <0.001). However, the site of infection in both groups generally was the bladder: 35 bladder infections over-all (33 per cent) and 15 cases of nephrostomy infection in the 41 patients with a nephrostomy tube (36 per cent). The most common organisms were E. coli, which was isolated more frequently in women, and group D streptococci, which was isolated more frequently in men. These organisms almost always were isolated in the bladder. This is an additional ecological argument in favor of the hypothesis that these infections are introduced mostly by the retrograde route during cystoscopy and insertion of the ureteral catheter.. Similar bacteriological results also have been obtained in bacteriological studies of endoscopic prostatic surgery. 7 Unfortunately, we have not found any confirmation of these findings in the literature, since to our knowledge all teams working in this field seem to use routine prophylactic antibiotic therapy. 1-4 This approach is open to discussion. Antibiotic therapy often is long-term, which carries a certain risk of selection of resistant organisms. It might be more logical to administer a single preoperative dose of an appropriate antibiotic that has been shown already to be effective in the prevention of urinary tract infections after transurethral resection of the prostate. 7 A randomized study of this type currently is underway in our department in a series of 100 patients. In any case, these infections never cause serious postoperative complications and in our series there were only 2 treatment failures in cases of proved infection (5 per cent), including 1 patient who had a residual stone, which seems to be as good a result as those obtained with prophylactic antibiotic therapy. We operated on 19 patients with renal stones associated with urinary tract infection. Provided that an antibiotic protocol
17
adapted to each organism is prescribed and that a nephrostomy tube is left in place routinely, we did not observe any severe septic complications. We only had 1 case oflong-term bacteriological failure; again, the patient had a residual stone. On the other hand, the 2 patients with Pseudomonas aeruginosa infection have sterile urine, with a followup of 4 months in 1 and 8 months in the other. Therefore, this technique is suitable for cases of infected stones as already has been reported. 4 • 6 However, we believe that it is important to ensure the patency of the excretory tract in these patients and to remember that bacteriological cure can be obtained only when the extraction is complete. CONCLUSION
Percutaneous nephrolithotomy is a clean but not sterile surgical technique. In patients without infection preoperatively and in the absence of prophylactic antibiotics, the procedure essentially responsible for introducing infection appears to be the intraoperative insertion of the ureteral catheter. The percentage of postoperative urinary tract infections in these patients was relatively high but with only moderate clinical consequences and a satisfactory response to treatment. As the ecology of these infections currently has been defined, it would seem logical to propose short-term antibiotic prophylaxis for this type of operation. The effective treatment of patients with urinary tract infection associated with renal stones preoperatively must include appropriate curative antibiotic therapy, a patent excretory pathway, and complete removal of all stones and stone fragments. REFERENCES
1. Alken, P., Hutschenreiter, G. and Gunther R.: Percutaneous kidney
stone removal. Eur. Urol., 8: 304, 1982. 2. Clayman, R. V., Surya, V., Miller, R. P., Castaneda-Zuniga, W. R., Smith, A. D., Hunter, D. H., Amplatz, K. and Lange, P. H.: Percutaneous nephrolithotomy: extraction of renal and ureteral calculi from 100 patients. J. Urol., 131: 868, 1984. 3. LeRoy, A. J., May, G. R., Segura, J. W. and Patterson, D. E.: Percutaneous ultrasonic lithotripsy. Rad. Clin. N. Amer., 22: 427, 1984. 4. Marberger, M., Stackl, W. and Hruby, M.: Percutaneous litholapaxy of renal calculi with ultrasound. Eur. Urol., 8: 236, 1982. 5. Vallancien, G., Capdeville, R., Charton, M., Veillon, B. and Brisset, J. M.: Ablation percutanee des calculs renaux. Presse Med., 12: 2997, 1983. 6. Smith, A. D. and Lee, W. J.: Percutaneous stone removal procedures including irrigation. Urol. Clin. N. Amer., 4: 719, 1983. 7. Charton, M., Dosne, B., Escovar, P., Kopf, A. and Brisset, J. M.: Traitement prophylactique minute des infections urinaires apres resection endoscopique de la prostate. Presse Med., 13: 545, 1984.
Copyright© 1986 by The 'lVilliaz:Gs 8r 'llilkins Co,
URINARY R VEILLON
M. CHARTON, G.
AND
J.M. BRISSET
From the Department of Uro-Nephrology, C, M, C, Porte de Choisy, Paris, France
ABSTRACT
Percutaneous extraction of renal stones is associated with a risk of infection, which sometimes can be severe as a result of the intraoperative introduction of a ureteral catheter, the itself and the fact that a nephrostomy tube sometimes is left in place. It generally is accepted that patients with a preoperative urinary tract infection should be covered during the operation by an appropriate antibiotic. However, the need for routine prophylactic antibiotic treatment in ,,~,u,v,,.vu with sterile urine preoperatively still is a subject of debate. We report the bacteriological results of 126 cases of percutaneous extraction of renal stones. Of the patients 107 had sterile urine preoperatively and deliberately did not receive prophylactic antibiotics so that the mechanisms of urinary tract infection after percutaneous nephrolithotomy could be studied. Of these patients 37 (35 per cent) suffered a postoperative urinary tract infection, usually owing to Escherichia coli, streptococcus or staphylococcus. The responsible organism was isolated in the bladder urine only in 22 cases, in the nephrostomy tube in 2 and in both sites in 13. Eleven patients (10 per cent) presented with a fever of or more. All of the infected patients received appropriate antibiotic therapy and there were 2 ,v,,vh'""'' failures on long-term followup (5 per cent). A total of 19 patients had a tract infection appropriate antibiotic therapy at I.east 24 hours for a minimum of 3 weeks. Five patients (26 per cent) presented a fever but there were no serious septic complications. All of the patients were discharged from the hospital with sterile urine and there was only 1 long-term bacteriological failure (5 per cent). Both with Pseudomonas infection were cured. The risk of clinical infection following 'nn~m,t-,'!.~,,u~,u the fact that 35 per cent of the patients have bacteriuria postoperatively, rrnrmorl<>rl a bacteriological examination is performed preoperatively and the patients with urinary tract infection are treated appropriately_ These results are in favor of short-term antibiotics adapted to the bacterial ecology. ?YOLlrnC.<'.>'C'>-,
Among the new techniques available to the urologist, percutaneous nephrolithotomy is a method in full expansion. 1- 6 Although the various surgical complications currently are well recognized, the infectious risks of an operation that involves the percutaneous insertion of material (nephroscopy and sometimes a nephrostomy often associated with manipulation Of the lOVVef tract C'mff(),Qf'rffm 1nt1•,>,C>n,P1"SiJ-1,7p H1Sertion of a retrograde ureteral <-a,,11tct,
an,,c,cnc•i-,
of residual stone following the first of 126 operations was performed. "'"'''""''"'"ci patients 65 per cent were men and 35 cent were women, with a mean age of 45 years. A ~"'~""=" or nephrological disease was noted in 78 µm,,,.,,,010, per cent of the men and 66 per cent of had tract infection in the past. A total of 39 ne, 1ha,nto underthe calculi. upon ,,v,w_c"""'U of a 1-stage dilation and extraction with the anesthesia, 6 A ureteral catheter was inserted in ""';c110,,m, at the u~,,,rn,ni,,i; operation. After tube was left in place in 48 patients urinary tract infection) depending on the local The nephrostomy tube was left in for an average of 3 days (range 1 to 7 The mean dimensions of the were 15 x 13 mm., and 53 per cent were located in the renal pelvis, 15 per cent in a renal calix, 7 per cent in the ureter and 25 per cent in multiple sites. Bacteriology. All of the patients had a preoperative cytobacteriological examination of the urine. A further examination was performed 2 days postoperatively on the bladder and nephrostomy urine, when a nephrostomy tube was present. Additional urine cultures were performed depending on the duration of the hospital stay. A bacteriological examination was
results of 126 cases of and attempt to pathway and implications of bacterial contamination of the urine in 107 patients who had sterile urine risks Although this type of operation has a inadequate asepsis, we deliberately chose not to administer any antibiotic prophylaxis to these patients to study accurately the bacterial ecology of the urine following percutaneous nephrolithotomy. We adopted this approach to define the most appropriate antibiotic prophylaxis for the prevention of infections occurring after this type of operation. .v,v~·~~·
PATIENTS AND METHODS
Between May 1983 and May 1984, 121 patients underwent percutaneous extraction of renal calculi. Of these patients 5 were operated on twice: 2 had a bilateral operation for calculi and 3 underwent a second operation on the same side because Accepted for publication August 23, 1985. 15
16
CHARTON AND ASSOCIATES
requested 1 month after the patient was discharged from the hospital for noninfected patients and 1 month after the end of treatment for those with urinary tract infection. We considered the bacteriuria to be significant when the bacterial count was 105 or more organisms per ml. urine. Blood cultures were performed when indicated by clinical context (fever and rigors). A total of 107 patients (75 men and 32 women) had sterile urine preoperatively and did not receive any prophylactic antibiotic therapy, since the aim of this study was to define the modalities of urinary tract infection after percutaneous nephrolithotomy. The patients who did not suffer an infection postoperatively did not receive any antibiotic treatment after they were discharged from the hospital. Patients with an infection were treated according to the urine culture and antibiotic sensitivity results for 20 days. Of the patients 7 men and 12 women had infected urine preoperatively, and were treated preoperatively and postoperatively according to the results of the antibiotic sensitivity tests. RESULTS
Patients with sterile urine preoperatiuely. There were 107 operations performed in patients with sterile urine preoperatively and 105 cases were interpretable bacteriologically. The patients with a positive cytobacteriological examination during the hospital stay were considered to have urinary tract infection. In 87 per cent of the patients the infection was diagnosed 2 days postoperatively and in 13 per cent there was a delay in the diagnosis. Four patients required secondary insertion of a ureteral catheter because of urine leak: 3 were already infected following the percutaneous extraction and 1 was infected following insertion of the catheter. Of 64 patients who did not have a nephrostomy tube 14 (22 per cent) had a postoperative urinary tract infection. In 41 patients the nephrostomy tube was maintained for 1 to 7 days and 23 (56 per cent) suffered infected bladder or nephrostomy urine. Over-all, 37 patients (35 per cent) suffered postoperative urinary tract infection. The infection often was detected in the bladder urine, and the predominant organisms were Escherichia coli and group D streptococci (table 1). Bladder origin of the postoperative infection also was observed in the patients with a nephrostomy tube, with an equal incidence regardless of the duration of the nephrostomy. The presence of a nephrostomy tube was associated with an increased incidence of urinary tract infection (23 of 41 patients or 56 per cent, versus 14 of 64 or 22 per cent, which is statistically significant at p <0.001). However, these results should be analyzed in detail: 8 of the 23 patients had infection of the bladder urine only and 2 had infection of the nephrostomy urine only. Of the 13 patients with bipolar infection 8 were infected with identical organisms (impossible to define the primary site of infection) and 5 had different organisms. Therefore, the nephrostomy was definitely responsible for only 7 infections among the 41 patients with a nephrostomy tube in situ (17 per cent). TABLE
1. Organisms detected after percutaneous nephrolithotomy in
patients with sterile urine preoperatively Site of Pos. Culture Bladder Nephrostomy Po st0P· No. Pts. No. Pts. Ureteral Catheter No. Pts. E.coli Streptococcus D Staphylococcus Streptococcus Klebsiella Acinetobacter Pseudomonas aeruginosa plus Streptococcus D E. coli plus Streptococcus D Staphylococcus plus Streptococcus D Totals
10 10
2 1
8
7
3
1
1 1
2
2
35
15
1
TABLE 2.
Postoperative infection in patients with sterile urine before nephrolithotomy and the course under treatment Bacteriology Sterile preop. Infected postop. Antibiotics effective Antibiotics failed Lost to followup
TABLE 3.
No. Pts. (%) 105 37 27 2 8
(35) (73) (5) (22)
Infected urine preoperatively
Bacteria E.coli Proteus Streptococcus D Providencia Citrobacter Pseudomonas aeruginosa Pseudomonas plus streptococcus plus staphylococcus
No. Pts. 9
4 2
1 1
1 1
With the same criteria of selection for the 105 patients who had intraoperative insertion of a retrograde ureteral catheter, we found that only 2 had infected nephrostomy urine and in 35 cases the bacteria were isolated in the bladder (22 patients had infected bladder urine only and 13 had bipolar infection, with an identical organism in 8 and different organisms in 5). The insertion of the retrograde ureteral catheter can be considered to be definitely responsible for the postoperative infection in 27 of 105 cases (26 per cent). Finally, in the 41 patients with a nephrostomy the urine obtained from this route was infected in 15 (36 per cent), while among the entire series the bladder urine was infected in 35 (33 per cent). We also demonstrated a statistically significant difference in the type of bacteria isolated after nephrolithotomy in men and women: there were more E. coli infections in women (8 of 15 versus 2 of 30, p <0.05, Yates' correction), and more streptococcal and staphylococcal infections in men (16 of 20 versus 6 of 15, p <0.02) with a higher incidence of infection in women (47 per cent) than in men (30 per cent) (pis not significant). Clinically, 11 patients (10 per cent) had a fever of more than 38.5C postoperatively. There were no cases of septic shock and all of the blood cultures performed remained sterile. Of the 37 patients with urinary tract infection 8 (22 per cent) were lost to followup after 1 month. Among the remaining patients we obtained 27 bacteriological cures (73 per cent) and 2 failures (5 per cent): 1 patient with E. coli infection and a residual stone, and 1 patient with bipolar Staphylococcal infection (bladder plus nephrostomy tube) (table 2). Of the 68 patients who were discharged from the hospital with sterile urine only 1 had a urinary tract infection at followup 1 month later. Patients with urinary tract infection preoperatiuely. There were 19 patients infected preoperatively, 14 of whom had definite chronic infection that had been detected on several occasions, suggesting an infected stone. Table 3 shows the bacteriological results of the preoperative urine specimens. In all of the patients appropriate antibiotic therapy was commenced at least 24 hours preoperatively and was continued for a minimum of 3 weeks postoperatively. Of the patients 5 suffered fever postoperatively. Only 1 of these patients had a nephrostomy tube. However, 3 of the patients had a residual calculus. There were no severe septic complications and the blood cultures of the 5 febrile patients remained sterile. All 4 patients (21 per cent) who were lost to followup at 1 month were discharged from the hospital with sterile urine under treatment. Among the remaining patients 14 (74 per cent) were cured of the infection, including 2 with Pseudomonas
URINARY TRACT INFECTION IN PERCUTANEOUS NEPHROLITHOTOMY
aeruginosa infection, and there was only 1 failure (5 per cent) (E. coli infection and residual stone). Of the total of 56 infections detected there were only 2 bacteriological failures of treatment (4 per cent) and no signs of clinical severity of these infections, whether they were acquired as a result of surgery or of operating on a primarily infected patient. DISCUSSION
The extraction of renal calculi raises 2 bacteriological issues: 1) what is the risk of infection with this type of operation in patients without preoperative urinary tract infection and 2) is it possible to operate safely on patients with renal calculi associated with urinary tract infection? Of the 105 patients with sterile urine preoperatively and in the absence of any prophylactic antibiotic therapy we detected 37 cases of urinary tract infection (35 per cent). Patients with a nephrostomy tube suffered an infection more often than those without a nephrostomy tube (p <0.001). However, the site of infection in both groups generally was the bladder: 35 bladder infections over-all (33 per cent) and 15 cases of nephrostomy infection in the 41 patients with a nephrostomy tube (36 per cent). The most common organisms were E. coli, which was isolated more frequently in women, and group D streptococci, which was isolated more frequently in men. These organisms almost always were isolated in the bladder. This is an additional ecological argument in favor of the hypothesis that these infections are introduced mostly by the retrograde route during cystoscopy and insertion of the ureteral catheter.. Similar bacteriological results also have been obtained in bacteriological studies of endoscopic prostatic surgery. 7 Unfortunately, we have not found any confirmation of these findings in the literature, since to our knowledge all teams working in this field seem to use routine prophylactic antibiotic therapy. 1-4 This approach is open to discussion. Antibiotic therapy often is long-term, which carries a certain risk of selection of resistant organisms. It might be more logical to administer a single preoperative dose of an appropriate antibiotic that has been shown already to be effective in the prevention of urinary tract infections after transurethral resection of the prostate. 7 A randomized study of this type currently is underway in our department in a series of 100 patients. In any case, these infections never cause serious postoperative complications and in our series there were only 2 treatment failures in cases of proved infection (5 per cent), including 1 patient who had a residual stone, which seems to be as good a result as those obtained with prophylactic antibiotic therapy. We operated on 19 patients with renal stones associated with urinary tract infection. Provided that an antibiotic protocol
17
adapted to each organism is prescribed and that a nephrostomy tube is left in place routinely, we did not observe any severe septic complications. We only had 1 case oflong-term bacteriological failure; again, the patient had a residual stone. On the other hand, the 2 patients with Pseudomonas aeruginosa infection have sterile urine, with a followup of 4 months in 1 and 8 months in the other. Therefore, this technique is suitable for cases of infected stones as already has been reported. 4 • 6 However, we believe that it is important to ensure the patency of the excretory tract in these patients and to remember that bacteriological cure can be obtained only when the extraction is complete. CONCLUSION
Percutaneous nephrolithotomy is a clean but not sterile surgical technique. In patients without infection preoperatively and in the absence of prophylactic antibiotics, the procedure essentially responsible for introducing infection appears to be the intraoperative insertion of the ureteral catheter. The percentage of postoperative urinary tract infections in these patients was relatively high but with only moderate clinical consequences and a satisfactory response to treatment. As the ecology of these infections currently has been defined, it would seem logical to propose short-term antibiotic prophylaxis for this type of operation. The effective treatment of patients with urinary tract infection associated with renal stones preoperatively must include appropriate curative antibiotic therapy, a patent excretory pathway, and complete removal of all stones and stone fragments. REFERENCES
1. Alken, P., Hutschenreiter, G. and Gunther R.: Percutaneous kidney
stone removal. Eur. Urol., 8: 304, 1982. 2. Clayman, R. V., Surya, V., Miller, R. P., Castaneda-Zuniga, W. R., Smith, A. D., Hunter, D. H., Amplatz, K. and Lange, P. H.: Percutaneous nephrolithotomy: extraction of renal and ureteral calculi from 100 patients. J. Urol., 131: 868, 1984. 3. LeRoy, A. J., May, G. R., Segura, J. W. and Patterson, D. E.: Percutaneous ultrasonic lithotripsy. Rad. Clin. N. Amer., 22: 427, 1984. 4. Marberger, M., Stackl, W. and Hruby, M.: Percutaneous litholapaxy of renal calculi with ultrasound. Eur. Urol., 8: 236, 1982. 5. Vallancien, G., Capdeville, R., Charton, M., Veillon, B. and Brisset, J. M.: Ablation percutanee des calculs renaux. Presse Med., 12: 2997, 1983. 6. Smith, A. D. and Lee, W. J.: Percutaneous stone removal procedures including irrigation. Urol. Clin. N. Amer., 4: 719, 1983. 7. Charton, M., Dosne, B., Escovar, P., Kopf, A. and Brisset, J. M.: Traitement prophylactique minute des infections urinaires apres resection endoscopique de la prostate. Presse Med., 13: 545, 1984.