- Email: [email protected]
Bacteria and septic complications in patients with perforated duodenal ulcers
Bacteria and Septic Complications in Patients With Perforated Duodenal Ulcers
John Boey, MD, Hong Kong John Wong, PhD (Syd), FRCSE, FRACS, Hong Kong Guan B. Ong, DSc, FRCS, FRACS, Hong Kong
Peritoneal spillage often appears alarmingly extensive in patients with perforated duodenal ulcers. In an effort to prevent infection, antibiotics are usually given even though peritoneal cultures taken at laparotomy are often sterile (I ,2] and the clinical significance of a positive culture uncertain. We therefore analyzed the relation between peritoneal cultures and septic complications in patients with perforated ulcers in order to provide guidelines for the appropriate use of antibiotics. Material
and Methods
Between November 1978 and February 1981, we prospectively studied 184 consecutive patients who underwent operation for perforated duodenal ulcers. There were 158 male and 26 female patients, with a mean age of 48.8 years (standard deviation 17.7). The median duration of perforation (interval between perforation and exploration) was 12.1 hours (range 3 hours to 10 days). All patients received parenteral antibiotics for at least 7 days, and cephalothin (Keflinm) alone was used in 94 percent of cases. These antibiotics were administered preoperatively in 90.8 percent, intraoperatively in 0.5 percent, and immediately postoperatively in 8.7 percent of patients. Simple closure of the perforation was performed in 116, truncal vagotomy with gastric drainage in 38, and proximal gastric vagotomy with omental patch repair in 30 patients. Saline lavage (minimum 3 liters) of the peritoneal cavity was performed in every case, and all wounds were closed primarily. No drain was inserted except in 11 patients who had gross peritoneal contamination. The latter was defined by either the inability to cleanse the peritoneal surface of fibrin and other debris despite copious saline lavage, or else the presence of an abscess. There were five postoperative deaths (excluding one suicide), and risk factors related to mortality are described elsewhere [3]. From the Department of Surgery, University of Hong Kong, Oueen Mary Hospital, Hong Kong. Requests for reprints should be addressed to Guan B. Ong, DSc. Department of Surgery, University of Hong Kong, Queen Mary Hospital, Hong Kong.
Volume 143, May 1992
Shortly after the start of this prospective study, peritoneal cultures were obtained routinely in all cases immediately after celiotomy and applied to horse Mood agar medium for aerobic cultures. Thus, microbioldgic data were available in 143 unselected cases. The relations between bacterial cultures, patient age, duration of perforation, length of operation, peritoneal contamination, and infectious complications were assessed. The corrected chisquare test, Fisher’s exact test., and the unpaired Student’s t test were used for statistical analysis [4]. Statistical significance (two-tailed p unless otherwise specified) was accepted at the 5 percent level.
Results Bacteriologic findings: Fifty-four organisms were isolated from 48 of the 143 patients (33.6 percent) who had peritoneal cultures taken. No more than two pathogens were cultured in any patient, and only half of the 54 bacteria cultured showed moderate or heavy growths. Bacterial growth was more likely whenever the duration of perforation exceeded 48 hours or when gross contamination (as defined earlier) was present (Table I). Among patients with positive cultures, light growth was found in 60 percent of the patients who underwent exploration within 2 days of perforation (n = 40), whereas heavy growth was found in every patient operated on after this period (n = 8) (p = 0.007). The single positive blood culture revealed Pseudomonas in a patient whose peritoneal culture grew Klebsiella. Wound infections were caused uniformly by gram-negative pathogens, but these differed consistently from those isolated by peritoneal swabs. Candida species (all albicans except for one parapsilosis strain), E. Coli, Pseudomonas, and Klebsiella were the organisms most frequently found. Antibiotic sensitivities are shown in Table II. There was no change in drug sensitivities or emergence of resistant strains during the period of study. Septic complications: Thirteen infections occurred in 8 of the 184 patients (4.3 percent): septic 635
Boey et al
80
TABLE I -n
-CULTURE 260 ti z
I-l
0
STERILE
95
m
POSITIVE
48
Relations Between Positive Peritoneal Cultures, Peritoneal Contamination,’ and Duration of Perforation Duration of Perforation <4a >4a Hours Hours
Patients (n)
CUJJN~TD~NE
41
Peritoneal culture Sterile Positive Not done Total Significance (p)+ Peritoneal contamination Minimal Severe Sianificance ID)+
-HOURS AFTER PERFORATION Figure 1. Relation between posit/we peritoneal cuftures and the duration of perforatfon. n = number of patfents in each group.
Organism
Antiblotlc Sensitivliles’
Total (n)
6 a 6 20 0.05t
158 6 9 11
:::
:::
...
l
Infections were significantly more common in late perforations (patients who underwent exploration more than 48 hours after onset) associated with either gross contamination or a positive peritoneal culture (Table III). Severe contamination and a positive culture, each more common in long-standing perforations (Figure 1 and Table I), predisposed to septic complications, but only in perforations with a dura-
of Organisms Cultured From Peritoneal Swabs
Ampiciliin
Cefamandoie
Trimethoorim
Tetracvciine
Gentamicin
Tobramvcin
Carbeniciiiin
...
...
...
...
...
...
...
...
8
l/l
616
818
414
216
618
313
l/2
4
l/l
l/l
...
O/l
011
212
l/l
l/l
O/l O/l 516
212 l/l ...
l/l ;;3
313 O/l 415
313 O/l 213
l/l l/l ...
&i l/l
..
...
...
..
...
..
...
212
212
l/2
212
...
l/l
l/l
18
Total
636
89 40 35 164
Peniciilin
Candida albicans Escherichia coii Pseudomonas Kiebsieiia Proteus Streptococcus viridans Staphylococcus aibus Staphyiococcus aureus Streptococcus fecaiis Miscellaneous
l
Moderate to Heavy Growth (n)
90 5 40 8 37 4 167 17 0.028x
As defined in text. + Corrected chi-square test. t One-tailed p.
shock and intraabdominal abscess were each present in four patients, and superficial wound infections developed in five. Six of these patients underwent exploration more than 2 days after perforation, and four of them had preoperative shock. All eight patients received preoperative cephalothin and had simple closure operations. Bacterial growth, which was always heavy, was obtained in six of the seven patients who had peritoneal cultures taken.
TABLE II
Peritoneal Contamination Minimal Severe
.
4 2 6
4;s
.
.
6
1
2
1
2
1
..
O/l
O/l
. .
111
2
0
...
...
. .
.,.
..
..
.
...
54
27
...
...
...
...
...
...
...
.
l/2
..
...
Number of sensitive organisms/number tested.
TheAmerkan Journal ef Surgery
Bacteria
TABLE III
in Perforated
Ulcers
Factors Related to Septic Complications In Perforated Ulcers
Duration of Perforation (hours)
n
Septic Complications (Patients [n]) Present Absent
Significance (P)
All oatients 248 >4a Patients without severe contamination <4a >4a Patients with severe contamination <4a >4a Patients with sterile culturest <4a >4a Patients with positive culturest <4a >4a
167 17
165 11
2 6
158 6
156 6
2 0
NS’
9 11
9 5
0 6
0.014+
90 5
a9 5
1 0
NS’
40 a
39 3
1 5
(0.001
l
l
Corrected chi-square test. + Fisher’s exact test. * Forty-one patients did not have peritoneal cultures. NS = not significant. l
tion greater than 48 hours (Table IV). These differences were not evident when a 24 hour time limit was applied. The degree of bacterial proliferation also influenced the likelihood of infection: no complications occurred in the 24 patients who had scanty growths, but 6 of the 24 with heavy growths had septic problems (p = 0.029, corrected chi-square test). The eight patients who had septic complications were generally older (mean age 63.1 years) than those who had a smooth postoperative course (mean age 47.9 years) (p
TABLE IV
Total
Volume 143, May 1982
Septic Complications (Patients [n]) Absent Present
Significance (P)
All Patients Cultures Sterile Positive Not done Contamination Minimal Severe
95 48 41
94 42 40
1 6 1
0.01’
164 20
162 14
2 6
l
Patients With Exploration Within 46 Hours
Comments
Overall, 66.4 percent of our patients had sterile peritoneal cultures, a rate that agrees with the reported range of 35 to 72 percent [1,2,5-71. Unlike some earlier investigators [5,6], we did find bacterial growth more often (61.5 percent) in patients who underwent exploration late after onset (more than 48 hours). Peritoneal cultures tend to be sterile, or at most scanty, shortly after perforation, and even up to 2 days after onset. Thereafter, bacterial proliferation and the development of peritonitis usually ensue. These observations support the view that normal stomach contents are essentially sterile [8], that only mild bacterial contamination exists in the initial phase after acute perforation, and that delayed surgical attention allows progressive transformation of a state of chemical irritation into fulminant bacterial peritonitis [1,9-l I].
Effect of Late Exploratlon (Mane Than 48 Hours) on Bacterial Cultures, perltoneal Contamination, and Septic Compllcatlons
Cultures Sterile Positive Not done Contamination Minimal Severe
90 40 37
a9 39 37
1 1 0
NS’
158 9
156 9
2 0
NS’
Patients With Exploration After 46 Hours Cultures Sterile Positive Not done Contamination Minimal Severe
5 a 4
5 3 3
0 5 1
0.04+
6 11
6 5
0 6
0.04t
Corrected chi-square test. + Fisher’s exact test, one-tailed p (two-tailed, p = 0.07). t Fisher’s exact test, two-tailed p. NS = not significant. l
637
Boey et al
What actually constitutes significant contamination that leads to postoperative infection? Visual estimates of peritoneal soiling are notoriously unreliable, but our classification of severe contamination does bear some degree of objectivity as well as predictive value. Infections are also more likely with positive peritoneal cultures, especially if bacterial growth is abundant. Nevertheless, the clinical consequence of a positive culture in itself is dubious: it neither predisposes to septic problems in patients who undergo exploration within 2 days of perforation (Table IV) nor does it correlate with pathogens responsible for subsequent infections. Wangensteen et al [7] also noted a similar discrepancy between organisms cultured at surgery and pathogens that produced postoperative abscesses. Late exploration, where severe peritoneal soiling and bacterial overgrowth are the rule, and old age are the main factors that significantly enhance the risk of infection in perforated ulcers. Only when perforations are long-standing (more than 48 hours) does a positive peritoneal culture or gross soiling indicate serious contamination. Infections occur in roughly 6 to 10 percent of patients with perforated duodenal ulcers [12-141. In this study, only two patients who were initially free of gross peritoneal contamination subsequently had wound infections. Five of 11 patients who had extensive soilage and who underwent exploration after 48 hours developed no abscess or wound complications. This relatively low incidence of infection (4.3 percent) cannot be attributed to a lower mean age of our patients or a shorter duration of perforation before surgery, as these factors were comparable to those in other series [1,7]. Possibly, the practice of routine administration of antibiotics (over 90 percent given preoperatively) and peritoneal lavage, also advocated by other workers [5,15], may protect against postoperative infection. This study suggests that perforated ulcers resemble clean-contaminated cases-a situation in which prophylactic antibiotics have been shown to reduce septic complications in emergency gastric and other gastrointestinal operations [16-231. We did not evaluate the role of antibiotics in a controlled manner, and these were given to all patients for at least a week. However, extrapolating from reported experiences with chemoprophylaxis in clean-contaminated operations, a three-dose regimen of antibiotics started preoperatively appears useful in all cases of suspected perforation. Candida, scarcely mentioned in previous culture reports, was the most common species cultured. This organism has been implicated in intraabdominal abscesses and systemic sepsis after perforation [2,7] and is known to be commonly present after vagotomy and other gastric procedures [8,24]; however, it was not incriminated in any of our septic cases. Gram-negative bacteria are the predominant pathogens that cause abscesses and wound
638
infections in perforated ulcers. Accordingly, based on available drug sensitivities, we prefer parenteral administration of either a cephalosporin or aminoglycoside agent. A longer course of antibiotics is indicated in patients with established sepsis or markedly delayed exploration. The first 2 days after ulcer perforation, a bacteriologically quiescent period, is usually free of serious contamination in most patients. This supports our previous conclusion [3] that in patients with perforated ulcers the operative risk becomes substantial roughly 48 hours after onset. Thus, the time limit within which immediate definitive operations may be considered in fit patients is likely to exceed the previously mentioned “golden period” of only 12 hours [25]. Summary To evaluate the clinical significance of bacterial contamination in perforated duodenal ulcers, we prospectively studied septic complications in 184 consecutive patients. All patients received parenteral antibiotics (over 90 percent preoperatively) for at least 7 days. Thirteen infections developed in eight patients (4.3 percent). Peritoneal cultures, performed in 143 unselected patients, were positive in 33.6 percent of cases. Bacterial growth occurred more often and in heavier amounts in patients who underwent exploration late (after 48 hours) and those who had gross peritoneal soilage. Candida and gram-negative organisms predominated, but there was no correlation with pathogens that produced abscesses or wound infections. Old age and late exploration significantly increased the risk of infection. Neither peritoneal soiling nor a positive culture was likely to be clinically important when exploration was performed within 2 days of perforation. We treat perforated ulcers as clean-contaminated cases, and recommend that three doses of prophylactic antibiotics be begun preoperatively in all patients. Acknowledgmenti We thank Drs. J. Koo, N.W. Lee, and P.H.M. Lam for their participation in the controlled trial of definitive ulcer operations. References 1. Greco RS, Cahow CE. Alternatives in the management of acute perforated duodenal ulcer. Am J Surg 1974;127:109-14. 2. Mark JBD. Factors influencing the treatment of perforated duodenal ulcer. Surg Gynecol Obstet 1969;129:325-30. 3. Boey J, Wong J, Ong GB. A prospective study of operative risk factors in perforated duodenal ulcers. Ann Surg 1982;195 (in press). 4. Siegel S. Nonparametric statistics for the behavioral sciences. New York: McGraw-Hill, 1956. 5. Hardy JD, Walker GR Jr, Conn JH. Perforated peptic ulcer. An analysis of 208 consecutive cases with emphasis on pathophysiologic changes and deaths. Ann Surg 1961; 153:Ql l-32. 6. Hamilton JE, Harbrecht PJ. Growing indications for vagotomy in perforated peptic ulcer. Surg Gynecol Obstet 1967;
The American Journal of Surgery
Bacteria
124~61-4. 7. Wangensteen SL, Wray RC. Golden GT. Perforated duodenal ulcer. Am J Surg 1972; 123:536-42. 8. Nichols RL, Smith JW. lntragastric microbial colonization in common disease states of the stomach and duodenum. Ann Surg 1975;182:557-61. 9. Harbrecht PJ, Hamilton JE. Reappraisal of simple suture of acute perforated peptic ulcer. Implications for definitive operation. Ann Surg 1960; 152: 1044-8. 10. Jarrett F, Donaldson GA. The ulcer diathesis in perforated duodenal ulcer disease. Am J Surg 1972;123:406-10. 11. Donaldson GA, Jarrett F. Perforated gastroduodenal ulcer disease at the Massachusetts General Hospital from 1952 to 1970. Am J Surg 1970;120:306-11. 12. Gray JG, Roberts AK. Definitive emergency treatment of perforated duodenal ulcer. Surg Gynecol Obstet 1976;143: 890-4. 13. McCaughsn JJ, Bowers R. Simple &sure for perforated peptic ulcer. Surgery 1957;42:476-83. 14. Maynard A DE L, Froix CJL, Dropeza G. Gsstrcduodenal perforation. Arch Surg 1968;97:96-104. 15. Smith L, Beehsn PJ. Definitive operations for perforated duodenal ulcers. Surg Gynecol Obstet 1969;129:465-9. 16. Bernard HR, Cole WR. The prophylaxis of surgical infection: the effect of prophylactic antimicrobial drugs on the incidence of infection following potentially contaminated op-
Volume 143, May 1982
in Perforated Ulcers
erations. Surgery 1964;56:151-7. 17. Polk HC Jr, Lopez-Mayor JF. Postoperative wound infection: a prospective study of determinant factors and prevention. Surgery 1969;66:97-103. 18. Evans C. Pollodc AV. The reduction of surgical wound infections by prophylactic parenteral cephaloridine. A controlled clinical trial. Br J Surg 1973;60:434-7. 19. Stokes EJ. Waterworth P, Franks V, Watson 6, Clark CG. Short-term rcutine antibiitic prophylaxis in surgery. Br J Surg 7974;61:739-42. 20. Condon RE. Rational use of prophylactic antibiotics in gastrointestinal surgery. Surg Clin North Am 1975;55:130918. 21. Stone HH, Hooper CA, Kolb LD, Geheber GE, Dawkins EJ. Antibiotic prophylaxis in gastric, biliary and colonic surgery. Ann Surg 1976; 184:443-52. 22. Chodak GW, Plaut ME. Use of systemic antibiotics for prophylaxis in surgery: a critical review. Arch Surg 1977; 112:326-34. 23. Lewis RT. Wound infection after gastrcduodenal operations: a 10 year review. Can J Surg 1977;20:435-40. 24. Brooks JR, Smii HF. Pease FB Jr. EIacteriologJrof the stomach immediately following vagotomy: the growth of Candida albicans. Ann Surg 1974; 179:859-62. 25. Nemaniach 61, Nicoloff DM. Perforated duodenal ulcer: longterm followup. Surgery 1970;67:727-34.
639
John Boey, MD, Hong Kong John Wong, PhD (Syd), FRCSE, FRACS, Hong Kong Guan B. Ong, DSc, FRCS, FRACS, Hong Kong
Peritoneal spillage often appears alarmingly extensive in patients with perforated duodenal ulcers. In an effort to prevent infection, antibiotics are usually given even though peritoneal cultures taken at laparotomy are often sterile (I ,2] and the clinical significance of a positive culture uncertain. We therefore analyzed the relation between peritoneal cultures and septic complications in patients with perforated ulcers in order to provide guidelines for the appropriate use of antibiotics. Material
and Methods
Between November 1978 and February 1981, we prospectively studied 184 consecutive patients who underwent operation for perforated duodenal ulcers. There were 158 male and 26 female patients, with a mean age of 48.8 years (standard deviation 17.7). The median duration of perforation (interval between perforation and exploration) was 12.1 hours (range 3 hours to 10 days). All patients received parenteral antibiotics for at least 7 days, and cephalothin (Keflinm) alone was used in 94 percent of cases. These antibiotics were administered preoperatively in 90.8 percent, intraoperatively in 0.5 percent, and immediately postoperatively in 8.7 percent of patients. Simple closure of the perforation was performed in 116, truncal vagotomy with gastric drainage in 38, and proximal gastric vagotomy with omental patch repair in 30 patients. Saline lavage (minimum 3 liters) of the peritoneal cavity was performed in every case, and all wounds were closed primarily. No drain was inserted except in 11 patients who had gross peritoneal contamination. The latter was defined by either the inability to cleanse the peritoneal surface of fibrin and other debris despite copious saline lavage, or else the presence of an abscess. There were five postoperative deaths (excluding one suicide), and risk factors related to mortality are described elsewhere [3]. From the Department of Surgery, University of Hong Kong, Oueen Mary Hospital, Hong Kong. Requests for reprints should be addressed to Guan B. Ong, DSc. Department of Surgery, University of Hong Kong, Queen Mary Hospital, Hong Kong.
Volume 143, May 1992
Shortly after the start of this prospective study, peritoneal cultures were obtained routinely in all cases immediately after celiotomy and applied to horse Mood agar medium for aerobic cultures. Thus, microbioldgic data were available in 143 unselected cases. The relations between bacterial cultures, patient age, duration of perforation, length of operation, peritoneal contamination, and infectious complications were assessed. The corrected chisquare test, Fisher’s exact test., and the unpaired Student’s t test were used for statistical analysis [4]. Statistical significance (two-tailed p unless otherwise specified) was accepted at the 5 percent level.
Results Bacteriologic findings: Fifty-four organisms were isolated from 48 of the 143 patients (33.6 percent) who had peritoneal cultures taken. No more than two pathogens were cultured in any patient, and only half of the 54 bacteria cultured showed moderate or heavy growths. Bacterial growth was more likely whenever the duration of perforation exceeded 48 hours or when gross contamination (as defined earlier) was present (Table I). Among patients with positive cultures, light growth was found in 60 percent of the patients who underwent exploration within 2 days of perforation (n = 40), whereas heavy growth was found in every patient operated on after this period (n = 8) (p = 0.007). The single positive blood culture revealed Pseudomonas in a patient whose peritoneal culture grew Klebsiella. Wound infections were caused uniformly by gram-negative pathogens, but these differed consistently from those isolated by peritoneal swabs. Candida species (all albicans except for one parapsilosis strain), E. Coli, Pseudomonas, and Klebsiella were the organisms most frequently found. Antibiotic sensitivities are shown in Table II. There was no change in drug sensitivities or emergence of resistant strains during the period of study. Septic complications: Thirteen infections occurred in 8 of the 184 patients (4.3 percent): septic 635
Boey et al
80
TABLE I -n
-CULTURE 260 ti z
I-l
0
STERILE
95
m
POSITIVE
48
Relations Between Positive Peritoneal Cultures, Peritoneal Contamination,’ and Duration of Perforation Duration of Perforation <4a >4a Hours Hours
Patients (n)
CUJJN~TD~NE
41
Peritoneal culture Sterile Positive Not done Total Significance (p)+ Peritoneal contamination Minimal Severe Sianificance ID)+
-HOURS AFTER PERFORATION Figure 1. Relation between posit/we peritoneal cuftures and the duration of perforatfon. n = number of patfents in each group.
Organism
Antiblotlc Sensitivliles’
Total (n)
6 a 6 20 0.05t
158 6 9 11
:::
:::
...
l
Infections were significantly more common in late perforations (patients who underwent exploration more than 48 hours after onset) associated with either gross contamination or a positive peritoneal culture (Table III). Severe contamination and a positive culture, each more common in long-standing perforations (Figure 1 and Table I), predisposed to septic complications, but only in perforations with a dura-
of Organisms Cultured From Peritoneal Swabs
Ampiciliin
Cefamandoie
Trimethoorim
Tetracvciine
Gentamicin
Tobramvcin
Carbeniciiiin
...
...
...
...
...
...
...
...
8
l/l
616
818
414
216
618
313
l/2
4
l/l
l/l
...
O/l
011
212
l/l
l/l
O/l O/l 516
212 l/l ...
l/l ;;3
313 O/l 415
313 O/l 213
l/l l/l ...
&i l/l
..
...
...
..
...
..
...
212
212
l/2
212
...
l/l
l/l
18
Total
636
89 40 35 164
Peniciilin
Candida albicans Escherichia coii Pseudomonas Kiebsieiia Proteus Streptococcus viridans Staphylococcus aibus Staphyiococcus aureus Streptococcus fecaiis Miscellaneous
l
Moderate to Heavy Growth (n)
90 5 40 8 37 4 167 17 0.028x
As defined in text. + Corrected chi-square test. t One-tailed p.
shock and intraabdominal abscess were each present in four patients, and superficial wound infections developed in five. Six of these patients underwent exploration more than 2 days after perforation, and four of them had preoperative shock. All eight patients received preoperative cephalothin and had simple closure operations. Bacterial growth, which was always heavy, was obtained in six of the seven patients who had peritoneal cultures taken.
TABLE II
Peritoneal Contamination Minimal Severe
.
4 2 6
4;s
.
.
6
1
2
1
2
1
..
O/l
O/l
. .
111
2
0
...
...
. .
.,.
..
..
.
...
54
27
...
...
...
...
...
...
...
.
l/2
..
...
Number of sensitive organisms/number tested.
TheAmerkan Journal ef Surgery
Bacteria
TABLE III
in Perforated
Ulcers
Factors Related to Septic Complications In Perforated Ulcers
Duration of Perforation (hours)
n
Septic Complications (Patients [n]) Present Absent
Significance (P)
All oatients 248 >4a Patients without severe contamination <4a >4a Patients with severe contamination <4a >4a Patients with sterile culturest <4a >4a Patients with positive culturest <4a >4a
167 17
165 11
2 6
158 6
156 6
2 0
NS’
9 11
9 5
0 6
0.014+
90 5
a9 5
1 0
NS’
40 a
39 3
1 5
(0.001
l
l
Corrected chi-square test. + Fisher’s exact test. * Forty-one patients did not have peritoneal cultures. NS = not significant. l
tion greater than 48 hours (Table IV). These differences were not evident when a 24 hour time limit was applied. The degree of bacterial proliferation also influenced the likelihood of infection: no complications occurred in the 24 patients who had scanty growths, but 6 of the 24 with heavy growths had septic problems (p = 0.029, corrected chi-square test). The eight patients who had septic complications were generally older (mean age 63.1 years) than those who had a smooth postoperative course (mean age 47.9 years) (p
TABLE IV
Total
Volume 143, May 1982
Septic Complications (Patients [n]) Absent Present
Significance (P)
All Patients Cultures Sterile Positive Not done Contamination Minimal Severe
95 48 41
94 42 40
1 6 1
0.01’
164 20
162 14
2 6
l
Patients With Exploration Within 46 Hours
Comments
Overall, 66.4 percent of our patients had sterile peritoneal cultures, a rate that agrees with the reported range of 35 to 72 percent [1,2,5-71. Unlike some earlier investigators [5,6], we did find bacterial growth more often (61.5 percent) in patients who underwent exploration late after onset (more than 48 hours). Peritoneal cultures tend to be sterile, or at most scanty, shortly after perforation, and even up to 2 days after onset. Thereafter, bacterial proliferation and the development of peritonitis usually ensue. These observations support the view that normal stomach contents are essentially sterile [8], that only mild bacterial contamination exists in the initial phase after acute perforation, and that delayed surgical attention allows progressive transformation of a state of chemical irritation into fulminant bacterial peritonitis [1,9-l I].
Effect of Late Exploratlon (Mane Than 48 Hours) on Bacterial Cultures, perltoneal Contamination, and Septic Compllcatlons
Cultures Sterile Positive Not done Contamination Minimal Severe
90 40 37
a9 39 37
1 1 0
NS’
158 9
156 9
2 0
NS’
Patients With Exploration After 46 Hours Cultures Sterile Positive Not done Contamination Minimal Severe
5 a 4
5 3 3
0 5 1
0.04+
6 11
6 5
0 6
0.04t
Corrected chi-square test. + Fisher’s exact test, one-tailed p (two-tailed, p = 0.07). t Fisher’s exact test, two-tailed p. NS = not significant. l
637
Boey et al
What actually constitutes significant contamination that leads to postoperative infection? Visual estimates of peritoneal soiling are notoriously unreliable, but our classification of severe contamination does bear some degree of objectivity as well as predictive value. Infections are also more likely with positive peritoneal cultures, especially if bacterial growth is abundant. Nevertheless, the clinical consequence of a positive culture in itself is dubious: it neither predisposes to septic problems in patients who undergo exploration within 2 days of perforation (Table IV) nor does it correlate with pathogens responsible for subsequent infections. Wangensteen et al [7] also noted a similar discrepancy between organisms cultured at surgery and pathogens that produced postoperative abscesses. Late exploration, where severe peritoneal soiling and bacterial overgrowth are the rule, and old age are the main factors that significantly enhance the risk of infection in perforated ulcers. Only when perforations are long-standing (more than 48 hours) does a positive peritoneal culture or gross soiling indicate serious contamination. Infections occur in roughly 6 to 10 percent of patients with perforated duodenal ulcers [12-141. In this study, only two patients who were initially free of gross peritoneal contamination subsequently had wound infections. Five of 11 patients who had extensive soilage and who underwent exploration after 48 hours developed no abscess or wound complications. This relatively low incidence of infection (4.3 percent) cannot be attributed to a lower mean age of our patients or a shorter duration of perforation before surgery, as these factors were comparable to those in other series [1,7]. Possibly, the practice of routine administration of antibiotics (over 90 percent given preoperatively) and peritoneal lavage, also advocated by other workers [5,15], may protect against postoperative infection. This study suggests that perforated ulcers resemble clean-contaminated cases-a situation in which prophylactic antibiotics have been shown to reduce septic complications in emergency gastric and other gastrointestinal operations [16-231. We did not evaluate the role of antibiotics in a controlled manner, and these were given to all patients for at least a week. However, extrapolating from reported experiences with chemoprophylaxis in clean-contaminated operations, a three-dose regimen of antibiotics started preoperatively appears useful in all cases of suspected perforation. Candida, scarcely mentioned in previous culture reports, was the most common species cultured. This organism has been implicated in intraabdominal abscesses and systemic sepsis after perforation [2,7] and is known to be commonly present after vagotomy and other gastric procedures [8,24]; however, it was not incriminated in any of our septic cases. Gram-negative bacteria are the predominant pathogens that cause abscesses and wound
638
infections in perforated ulcers. Accordingly, based on available drug sensitivities, we prefer parenteral administration of either a cephalosporin or aminoglycoside agent. A longer course of antibiotics is indicated in patients with established sepsis or markedly delayed exploration. The first 2 days after ulcer perforation, a bacteriologically quiescent period, is usually free of serious contamination in most patients. This supports our previous conclusion [3] that in patients with perforated ulcers the operative risk becomes substantial roughly 48 hours after onset. Thus, the time limit within which immediate definitive operations may be considered in fit patients is likely to exceed the previously mentioned “golden period” of only 12 hours [25]. Summary To evaluate the clinical significance of bacterial contamination in perforated duodenal ulcers, we prospectively studied septic complications in 184 consecutive patients. All patients received parenteral antibiotics (over 90 percent preoperatively) for at least 7 days. Thirteen infections developed in eight patients (4.3 percent). Peritoneal cultures, performed in 143 unselected patients, were positive in 33.6 percent of cases. Bacterial growth occurred more often and in heavier amounts in patients who underwent exploration late (after 48 hours) and those who had gross peritoneal soilage. Candida and gram-negative organisms predominated, but there was no correlation with pathogens that produced abscesses or wound infections. Old age and late exploration significantly increased the risk of infection. Neither peritoneal soiling nor a positive culture was likely to be clinically important when exploration was performed within 2 days of perforation. We treat perforated ulcers as clean-contaminated cases, and recommend that three doses of prophylactic antibiotics be begun preoperatively in all patients. Acknowledgmenti We thank Drs. J. Koo, N.W. Lee, and P.H.M. Lam for their participation in the controlled trial of definitive ulcer operations. References 1. Greco RS, Cahow CE. Alternatives in the management of acute perforated duodenal ulcer. Am J Surg 1974;127:109-14. 2. Mark JBD. Factors influencing the treatment of perforated duodenal ulcer. Surg Gynecol Obstet 1969;129:325-30. 3. Boey J, Wong J, Ong GB. A prospective study of operative risk factors in perforated duodenal ulcers. Ann Surg 1982;195 (in press). 4. Siegel S. Nonparametric statistics for the behavioral sciences. New York: McGraw-Hill, 1956. 5. Hardy JD, Walker GR Jr, Conn JH. Perforated peptic ulcer. An analysis of 208 consecutive cases with emphasis on pathophysiologic changes and deaths. Ann Surg 1961; 153:Ql l-32. 6. Hamilton JE, Harbrecht PJ. Growing indications for vagotomy in perforated peptic ulcer. Surg Gynecol Obstet 1967;
The American Journal of Surgery
Bacteria
124~61-4. 7. Wangensteen SL, Wray RC. Golden GT. Perforated duodenal ulcer. Am J Surg 1972; 123:536-42. 8. Nichols RL, Smith JW. lntragastric microbial colonization in common disease states of the stomach and duodenum. Ann Surg 1975;182:557-61. 9. Harbrecht PJ, Hamilton JE. Reappraisal of simple suture of acute perforated peptic ulcer. Implications for definitive operation. Ann Surg 1960; 152: 1044-8. 10. Jarrett F, Donaldson GA. The ulcer diathesis in perforated duodenal ulcer disease. Am J Surg 1972;123:406-10. 11. Donaldson GA, Jarrett F. Perforated gastroduodenal ulcer disease at the Massachusetts General Hospital from 1952 to 1970. Am J Surg 1970;120:306-11. 12. Gray JG, Roberts AK. Definitive emergency treatment of perforated duodenal ulcer. Surg Gynecol Obstet 1976;143: 890-4. 13. McCaughsn JJ, Bowers R. Simple &sure for perforated peptic ulcer. Surgery 1957;42:476-83. 14. Maynard A DE L, Froix CJL, Dropeza G. Gsstrcduodenal perforation. Arch Surg 1968;97:96-104. 15. Smith L, Beehsn PJ. Definitive operations for perforated duodenal ulcers. Surg Gynecol Obstet 1969;129:465-9. 16. Bernard HR, Cole WR. The prophylaxis of surgical infection: the effect of prophylactic antimicrobial drugs on the incidence of infection following potentially contaminated op-
Volume 143, May 1982
in Perforated Ulcers
erations. Surgery 1964;56:151-7. 17. Polk HC Jr, Lopez-Mayor JF. Postoperative wound infection: a prospective study of determinant factors and prevention. Surgery 1969;66:97-103. 18. Evans C. Pollodc AV. The reduction of surgical wound infections by prophylactic parenteral cephaloridine. A controlled clinical trial. Br J Surg 1973;60:434-7. 19. Stokes EJ. Waterworth P, Franks V, Watson 6, Clark CG. Short-term rcutine antibiitic prophylaxis in surgery. Br J Surg 7974;61:739-42. 20. Condon RE. Rational use of prophylactic antibiotics in gastrointestinal surgery. Surg Clin North Am 1975;55:130918. 21. Stone HH, Hooper CA, Kolb LD, Geheber GE, Dawkins EJ. Antibiotic prophylaxis in gastric, biliary and colonic surgery. Ann Surg 1976; 184:443-52. 22. Chodak GW, Plaut ME. Use of systemic antibiotics for prophylaxis in surgery: a critical review. Arch Surg 1977; 112:326-34. 23. Lewis RT. Wound infection after gastrcduodenal operations: a 10 year review. Can J Surg 1977;20:435-40. 24. Brooks JR, Smii HF. Pease FB Jr. EIacteriologJrof the stomach immediately following vagotomy: the growth of Candida albicans. Ann Surg 1974; 179:859-62. 25. Nemaniach 61, Nicoloff DM. Perforated duodenal ulcer: longterm followup. Surgery 1970;67:727-34.
639