Bacterial Studies of Peritoneal Cavity and Postoperative Surgical Drainage Following Perforated Appendix in Children

ITZHAK BROOK, M.D., M.S.

This study reports bacterial specimens obtained from 112 From the Departments of Infectious and children presenting with a ruptured appendix. Additional Clinical Microbiology Laboratories 4 samples were studied from 11 of these patients who developed Children's Hospital National Medical Center a postoperative surgical draining wound. Bacterial growth and George Washington University School of occurred in 100 peritoneal fluid specimens. Anaerobic Medicine, Washington, D.C. bacteria alone were present in 14 specimens, aerobes alone in 12, and mixed aerobic an anerobic flora in 74 specimens. There were 144 aerobic isolates (1.4 per specimen). The predominant isolates were: E. coli (57 specimens); alpha- is involved. Characteristically, the more types of hemolytic steptococcus (16 specimens); gamma-hemolytic bacteria that can be isolated, the graver the morbidity. streptococcus (15 specimens); Group D streptococcus (12 The specific micro-organisms involved in peritonitis specimens); and P. aeruginosa (9 specimens). There were 301 are generally those of the normal flora of the gastro- anaerobic isolates (three per specimen). The predominant intestinal tract where isolates were: 157 Bacteroides spp. (including 92 B. fragilis anaerobic bacteria outnumber group and 26 B. melaninogenicus group); 62 gram-positive aerobes in the ratio 1:1,000.5 anaerobic cocci (including 30 Peptococcus sp.; 29 Peptostrepto- Studies in adults demonstrated the presence of coccus sp.); 27 Fusobactenium sp.; and 16 Clostridium sp. mixed aerobic and anaerobic flora in the peritoneal B. fragilis and Peptococcus sp. occurred in 23 patients. Beta cavity of patients with ruptured appendix6 and showed lactamase production was detectable in 98 isolates recovered from 74 patients. These included all isolates of B. fragilis and that these organisms may occasionally be recovered six of the 23 Bacteroides sp. Forty-nine organisms (16 aerobic from the postoperative wound.7 and 33 anaerobic) were recovered from the draining . A few studies of the bacterial flora of the peritoneal The predominant organisms were: B. fragilis (8 specimens); cavity and postoperative wounds following perforated E. coli (6 specimens); Peptostreptococcus sp. (5 specimens); appendix in children have been conducted;8 9 however, and three specimens each of P. aeruginosa and Peptococcus the sp. Most of these isolates were also recovered from the techniques of specimen collection and identification peritoneal cavity of the patients. These findings demonstrate on anaerobic micro-organisms were not described and the polymicrobial aerobic and anaerobic nature of peritoneal the identification of the organisms was not complete. cavity and postoperative wound flora in children with per- The purpose of this study was to quantify the forated appendix, and demonstrate the presence of beta aerobic and anaerobic flora of the peritoneal fluid and lactamase-producing organisms in three-fourths of the patients. the postoperative wound in children who undergo following perforation of the appendix. P ERFORATED APPENDICITIS IS THE MOST common Materials and Methods cause of peritonitis in children. Despite a signifi- cant decrease in One hundred twelve patients (67 males and 45 mortality, this still produces females) undergoing for a high morbidity2-4 secondary to infectious compli- appendectomy perforation of' cation. Peritonitis the appendix, between September 1977 to December following appendicitis is usually a 1979, were in synergistic infection in which more than one included the study. Their ages ranged organism from 1 month to 16 years (mean: 7 years and 6 months). Postoperative wound drainage (defined as the presence Reprint requests: Itzhak Brook, M.D., M.S. Infectious and Clinical Microbiology Laboratory, Children's Hospital, National of discharge of serous or purulent material from the Medical Center, Ill Michigan Avenue, N.W., Washington, DC operation wound with or without associated inflam- 20010. mation of the skin edges) occurred in 11 (10%) of the Submitted for publication: February 19, 1980. children.

0003-4932/80/0800/0208 $00.75 © J. B. Lippincott Company 208 Vol. 192 * No. 2 BACTERIAL STUDIES IN CHILDREN 209 Collection of Specimens anaerobic gram-positive cocci (30 Peptococcus sp., 29 peptostreptococcus sp.), 27 Fusobacterium sp., Peritoneal cavity cultures were obtained from the and 16 Clostridium sp. appendicea fossa before removal of the appendix. Mixed bacterial flora were present in 91 patients, Cultures of the postoperative wound were obtained by where the average number of isolates per specimen swabbing or suctioning of the purulent fluid or pus. was 4.8. No consistent patterns of bacterial com- All specimens were sent to the laboratory using Post-A- binations were noted, except the one between B. Col® swabs (BBL, Cockeysville, MD) or a corked fragilis and E. coli in 39 patients, and B. fragilis and syringe and needle. The time between specimen col- Peptococccus sp. in 23 instances. Only one bacterial lection and inoculation never exceeded 60 minutes. isolate was recovered in nine instances. These included Sheep blood, chocolate, and MacConkey's agar 3 E. coli, 2 B. fragilis, and one of the following: B. vul- plates were inoculated for aerobes. The plates were gatus, K. pneumoniae, Group A streptococcus and incubated at 37 C aerobically (MacKonkey's) or under Group D streptococcus. 5% C02, and examined at 24 and 48 hours. For Beta lactamase production was detectable in 98 anaerobes, the material was plated on the bench onto isolates recovered from 74 patients. These included prereduced vitamin K,-enriched brucella blood agar, all isolates ofB. fragilis group and five of the 23 (22%) an anaerobic blood agar plate containing kanamycin Bacteroides sp. None of the isolates of B. mela- (75 gg/ml) and vancomycin (7.5 ,g/ml), anaerobic ninogenicus was beta lactamase-producing. blood plate containing phenylethyl alcohol, and onto enriched thioglycolate broth (containing hemin, sodium Culture of Wound Drainage bicarbonate, and vitamin Kj).'0 The plates were in- cubated in GasPakgjars (BBL, Cockeysville, MD) and Bacterial growth was noted in all of the 11 draining examined at 48 and 96 hours. The thioglycolate broth wounds. In all but one of these instances, bacterial was incubated for 14 days. Anaerobes were identified growth was also observed in the peritoneal fluid. by techiques previously described.10 Aerobic bacteria Forty-nine organisms were recovered. Anaerobic bac- were identified using conventional methods." Beta teria alone were present in one patient, aerobes alone lactamase activity was determined on all gram-negative in one, and mixed aerobic and anaerobic flora were anaerobic organisms, using the chromogenic cephalo- present in nine. sporin analog 87/312 methodology.12 There were 16 aerobic isolates (1.5 per specimen). Each patient was treated with parenteral anti- These included six E. coli and three of each of Gamma- microbial therapy for five to seven days. The anti- hemolytic streptococcus and P. aeruginosa. biotics used were clindamycin (20 mg/kg/day) and There were 33 anaerobic isolates (3 per specimen). gentamicin (5 mg/kg/day). These included eight B. fragilis group, five Pepto- streptococcus sp. and three Peptococcus sp. P. acnes Results was recovered in two instances and in one of these it was the only organism recovered. Peritoneal Cavity The isolates recovered from the infected wounds, Bacteria grew in 100 of 112 specimens. Anaerobic except both isolates of P. acnes and one of each P. bacteria alone were present in 14 specimens, aerobes aeruginosa and S. epidermidis were also recovered alone in 12, and mixed aerobic and anaerobic flora from the peritoneal cavity of these patients. were present in 74 specimens. Thus, 88% of specimens yielded anaerobic organisms while 86% showed Discussion aerobic organisms. Our results clearly demonstrate the presence of Four hundred forty-five bacterial isolates were re- mixed aerobic and anaerobic flora in the peritoneal covered from the 100 culture-positive specimens (4.45 cavity and in the drainage of the surgical wounds that per specimen) (Table 1), including 144 aerobes (1.4 per follow perforation of the appendix in children. We ob- specimen). The most frequently isolated aerobes were tained anaerobic organisms in 88% and aerobes in 86% in E. coli (57 isolates), alpha-hemolytic streptococci,'6 of peritoneal cultures, a finding which concurs to the gamma-hemolytic steptococcus,'5 Group D strepto- data of previous reports in children8'9 and adults.6"3 coccus,12 and P. aeruginosa.9 There were 301 We recovered anaerobes in 91% of the exudates of anaerobes isolated (three per specimen). One hundred the draining wound which corresponds to findings in fifty-seven of the isolates were Bacteroides spp. adult patients.7 These included 92 B. fragilis group and 26 B. mela- The recovery of H. influenzae type B in three in- ninogenicus group. Fifty-eight of the isolates were stances, warrants special concern. This organism is 210 BROOK Ann. Surg. * August 1980 TABLE 1. Organisms Isolatedfrom Peritoneal Fluidfrom 100 Patients with Perforated Appendix and 11 with Postoperative Wound Infection No. of No. of Aerobic and Facultative Isolates Anaerobic Isolates Isolates Gram-positive cocci Gram-positive cocci Alpha-hem. streptococcus 16(1) Peptococcus spp. 13(2) Gamma-hem. streptococcus 15(3) Peptococcus magnus 10 Group D streptococcus 12(1) Peptococcus asaccharolyticus 3(1) Group A, beta-hem. strep. 3 Peptococcus saccharolyticus 2 Group C, beta-hem. strep. 2 Peptococcus prevoti 2 Group F, beta-hem. strep. 3 Peptostreptococcus spp. 6(1) Staphylococcus epidermidis 2(1) Peptostreptococcus micros 13(1) Peptostreptococcus intermidus 2 Streptococcus constellatus 3(2) Streptococcus intermidius 5(1) Microaerophilic streptococcus 3(1) Gram-negative cocci Veillonella parvulla 2 Veillonella alcalesence Gram-positive bacilli Gram-positive bacilli Diptheroid sp. 4(1) Bifidobacterium spp. 3(1) Bifidobacterium adolescentis 4 Eubacterium spp. 10(2) Eubacterium lentum 7 Lactobacillus sp. 1 Propionibactrium acnes 9(2) Actinomyces adontalyticium 2 Clostridium spp. 5(1) Clostridium perfrigens 3 Clostridium ramosum 4 Clostridium difJicille 1 Clostriium butyricum 1 Clostridium innocuum 2(1) Gram-negative bacilli Gram-negative bacilli Haemophilus influenzae Type B 3 Fusobacterium spp. 6(1) Haemophilus sp. 1 Fusobacterium varium 3 Eikenella corrodens 2 Fusobacterium necrophorum 3 Pseudomonas aeruginosa 9(3) Fusobacterium nucleatum 15(2) Escherichia coli 57(6) Bacteroides spp. 23(2) Klebsiella pneumoniae 7 Bacteroides melaninogenicus 21(2) Klebsiella oxytoca 2 Bacteroides melaninogenicus ss. asaccarolyticus 2 Enterobacter cloacae 3(1) Bacteroides melaninogenicus ss. intermedius 3 Moraxella sp. 1 Bacteroides rumin ss. rumin 1 Proteus vulgatus 2 Bacteroides rumin ss. brevis 5(1) Bacteroides ovatus 10(1) Bacteroides fragilis B 54(5) Bacteroides vulgatus fragilis 7(1) Bacteroides distasonis group 3 Bacteroides thetaiotaomicron 13(1) Bacteroides uniformis 10 Bacteroides splanchnicus 3(l) Total number of aerobes and facultatives 144(16) Total number of anaerobes 301(33) Numbers in parentheses indicate number of organisms recovered from postoperative wound infection in 11 patients. not usually looked for in peritoneal fluid and appro- reported in adult patients and may be due to differences priate culture media containing X and V factors are in colonization of the in children. usually not employed.11 This organism is more In contrast to previous investigators, we have at- prevalent as a pathogen in pediatric patients, and has tempted to identify all anaerobic strains according to recently shown increased resistance to ampicillin.14 their subspecies. This has led to some interesting ob- More isolates of this organism might be recovered servations. We have found that B. melaninogenicus from peritoneal fluid, if appropriate media were used. is one of the predominant isolates recovered from the We recovered no S. aureus and isolated Proteus infected peritoneal fluids. Although all isolates of this sp. only in two instances. This contrasts to findings organism do not produce beta lactamase and are sus- Vol. 192 * No. 2 BACTERIAL STUDIES IN CHILDREN 211 ceptible to penicillin, growing numbers of strains have oped intra-abdominal abscesses caused by anaerobes. recently been reported as resistant.15 This may be of Evaluation of the effect clindamycin, gentamicin, or clinical concern if some gastrointestinal isolates ac- a combination of both showed that the untreated con- quire the ability to produce beta-lactamase. Some ofthe trol or clindamycin treated had an identical mortality clostridal isolates such as C. difficile and C. perfringens rate of about 35%. However, gentamicin, singly or can be toxin producers.13 combined with clindamycin, led to greater than 90% Since anaerobes vary considerably in their sus- survival. This suggests that the early mortality in the ceptibility to oxygen,13 improper techniques of trans- peritonitis and septicemia phase is due to gentamicin- portation of specimens and delays in inoculation can sensitive cloriform bacteria. result in deletion of virulent strains. Results of im- The effect of treatment on abscess formation was properly performed techniques can produce misleading entirely different. All untreated animals that survived, information. Stone, et al.6 demonstrated that a pro- as well as those who were treated with gentamicin longed exposure to oxygen of peritoneal specimen ob- alone, developed abscesses. However, clindamycin tained at surgery resulted in decrease in the number alone or in combination with gentamicin was associated of anaerobic strains isolated. It is therefore important with a markedly reduced incidence and size of the that any specimen to be processed for anaerobes should abscesses which occurred in only 5% of the animals. be collected quickly and with minimal exposure to These findings indicate that anaerobes may be respon- oxygen. sible for complications such as intra-abdominal abscess Peritonitis is an excellent example of a synegistic formation following abdominal perforation and show infection between aerobic and anaerobic micro- that optimal treatment of intestinal performation re- organisms. The two types of bacteria have opposite quires in addition to surgical drainage, a drug which oxygen requirements, and the alteration each causes will control both aerobic and anaerobic bacteria. in its environment allows for the rapid proliferation The treatment of abdominal infection should always of their partners.'6-18 The results of appropriate cul- include surgical correction and drainage. The medical ture techniques have consistently documented that the therapy should supplement the surgical approach by great majority of intra-abdominal are based attempting to eradicate both aerobic and anaerobic on this symbiotic relationship.19-21 The more types of micro-organisms. Aminoglycosides are effective in bacteria isolated from patients with peritonitis, the eliminating enteric gram-negative rods, although grow- higher the postsurgical morbidity.22'23 ing numbers of strains are resistant to some of these Like other investigators, we have found E. coli and B. drugs. Antibiotics effective against B. fragilis group fragilis to be the organisms most frequently isolated from include chloramphenicol, carbenicillin, ticarcillin, peritoneal fluid and surgical wounds as components clindamycin, cefoxitin, and metronidazole.26 Since of a mixed infection.68 The recovery of multiple some strains of B. fragilis may acquire resistance bacteria from these sites poses important questions, to one of these antibiotics13 susceptibility testing of concerning which organisms require specific therapy the organism should be performed in serious infections. and whether treatment should incorporate drugs to cover both aerobic and the anaerobic bacteria. There Acknowledgments is general agreement that aminoglycosides are appro- The author gratefully acknowledges the assistance of L. Calhoun, priate treatment for coliforms. The major contro- L. Harteker, and of the staff of the Operating Room at Children's versy concerns the necessity of treating all the an- Hospital National Medical Center. aerobic bacteria, including B. fragilis group, in pa- References tients with peritonitis. A recent study2l compared 1. Grosfield JL, Weinberger M, Clatworthy HW, Jr. Acute two groups ofpatients following trauma to the abdomen appendicitis in the first two years of life. J Pediatr Surg 1973; and perforation. One group was treated with gentami- 8:285. 2. Boles ET, Jr, Ireton RJ, Clatworthy HW, Jr. Acute appendicitis cin and a cephalosporin which was not effective against in children. Arch Surg 1959; 79:447. most isolates of B fragilis, and the other group was 3. Landsden FT. Acute appendicitis in children. Am J Surg 1963; treated with gentamicin and clindamycin. More com- 206:938. 4. Maddox JR, Jr, Johnson WW, Sergeant CK. in plications occurred in the cephalosporin treated group, a Children's Hospital. Arch Surg 1964; 89:223. showing the efficacy of preventing complications such 5. Gorbach SL. Intestinal microflora. Gastroenterology 1971; as intra-abdominal abscesses in patients treated for 60:1110. 6. Stone HH, Kolb LD, Geheber CE. Incidence and significance of anaerobic bacteria including B. fragilis. interperitoneal anaerobic bacteria. Ann Surg 1975; 181:705. In another study,24'25 peritonitis was induced in rats 7. Sanderson PJ, Wren MWP, Baldwin AWF. Anaerobic organisms by inserting gelatin capsules containing fecal contents in postoperative wounds. J Clin Pathol 1979; 32:143-147. 8. Marchildon MB, Dudgeon DL. Perforating appendicitis: a into their abdominal cavity. The animals that survived current experience in Children's Hospital. Ann Surg 1977; the initial septisemic stage caused by coliforms devel- 185:84. 212 BROOK Ann. Surg. * August 1980 9. Stone JH. Bacterial flora of appendicitis in children. J Pediatr 18. Stone HH, Martin JD, Jr. Synergistic necrotizing cellulitis. Ann Surg 1976; 11:37. Surg 1972; 175:702. 10. Sutter VL, Vargo VL, Finegold SM. Wadsworth Anaerobic 19. Gorbach SL, Bartlett JG. Anaerobic infections. N Eng J Med Bacteriology Manual. Second Edition, UCLA Extension 1974; 290:1177, 1237 & 1269. Division, Los Angeles. 1975. 20. Thadepalli H, Gorbach SL, Broido P, et al. Prospective study of 11. Lennette EH, Spaulding EH, Traunt JR. Manual of Clinical infections in penetrating abdominal trauma. Am J Sutr 1972; Microbiology. Second Edition. Washington, DC., American 25:1405. Society for Microbiology. 1974. 21. Thadepalli H, Gorbach SL, Broido PW, et al. Abdominal trauma, 12. O'Callagham CH, Morris A, Kirby SM, et al. Novel method for anaerobes and antibiotics. Surg Gynecol Obstet 1973; 137:270. detection of beta-lactamase by using a chromatogenic 22. Meleney FL, Olpp J, Harvey HD, et al. Peritonitis. II. cephalosporin substrate. Antimicrob Agents Chemother. Synergism of bacteria commonly found in peritoneal 1972; 1:283. exudates. Arch Surg 1932; 25:709. 13. Finegold SM. Anaerobic Bacteria in Human Disease. New York, 23. Altemeier WA. The bacterial flora of acute perforated ap- Academic Press. 1977. pendicitis with peritonitis. Ann Surg 1938; 107:517. 14. Syrlopoulou V, Scheifelle D, Smith AL, et al. Increasing 24. Weinstein WM, Onderdonk AB, Bartlett JG, et al. Experimental incidence of ampicillin resistance in Haemophilus influenzae. intraabdoninal abscesses in rats. I. Development of an J Pediatr 1978; 92:899. experimental model. Infect Immun 1974; 10: 1250. 15. Murray PR, Rosenblatt JE. Penicillin resistant and penicillinase 25. Onderdonk AB, Weinstein WM, Sullivan NM, et al. Experi- production in clinical isolates of Bacteroides melaninogenicus. mental intraabdominal abscesses in rats. II. Quantitative Antimicro Agents Chemother. 1977; 22:605. bacteriology of infected animals. Infect Immun 1974; 16. Meleney FL. Hemolytic streptococcus . Arch Surg 10:1256-1259. 1924; 9:317. 26. Sutter VL, Finegold SM. Susceptibility of anaerobic bacteria 17. Rea WJ, Wyrick WJ, Jr. Necrotizing fasciitis. Ann Surg 1970; to 23 antimicrobial agents. Antimicrob Agent Chemother 172:957, 1970. 1976; 10:736.