The most important lesson to be learned from this and other studies is not the management of these patients, because this has to be tailored to the needs of the individual, but that cholecystectomy should be done more frequently in cases of recurrent chronic cholecystitis. We are indebted to Ms. Anna Fields of the Dr. George Williamson Medical Library and Ms. Mary Ironside of the medical records department, Ottawa Civic Hospital for their assistance in preparing this paper.

1. COWLEY LL, Wooo VW: Perforation of the gallbladder. Clin Med 73: 37, 1966 2. NIEMEIER DW: Acute free perforation of the gallbladder. Ann Surg 99: 922, 1934 3. WALLACE R, ALLEN AW: Acute cholecystitis. Arch Surg 43: 762, 1941 4. GLENN F, MooRE SW: Gangrene and perforation of the wall of the gallbladder. Arch Surg 44: 677, 1942

9. ELIASosI EL, MCLAUGHUN CW: Perforation of the galibladder. Ann Surg 99: 914, 1934 10. Cowizy LL, HARKINs uN: Perforation of the galibladder. Surg Gynecol Obstet 77: 661, 1943 11. SCHAEFFER RL: Perforation of the gallbladder. Penn Med J 45: 566, 1942 12: CAVE HW: Immediate or delayed treatment of acute cholecystitis: liver shock and deatn. Surg Gynecol Obstet 66: 308, 1938 13. JOHNSTONE GA, OSTENDOUPH JE: Cholecystitis with perforation. Arch Surg 53: 1, 1946

5. PINES B, RASINOVITCH G: Perforation of the

14. STROHL EL, DIFFENBAUGH WG, BAKER JH,

Referenc.s

gallbladder in acute cholecyatitia. Ann Surg 140: 170, 1954

6. Moasa L, KRYNSKI B, WRIGHT AR: Acute

perforation of the gallbladder. Am I Surg 94: 772, 1957 7. GALLAGHER WB: Perforation of the gallbladder. Am I Surg 100: 407, 1960 8. STONE WW, DOUGLASS FM: Perforation of the gallbladder. Am I Surg 45: 301, 1939

et al: Gangrene and perforation of the gallbladder (abstr). Surg Gynecol Obstet 114: 1,

1962 15. VALLEFO MC, DICKIE HG: Perforation of the galibladder In the aged. West Va Med / 61:

279, 1965 16. MACDONALD JA: Early cholecystectomy for acute cholecystitis. Can Med Anoc 1 111: 796, 1974

Clindamycin plus gentamicin as expectant therapy for presumed mixed infections ANTHONY W. CHOW, MD, FRCP[C], FACP; JEREMY K. OTA, MD; LUCIEN B. GUZE, MD, FACP

The prevalence of obligate anaerobes was studied prospectively in 60 patients with severe sepsis of intra-abdominal, soft tissue, female genital or oropuimonary origin. In addition, the efficacy of clindamycin (for anaerobes) plus gentamicin (for aerobic bacteria, especially coliforms) as initial empiric therapy in these patients was evaluated. Among 54 patients with cultural proof of infection, anaerobic pathogens were recovered from 520/o. Nineteen patients had bacteremia; Bacteroides fragils and Kiebsiella pneumoniae were the most prevalent pathogens, being isolated in five patients each. infection was eradicated in 56 of the 60 patients (93/o). Mortality related to sepsis was 7/. in the entire group, 16% in patients with bacteremia and 20/o In patients without bacteremia. Eighty-five percent of aerobic isolates tested were susceptible in vitro to either gentamicin or clindamycin; 970/0 of anaerobic isolates were inhibited by 5 ,sg/mi of clindamycin. Une etude prospective de Ia pr.vaience des bact6ries ana.robies obligatoires a 6t6 faite chez 60 patients atteints d'infections graves ayant pour origine i'abdomen, las tissus mous, las organes genitaux f.minins ou les tissus oropuimonaires. De plus, on a 6vaiu6 iefficacit. de Ia clindamycine (pour les ana6robies) et de Ia gentamicine From the department of medicine, Harbor General Hospital, Torrance, California; research and medical services, Veterans Administration Hospital (Wadsworth, Los Angeles; and school of medicine University of California, Los Angeles Reprint requests to: Dr. Anthony W. Chow, Division of infectious disease, Harbor General Hospital, 1000 W Carson St., Torrance, CA 90509, USA

(pour Des bact6rlos a6robles at sp6cialemant las coliformas) utllls6es empiriquemant comma traitemant dattaque. Parml 54 patients dont l'infoction a 6t6 d6montr6e par culture on a retrouv6 des bact6rios ana6roblas pathoganes chaz 52.'/.. Dix-neuf patients souffralent do bact6r6mle; Bacteroides fragilis at KIebsieIIa pneumoniae ont ete las bacteria. las plus fr6quemment retrouv6as alors qu'allas ont .t6 isoi6as chez clnq patients chacuna. L'infection a 6t6 enray6a chez 56 des 60 patients (93/o). La mortalit6 due a linfection a 6t. do 70/0 pour Ia groupe complet, do l60/o chez las patients souffrant do bact6r6mle at do 2/o choz las sujats no presentant pas do bact.remie. Ouatro-vingt-cinq pourcent des souches a6robies test6es 6taiont sonsibles in vitro, soit & Ia gontamicino, soit & Ia clindamycine, alors quo 970/* des ana6robies 6talent inhibdos par 5 jsg/ml do clindamycine.

Recent technologic advances have greatly facilitated the isolation and identification of obligate anaerobes from clinical material.1 However, with few exceptions,24 the prevalence of anaerobic bacteria in specific infections has not been adequately investigated by prospective study. For this reason, although antibiotics efficacious against anaerobic bacteria (such as clindamycin and chloramphenicol) are readily available, the necessity to include these agents in the initial empiric treatment of suspected or presumed sepsis remains controversial.5 We report the results of a prospective evaluation of the prevalence of obligate anaerobes in presumed sepsis of intraabdominal, soft tissue, female genital or

oropulmonary origin. Fastidious techniques were used to isolate both aerobic and anaerobic microorganisms, and special precautions were taken to avoid contamination of specimens by normal flora. In addition, the efficacy of dindamycin plus gentamicin as initial empiric therapy in these infections was examined prospectively. Clindamycin was chosen because of its demonstrated efficacy, both in vitro and in vivo, against anaerobic pathogens610 and the lack of antagonism, at least in vitro, between clindamycin and gentamicin against both aerobic and anaerobic bacteria.11'12 Patients and methods Patient population Sixty patients admitted consecutively to hospital in 1973 with presumed sepsis of intra-abdominal, soft tissue, female genital or oropulmonary origin were studied. The mean age was 41 years for the entire group, 35 years for the 19 females and 43 years for the 41 males. Informed consent was obtained from all patients. All were seriously ill, requiring immediate institution of empiric antibiotic therapy prior to availability of cultural data. Ten patients had received prior antibiotic therapy (three, penicillin; four, ampicillin; and three, cephalothin) within 10 days of combination therapy but were included in the study because of continued clinical deterioration or relapse despite use of these antibiotics. Patients who had received prior therapy with chloramphenicol, tetracycline, lincomycm or other antibiotics were excluded from the study. Pregnant women, infants and patients with known toxic or

CMA JOURNAL/DECEMBER 18, 1976/VOL. 115 1225

Table I-Distribution and types of primary infection in 60 critically ill patients with documented or presumed sepsis treated empirically with clindamycin and gentamicin Primary site of infection Gastrointestinal Peritonitis with Ruptured appendix Intestinal perforatlon. Intestinal gangrene Cirrhosis

Pathogens Isolated; no. of patients* Total no. of patients Aerobic Anaerobic Both Total 39 17 (6) 5 (4) 13 (1) 35 (11)

Intra-abdominal absces: Hepatic Appendiceal Pancreatic Paracollc Perlnephric Perirectal 2 1 - - 2 Softtissue Decubiti or gangrene Wound sepsis Popliteal abscess Suppurative thrombophlebitis

8

Female genital Tubs-ovarian or pelvic abscess Postpartum endometritls

7

Oropulmonary Pneumonia andlor empyema Paratracheal abscess Submandibular abscess

6

10 11 3 1

4 (1) 4 (1) 1 (1)

(1) 3 (1) 1 (1) -

5 (1) 2 2 -

4 3 3 1 1

2(2) 1 3 (1) 1 -

1 (1) -

2 1 -

4(2) 3 (1) 3 (1) 1 -

4 2 1 1

3 (4) 2 (3) 1 (1)

2 2 1 -

8 (4) 4 (3) 2 1 1 (1)

6 1

2 (2) 1 (1) 2 (2) 1 (1)

2

2 -

5 (3) 4 (2) 1 (1)

4 1 1

4 (1) 1 4 (1) -

1

1 -

6 (1) 4 (1) 1 1

Total 60 *No. with bacteremia In parenthesis. tFollowing trauma or gastrointestinal surgery.

26 (13)7 (5)

-

1

5

21(1)

9 9 3 1

(3) (2) (1) (1)

54 (19)

Table Il-Relative frequency of bacteria isolated from blood of 19 patients No. land %l of Organisms patients* Bacteroides fragilis St (3) [261 Kiebsiella pneumoniae 5 (4) 1261 Staphylococcus aureus 3 (3) 1161 Bacteroides pneumosintes 2 (1) [111 Escherichia coli 2 (1) [111 Streptococcus, non-group D 2 (1) [111 Proteus mirabiis 1 (1) [51 Enterobacter agglomerans 1 (1) [51 Pseudomonas cepacia 1 (1) [51 Streptococcus, group D 1 151 Clostridium ramosum 1 [51 Acidaminococcus fermentans 1 151 *No. of patients with pure cultures in parenthesis. tTwo patients had two different subspecies of B. fragiis.

Table IV..Sensitivity* of aerobic isolates to either gentamicin or clindamycin by agar-diffusion technique

Organisms Escherichia coli Kiebsiella

Gentamicin (10-.g disc) No. No. tested sensitive 23 23 18 18

Organisms Staphylococcus Streptococcus, group D Streptococcus, non-group D Lactobacillus Neisseria sp.

Proteus 10 10 Pseudomonas 4 4 Enterobacter 3 3 Serratia 3 3 Citrobacter 1 1 Total 62 62 *No. sensitive to either gentamicin or clindamycin, 88/103 (85%).

Clindamycin (2-ILg disc) No. No. tested sensitive 17 15 12

1

10 1 1

9 1 0 26

41

Table Ill-Relative frequency of bacteria isolated from primary sites of infection and blood in 54 patients*

Organisms Aerobic isolates Escherichia coli Staphylococcus Streptococcus, group D Kiebsiella pneumoniae Streptococcus, non-group D Miscellaneous Proteus Pseudomonas Enterobacter Serratia

Gastrointestinal 69 (9) 18 (2) 9 (2) 11(1) 7 (2) 6 (1) 6 4

3 3 (1) 2

Site of infection; no. of isolatest Soft Female tissue genital Oropulmonary 23 (4) 11(2) 6 (1) 3 2 1 3 3 (1) 2 3 1 4 (2) 1 1 (1) 5 (1) 1 1 2 1 3(1) 1 1 (1) 1

Total 109 (16) 24 (2) 17 (3) 15 (1) 13 (5) 13 (2) 9 8(1) 4 (1) 3 (1) 3

% of patients

44 31 28 24 24 17 15 7 6 6

Anaerobic isolates 29 (7) 11 4 (2) 3 47 (9) Bacteroides, not fragiis 7 (1) 1 1 (1) 1 10 (2) 19 B. fragilis 7 (5) 2 9§(5) 17 Nonsporulating gram-positive bacilli 4 2 1 7 13 Clostridium 4 (1) 2 6 (1) 11 Peptococcus 4 1 1 6 11 Peptostreptococcus 1 3 1 5 9 Fusobacterium 1 1 2 4 Butyrivibrio 1 1 2 Acidaminococcus 1 (1) 1 (1) 2 Total 98 (16) 34 (4) 15 (4) 9 (1) 156. (25) *Specimens potentially contaminated by commensal flora of mucocutaneous surfaces (i.e., throat swabs, expectorated sputum, bronchoscopic and nasotracheal aspirates, vaginal secretions, feces, colostomy effluent and superficial wound swabs, etc.) were not evaluated. tNo. of bacteremic isolates in parenthesis. IIncludes 1 Citrobacter, 2 Lactobacillus, 1 Neisseria sp., 1 diphtheroid, 1 Micrococcus and 3 Candida sp. Two patients had two different subspecies of B. fragilis isolated from blood.

antibiotic therapy. The remaining pa- generalized maculopapular and pruritic tient acquired Pseudomonas pneumonia rash developed in a third. All these following gunshot wounds to the abdo- manifestations resolved promptly folmen and failed to respond despite anti- lowing discontinuation of antibiotic biotic and intensive supportive therapy. therapy. Superinfection was noted in four paAntibiotic toxicity and superinfection tients: two with urinary tract infection Side effects possibly related to din- due to Candida sp., one with cellulitis damycin-gentamicin therapy developed due to Candida and Pseudomonas sp. in 16 patients (27%). Nephrotoxicity at the intravenous catheter site, and requiring modification of the gentami- one with hospital-acquired Pseudomocm dosage was noted in five patients; nas pneumonia. Except in the last pain four such patients who survived, tient, all symptoms resolved following renal function returned to normal after discontinuation of antibiotic therapy. cessation of antibiotic therapy. Super- No hepatic toxicity or colitis was enficial thrombophlebitis, observed in countered during clindamycin-gentaeight patients, necessitated frequent ro- micin therapy. tation of intravenous sites for continued administration of antibiotics. Eosino- Discussion philia occurred in two patients and a The importance of obligate anaerobes, in addition to aerobes, in serious infections of intra-abdominal, soft tisTable V-Sensitivity of anaerobic isolates to clindamycin by agar-dilution technique sue, female genital or oropulmonary Minimum inhibitory concentration (p.g/ml); origin is demonstrated clearly in this no. of isolates prospective study. Among 54 patients No. with cultural proof of infection, obliOrganisms tested