Surg Today (2004) 34:725–731 DOI 10.1007/s00595-003-2810-1

Original Articles Recovery of the Susceptibility of Isolated Bacterium Achieved by Giving Long-Established Antibiotics as Prophylaxis Against Postoperative Infections Shinya Kusachi, Yoshinobu Sumiyama, Yoichi Arima, Yuichi Yoshida, Yoichi Nakamura, Hidenori Tanaka, Jiro Nagao, Yoshihisa Saida, Manabu Watanabe, and Junko Sato Third Department of Surgery, Toho University School of Medicine, 2-17-6 Ohashi, Meguro-ku, Tokyo 153-8515, Japan

Abstract Purpose. To evaluate the effectiveness of operative antibiotic therapy we changed the antibiotics and examined the susceptibility of the postoperative infection. Methods. We studied 4 593 patients who underwent gastrointestinal surgery during the last 12.5 years, and examined the changes in intraoperative antibiotics, the return of bacteria isolated from infectious sites, the trends in frequency of Methicillin-resistant Staphylococcus aureus (MRSA) isolation, and changes in the antibiotic susceptibility of Pseudomonas aeruginosa and Bacterioides fragilis. We changed the antibiotics to Cefazolin (CEZ) for upper gastrointestinal tract surgery and cholecystectomy, and to Cefotiam (CTM) for colonic, liver, and pancreatic surgery. We also reduced the period of drug administration. Results. The rate of MRSA infections decreased, the rate of P. aeruginosa infections was always under 20% and the rate of B. fragilis infections increased. After the guidelines of drug susceptibility were prepared, the minimum inhibitory concentrations (MICs) of Cefsulodin (CFS) and Piperacillin (PIPC) for P. aeruginosa and of Latamoxef (LMOX) and PIPC for B. fragilis, decreased. Conclusion. The use of antibiotics considered comparatively old for the prophylaxis of postoperative infection prevented the emergence of MRSA-like resistant strains, influenced changes in Gram-negative bacteria drug susceptibility, and led to an overall reduction in multiple drug resistance. Key words Methicillin-resistant Staphylococcus aureus · Pseudomonas aeruginosa · Bacteroides fragilis · Imipenem · SSI

Reprint requests to: S. Kusachi Recieved: January 30, 2003 / Accepted: November 4, 2003

Introduction Cephem antibiotics were developed in earnest from the 1970s, and were successively introduced for administration at the time of surgery to prevent postoperative infections. When first-generation cephems appeared, Escherichia coli and Klebsiella spp. were often isolated from postoperative infection sites, but with the emergence of second-generation cephems, the frequency of Pseudomonas aeruginosa and Bacteroides fragilis isolates from postoperative infection sites increased. When third-generation cephems emerged, the frequency of Staphylococcus aureus isolates increased. It soon became obvious that changing the antimicrobial drug in the operation room would change the postoperative infection site isolates. Moreover, the antibiotics used at the time of surgery changed with trends in the medical industry and there are reports on the return of postoperative infection site isolates during that time. Recently, some medical departments have started giving the types of antibiotics used many years ago, but to our knowledge, this is the only report that investigates postoperative infection site isolates in this respect. In 1990, we made our own independent guidelines for the use of antibiotics during surgery in an attempt to decrease the incidence of MRSA infections, monitored by an Infection Control Doctor. Our guidelines on antimicrobial drugs in the surgical period were drawn up, and consisted of a change to Cefazolin (CEZ) for upper gastrointestinal tract surgery and cholecystectomy, and to Cefotiam (CTM) for lower gastrointestinal tracts liver, and pancreatic surgery. Moreover the period of drug administration was reduced. This resulted in a subsequent significant decrease in MRSA isolates. During the last 12.5 years, we studied cases of gastrointestinal surgery before and after the enactment of the guidelines, and investigated the changes in antibiotics given during the surgical period

S. Kusachi et al.: Recovery of Susceptibility of Isolated Bacterium

726

Table 1. Operation sites and diseases found in the patients studied Term Organ

Disease

A

B

C

D

E

Tumor Others Tumor Perforation Tumor Others Stone Tumor Tumor Other Tumor Perforation Others

21 1 194 33 7 4 137 7 12

19 1 166 18 11

18 1 181 18 15

15 4 159 31 18

16 0 178 21 15

156 9 12

203 7 16

277 8 14

Appendix Others

135 17 2 153 110

174 18 3 217 95

216 15 3 129 83

201 7 17 3 228 16 2 125 92

Total

833

899

905

918

1043 4598

Esophagus Stomach Liver Gallbladder Pancreas Colon

211 10 5 182 106

The number of surgical patients increased, but there was no significant difference in the incidence of various disorders

and the return of bacteria isolated from postoperative infection sites. The study focused on the trends in the frequency of MRSA isolation, and the changes in the antibiotic susceptibility of P. aeruginosa and B. fragilis.

of the National Committee for Clinical Laboratory Standards (NCCLS). An emergency department was established at this clinic in 1998, during the study period, so the number of emergency cases increased thereafter. Testing was done using the chi-squared test, and differences were considered significant at P ⬍ 0.05.

Results Changes in Antimicrobial Drugs The antimicrobial drugs used during the surgical period in Term A were the third-generation cephems. LMOX and Ceftizoxime (CZX) and the second-generation cephems CTM and Cefmetazole (CMZ), which were administered in combination with the aminoglycosides Amikacin (AMK), to prevent postoperative infection. The administration period was 11.4 ⫾ 3.1 dyas (Fig. 1). In Term B and subsequent periods, the main drugs used were the first-generation cephem CEZ and the second-generation cephems CTM or CMZ and the period of administration gradually decreased. In Term E, the mean period of administration was significantly reduced to 4.3 ⫾ 1.9 days. Furthermore, the types and doses of antimicrobial drugs administered annually were also reduced significantly from an initial 12 drugs and a cumulative dosage of 33 200 g to 6 drugs, including 3 main ones, and 11 700 g. Changes in the Incidence of Postoperative Infection

Materials and Methods We prospectively studied 4 598 patients who underwent gastrointestinal (GI) surgery between September 1987 and February 2000 (Table 1). The patients were grouped into surgical periods of 2.5 years each, ranging from Term A to Term E, and comparatively investigated. In Term A, the attending physician decided on the antimicrobial drugs, but from Term B we established our own guidelines. For upper GI tract surgery, CEZ was given to prevent postoperative infection, and commenced immediately before surgery. CTM was selected for lower GI tract surgery, and given by the same method. The items investigated were changes in the types, dosage, and administration period of antimicrobial drugs, postoperative infection rates and infection site trends, and the return of isolates to the clinical setting. We especially compared the minimum inhibitory concentrations (MICs) of typical antimicrobial drugs used for P. aeruginosa and B. fragilis. MIC measurements were based on the standards of the Japan Society of Chemotherapy and the MIC standard values were those

The overall rate of postoperative infection after gastrointestinal surgery decreased slightly from 12.9% in Term A to 12.0% in Term D; however, in Term E the postoperative infection rate increased to 13.7% because of the increase in emergency surgery, but the change was not significant (Table 2). Changes in the Bacterium Isolated from Infectious Sites A comparison of the postoperative infection site isolates from Term A to Term E revealed that the rate of MRSA and Klebsiella decreased. Pseudomonas aeruginosa was the most frequent isolate throughout all of the periods. Although its isolation rate decreased from 27.1% in Term A to 22.3% in Term E, it was always above 20%. The frequency of E. coli, and Enterococcus isolation did not change remarkably throughout the study period but the frequency of B. fragilis isolation gradually increased (Fig. 2).

There was no significant difference in the overall postoperative infection rate in the total study period, but the rate of MRSA infection decreased significantly from Term B to Term D. The MRSA infection rate increased significantly during Term E, after the emergency room opened in our hospital SSI, surgical site infection; Sup., superficial; Resp., respiratory; Cath., catheter; UTI, urinary tract infection; E.C., Escherichia coli; MRSA, methicillin-resistant Staphylococcus aureus

13.70% (143/1043) 11.10% (116/1043) 7.00% (73/1043) 3.20% (33/1043) (†6) 1.10% (11/1043) (†5) 1.80% (19/1043) 1.10% (11/1043) 0.20% (2/1043) 2.70% (28/1043) (†6) 0.60% (6/1043) 12.00% (110/918) 9.60% (88/918) 6.90% (63/918) 2.30% (25/918) (†4) 1.00% (9/918) (†4) 2.30% (21/918) 1.40% (13/918) 0 0.40% (4/918) (†3) 0.40% (4/918) 11.10% (100/899) 8.60% (77/899) 6.50% (59/899) 2.00% (18/899) (†3) 0.70% (7/899) (†3) 1.70% (15/899) 0.90% (8/899) 0 0.20% (2/899) (†2) 0.30% (3/899) Incidence of postoperative infection SSI Sup. SSI Deep SSI Resp. Cath. UTI E.C. MRSA infection Fatal rate

12.90% (108/833) 8.90% (74/833) 6.40% (53/833) 2.60% (22/833) (†4) 1.70% (14/833) (†4) 2.00% (17/833) 1.20% (10/833) 1.20% (10/833) (†1) 4.10% (34/833) (†4) 0.60% (5/833)

10.90% (99/905) 8.20% (74/905) 5.70% (52/905) 2.40% (22/905) (†3) 0.80% (7/905) (†2) 2.10% (19/905) 1.20% (11/905) 0 0.20% (2/905) (†2) 0.30% (3/905)

E D C B A

Table 2. Incidence and sites of postoperative infections

Term

S. Kusachi et al.: Recovery of Susceptibility of Isolated Bacterium

727

Drug Susceptibility of Isolates Changes in Drug Susceptibility of P. aeruginosa The MIC of CFS for P. aeruginosa was compared between Term A before the change in antibiotics, and the subsequent periods. The CFS resistance was at least 32 µg/ml based on the NCCLS, so an assessment at this value showed that in Term A, 30% of infections were resistant strains, but in Term B and subsequent periods, the MIC recovered, and the percentage of resistant strains decreased significantly, to about 10% (Fig. 3) The resistance of P. aeruginosa to PIPC was at least 128 µg/ml. Therefore, in Term A, 10% were resistant strains, but in Term B and subsequent periods, all infections comprised susceptible strains (Fig. 4). Changes in B. fragilis Drug Susceptibility The MIC distribution for B. fragilis was largely dependent on the time of isolation, so we compared the strains isolated soon after postoperative infection and the strains isolated after various antimicrobial drugs were given for postoperative infection. Most of the early isolates were from wound infections in patients who underwent elective surgery, whereas late phase isolates were from peritoneal abscess after suture failure or emergency surgery. The resistance to LMOX according to the NCCLS was 32 µg/ml, with no more than 10% of strains in the late phase isolates, but 50% of the late phase isolates were resistant strains (Fig. 5). The resistance to PIPC according to the NCCLS was 64 µg/ml, so all early isolates were susceptible. Even among the late isolates, no more than 10% were resistant strains (Fig. 6). The resistance to Imipenem (IPM) was 32 µg/ml, so both the early and late phase isolates were susceptible. However, among the late phase isolates, the cumulative MIC moved to the left, so the strains were highly susceptible (Fig. 7).

Discussion In Japan, the number of nosocomial MRSA infections increased rapidly from the late 1980s.1,2 MRSA enteritis after gastrointestinal surgery attracted much attention because of its rapid onset and fatal outcome if treatment was delayed.3 According to a questionnaire survey sent to 252 institutions, the method of administering postoperative infection prophylaxis antibiotics was one of the causes of MRSA infection.4 Thus, in September 1990, guidelines on the use of antimicrobial drugs in the surgical period according to the organ operated on were drawn up and put into practice.5 Before the enactment of these guidelines, the antimicrobial drugs used

S. Kusachi et al.: Recovery of Susceptibility of Isolated Bacterium

728

CTM

CEZ

Fig. 1. Changs in the antibiotics given drugs during the operative phase. In Term A, 11 drugs were given for an average of 11.4 days and the total dose was 33 200 g. After Term B, when antibiotics were changed, four major drugs were given and their doses were decreased gradually. After Term D, both the duration and total dose were decreased significantly, and more than 70% of these antibiotics were either CEZ or CTM. Term A: September 1987–February 1990; Term B: March 1990–August 1992; Term C: September 1992–February 1995;

Term D: March 1995–August 1997; Term E: September 1997– February 2000. CTM, Cefotiam; CMZ, Cefmetazole; LMOX, Latamoxef; CMX, Cefmenoxime; CZX, Ceftizoxime; CFS, Cefsulodin; CTX, Cefotaxime; FOM, Fosfomycin; AMK, Amikacin; PIPC, Piperacillin; VCM, Vancomycin; CZON, Cefzonam; PAPM/BP, Panipenem/Betamipron; CEZ, Cefazolin; FMOX, Flomoxef; MSSA, methicillin-sensitive Staphylococcus aureus

Fig. 2. Changes in isolated bacterium. After Term B, the frequency of MRSA and Klebsiella spp. isolation decreased, but the frequency of Pseudomonas aeruginosa and Bacteroides fragilis isolation increased. MRSA, methicillin-resistant Sta-

phylococcus aureus; E. coli, Escherichia coli; E. cloacae, Enterobacter cloacae; S. epidermidis, Staphylococcus epidermidis

S. Kusachi et al.: Recovery of Susceptibility of Isolated Bacterium

729

Fig. 3. Changes in minimum inhibitory concentrations (MICs) of P. aeruginosa against Cefsulodin (CFS). The susceptibility of P. aeruginosa by The National Committee for Clinical Laboratory Standards (NCCLS) to CFS was 32 µg/ml. In Term A, 30% of P. aeruginosa was resistant and after Term B that fraction decreased to 10%. The MIC accumulation curve was shifted to the left after Term B, in comparison with Term A, and drug sensitivity was recovered

Fig. 4. Changes in MICs of P. aeruginosa against Piperacillin (PIPC). The tolerable concentration of P. aeruginosa by NCCLS for PIPC was 128 µg/ml. In Term A 10% of P. aeruginosa was resistant and after Term B there were no resistant bacteria. The MIC accumulation curve was shifted to the left after Term B, in comparison with Term A, and drug sensitivity was recovered

Fig. 5. Changes in MICs of B. fragilis against Latamoxef (LMOX). The tolerable concentration of B. fragilis by NCCLS for LMOX was 32 µg/ml. In Term A, 50% of B. fragilis was resistant and after Term B that fraction decreased to 10%. The MIC accumulation curve was shifted to the left after Term B, in comparison with Term A, and drug sensitivity was recovered

730

S. Kusachi et al.: Recovery of Susceptibility of Isolated Bacterium

Fig. 6. Changes in MICs of B. fragilis against PIPC. The tolerable concentration of B. fragilis by NCCLS for PIPC was 64 µg/ ml. In Term A, 10% of B. fragilis was resistant to PIPC but after Term B there were no resistant bacteria. The MIC accumulation curve was shifted to the left after Term B, in comparison with Term A, and drug sensitivity recovered

Fig. 7. Changes in MICs of B. fragilis against Imipenem (IPM). The tolerable concentration of B. fragilis by NCCLS for IPM was 32 µg/ml. In Term A, there were no resistant bacteria. The MIC accumulation curve was shifted to the left after Term B, in comparison with Term A, and drug sensitivity was recovered

for preventing postoperative infection were thirdgeneration cephems and second-generation cephems, or these drugs combined with aminoglycosides. Moreover, the period of administration was at least 10 days. After the enactment of the guidelines, the central drugs used were first-generation and second-generation cephems and the administration period was gradually reduced. The types and dosage of antimicrobial drugs were also decreased significantly, from 12 types and 33 200 g before the change to 6 drugs, including 3 main ones, and 11 700 g. Despite the increase in emergency admissions in Term E the overall rate of infection after gastrointestinal surgery did not increase significantly during this period.

Regarding the selection of prophylactic drugs for postoperative infection, Nichols et al. recommended that the primary choice be first-generation cephams.4 Based on fundamental research using rats with MRSA, Aoyagi suggested that selecting antimicrobials that severely upset the intestinal bacterial flora could promote MRSA by reducing microbial substitution.6 Moreover, the decrease in E. coli and Enterococcus faecalis verified the increase in MRSA in the enteric canal. Accordingly, when there was a high risk of postoperative infection with MRSA, even with aseptic or semi-contaminated surgery, first-generation cephems were indicated for upper gastrointestinal surgery involving relatively minor contamination. In Western countries, antibiotics

S. Kusachi et al.: Recovery of Susceptibility of Isolated Bacterium

are generally given only during surgery.5 We reduced the administration period from 11.3 days in Term A to 4.3 days in Term E, but this is still considerably different from the Western guidelines, although surgery itself differs between Japan and the West. In Japan, where stomach cancer is common, extended dissection is often performed, resulting in large defects in the peritoneum and loss of the potent infection preventative mechanism of the peritoneum. Moreover, extensive damage to the peritoneum causes peritoneal effusion, increasing the intra-abdominal dead space, and resulting in deep surgical site infection. For these reasons, the Western view of only giving prophylactic antibiotics during surgery will need to be further investigated to determine if it is applicable in Japan. Cephem antibiotics were developed in earnest from the 1970s, and were successively introduced at the time of surgery for the purpose of preventing postoperative infections. When first-generation cephems appeared, E. coli and Klebsiella were frequently isolated from postoperative infection sites, but with the emergence of second-generation cephems, the frequency of both P. aeruginosa and B. fragilis isolates from postoperative infection sites increased. Then when third-generation cephems emerged, the frequency of S. aureus isolates increased. It became obvious that changing the antimicrobial drug administered in the operating room and surgical ward would change the postoperative infection site isolates. In fact, the antibiotics given at the time of surgery changed with trends in the medical industry and there are reports on the return of postoperative infection site isolates during that time. However, in some medical departments there has been a return to using the types of antibiotics commonly used decades ago, but as far as we can ascertain, our report is the first one that investigates the postoperative infection site isolates in relation to this change. We sought to reduce MRSA infection by altering the antimicrobial drugs used in the operation room, but changes also occurred in Gramnagative bacteria susceptibility as the number of cases

731

of MRSA decreased. After the guidelines were introduced, the MICs of CFS and PIPC for P. aeruginosa, and the MICs of LMOX and PIPC for B. fragilis decreased. This showed that the selection of therapy in the clinical setting had become easier, and also highlighted the rediscovered usefulness of drugs that had temporarily been disregarded, and the unnecessary development of new drugs. Thus, the use of drugs considered comparatively old as prophylaxis against postoperative infection prevented the emergence of MRSA-like resistant strains, influenced changes in Gram-negative bacteria drug susceptibility, and led to a reduction in multiple drug resistance. This not only improved the prognosis of patients with postoperative infections and helped individual patients, but also brought about social benefits, such as curtailing spiraling medical expenses accompanying already limited new antibacterial drug development funds and the emergence of multiple drug resistant strains.

References 1. Kusachi S, Sumiyama Y, Nagao J, Kawai K, Arima Y, Yoshida Y, et al. New methods of control against postoperative methicillinresistant Staphylococcus aureus infection. Surg Today 1999;29:724– 9. 2. Takesue Y, Yokoyama T, Kodama T. Methicillin-resistant Staphylococcus aureus in nosocoimial infection in the surgical ward and operating room. Hiroshima J Med Sci 1989;38:183–6. 3. Hori K, Yura J, Shinagawa N, Sakurai S, Mashita K, Mizuno A. Postoperative enterocolitis and the current status of MRSA enterocolitis: the result of a questionnaire survey in Japan. J Jpn Assoc Infect Dis 1990;63:701–7. 4. Nichols RL. Surgical antibiotic prophylaxis. Med Clin North Am 1995;79:509–22. 5. Sumiyama Y, Yokoyama T. Issue related to the use of antibacterial agents in the field of surgery of digestive organ (in Japanese with English abstact). Jpn J Gastroenterol Surg 1994;27:2358– 67. 6. Aoyagi K. Experimental studies on the selection of antibacterial agents for the prevention of postoperative infection following colon surgery by using rats (in Japanese with English abstract). J Jpn Soc Coloproctol 1995;48:979–91.