When flexible gastrointestinal endoscopic procedures were introduced more than 20 years ago, concern was raised for the development of transient bacteremia leading to serious infections, including endocarditis, cholangitis, central venous line infection, or infection of prostheses or grafts (1–3). However, several types of data indicate that the risk may not be great. Transient bacteremia occurs with many activities and procedures but is usually cleared rapidly by healthy people (4). For example, the frequency of bacteremia associated with chewing hard candy is 17% (5) and with routine rectal examination is 4% (6). The risk of bacteremia for dental procedures is high (30%–90%) (7), but the cumulative burden of bacteremia is as much as 1,000 times greater for mastication than for dental extraction (8). Using case-control methodology, no increased risk of endocarditis has been found in persons with cardiac abnormalities undergoing dental procedures (9). These findings have led some to question the need for antimicrobial treatment for routine dental procedures (8–11).
The frequency of bacteremia associated with upper or lower gastrointestinal endoscopy with or without biopsy is low, ranging from about 2% to 5% (1,2,12). The reported risk of bacteremia with endoscopic retrograde cholangiopancreatography (ERCP) is similar (5%–6%) to that for upper and lower endoscopy unless an obstructed biliary ductal system is present (2,13–15). A substantially higher rate of bacteremia has been reported for variceal sclerotherapy (31%) (16–18) and esophageal dilation (20%–40%) (19–22). However, infections associated with gastrointestinal endosopic procedures are rare, including those performed in children (12); very few published reports exist that document infective endocarditis or other serious infection as a sequela of endoscopy (2,8,23,24). Indeed, the incidence of bacterial endocarditis has not increased despite the estimated millions of flexible endoscopic procedures performed in adults and children each year (2).
Criteria for the use of prophylactic antibiotics have been developed for endoscopic procedures based on the associated risk of developing bacteremia (1,2). Procedures considered to be of low risk include esophagogastroduodenoscopy (EGD) with or without biopsy, flexible sigmoidoscopy and colonoscopy with or without biopsy, and ERCP in the absence of ductal obstruction (1,2). Higher-risk endoscopic procedures include esophageal stricture dilation, esophageal sclerotherapy, and ERCP of an obstructed biliary tree. Prospective, randomized, controlled data support the use of prophylactic antibiotics for the placement of percutaneous endoscopic gastrostomy (PEG) tubes (25).
The indiscriminate use of prophylactic antibiotics is not without risk and can result in excess cost and the potential for complications such as allergic drug reactions and antibiotic-associated colitis (9). In addition, no prospective, controlled studies have documented that the use of prophylactic antibiotics prevents infective endocarditis (2,9).
When the use of antibiotic prophylaxis for endoscopic procedures is considered, the following factors should be evaluated: 1) the degree to which the patient's underlying condition creates a risk of severe infection, 2) the risk of bacteremia associated with the procedure, 3) the potential adverse reactions of the prophylactic antimicrobial agent, and 4) cost–benefit aspects of the prophylaxis (1). The American Heart Association (AHA) divides cardiac disease into high-, moderate-, and negligible-risk conditions based on potential outcome if endocarditis develops (Table 1) (1). Even fewer data are available to assess the risk of infection for conditions such as immunodeficiency or the presence of a foreign body such as a central venous line or ventriculo-peritoneal shunt.
No specific guidelines have been developed for the use of prophylactic antibiotics for endoscopic procedures in children. We surveyed academic pediatric centers in the United States and Canada to determine the current practices of these groups and to compare these practices with the recommendations of the AHA and the American Society of Gastrointestinal Endoscopy (ASGE).
Representatives from 15 pediatric gastroenterology programs at academic centers in the United States and Canada were asked to complete a written questionnaire detailing the antibiotic prophylaxis practices for their group. One representative completed the questionnaire, but the responses often represented a consensus of the practitioners at each center. The centers surveyed included three in Canada (B.C. Children's Hospital, Vancouver; Hospital for Sick Children, Toronto; and Hospital Ste. Justine, Montreal) and 12 from the United States (Children's Hospital and Massachusetts General Hospital, Boston; Children's Hospital of Philadelphia; Miami Children's Hospital; University of Alabama; Children's Hospital, Cincinnati; J. Whitcomb Riley Hospital for Children, Indianapolis; Baylor University; Children's Hospital, Denver; the University of California, Los Angeles; and the University of California, San Francisco). These centers were selected because they have large clinical pediatric gastrointestinal practices that have active endoscopy services and represent a wide geographic cross-section of programs.
The questionnaire sought information on antibiotic prophylaxis for six conditions and six procedures, which yielded 36 categories. The six conditions evaluated included three related to congenital heart disease as well as immune compromise, the presence of a central venous line, and the presence of a ventriculo-peritoneal shunt. Six procedures were evaluated, including EGD with biopsy, flexible sigmoidoscopy or colonoscopy with biopsy, ERCP, esophageal sclerotherapy, esophageal dilation, and PEG tube placement.
Responses were compared with the guidelines for prophylaxis established by the AHA and the ASGE (1,2). The cardiac conditions are grouped into those with high, moderate, and negligible risk of infectious complication (Table 1) (1). For the three noncardiac conditions considered, the AHA has no recommendations, and the ASGE lists recommendations only for immunocompromised patients (2). The ASGE recommendations are influenced more by the nature of the procedure than the condition; because data on infection risk are few, routine prophylaxis is not recommended even for highrisk cardiac conditions if the procedure has a low risk of infection (2).
Procedures considered by the AHA and ASGE to have a low risk of associated infection include EGD with or without biopsy, flexible sigmoidoscopy or colonoscopy with or without biopsy, and ERCP with no evidence of obstruction (Table 1). They recommend no routine antibiotic prophylaxis for lower-risk procedures and feel that evaluation should be performed on a case-by-case basis (1,2). Procedures with a higher risk of infectious complication include esophageal dilation, esophageal sclerotherapy, and placement of a PEG (1,2).
The six conditions and six procedures yielded 36 categories for evaluation; the responses from the 15 pediatric gastroenterology programs are summarized in Tables 2–7. There was no difference in the pattern of response by the country or by the regions within a country of the programs. Most centers indicated that they did not have a written policy and that their responses were often based on a consensus of the endoscopists at their institution.
The responses indicated that the nature of the procedure and the child's condition both were important determinants as to whether prophylactic antibiotics were used. The majority of centers reported routine use of antibiotic prophylaxis in about half (17) of the response categories, which represented three specific situations. The first was cardiac conditions with a high risk of bacterial endocarditis in which the majority of centers used prophylactic antibiotics for all six procedures performed (Tables 2–7). The second was the procedure of PEG placement, in which antibiotic prophylaxis was used by the majority of centers for all six conditions (Table 7). The majority of centers also routinely gave antibiotic prophylaxis to children with cardiac conditions with a moderate risk of bacterial endocarditis for all procedures except for EGD with biopsy, where 9 of the 15 centers do not (Table 2).
In contrast, the majority of responding programs did not use antibiotic prophylaxis for the three low-risk procedures or for the negligible-risk cardiac conditions. The pattern of prophylaxis use for the three higher-risk procedures is found in Tables 5–7. The majority of the centers did not use prophylaxis for dilation or sclerotherapy for the three clinical conditions for which few data on risk are available: patients with immunocompromise, central lines, and ventriculo-peritoneal shunts.
Unanimity of opinion was rarely reported. In only two response categories did all the centers agree on antibiotic prophylaxis: use of antibiotics for ERCP and PEG placement in patients with high cardiac risk for endocarditis.
The results indicate that the majority of these academic pediatric centers follow the guidelines recommended by the AHA and the ASGE and use antibiotic prophylaxis only in selected situations. The specific clinical conditions in which the majority of the centers routinely use antibiotic prophylaxis are for children with cardiac lesions that are at high or moderate risk for bacterial endocarditis. The only procedure in which a majority of the centers use antibiotic prophylaxis for all underlying clinical conditions is PEG placement, which is the only procedure for which randomized, controlled data exist to support its use (25).
From the practices reported by participating centers, only a small proportion of pediatric patients undergoing endoscopy receive prophylaxis. Children who have high- or moderate-risk cardiac conditions and those receiving PEGs represent approximately half of the categories evaluated in this survey, but they represent far fewer patients than those without cardiac disease and those undergoing endoscopy, flexible sigmoidoscopy, or colonoscopy.
The practice of limiting the use of prophylactic antibiotics to high-risk groups is consistent with the current recommendation to limit prophylaxis for dental procedures to those with a high risk of infectious complications when performed in patients with prosthetic valves or previous endocarditis (9–11). Adoption of the practice of limiting antibiotic prophylaxis has the potential to reduce the cost and morbidity of many dental and gastrointestinal procedures and to help reduce the growing problem of antibiotic-resistant organisms.
Few data exist on the need for antibiotic prophylaxis for children with immunocompromise or presence of foreign bodies, including central venous lines and ventriculo-peritoneal shunts. The ASGE does not recommend routine use of antibiotic prophylaxis for immunocompromised patients undergoing endoscopy, flexible sigmoidoscopy or colonoscopy, and ERCP without ductal dilation (2). No recommendations are available for the children with central venous lines or ventriculo-peritoneal shunts. The practices of the participating academic centers indicate that routine antibiotic prophylaxis may not be required for these procedures. The only exception in these populations was the routine use prophylaxis for PEG placement.
Caution should be used in interpreting these results. No data on the rate of infectious complications of endoscopic procedures at the participating centers was reviewed for this study. The responses represent the reported practices of the endoscopists at the institutions, and no quantitation or validation of practices at the centers was sought. However, the fact that large, busy clinical gastrointestinal services often do not report the use of antibiotic prophylaxis for many conditions and procedures indicates that infectious complications are likely to be uncommon.
We conclude that the spirit of the AHA and ASGE guidelines are followed closely by the pediatric gastrointestinal centers in the United States and Canada that participated in this study. Most of these centers use prophylactic antibiotics only in selected cases, which represent a small proportion of all children undergoing endoscopic procedures.
1. Dajani AS, Taubert KA, Wilson W, et al. Prevention of bacterial endocarditis: Recommendations by the American Heart Association. JAMA 1997; 277:1794–801.
2. .Antibiotic prophylaxis for gastrointestinal endoscopy
ASGE Publication No. 1027, May 1995.
3. Antibiotic prophylaxis of infective endocarditis: Recommendations from the Endocarditis Working Party of the British Society for Antimicrobial Chemotherapy. Lancet 1990; 335:88–9.
4. Botoman AV, Surawicz CM. Bacteremia with gastrointestinal endoscopic procedures. Gastrointest Endosc 1986; 32:342–6.
5. Kaye D, ed. Infective endocarditis.
Baltimore: University Park Press; 1976:248–9.
6. Hoffman BI, Kobasa W, Kaye D. Bacteremia after rectal examination. Ann Intern Med 1978; 88:658–9.
7. Everett ED, Hirschmann JV. Transient bacteremia and endocarditis prophylaxis: A review. Medicine 1977; 56:61–77.
8. Ulualp K, Condon RE. Antibiotic prophylaxis for scheduled operative procedures. Infect Dis Clin North Am 1992; 6:613–25.
9. Strom BL, Abrutyn E, Berlin JA, et al. Dental and cardiac risk factors for infective endocarditis: A population-based, case-control study. Ann Intern Med 1998; 129:761–9.
10. Van der Meer JT, Thompson J, Valkenburg HA, Michel MF. Epidemiology of bacterial endocarditis in the Netherlands: II. Antecedent procedures and use of prophylaxis. Arch Intern Med 1992; 152:1869–73.
11. Durack DT. Antibiotics for prevention of endocarditis during dentistry: Time to scale back? Ann Intern Med 1992; 129:829–30.
12. Byrne WJ, Euler AR, Campbell M, Eisenach KD. Bacteremia in children following upper gastrointestinal endoscopy or colonoscopy. J Pediatr Gastroenterol Nutr 1982; 1:551–3.
13. Low D, Shoenut P, Kennedy J, et al. Risk of bacteremia with endoscopic sphincterotomy. Can J Surg 1987; 30:421–3.
14. Low DE, Micflikier AB, Kennedy JK, et al. Infectious complication of endoscopic retrograde cholangiopancreatography. Arch Intern Med 1980; 140;1076–7.
15. Dutta S, Cox M, Williams R, et al. Prospective evaluation of the risk of bacteremia and the role of antibiotics in ERCP. Clin Gastroenterol 1983; 5:325–9.
16. Cohen LB, Korsen MA, Scherl EJ, et al. Bacteremia after endoscopic injection sclerosis. Gastrointest Endosc 1983; 29:198– 200.
17. Camara DS, Gruber M, Barde CJ, et al. Transient bacteremia following endoscopic injection sclerotherapy of esophageal varices. Arch Intern Med 1983; 143:1350–2.
18. Snady H, Korsten MA, Waye JD. The relationship of bacteremia to the length of the injection needle in endoscopic variceal sclerotherapy. Gastrointest Endosc 985; 21:243–6.
19. Dellipiani AW. The incidence of bacteremia after outpatient Hurst bougienage in the management of benign esophageal stricture. Endoscopy 1983; 31:265–7.
20. Shorvon PJ, Eykyn SJ, Cotton PB. Gastrointestinal instrumentation, bacteremia, and endocarditis. Gut 1983; 24:1078–93.
21. Zucarro G, Richter JE, Rice TW, et al. Viridans Streptococcal bacteremia after esophageal stricture dilation. Gastrointest Endosc 1998; 48:568–73.
22. Nelson DB, Sanderson SJ, Azar MM. Bacteremia with esophageal dilation. Gastrointest Endosc 1998; 48:563–7.
23. Pritchard T, Foust R, Cantey R, Leman R. Prosthetic valve endocarditis due to cardiobacterium hominis occurring after upper gastrointestinal endoscopy. Am J Med 1991; 90:516–8.
24. Logan R, Hastings J. Bacterial endocarditis: A complication of gastroscopy. Br Med J 1988; 296:1107.
25. Jain NK, Larson DE, Schroeder KW, et al. Antibiotic prophylaxis for percutaneous endoscopic gastrostomy: A prospective, randomized, double-blind clinical trial. Ann Intern Med 1987; 107:824–8.