Original Contribution
Prophylactic Antibiotics for Hemorrhoidectomy: Are They Really Needed? Daniel W. Nelson, D.O.1 • Brad J. Champagne, M.D.2 • David E. Rivadeneira, M.D.3 Brad R. Davis, M.D.4 • Justin A. Maykel, M.D.5 • Howard M. Ross, M.D.6 Eric K. Johnson, M.D.1 • Scott R. Steele, M.D.1 1 Department of Surgery, Madigan Army Medical Center, Tacoma, Washington 2 Division of Colorectal Surgery, Case Medical Center, Cleveland, Ohio 3 Division of Colorectal Surgery, Huntington Hospital/North Shore-LIJ Health System, Great Neck, New York 4 Department of Surgery, University of Cincinnati, Cincinnati, Ohio 5 Division of Colorectal Surgery, University of Massachusetts Memorial Medical Center, Worcester, Massachusetts 6 Division of Colorectal Surgery, Meridian Health System, Tinton Falls, New Jersey
BACKGROUND: Hemorrhoidectomy is considered by many to be a contaminated operation that requires antibiotic prophylaxis to lower the incidence of surgical site infection. In reality, little evidence exists to either support or refute the use of antibiotic prophylaxis in this setting. OBJECTIVE: This study aimed to determine if antibiotic prophylaxis is associated with reduced incidence of postoperative surgical site infection following hemorrhoidectomy. DESIGN: This is a retrospective database review. SETTING: This study was conducted at multiple institutions. PATIENTS: All patients undergoing hemorrhoidectomy with minimum 3-month follow-up were included. MAIN OUTCOME MEASURES: The primary outcome measure was the incidence of postoperative surgical site infection.
Disclaimer: The results and opinions expressed in this article are those of the authors, and do not reflect the opinions or official policy of the United States Army or the Department of Defense. Financial Disclosure: None reported. Poster presentation at the meeting of The American Society of Colon and Rectal Surgeons, Phoenix, AZ, April 27 to May 1, 2013. Correspondence: Scott R. Steele, M.D., Department of Surgery, Madigan Army Medical Center, 9040-A Fitzsimmons Ave, Tacoma, WA 98431. E-mail: [email protected] Dis Colon Rectum 2014; 57: 365–369 DOI: 10.1097/DCR.0b013e3182a0e522 © The ASCRS 2014 Diseases of the Colon & Rectum Volume 57: 3 (2014)
RESULTS: Eight hundred fifty-two patients met the inclusion criteria (50.1% female; mean age, 50.0 ± 13.7 years). The prevalence of preoperative risk factors for surgical site infection included 7.7% with a smoking history, 2.5% with diabetes mellitus, 0.8% receiving steroids, and 0.2% with Crohn’s disease. Surgery was performed predominately for 3-column prolapsed internal and mixed internal/external hemorrhoidal disease. All surgeries performed were closed hemorrhoidectomies. Antibiotic prophylaxis was used in a fewer number of cases (41.3% vs 58.7%). Overall, there were only 12 documented postoperative infections identified, producing an overall incidence of 1.4%. Of those patients who developed postoperative surgical site infections, 9 (75%) did not receive antibiotic prophylaxis (p = 0.25). On multivariate regression analysis, no perioperative risk factor was associated with an increased risk of developing a posthemorrhoidectomy surgical site infection. Conversely, there were no adverse antibiotic-related complications such as Clostridium difficile colitis or antibiotic-associated diarrhea in those receiving antibiotic prophylaxis. LIMITATIONS: This study was limited by the retrospective nature of the analysis. CONCLUSIONS: Postoperative surgical site infection is an exceedingly rare event following hemorrhoidectomy. Antibiotic prophylaxis does not reduce the incidence of postoperative surgical site infection, and its routine use appears unnecessary. KEYWORDS: Hemorrhoid; Hemorrhoidectomy; Surgical site infection; Complication; Antibiotic prophylaxis.
365
366
H
emorrhoidal concerns affect more than 10 million Americans annually, making hemorrhoids one of the most common pathologies encountered in surgical practice.1 It has been estimated that nearly 10% of patients presenting with hemorrhoidal symptoms will require operative therapy.2 Practice parameters have been established by the Standards Practice Task Force of the American Society of Colon and Rectal Surgeons for the management of hemorrhoids; however, these guidelines do not provide recommendations for or against the use of antimicrobial prophylaxis before the operative treatment of hemorrhoids.3 Hemorrhoidectomy is generally considered a cleancontaminated (class 2) operation. However, because of the inherent presence of critical indices of bacterial colonization of anal wounds following elective hemorrhoidectomy,4 one could make the case that this represents a contaminated (class 3) operation; and, therefore, the administration of antibiotic prophylaxis to reduce the risk of postoperative surgical site infection (SSI) might make sense. The use of prophylactic antibiotics in hemorrhoidectomy is probably also driven in large part by the theoretical, albeit low incidence of severe life-threatening postoperative perineal sepsis.5–7 However, the actual reported frequency of postoperative infectious complications following hemorrhoidectomy appears to be exceedingly low at 0.5% to 1.2%.8 Furthermore, there is a paucity of evidence to either support or refute the use of antibiotic prophylaxis before elective operative treatment of hemorrhoids. Therefore, the objective of this study was to examine a large, multi-institutional cohort of patients undergoing elective hemorrhoidectomy to determine if antibiotic prophylaxis was associated with a decreased incidence of SSI.
MATERIALS AND METHODS We conducted a multi-institutional retrospective database review from 2006 to 2012. All patients undergoing closed hemorrhoidectomy with a minimum of 3 months followup were included for analysis. The authors, representing 6 institutions, contributed data collected from their own operative records. Data were then assembled into a single database and analyzed. All data analysis was performed at Madigan Army Medical Center. This study was reviewed and approved by the host institutional review board. All data were collected retrospectively and included available preoperative, operative, and postoperative follow-up data. Specific points of interest included patient demographics, whether or not preoperative antibiotic prophylaxis was administered, and the incidence of postoperative SSI. The decision to prescribe antibiotic prophylaxis varied among surgeons and institutions. All surgeons either systematically prescribed antibiotic prophylaxis or uniformly withheld antibiotics. This decision was based primarily on the previous training experience of each
Nelson et al: Antibiotic Prophylaxis for Hemorrhoids
individual surgeon. When prophylactic antibiotics were administered, typical regimens consisted of a single dose, 30 minutes before the incision, of either single- or dualorganism coverage for typical anorectal flora. Regimens consisted of cefoxitin or cefazolin for gram-positive and -negative organisms and clindamycin or metronidazole for anaerobic enteric organisms. Surgical site infections were classified as superficial wound infections in patients presenting before or during routine follow-up with fever, persistent or worsening pain, and signs of local inflammation, drainage, and/or wound dehiscence. Wounds were further characterized as deep SSIs if they required incision and drainage. In addition, frequencies of known perioperative risk factors for SSI such as current history of smoking, diabetes mellitus, Crohn’s disease, and active steroid use were collected and analyzed. Categorical variables were compared by use of the χ2 or Fisher exact test. Binary logistic regression was used to determine the risk of developing a postoperative SSI associated with known perioperative risk factors. Significance was set at a p value of less than 0.05 with a 95% CI. All statistics were performed by using SPSS version 11.1 (SPSS, Inc. Chicago, IL).
RESULTS In total, 852 patients were identified from the contributing authors’ 6 institutions including years 2006 and 2012. The mean age of the group was 50.0 ± 13.7 years, and 50.1% were women. The demographics for all patients included in the study are shown in Table 1. For patients undergoing elective closed hemorrhoidectomy, the majority of patients presented with internal TABLE 1. Demographic data for all patients undergoing elective hemorrhoidectomy (n = 852) Variable Age, mean ± SD Sex Male Female Comorbidities History of diabetes mellitus History of Crohn’s disease Recent or current steroid use History of smoking Preoperative concomitant infection Hemorrhoid diagnosis Internal External Mixed Incomplete data Operative method Closed hemorrhoidectomy Preoperative enema Antibiotic prophylaxis
% (n) 50.0 ± 13.7 49.9 (425) 50.1 (427) 2.5 (21) 0.2 (2) 0.8 (7) 7.7 (66) 0.4 (3) 48.0 (409) 4.1 (35) 43.3 (369) 4.6 (39) 100 (852) 23.5 (200) 41.3 (352)
367
Diseases of the Colon & Rectum Volume 57: 3 (2014)
TABLE 2. Univariate analysis of all perioperative risk factors for posthemorrhoidectomy SSI
Variable Sex Male Female Comorbidities History of diabetes mellitus History of Crohn’s disease Recent or current steroid use History of smoking Hemorrhoid diagnosis Internal External Mixed Preoperative enema Antibiotic prophylaxis
Absence of Presence of postoperative postoperative SSI, % (n) SSI, % (n)
Variable p 0.57
98.8 (420) 98.4 (436)
1.2 (5) 1.6 (7)
100 (21) 100 (2) 100 (7) 98.5 (65)
0 (0) 0 (0) 0 (0) 1.5 (1)
99.0 (405) 100 (35) 97.8 (361) 98.5 (197) 99.1 (349)
1.0 (4) 0 (0) 2.2 (8) 1.5 (3) 0.9 (3)
TABLE 3. Multivariable regression analysis of all perioperative risk factors for posthemorrhoidectomy SSI
0.67 0.90 0.80 0.53 0.30
Sex (male) Comorbidities History of diabetes mellitus History of Crohn’s disease Recent or current steroid use History of smoking Hemorrhoid diagnosis Internal External Mixed Preoperative enema Antibiotics prophylaxis
OR
95% CI
p
0.558
0.090–3.471
0.53
0.000 4.103 0.000 8.874
0 0 0 0.455–173.171
1.00 1.00 1.00 0.15
0.000 0.000 1.019 1.934 0.561
0 0 0.147–7.053 0.211–17.709 0.036–8.755
1.00 1.00 0.99 0.56 0.68
SSI = surgical site infection.
0.65 0.25
SSI = surgical site infection.
hemorrhoidal disease (48.0%) nonresponsive to conservative measures. An additional 43.3% of patients presented with mixed internal/external hemorrhoids, and a small minority of patients underwent elective hemorrhoidectomy for isolated external disease (4.1%). The prevalence of preoperative risk factors for SSI included 7.7% with a current smoking history, 2.5% with diabetes mellitus, 0.8% receiving steroids, and 0.2% with Crohn’s disease. Preoperative preparation with enemas was performed in 23.5% of patients. Three patients (0.4%) were documented as having evidence of concomitant infection preoperatively and were excluded from the final analysis. Antibiotic prophylaxis was used in a slight minority of cases (41.3% vs 58.7%). There were no differences with regard to preoperative risk factors for SSI between those receiving and not receiving antibiotic prophylaxis (all p > 0.05). Overall, there were only 12 documented postoperative SSIs identified, producing an overall incidence of 1.4%. There was 1 deep incisional SSI (0.1%), which consisted of a perirectal abscess that required incision and drainage. The remaining 11 (1.3%) were characterized as superficial incisional SSIs that were treated with antibiotics alone. There were no reported cases of perineal or pelvic sepsis. Of those patients who developed postoperative SSIs, 9 (75%) did not receive antibiotic prophylaxis; however, this difference was not significant (p = 0.25). On univariate analysis, no perioperative risk factor was associated with a greater likelihood of developing a postoperative SSI (Table 2). In addition, on multivariable regression analysis, no perioperative risk factor was associated with increased risk of developing a posthemorrhoidectomy wound infection (Table 3). Conversely, there were no adverse antibiotic-related complications such as Clostridium difficile colitis or antibiotic-associated diarrhea reported in those receiving antibiotic prophylaxis.
DISCUSSION A variety of office-based (sclerotherapy, rubber band ligation, cryotherapy) and surgical (open, closed, and stapled hemorrhoidectomy) procedures are available options for the management of hemorrhoid disease. Although it is difficult to ascertain the true incidence of postoperative SSI following these procedures, in part, because of varying definitions in the literature, all of these methods have uniformly been associated with low rates of postoperative SSI.9–14 This may be quite surprising when one considers the diverse fecal flora intimately associated with the posthemorrhoidectomy wound. In 1991, de Paula et al4 examined the anal wounds of 20 patients that underwent open hemorrhoidectomy and documented the presence of progressively increasing colonization counts during serial cultures up to critical indices of colonization. The most common bacterium identified was Escherichia coli, followed by Staphylococcus aureus and Staphylococcus epidermidis. Pseudomonas aeruginosa, Enterococcus faecalis, Klebsiella pneumonia, Proteus vulgaris, and Proteus mirabilis were also identified in fewer numbers. The diverse flora and high bacterial loads, however, did not impair appropriate healing, and subsequent investigations have suggested that local immunoinflammatory defenses may play a role in wound protection and healing following hemorrhoidectomy.15 Transient bacteremia and postoperative fever has also been well described following anorectal examination and hemorrhoidectomy procedures. In 1 series of 36 patients undergoing open hemorrhoidectomy, transient bacteremia was documented in 8.3% of cases.16 In another series of 50 patients, bacteremia was identified in 2% of patients undergoing proctoscopy and in 8% following sclerotherapy treatment of hemorrhoids.17 Finally, in a series of 205 patients randomly assigned to stapled or diathermy hemorrhoidectomy, bacteremia was documented in 11% and 5% of patients. There were no serious complications or
368
wound infections reported in any of these studies, and the bacteremia resolved without treatment. It has also been postulated that reduction of bacterial loads at posthemorrhoidectomy wound sites may improve wound healing as well as reduce postoperative discomfort. As such, several studies have examined the effects of antibiotic administration in the perioperative period on wound healing and postoperative pain. Both oral18,19 and topical20,21 metronidazole preparations have been trialed. Based on these studies, evidence appears to suggest that perioperative antibiotic administration does have the beneficial effect of reducing posthemorrhoidectomy pain and improving wound healing.19–21 However, in a well-constructed retrospective, matched-pair case-control series of 176 patients, no benefit was noted with the administration of perioperative antibiotics compared with control with regard to overall postoperative complications.18 Furthermore, the authors found identical rates of wound dehiscence (2.25%) in each group at 2 weeks postoperatively, with all wounds healed by 4 weeks, and no reported wound infections. The frequent use of antibiotic prophylaxis before hemorrhoidectomy can largely be attributed to these findings, as well as the theoretical risk of potentially catastrophic complications such as perineal or pelvic sepsis. In addition, perioperative pathways and attempts to comply with Surgical Case Improvement Project guidelines often drive practice patterns. In the present series, the decision to give or withhold antibiotics was left to the attending surgeon and largely based on previous training or practice patterns of the group. Of note, there was no difference in infection rates between institutions where antibiotics were a routine part of hemorrhoid surgery and those where antibiotics were withheld. Furthermore, there was also no difference in infection rates at 1 institution where routine antibiotics were initially given for the first few years out of fellowship training, and then subsequently withheld for the past several years. Although prophylactic antibiotics have been shown to be beneficial in elective colorectal surgery, this is almost exclusively in abdominal and proctectomy cases, and not for anorectal disease.22,23 In reality, a paucity of evidence exists to support or refute the claim that prophylactic antibiotics actually reduce the already exceedingly low rate of postoperative SSI following hemorrhoidectomy. To our knowledge, no study has ever specifically examined the effect of preoperative antibiotic administration on the rate of postoperative SSI. Ultimately, our goals were to both determine if administration of prophylactic antibiotics reduced the incidence of postoperative SSI and, similarly, if withholding antibiotics leads to increased complications or changes the spectrum of complications. Based on our collective anecdotal experience, we were not surprised to find that the overall rate of SSI was exceedingly low at 1.4%. Furthermore, there were no cases of pelvic sepsis or major com-
Nelson et al: Antibiotic Prophylaxis for Hemorrhoids
plications in those cases that did not receive antibiotics. Although 75% (9 of 12) of SSIs were in patients who did not receive antibiotic prophylaxis, the rates of SSI between the 2 groups were not statistically significant with overall infection rates of 1.8% in those not receiving prophylactic antibiotics compared with 0.9% in those patients that did receive antibiotic prophylaxis (p = 0.25). To further clarify the role of prophylactic antibiotics in hemorrhoidectomy and determine if a meaningful difference in postoperative SSI rates exists between those receiving and those not receiving preoperative antibiotics, a prospective randomized trial would be necessary. Based on the rates of SSI in this study of 0.9% (receiving antibiotics) and 1.8% (not receiving antibiotics), with a type 1 error of 0.05 and power of 80%, detecting a difference of 1% would require an estimated sample size of 2600 patients in each study arm to adequately address this issue. For a slightly higher power of 90%, 3500 patients in each group would be necessary. Therefore, although the present study may be somewhat underpowered, it represents a significant population in which a realistic sample size may not be feasible. Furthermore, the current study is ~8-fold larger than any previously published experience and represents the only multi-institutional series examining the use of prophylactic antibiotics in hemorrhoid surgery. This study has certain limitations. First, as with any retrospective database review, it is subject to the inherent biases that accompany such a study design. Definitions used to define surgical site were broad in an attempt to be more inclusive rather than less inclusive. Clearly, persistent pain, signs of local inflammation, and wound dehiscence can occur in the absence of infection, which when included may have overestimated the incidence of postoperative SSI. Patients included in this study encompassed the full spectrum of hemorrhoid disease including all grades of internal, mixed internal/external, and isolated external hemorrhoids. Because of the variability, there exists an inherent risk of introducing a small degree of bias of the cohort as a whole. However, as multivariate regression analysis suggests, when stratified by hemorrhoidal diagnosis alone there was still no difference in the rate of postoperative SSI. In addition, a variety of operative procedures exist for the management of hemorrhoid disease. These procedures may not be entirely comparable, and some may pose a higher risk of postoperative infection. In this study, only closed hemorrhoidectomy was considered. However, open versus closed methods for hemorrhoid surgery has been previously shown to have equivalent results with regard to overall complications.24,25 In addition, the study populations were not equally distributed between those that received prophylactic antibiotics and those that did not. Finally, we acknowledge that patients could have presented to an outside hospital with an infection that may alter the overall rates of SSI; however, all patients were seen in clinic
Diseases of the Colon & Rectum Volume 57: 3 (2014)
that included a p hysical e xamination, and these were generally “closed” health care systems where follow-up was captured. Yet, despite these limitations, this study does represent the largest retrospective examination of the role of prophylactic antibiotics for hemorrhoidectomy.
CONCLUSION We found no difference in postoperative complications between patients receiving and not receiving prophylactic antibiotics for hemorrhoidectomy. No perioperative risk factor was associated with an increased risk of developing a postoperative SSI. Based on the results of this study, a prospective case-matched controlled study is warranted to further clarify the role of prophylactic antibiotics for elective hemorrhoidectomy and to be able to provide the clinician with higher-level evidence-based recommendations. ACKNOWLEDGMENTS The authors thank Aneel Damle (University of Massachusetts Memorial Medical Center), Christopher Porta, Derek McVay, and Avery Walker (Madigan Army Medical Center) for their diligent efforts and a ssistance with data collection. REFERENCES 1. Johanson JF, Sonnenberg A. The prevalence of hemorrhoids and chronic constipation. An epidemiologic study. Gastroenterology. 1990;98:380–386. 2. Bleday R, Pena JP, Rothenberger DA, Goldberg SM, Buls JG. Symptomatic hemorrhoids: current incidence and complications of operative therapy. Dis Colon Rectum. 1992;35:477–481. 3. Rivadeneira DE, Steele SR, Ternent C, Chalasani S, Buie WD, Rafferty JL; Standards Practice Task Force of The American Society of Colon and Rectal Surgeons. Practice parameters for the management of hemorrhoids (revised 2010). Dis Colon Rectum. 2011;54:1059–1064. 4. de Paula PR, Speranzini MB, Hamzagic HC, et al. Bacteriology of the anal wound after open hemorrhoidectomy: qualitative and quantitative analysis. Dis Colon Rectum. 1991;34:664–669. 5. McCloud JM, Jameson JS, Scott AN. Life-threatening sepsis following treatment for haemorrhoids: a systematic review. Colorectal Dis. 2006;8:748–755. 6. Schulte T, Fändrich F, Kahlke V. Life-threatening rectal necrosis after injection sclerotherapy for haemorrhoids. Int J Colorectal Dis. 2008;23:725–726. 7. Sim HL, Tan KY, Poon PL, Cheng A, Mak K. Life-threatening perineal sepsis after rubber band ligation of haemorrhoids. Tech Coloproctol. 2009;13:161–164.
369
8. Guy RJ, Seow-Choen F. Septic complications after treatment of haemorrhoids. Br J Surg. 2003;90:147–156. 9. Sielezneff I, Salle E, Lécuyer J, Brunet C, Sarles JC, Sastre B. Early postoperative morbidity after hemorrhoidectomy using the Milligan-Morgan technic. A retrospective studies of 1,134 cases [in French]. J Chir (Paris). 1997;134:243–247. 10. Arbman G, Krook H, Haapaniemi S. Closed vs. open hemorrhoidectomy–is there any difference? Dis Colon Rectum. 2000;43:31–34. 11. Dixon MR, Stamos MJ, Grant SR, et al. Stapled hemorrhoidectomy: a review of our early experience. Am Surg. 2003;69:862–865. 12. Brusciano L, Ayabaca SM, Pescatori M, et al. Reinterventions after complicated or failed stapled hemorrhoidopexy. Dis Colon Rectum. 2004;47:1846–1851. 13. Iyer VS, Shrier I, Gordon PH. Long-term outcome of rubber band ligation for symptomatic primary and recurrent internal hemorrhoids. Dis Colon Rectum. 2004;47:1364–1370. 14. Pessaux P, Tuech JJ, Laurent B, et al. Morbidity after stapled haemorrhoidectomy: long-term results about 140 patients and review of the literature [in French]. Ann Chir. 2004;129:571–577. 15. de Paula PR, Matos D, Franco M, et al. Why do anal wounds heal adequately? A study of the local immunoinflammatory defense mechanisms. Dis Colon Rectum. 2004;47:1861–1867. 16. Bonardi RA, Rosin JD, Stonesifer GL Jr, Bauer FW. Bacteremias associated with routine hemorrhoidectomies. Dis Colon Rectum. 1976;19:233–236. 17. Adami B, Eckardt VF, Suermann RB, Karbach U, Ewe K. Bacteremia after proctoscopy and hemorrhoidal injection sclerotherapy. Dis Colon Rectum. 1981;24:373–374. 18. Wesarachawit W, Pattana-arun J. Antibiotics and early post operative complications of closed hemorrhoidectomy: a retrospective matched pair study. J Med Assoc Thai. 2007;90:1828–1832. 19. Al-Mulhim AS, Ali AM, Al-Masuod N, Alwahidi A. Post hemorrhoidectomy pain: a randomized controlled trial. Saudi Med J. 2006;27:1538–1541. 20. Ala S, Saeedi M, Eshghi F, Mirzabeygi P. Topical metronidazole can reduce pain after surgery and pain on defecation in postoperative hemorrhoidectomy. Dis Colon Rectum. 2008;51:235–238. 21. Nicholson TJ, Armstrong D. Topical metronidazole (10 percent) decreases posthemorrhoidectomy pain and improves healing. Dis Colon Rectum. 2004;47:711–716. 22. Nguyen N, Yegiyants S, Kaloostian C, Abbas MA, Difronzo LA. The Surgical Care Improvement project (SCIP) initiative to reduce infection in elective colorectal surgery: which performance measures affect outcome? Am Surg. 2008;74:1012–1016. 23. Kobayashi M, Mohri Y, Inoue Y, Okita Y, Miki C, Kusunoki M. Continuous follow-up of surgical site infections for 30 days after colorectal surgery. World J Surg. 2008;32:1142–1146. 24. You SY, Kim SH, Chung CS, Lee DK. Open vs. closed hemorrhoidectomy. Dis Colon Rectum. 2005;48:108–113. 25. Sohn VY, Martin MJ, Mullenix PS, Cuadrado DG, Place RJ, Steele SR. A comparison of open versus closed techniques using the Harmonic Scalpel in outpatient hemorrhoid surgery. Mil Med. 2008;173:689–692.
Prophylactic Antibiotics for Hemorrhoidectomy: Are They Really Needed? Daniel W. Nelson, D.O.1 • Brad J. Champagne, M.D.2 • David E. Rivadeneira, M.D.3 Brad R. Davis, M.D.4 • Justin A. Maykel, M.D.5 • Howard M. Ross, M.D.6 Eric K. Johnson, M.D.1 • Scott R. Steele, M.D.1 1 Department of Surgery, Madigan Army Medical Center, Tacoma, Washington 2 Division of Colorectal Surgery, Case Medical Center, Cleveland, Ohio 3 Division of Colorectal Surgery, Huntington Hospital/North Shore-LIJ Health System, Great Neck, New York 4 Department of Surgery, University of Cincinnati, Cincinnati, Ohio 5 Division of Colorectal Surgery, University of Massachusetts Memorial Medical Center, Worcester, Massachusetts 6 Division of Colorectal Surgery, Meridian Health System, Tinton Falls, New Jersey
BACKGROUND: Hemorrhoidectomy is considered by many to be a contaminated operation that requires antibiotic prophylaxis to lower the incidence of surgical site infection. In reality, little evidence exists to either support or refute the use of antibiotic prophylaxis in this setting. OBJECTIVE: This study aimed to determine if antibiotic prophylaxis is associated with reduced incidence of postoperative surgical site infection following hemorrhoidectomy. DESIGN: This is a retrospective database review. SETTING: This study was conducted at multiple institutions. PATIENTS: All patients undergoing hemorrhoidectomy with minimum 3-month follow-up were included. MAIN OUTCOME MEASURES: The primary outcome measure was the incidence of postoperative surgical site infection.
Disclaimer: The results and opinions expressed in this article are those of the authors, and do not reflect the opinions or official policy of the United States Army or the Department of Defense. Financial Disclosure: None reported. Poster presentation at the meeting of The American Society of Colon and Rectal Surgeons, Phoenix, AZ, April 27 to May 1, 2013. Correspondence: Scott R. Steele, M.D., Department of Surgery, Madigan Army Medical Center, 9040-A Fitzsimmons Ave, Tacoma, WA 98431. E-mail: [email protected] Dis Colon Rectum 2014; 57: 365–369 DOI: 10.1097/DCR.0b013e3182a0e522 © The ASCRS 2014 Diseases of the Colon & Rectum Volume 57: 3 (2014)
RESULTS: Eight hundred fifty-two patients met the inclusion criteria (50.1% female; mean age, 50.0 ± 13.7 years). The prevalence of preoperative risk factors for surgical site infection included 7.7% with a smoking history, 2.5% with diabetes mellitus, 0.8% receiving steroids, and 0.2% with Crohn’s disease. Surgery was performed predominately for 3-column prolapsed internal and mixed internal/external hemorrhoidal disease. All surgeries performed were closed hemorrhoidectomies. Antibiotic prophylaxis was used in a fewer number of cases (41.3% vs 58.7%). Overall, there were only 12 documented postoperative infections identified, producing an overall incidence of 1.4%. Of those patients who developed postoperative surgical site infections, 9 (75%) did not receive antibiotic prophylaxis (p = 0.25). On multivariate regression analysis, no perioperative risk factor was associated with an increased risk of developing a posthemorrhoidectomy surgical site infection. Conversely, there were no adverse antibiotic-related complications such as Clostridium difficile colitis or antibiotic-associated diarrhea in those receiving antibiotic prophylaxis. LIMITATIONS: This study was limited by the retrospective nature of the analysis. CONCLUSIONS: Postoperative surgical site infection is an exceedingly rare event following hemorrhoidectomy. Antibiotic prophylaxis does not reduce the incidence of postoperative surgical site infection, and its routine use appears unnecessary. KEYWORDS: Hemorrhoid; Hemorrhoidectomy; Surgical site infection; Complication; Antibiotic prophylaxis.
365
366
H
emorrhoidal concerns affect more than 10 million Americans annually, making hemorrhoids one of the most common pathologies encountered in surgical practice.1 It has been estimated that nearly 10% of patients presenting with hemorrhoidal symptoms will require operative therapy.2 Practice parameters have been established by the Standards Practice Task Force of the American Society of Colon and Rectal Surgeons for the management of hemorrhoids; however, these guidelines do not provide recommendations for or against the use of antimicrobial prophylaxis before the operative treatment of hemorrhoids.3 Hemorrhoidectomy is generally considered a cleancontaminated (class 2) operation. However, because of the inherent presence of critical indices of bacterial colonization of anal wounds following elective hemorrhoidectomy,4 one could make the case that this represents a contaminated (class 3) operation; and, therefore, the administration of antibiotic prophylaxis to reduce the risk of postoperative surgical site infection (SSI) might make sense. The use of prophylactic antibiotics in hemorrhoidectomy is probably also driven in large part by the theoretical, albeit low incidence of severe life-threatening postoperative perineal sepsis.5–7 However, the actual reported frequency of postoperative infectious complications following hemorrhoidectomy appears to be exceedingly low at 0.5% to 1.2%.8 Furthermore, there is a paucity of evidence to either support or refute the use of antibiotic prophylaxis before elective operative treatment of hemorrhoids. Therefore, the objective of this study was to examine a large, multi-institutional cohort of patients undergoing elective hemorrhoidectomy to determine if antibiotic prophylaxis was associated with a decreased incidence of SSI.
MATERIALS AND METHODS We conducted a multi-institutional retrospective database review from 2006 to 2012. All patients undergoing closed hemorrhoidectomy with a minimum of 3 months followup were included for analysis. The authors, representing 6 institutions, contributed data collected from their own operative records. Data were then assembled into a single database and analyzed. All data analysis was performed at Madigan Army Medical Center. This study was reviewed and approved by the host institutional review board. All data were collected retrospectively and included available preoperative, operative, and postoperative follow-up data. Specific points of interest included patient demographics, whether or not preoperative antibiotic prophylaxis was administered, and the incidence of postoperative SSI. The decision to prescribe antibiotic prophylaxis varied among surgeons and institutions. All surgeons either systematically prescribed antibiotic prophylaxis or uniformly withheld antibiotics. This decision was based primarily on the previous training experience of each
Nelson et al: Antibiotic Prophylaxis for Hemorrhoids
individual surgeon. When prophylactic antibiotics were administered, typical regimens consisted of a single dose, 30 minutes before the incision, of either single- or dualorganism coverage for typical anorectal flora. Regimens consisted of cefoxitin or cefazolin for gram-positive and -negative organisms and clindamycin or metronidazole for anaerobic enteric organisms. Surgical site infections were classified as superficial wound infections in patients presenting before or during routine follow-up with fever, persistent or worsening pain, and signs of local inflammation, drainage, and/or wound dehiscence. Wounds were further characterized as deep SSIs if they required incision and drainage. In addition, frequencies of known perioperative risk factors for SSI such as current history of smoking, diabetes mellitus, Crohn’s disease, and active steroid use were collected and analyzed. Categorical variables were compared by use of the χ2 or Fisher exact test. Binary logistic regression was used to determine the risk of developing a postoperative SSI associated with known perioperative risk factors. Significance was set at a p value of less than 0.05 with a 95% CI. All statistics were performed by using SPSS version 11.1 (SPSS, Inc. Chicago, IL).
RESULTS In total, 852 patients were identified from the contributing authors’ 6 institutions including years 2006 and 2012. The mean age of the group was 50.0 ± 13.7 years, and 50.1% were women. The demographics for all patients included in the study are shown in Table 1. For patients undergoing elective closed hemorrhoidectomy, the majority of patients presented with internal TABLE 1. Demographic data for all patients undergoing elective hemorrhoidectomy (n = 852) Variable Age, mean ± SD Sex Male Female Comorbidities History of diabetes mellitus History of Crohn’s disease Recent or current steroid use History of smoking Preoperative concomitant infection Hemorrhoid diagnosis Internal External Mixed Incomplete data Operative method Closed hemorrhoidectomy Preoperative enema Antibiotic prophylaxis
% (n) 50.0 ± 13.7 49.9 (425) 50.1 (427) 2.5 (21) 0.2 (2) 0.8 (7) 7.7 (66) 0.4 (3) 48.0 (409) 4.1 (35) 43.3 (369) 4.6 (39) 100 (852) 23.5 (200) 41.3 (352)
367
Diseases of the Colon & Rectum Volume 57: 3 (2014)
TABLE 2. Univariate analysis of all perioperative risk factors for posthemorrhoidectomy SSI
Variable Sex Male Female Comorbidities History of diabetes mellitus History of Crohn’s disease Recent or current steroid use History of smoking Hemorrhoid diagnosis Internal External Mixed Preoperative enema Antibiotic prophylaxis
Absence of Presence of postoperative postoperative SSI, % (n) SSI, % (n)
Variable p 0.57
98.8 (420) 98.4 (436)
1.2 (5) 1.6 (7)
100 (21) 100 (2) 100 (7) 98.5 (65)
0 (0) 0 (0) 0 (0) 1.5 (1)
99.0 (405) 100 (35) 97.8 (361) 98.5 (197) 99.1 (349)
1.0 (4) 0 (0) 2.2 (8) 1.5 (3) 0.9 (3)
TABLE 3. Multivariable regression analysis of all perioperative risk factors for posthemorrhoidectomy SSI
0.67 0.90 0.80 0.53 0.30
Sex (male) Comorbidities History of diabetes mellitus History of Crohn’s disease Recent or current steroid use History of smoking Hemorrhoid diagnosis Internal External Mixed Preoperative enema Antibiotics prophylaxis
OR
95% CI
p
0.558
0.090–3.471
0.53
0.000 4.103 0.000 8.874
0 0 0 0.455–173.171
1.00 1.00 1.00 0.15
0.000 0.000 1.019 1.934 0.561
0 0 0.147–7.053 0.211–17.709 0.036–8.755
1.00 1.00 0.99 0.56 0.68
SSI = surgical site infection.
0.65 0.25
SSI = surgical site infection.
hemorrhoidal disease (48.0%) nonresponsive to conservative measures. An additional 43.3% of patients presented with mixed internal/external hemorrhoids, and a small minority of patients underwent elective hemorrhoidectomy for isolated external disease (4.1%). The prevalence of preoperative risk factors for SSI included 7.7% with a current smoking history, 2.5% with diabetes mellitus, 0.8% receiving steroids, and 0.2% with Crohn’s disease. Preoperative preparation with enemas was performed in 23.5% of patients. Three patients (0.4%) were documented as having evidence of concomitant infection preoperatively and were excluded from the final analysis. Antibiotic prophylaxis was used in a slight minority of cases (41.3% vs 58.7%). There were no differences with regard to preoperative risk factors for SSI between those receiving and not receiving antibiotic prophylaxis (all p > 0.05). Overall, there were only 12 documented postoperative SSIs identified, producing an overall incidence of 1.4%. There was 1 deep incisional SSI (0.1%), which consisted of a perirectal abscess that required incision and drainage. The remaining 11 (1.3%) were characterized as superficial incisional SSIs that were treated with antibiotics alone. There were no reported cases of perineal or pelvic sepsis. Of those patients who developed postoperative SSIs, 9 (75%) did not receive antibiotic prophylaxis; however, this difference was not significant (p = 0.25). On univariate analysis, no perioperative risk factor was associated with a greater likelihood of developing a postoperative SSI (Table 2). In addition, on multivariable regression analysis, no perioperative risk factor was associated with increased risk of developing a posthemorrhoidectomy wound infection (Table 3). Conversely, there were no adverse antibiotic-related complications such as Clostridium difficile colitis or antibiotic-associated diarrhea reported in those receiving antibiotic prophylaxis.
DISCUSSION A variety of office-based (sclerotherapy, rubber band ligation, cryotherapy) and surgical (open, closed, and stapled hemorrhoidectomy) procedures are available options for the management of hemorrhoid disease. Although it is difficult to ascertain the true incidence of postoperative SSI following these procedures, in part, because of varying definitions in the literature, all of these methods have uniformly been associated with low rates of postoperative SSI.9–14 This may be quite surprising when one considers the diverse fecal flora intimately associated with the posthemorrhoidectomy wound. In 1991, de Paula et al4 examined the anal wounds of 20 patients that underwent open hemorrhoidectomy and documented the presence of progressively increasing colonization counts during serial cultures up to critical indices of colonization. The most common bacterium identified was Escherichia coli, followed by Staphylococcus aureus and Staphylococcus epidermidis. Pseudomonas aeruginosa, Enterococcus faecalis, Klebsiella pneumonia, Proteus vulgaris, and Proteus mirabilis were also identified in fewer numbers. The diverse flora and high bacterial loads, however, did not impair appropriate healing, and subsequent investigations have suggested that local immunoinflammatory defenses may play a role in wound protection and healing following hemorrhoidectomy.15 Transient bacteremia and postoperative fever has also been well described following anorectal examination and hemorrhoidectomy procedures. In 1 series of 36 patients undergoing open hemorrhoidectomy, transient bacteremia was documented in 8.3% of cases.16 In another series of 50 patients, bacteremia was identified in 2% of patients undergoing proctoscopy and in 8% following sclerotherapy treatment of hemorrhoids.17 Finally, in a series of 205 patients randomly assigned to stapled or diathermy hemorrhoidectomy, bacteremia was documented in 11% and 5% of patients. There were no serious complications or
368
wound infections reported in any of these studies, and the bacteremia resolved without treatment. It has also been postulated that reduction of bacterial loads at posthemorrhoidectomy wound sites may improve wound healing as well as reduce postoperative discomfort. As such, several studies have examined the effects of antibiotic administration in the perioperative period on wound healing and postoperative pain. Both oral18,19 and topical20,21 metronidazole preparations have been trialed. Based on these studies, evidence appears to suggest that perioperative antibiotic administration does have the beneficial effect of reducing posthemorrhoidectomy pain and improving wound healing.19–21 However, in a well-constructed retrospective, matched-pair case-control series of 176 patients, no benefit was noted with the administration of perioperative antibiotics compared with control with regard to overall postoperative complications.18 Furthermore, the authors found identical rates of wound dehiscence (2.25%) in each group at 2 weeks postoperatively, with all wounds healed by 4 weeks, and no reported wound infections. The frequent use of antibiotic prophylaxis before hemorrhoidectomy can largely be attributed to these findings, as well as the theoretical risk of potentially catastrophic complications such as perineal or pelvic sepsis. In addition, perioperative pathways and attempts to comply with Surgical Case Improvement Project guidelines often drive practice patterns. In the present series, the decision to give or withhold antibiotics was left to the attending surgeon and largely based on previous training or practice patterns of the group. Of note, there was no difference in infection rates between institutions where antibiotics were a routine part of hemorrhoid surgery and those where antibiotics were withheld. Furthermore, there was also no difference in infection rates at 1 institution where routine antibiotics were initially given for the first few years out of fellowship training, and then subsequently withheld for the past several years. Although prophylactic antibiotics have been shown to be beneficial in elective colorectal surgery, this is almost exclusively in abdominal and proctectomy cases, and not for anorectal disease.22,23 In reality, a paucity of evidence exists to support or refute the claim that prophylactic antibiotics actually reduce the already exceedingly low rate of postoperative SSI following hemorrhoidectomy. To our knowledge, no study has ever specifically examined the effect of preoperative antibiotic administration on the rate of postoperative SSI. Ultimately, our goals were to both determine if administration of prophylactic antibiotics reduced the incidence of postoperative SSI and, similarly, if withholding antibiotics leads to increased complications or changes the spectrum of complications. Based on our collective anecdotal experience, we were not surprised to find that the overall rate of SSI was exceedingly low at 1.4%. Furthermore, there were no cases of pelvic sepsis or major com-
Nelson et al: Antibiotic Prophylaxis for Hemorrhoids
plications in those cases that did not receive antibiotics. Although 75% (9 of 12) of SSIs were in patients who did not receive antibiotic prophylaxis, the rates of SSI between the 2 groups were not statistically significant with overall infection rates of 1.8% in those not receiving prophylactic antibiotics compared with 0.9% in those patients that did receive antibiotic prophylaxis (p = 0.25). To further clarify the role of prophylactic antibiotics in hemorrhoidectomy and determine if a meaningful difference in postoperative SSI rates exists between those receiving and those not receiving preoperative antibiotics, a prospective randomized trial would be necessary. Based on the rates of SSI in this study of 0.9% (receiving antibiotics) and 1.8% (not receiving antibiotics), with a type 1 error of 0.05 and power of 80%, detecting a difference of 1% would require an estimated sample size of 2600 patients in each study arm to adequately address this issue. For a slightly higher power of 90%, 3500 patients in each group would be necessary. Therefore, although the present study may be somewhat underpowered, it represents a significant population in which a realistic sample size may not be feasible. Furthermore, the current study is ~8-fold larger than any previously published experience and represents the only multi-institutional series examining the use of prophylactic antibiotics in hemorrhoid surgery. This study has certain limitations. First, as with any retrospective database review, it is subject to the inherent biases that accompany such a study design. Definitions used to define surgical site were broad in an attempt to be more inclusive rather than less inclusive. Clearly, persistent pain, signs of local inflammation, and wound dehiscence can occur in the absence of infection, which when included may have overestimated the incidence of postoperative SSI. Patients included in this study encompassed the full spectrum of hemorrhoid disease including all grades of internal, mixed internal/external, and isolated external hemorrhoids. Because of the variability, there exists an inherent risk of introducing a small degree of bias of the cohort as a whole. However, as multivariate regression analysis suggests, when stratified by hemorrhoidal diagnosis alone there was still no difference in the rate of postoperative SSI. In addition, a variety of operative procedures exist for the management of hemorrhoid disease. These procedures may not be entirely comparable, and some may pose a higher risk of postoperative infection. In this study, only closed hemorrhoidectomy was considered. However, open versus closed methods for hemorrhoid surgery has been previously shown to have equivalent results with regard to overall complications.24,25 In addition, the study populations were not equally distributed between those that received prophylactic antibiotics and those that did not. Finally, we acknowledge that patients could have presented to an outside hospital with an infection that may alter the overall rates of SSI; however, all patients were seen in clinic
Diseases of the Colon & Rectum Volume 57: 3 (2014)
that included a p hysical e xamination, and these were generally “closed” health care systems where follow-up was captured. Yet, despite these limitations, this study does represent the largest retrospective examination of the role of prophylactic antibiotics for hemorrhoidectomy.
CONCLUSION We found no difference in postoperative complications between patients receiving and not receiving prophylactic antibiotics for hemorrhoidectomy. No perioperative risk factor was associated with an increased risk of developing a postoperative SSI. Based on the results of this study, a prospective case-matched controlled study is warranted to further clarify the role of prophylactic antibiotics for elective hemorrhoidectomy and to be able to provide the clinician with higher-level evidence-based recommendations. ACKNOWLEDGMENTS The authors thank Aneel Damle (University of Massachusetts Memorial Medical Center), Christopher Porta, Derek McVay, and Avery Walker (Madigan Army Medical Center) for their diligent efforts and a ssistance with data collection. REFERENCES 1. Johanson JF, Sonnenberg A. The prevalence of hemorrhoids and chronic constipation. An epidemiologic study. Gastroenterology. 1990;98:380–386. 2. Bleday R, Pena JP, Rothenberger DA, Goldberg SM, Buls JG. Symptomatic hemorrhoids: current incidence and complications of operative therapy. Dis Colon Rectum. 1992;35:477–481. 3. Rivadeneira DE, Steele SR, Ternent C, Chalasani S, Buie WD, Rafferty JL; Standards Practice Task Force of The American Society of Colon and Rectal Surgeons. Practice parameters for the management of hemorrhoids (revised 2010). Dis Colon Rectum. 2011;54:1059–1064. 4. de Paula PR, Speranzini MB, Hamzagic HC, et al. Bacteriology of the anal wound after open hemorrhoidectomy: qualitative and quantitative analysis. Dis Colon Rectum. 1991;34:664–669. 5. McCloud JM, Jameson JS, Scott AN. Life-threatening sepsis following treatment for haemorrhoids: a systematic review. Colorectal Dis. 2006;8:748–755. 6. Schulte T, Fändrich F, Kahlke V. Life-threatening rectal necrosis after injection sclerotherapy for haemorrhoids. Int J Colorectal Dis. 2008;23:725–726. 7. Sim HL, Tan KY, Poon PL, Cheng A, Mak K. Life-threatening perineal sepsis after rubber band ligation of haemorrhoids. Tech Coloproctol. 2009;13:161–164.
369
8. Guy RJ, Seow-Choen F. Septic complications after treatment of haemorrhoids. Br J Surg. 2003;90:147–156. 9. Sielezneff I, Salle E, Lécuyer J, Brunet C, Sarles JC, Sastre B. Early postoperative morbidity after hemorrhoidectomy using the Milligan-Morgan technic. A retrospective studies of 1,134 cases [in French]. J Chir (Paris). 1997;134:243–247. 10. Arbman G, Krook H, Haapaniemi S. Closed vs. open hemorrhoidectomy–is there any difference? Dis Colon Rectum. 2000;43:31–34. 11. Dixon MR, Stamos MJ, Grant SR, et al. Stapled hemorrhoidectomy: a review of our early experience. Am Surg. 2003;69:862–865. 12. Brusciano L, Ayabaca SM, Pescatori M, et al. Reinterventions after complicated or failed stapled hemorrhoidopexy. Dis Colon Rectum. 2004;47:1846–1851. 13. Iyer VS, Shrier I, Gordon PH. Long-term outcome of rubber band ligation for symptomatic primary and recurrent internal hemorrhoids. Dis Colon Rectum. 2004;47:1364–1370. 14. Pessaux P, Tuech JJ, Laurent B, et al. Morbidity after stapled haemorrhoidectomy: long-term results about 140 patients and review of the literature [in French]. Ann Chir. 2004;129:571–577. 15. de Paula PR, Matos D, Franco M, et al. Why do anal wounds heal adequately? A study of the local immunoinflammatory defense mechanisms. Dis Colon Rectum. 2004;47:1861–1867. 16. Bonardi RA, Rosin JD, Stonesifer GL Jr, Bauer FW. Bacteremias associated with routine hemorrhoidectomies. Dis Colon Rectum. 1976;19:233–236. 17. Adami B, Eckardt VF, Suermann RB, Karbach U, Ewe K. Bacteremia after proctoscopy and hemorrhoidal injection sclerotherapy. Dis Colon Rectum. 1981;24:373–374. 18. Wesarachawit W, Pattana-arun J. Antibiotics and early post operative complications of closed hemorrhoidectomy: a retrospective matched pair study. J Med Assoc Thai. 2007;90:1828–1832. 19. Al-Mulhim AS, Ali AM, Al-Masuod N, Alwahidi A. Post hemorrhoidectomy pain: a randomized controlled trial. Saudi Med J. 2006;27:1538–1541. 20. Ala S, Saeedi M, Eshghi F, Mirzabeygi P. Topical metronidazole can reduce pain after surgery and pain on defecation in postoperative hemorrhoidectomy. Dis Colon Rectum. 2008;51:235–238. 21. Nicholson TJ, Armstrong D. Topical metronidazole (10 percent) decreases posthemorrhoidectomy pain and improves healing. Dis Colon Rectum. 2004;47:711–716. 22. Nguyen N, Yegiyants S, Kaloostian C, Abbas MA, Difronzo LA. The Surgical Care Improvement project (SCIP) initiative to reduce infection in elective colorectal surgery: which performance measures affect outcome? Am Surg. 2008;74:1012–1016. 23. Kobayashi M, Mohri Y, Inoue Y, Okita Y, Miki C, Kusunoki M. Continuous follow-up of surgical site infections for 30 days after colorectal surgery. World J Surg. 2008;32:1142–1146. 24. You SY, Kim SH, Chung CS, Lee DK. Open vs. closed hemorrhoidectomy. Dis Colon Rectum. 2005;48:108–113. 25. Sohn VY, Martin MJ, Mullenix PS, Cuadrado DG, Place RJ, Steele SR. A comparison of open versus closed techniques using the Harmonic Scalpel in outpatient hemorrhoid surgery. Mil Med. 2008;173:689–692.
