European Heart Journal (2004) 25, 1341–1362 ESC Expert consensus document Expert consensus document on b-adrenergicreceptor blockers The Task Force on Beta-Blockers of the European Societyof Cardiology Task Force Members, Jos on, Chairperson* (Spain), Karl Swedberg (Sweden), John McMurray (UK), Juan Tamargo (Spain), Aldo P. Maggioni (Italy),Henry Dargie (UK), Michal Tendera (Poland), Finn Waagstein (Sweden), Jan Kjekshus(Norway), Philippe Lechat (France), Christian Torp-Pedersen (Denmark)
Topical metronidazole (10 percent) decreases posthemorrhoidectomy pain and improves healing
Topical Metronidazole (10 Percent)Decreases Posthemorrhoidectomy Painand Improves Healing Thomas J. Nicholson, M.D., David Armstrong, M.D.
Georgia Colon and Rectal Surgical Clinic, Atlanta, Georgia INTRODUCTION: Oral metronidazole has been previously improved, compared with that of carrier controls. [Key demonstrated to decrease postoperative pain after open dia- words: Hemorrhoidectomy; Metronidazole; Internal hemor- thermy hemorrhoidectomy. The current study investigates rhoids; Postoperative pain] the efficacy of topical metronidazole (10 percent) in reduc-ing postoperative pain and promoting wound healing afterHarmonic Scalpel® hemorrhoidectomy. METHODS: A pro-spective, randomized trial was conducted to compare post- T he single most important challenge after surgical hemorrhoidectomy is the management of post- hemorrhoidectomy pain and wound healing with use of operative pain. Postoperative pain likely has two ma- topical metronidazole (10 percent) vs. placebo carrier, ap-plied to the surgical site. Surgical indications included grade jor components: first, discomfort from the surgical in- 3 or 4 internal or external hemorrhoidal disease, with or cision in the uniquely sensitive anoderm and perianal without a fissure-in-ano. Pain was assessed using a visual skin and second, discomfort from tissue inflammation analog score (VAS) preoperatively and on postoperativedays 1, 2, 7, 14, and 28. Twenty-four-hour narcotic use resulting from bacterial infiltration of the wound. De- (hydrocodone 10 mg) was recorded on postoperative days creasing pain from the surgical incision itself can be 1, 2, 7, 14, and 28. Digital photographs of the surgical site accomplished by minimizing surgical trauma, for ex- were taken at 14 days postoperatively. The photographs ample, by utilizing the Harmonic Scalpel® (Ethicon were independently ranked by three blinded observers ac-cording to a) postoperative edema, b) primary vs. secondary Endo-Surgery, Inc., Cincinnati, OH), which dimin- healing, and c) overall wound healing. RESULTS: Twenty ishes thermal injury to the subjacent tissue.1,2 De- patients were randomized in a prospective manner, ten to creased postoperative pain after Harmonic Scalpel® the topical 10 percent metronidazole group and ten to theplacebo carrier group. Patients in the topical metronidazole hemorrhoidectomy compared with that in electrocau- group experienced significantly less postoperative pain at tery controls was reported by the senior author,3 and day 7 (VAS ± SEM, 3.4 ± 0.4 vs. 6.3 ± 0.5; P = 0.002) and day has since been confirmed by other studies.4,5 14 (1.0 ± 0.4 vs. 3.2 ± 0.7, P = 0.02). There was no statisticaldifference in narcotic analgesic requirements between Prevention of bacterial colonization of the hemor- groups. In the metronidazole group, postoperative edema rhoidectomy site has received less attention. Perfor- was ranked significantly lower (mean score, 3.0 vs. 7.0, P < mance of open vs. closed hemorrhoidectomy has 0.01) and overall wound healing ranked significantly better been claimed to prevent secondary bacterial infec- (4.0 vs. 7.0, P = 0.03) than in controls. CONCLUSION: Topi-cal 10 percent metronidazole significantly reduces post- tion,6 but has not been conclusively shown to de- hemorrhoidectomy discomfort at days 7 and 14 postopera- crease postoperative pain.7,8 More recently, use of tively. Postoperative edema is reduced and overall healing is preoperative and postoperative antibiotics (metroni-dazole) has been studied to determine the effect of Correspondence to: David Armstrong, M.D., Georgia Colon and decreasing bacterial colonization of the surgical site.9 Rectal Surgical Clinic, 5555 Peachtree Dunwoody Road, Atlanta,Georgia 30342, e-mail: [email protected] In 1998, Carpenti et al.9 demonstrated in a prospec- Dis Colon Rectum 2004; 47: 711–716 tive, randomized trial that oral metronidazole sig- nificantly decreased postoperative pain after open The American Society of Colon and Rectal SurgeonsPublished online: 2 April 2004 diathermy hemorrhoidectomy. Results from a subse- NICHOLSON ET AL Dis Colon Rectum, May 2004 quent trail by Balfour et al.,10 who studied treatment rier (petrolatum cream) alone were obtained from a with oral metronidazole after closed hemorrhoidec- compounding pharmacy (Monfort's Compounding tomy, showed no difference between groups. How- Center, Lawrenceville GA). Study patients applied ap- ever, low pain scores and an insufficient number of proximately 2.5 cc of 10 percent metronidazole cream patients in the series may have resulted in a Type II to the surgical site three times daily after a sitz bath or error. The current study examines the influence of warm soak. Control patients applied the same quan- topical metronidazole (10 percent) on postopera- tity of the inert carrier three times a day after a sitz tive pain and healing after closed Harmonic Scalpel® bath or warm soak. All patients were supplied with a hemorrhoidectomy. The premise for topical ap- standard narcotic analgesic (hydrocodone, 10 mg by plication of metronidazole to the surgical site was mouth every 4–6 hours as needed; 25 tablets dis- to optimize local tissue concentrations, avoid sys- pensed), instructed to take a fiber supplement (Kon- temic side effects, and improve bioavailability. An syl, Konsyl Pharmaceuticals, Edison, NJ) twice daily average application volume of 2.5 ml of 10 percent and 30 cc mineral oil once daily, and maintain a high metronidazole cream contains 250 mg of metronida- fluid intake.
zole. This concentration was chosen to replicate the Postoperative pain was evaluated using a visual standard oral metronidazole dosage. Previous phar- analog score (VAS), which was recorded by the pa- macokinetic studies have demonstrated a very low tient on days 1, 2, 7, 14, and 28. Patients in both degree of systemic absorption from topical metroni- groups ranked the level of pain from 0 (no pain) to 10(very severe pain). Twenty-four-hour narcotic analge- dazole application: absorption of aqueous forms of sic requirement (number of hydrocodone pills) was topical metronidazole (MetroLotion® 0.75 percent) is also recorded by the patient on days 1, 2, 7, 14, and approximately 100 times less than oral administra- 28. Patients were evaluated at two and four weeks postoperatively, and the completed data sheets werecollected at the four-week visit. Pain scores at each PATIENTS AND METHODS time interval were compared between groups withWilcoxon's rank-sum test (nonparametric analysis ofranked data). Twenty-four-hour analgesic require- The study population included 20 patients with ments on each study day were compared between grade 3 or 4 hemorrhoidal disease, with significant groups with a two-sample t-test (parametric analysis external components, and with or without a fissure- of quantitative data).
in-ano. Patients with a neurologic deficit, chronic pain Wound healing was evaluated at two weeks by tak- syndrome, or currently taking narcotic analgesics ing a high-quality color digital photograph of the sur- were excluded. All patients gave informed consent gical site. At the end of the study, three surgeons before involvement in the study. All patients under- independently ranked the surgical incisions in a went a closed three-quadrant Harmonic Scalpel® double-blinded manner. Each photograph was hemorrhoidectomy under general anesthesia, with graded on a scale of 1 to 10, in three different catego- the patient in prone jackknife position. All procedures ries: A, postoperative incisional edema (score: 1, ab- were performed by the same surgeon (D.N.A.). The sence of edema, to 10, extensive edema); B, primary technique of Harmonic Scalpel® hemorrhoidectomy vs. secondary wound healing (score: 1, well-healed has been described elsewhere.1,3 In patients requiring incision, primary intention healing, to 10, wide open fissurectomy and sphincterotomy, the flat blade of the incision, secondary intention healing); and C, overall Harmonic Scalpel® was used to cauterize the fissure, wound healing (1, overall well-healed incision, to 10, and a left lateral internal sphincterotomy was per- overall poor healing). For each category (A, B, and C), formed, extending to the proximal extent of the fis- each of the three surgeons ranked all photographs and a mean rank for each photograph was calculated.
Before surgery was performed, patients were pro- Thus, all photographs had three mean ranks (one spectively randomized into one of two groups. Ran- from each observer) for categories A, B, and C. Rank- domization was performed in a single-blind manner ing in categories A, B, and C were then compared and determined by witnessed coin toss. Identical between metronidazole and control groups with the tubes of metronidazole 10 percent in an inert carrier Mann-Whitney U test (nonparametric analysis of un- (petrolatum cream) and tubes containing the inert car- paired ranked data).
TOPICAL METRONIDAZOLE Figure 1. Pain scores (visual analog scores (VAS)) before
Figure 2. Posthemorrhoidectomy narcotic analgesic re-
and after hemorrhoidectomy in topical metronidazole quirements. There was no statistical difference in narcotic group and controls. Metronidazole patients reported sig- requirements between the metronidazole patients and nificantly less pain at days 7 and 14, compared with con- controls (mean ± SEM).
trols (mean ± SEM; **P = 0.002, *P = 0.02, respectively,Wilcoxon's rank-sum test).
0.03) than in controls. There was no difference be- tween groups when incisions were ranked accordingto category B, primary vs. secondary healing (P > Twenty patients were prospectively enrolled in the 0.05). Interobserver variation for all categories was study during the six-month study period. Ten patients 88.5 percent. Consistency was highest in category B received 10 percent metronidazole cream and ten pa- (primary vs. secondary healing, 93 percent consis- tients received the inert carrier. None of the patients tency), even though no difference was identified be- were lost to follow-up, and data collection was com- tween groups. Interobserver consistency was 88 per- plete. The metronidazole and control groups were cent in category A (wound edema) and 83.5 percent comparable in terms of age (mean ± standard error of in category C (overall healing).
the mean (SEM), 47.7 ± 3.2 years vs. 48.5 ± 3.1 years)and gender (males/females, 6:4 vs. 7:3). Three pa- tients in the metronidazole group and four patients inthe control group had an associated fissure-in-ano The role of bacterial colonization in posthemor- and underwent fissurectomy and sphincterotomy in rhoidectomy pain is unknown.12 Leaving the surgical addition to hemorrhoidectomy.
site open to facilitate drainage and prevent bacterial There was no significant difference between infection has been proposed to decrease postopera- groups in preoperative pain scores or pain scores on tive pain. In one prospective study of open vs. closed days 1 and 2. Patients in the topical metronidazole hemorrhoidectomy, Ho et al.6 reported faster healing group experienced significantly less postoperative times (4.9 weeks) after open procedures, compared pain at day 7 (VAS ± SEM, 3.4 ± 0.4 vs. 6.3 ± 0.5; P = with 6.9 weeks after closed hemorrhoidectomy. By 0.002) and day 14 (1.0 ± 0.4 vs. 3.2 ± 0.7; P = 0.02).
contrast, proponents of the closed technique7,8 claim There was no significant difference between groups less postoperative pain, fewer complications, and on day 28 (Fig. 1).
shorter hospital stays, compared with open hemor- There was no significant difference in narcotic re- rhoidectomy. With many additional variables at quirements between metronidazole and control work,7,8 the role of bacterial colonization in posthem- groups (P = 0.32; Fig. 2). To demonstrate significantly orrhoidectomy pain is difficult to determine. Practi- lower narcotic analgesic requirements in the metroni- cally every posthemorrhoidectomy incision appears dazole group at day 7 (and reach a power factor of edematous during the first few days and weeks after 0.8), a total of 19 patients would have been required surgery, and much of this tissue edema is almost cer- in each group.
tainly the result of factors other than bacterial infiltra- Wound healing in the metronidazole group was significantly better than in controls when ranked ac- Bacteriological studies of posthemorrhoidectomy cording to category A, postoperative edema (mean incisions provide no clarification of the role of bacte- score, 3.0 vs. 7.0; P < 0.01). In addition, category C, ria in postoperative pain, edema, or healing. Perianal overall wound healing, ranked significantly better in abscess, cellulitis, and gangrene are remarkably rare the metronidazole group (mean score, 4.0 vs. 7.0; P = after hemorrhoidectomy: Retrospective studies report NICHOLSON ET AL Dis Colon Rectum, May 2004 abscess or fistula in only 0 to 2 percent of otherwise increases bioavailability but also avoids systemic side healthy patients.1,6–8,13–16 Even with this low inci- effects. Petrolatum was chosen as the carrier medium, dence, most of these cases probably result from tech- because this is the most common carrier vehicle used nical failures, rather than exclusively from bacterial for topical anorectal preparations, the medium is re- invasion of the surgical site. De Paula et al.17 exam- tained at the site of application for substantial period ined the bacterial flora of posthemorrhoidectomy in- of time, and it has no adverse pharmacological inter- cisions for up to four weeks after surgery: all hemor- action with metronidazole.
rhoidectomy incisions were colonized by aerobic In the current study, no difference in postoperative bacteria by day 20, yet all healed without obvious pain was found on days 1 or 2, but the metronidazole signs of sepsis. Surprisingly, no anaerobic organisms group experienced significantly less pain by days 7 were isolated from any incision at any time in the and 14. These findings are consistent with Carapeti's study. Conversely, Brook and Frazier18 cultured study, in which a similar visual analog score was used, needle aspirates of "infected" hemorrhoids in 19 pa- which reported no difference between groups on tients and isolated anaerobes alone in 6 patients (32 days 1 through 4, but significantly less pain on days 5, percent), mixed anaerobes and aerobes in 12 patients 6, and 7.9 Visual analog scores were not continued (63 percent), and aerobes alone in only 1 patient (5 beyond day 7 in the Carapeti study. These findings percent). In both these studies, the predominant aer- suggest that pain in the first few days after hemor- obe isolated was E. coli, and the most common an- rhoidectomy is purely surgical in nature and unrelated aerobes were B. fragilis and Peptostreptococcus. to bacterial infiltration or tissue edema. In the current Metronidazole, the most common antibiotic used study, there was no statistical difference in narcotic against anaerobic bacteria, has a long-recognized and analgesic requirements between groups, although very useful role in treating a wide variety of anorectal metronidazole patients tended to require fewer nar- conditions.19–21 Its efficacy may be in part bacteri- cotics on days 7 and 14. After day 2, narcotic require- cidal, in addition to its lesser-understood anti- ments were so low that any statistical difference be- inflammatory actions.22 This latter property is used in tween groups would have required a larger number treatment of other inflammatory skin conditions, such of patients (specifically, 19 in each group to demon- as rosacea.23,24 Metronidazole is frequently the first- strate a statistical difference at day 7). The failure to line medication used in treating anorectal Crohn's dis- demonstrate a difference in narcotic requirements ease,19,21 a difficult and intractable anorectal condi- may therefore be a Type II error.
tion that frequently results in proctocolectomy and a Evaluation of wound healing in the current series is a new and previously unreported modality of evalu- In 1998, Carapeti et al.9 reported a potential role for ating posthemorrhoidectomy healing. By taking stan- oral metronidazole in diminishing postoperative pain dard digital photographs at a standard two-week after open diathermy hemorrhoidectomy. This pro- postoperative interval and evaluating these by three spective, double-blind study demonstrated a signifi- blinded observers, an impartial and semiquantitative cant decrease in postoperative pain on days 5 to 7 in measure of healing was established. Of note is the metronidazole patients, greater patient satisfaction, finding that incisional edema was consistently ranked and earlier return to work, compared with placebo less in the metronidazole group, compared with car- rier controls' rankings. Tissue edema may result from The rationale for using topical metronidazole in secondary bacterial infiltration or it may be an exag- posthemorrhoidectomy pain relief was to increase gerated manifestation of the normal healing process.
drug bioavailability. The greater tissue concentrations The diminished edema in the metronidazole group would, in theory, potentiate the anti-inflammatory may have been a result of the drug's bactericidal or and antimicrobial actions, compared with an equiva- anti-inflammatory actions. Posthemorrhoidectomy lent oral dose. The final 10 percent topical metroni- edema usually manifested as raised and even "rolled- dazole concentration was chosen to reproduce the over" incisional margins, often with "fleshy" skin tags standard 250-mg oral dosage (since an average 2.5 cc and tissue induration (Fig. 3). In the metronidazole application contains 250 mg metronidazole). Repeat- group, the tissue margins appeared consistently flat, ing the topical application three times a day again well defined, and with an absence of tissue edema, reproduces the standard oral regimen. The conver- and skin tags were less prominent (Fig. 4). The di- sion of an equivalent oral to topical dose not only minished postoperative pain in the metronidazole
TOPICAL METRONIDAZOLE Figure 4. A, B. Metronidazole patients: posthemorrhoid-
ectomy incisions, two weeks after three-quadrant closed
hemorrhoidectomy. Incisions appear healthy, even at two
weeks after surgery (note persistent chromic catgut suture
in a). The incision margins are flat, the suture line is intact,
Figure 3. A, B. Control patients: posthemorrhoidectomy
and no edematous skin tags are noted. Both patients surgical incisions, two weeks after three-quadrant closed ranked consistently low for incision edema and ranked hemorrhoidectomy. Note extensive incisional edema, with high for overall healing.
"rolled-over" incision margins and "fleshy" skin tags. Bothpatients ranked consistently high for incision edema and The beneficial role of topical 10 percent metroni- ranked low for overall healing.
dazole may be antibacterial or anti-inflammatory innature. Topical application improves bioavailability group may be directly or indirectly related to the di- and avoids side effects frequently seen with systemic minished tissue edema, because both variables were recorded at the same two-week interval after surgery.
Overall wound healing in the metronidazole group was also ranked significantly better than by controls, The current study demonstrates the efficacy of topi- again reflecting the relative absence of tissue edema, cal 10 percent metronidazole in diminishing postop- less prominent skin tags, and "cleaner" incisions. Pri- erative pain and improving wound healing after Har- mary vs. secondary healing was similar between monic Scalpel® hemorrhoidectomy.
groups. This factor is primarily caused by local dis-ruption of the incisions, a frequent finding after closed hemorrhoidectomy, or to the large size or 1. Armstrong DN, Frankum C. Schertzer, Ambroze WL, ME grade of the hemorrhoids. Since grade 4 hemorrhoids Orangio GR. Harmonic Scalpel® hemorrhoidectomy: were equally distributed in both metronidazole and five hundred consecutive cases. Dis Colon Rectum control groups, a similar rate of mechanical wound disruption may be expected in both groups.
2. McCarus SD. Physiologic mechanism of the ultrasoni- NICHOLSON ET AL Dis Colon Rectum, May 2004 cally activated Harmonic Scalpel®. J Am Assoc Gynecol 13. Seow-Choen F, Low HC. Prospective randomized study Laparosc 1996;3:601–8 of radical versus four piles haemorrhoidectomy for 3. Armstrong DN, Ambroze WL. Schertzer ME Orangio symptomatic large circumferential prolapsed piles. Br J GR. Harmonic Scalpel® vs. electrocautery hemorrhoid- Surg 1995;82:188–9 ectomy: a prospective evaluation. Dis Colon Rectum 14. Lacerda-Filho A, Cunha-Melo JR. Outpatient haemor- rhoidectomy under local anaesthesia. Eur J Surg 1997; 4. Chung CC, Ha JP, Tai YP, Tsang WW, Li MK. Double blind, randomized trial comparing Harmonic Scalpel™ 15. Buls JG, Goldberg SM. Modern management of hemor- hemorrhoidectomy, bipolar scissors hemorrhoidectomy rhoids. Surg Clin North Am 1978;58:469–78 and scissors excision: ligation technique. Dis Colon 16. Bleday R, Pena JP, Rothenberger DA, Goldberg SM, Rectum 2002;45:789–94 Buls JG. Symptomatic hemorrhoids: current incidence 5. Dreznik Z, Ramadan E. Harmonic scalpel hemorrhoid- and complications of operative therapy. Dis Colon Rec- ectomy: preliminary result of a new alternative method.
tum 1992;35:477–81 Dis Colon Rectum 2001;44:A36 17. de Paula PR, Speranzini MB, Hamzagic HC, et al. Bac- 6. Ho YH, Seow-Choen F, Tan M, Leong AF. Randomized teriology of the anal wound after open hemorrhoidec- controlled trial of open and closed haemorrhoidec- tomy: qualitative and quantitative analysis. Dis Colon tomy. Br J Surg 1997;84:1729–30 Rectum 1991;34:664–9 7. Neto JA, Quilici FA, Cordeiro F, Reis JA. Open versus 18. Brook I, Frazier EH. Aerobic and anaerobic microbiol- semi-open hemorrhoidectomy: a random trial. Int Surg ogy of infected haemorrhoids. Am J Gastrenterol 1996; 8. Hosch SB, Knoefel WT, Pichlmeier U, et al. Surgical 19. Jakobovits J, Schuster MM. Metronidazole therapy for treatment of piles: prospective, randomized study of Crohn's Disease associated fistulae. Am J Gastroenterol Parks vs. Milligan-Morgan hemorrhoidectomy. Dis Co- lon Rectum 1998;41:159–64 20. Brandt LJ, Bernstein LH, Baley SJ, et al. Metronidazole 9. Carapeti EA, Kamm MA, McDonald PJ, Phillips RK.
therapy for perineal Crohn's disease: a follow up study.
Double blind randomized controlled trial the effect of metronidazole on pain after day case hemorrhoidec- 21. McClane SJ, Rombeau JL. Anorectal Crohn's disease.
tomy. Lancet 1998;351:169–72 Surg Clin North Am 2001;81:169–83 10. Balfour L, Stojkovic SG, Botterill ID, Burke DA, Finan P, 22. Metronidazole. In: American hospital formulary service Sagar PM. A randomized, double-blind trial of the effect drug information. Bethesda: American Society of Health of metronidazole on pain after closed hemorrhoidec- System Pharmacists, 2002:864–75.
tomy. Dis Colon Rectum 2002;45:1186–91 23. Dahl MV, Katz HI, Krueger GG, et al. Topical metroni- 11. Metrocream. In: Physicians desk reference 2003. 57th dazole maintains remissions of rosacea. Arch Dermatol ed. Montvale: Thomson PDR, 2003:1383–4.
12. Guy RJ, Seow-Choen F. Septic complications after treat- 24. Gupta AK, Chaudhry M. Topical metronidazole for ro- ment of hemorrhoids. Br J Surg 2003;90:147–56 sacea. Skin Therapy Lett 2002;7:1–3
Making sense of Making sense of antipsychotics This booklet is for anyone who has been prescribed antipsychotic drugs, or who thinks they may be offered them, and also for their friends and family. It explains why these drugs may be prescribed, what their effects are, and when to avoid them. Note:Antipsychotics are sometimes called ‘neuroleptics' or ‘major tranquillisers'.