Topical Application Of Antibiotics Research Articles

Selective Digestive Tract Decontamination—Will It Prevent Infection with Multidrug-Resistant Gram-Negative Pathogens but Still Be Applicable in Institutions where Methicillin-Resistant Staphylococcus aureus and Vancomycin-Resistant Enterococci Are Endemic?

The purposes of selective decontamination of the digestive tract are to treat infections that may be incubating at the time a patient is admitted to an intensive care unit (ICU), by intravenous administration of antibiotics during the first days of a stay in the ICU, and to prevent ICU-acquired infections, by topical application of antibiotics in the oropharynx and the gastrointestinal tract. Despite multiple trials in which a considerable reduction in the incidence of ventilator-associated pneumonia was demonstrated, major objections against the routine use of selective decontamination of the digestive tract have included a lack of demonstrated reductions in mortality rates and in length of stay (in individual trials), a lack of cost-efficacy data, and the threat of selection of multidrug-resistant bacteria. Recently, 2 controlled, randomized studies reported significant reductions in mortality rates among patients in ICUs who underwent selective decontamination of the digestive tract in combination with reduced selection of antibiotic-resistant pathogens. However, those studies were performed in settings where levels of antibiotic resistance are low, and some methodological issues remain unresolved. If these beneficial results are confirmed, the question of how to balance these benefits against the expected enhanced selection of methicillin-resistant Staphylococcus aureus, vancomycin-resistant enterococci, and, possibly, multidrug-resistant gram-negative bacteria will emerge.

TIME FOR A CHANGE? YES, BUT NOT TO CIPROFLOXACIN

5 January 2006 Dear Editor, We read with interest the opinion of Coates1 about ear drops and tympanic membrane perforations. While we strongly agree that it is ‘time for a change’ away from topical aminoglycosides, we do not agree that the change should be to fluoroquinolones. Fluoroquinolones have established themselves as extremely valuable agents for the treatment of serious bacterial infections in humans, including those caused by bacteria resistant to other classes of drugs. Their value lies not only in these properties, but in our ability to use them orally and allow outpatient treatment for such infections, either as follow on from initial hospital treatment, or from the outset. Reflecting this perspective, the Pharmaceutical Benefits Scheme allows access to fluoroquinolones that is limited to ‘Authority required’ (except for ciprofloxacin and gonorrhoea which is a ‘Restricted benefit’). This is true for both oral and eye drop formulations, and for all fluoroquinolones. In addition, Australia has been one of the few countries to successfully prevent the introduction of fluoroquinolones into food animal production, thereby reducing resistance coming through the food chain to humans. Although arguments are sometimes made to the contrary, resistance can be selected with topical therapy despite locally high concentrations. If the opposite were true, then we would not be observing mupirocin resistance, because this agent is only used topically in ‘high concentration’. Unfortunately mupirocin resistance is now well documented, especially in areas of intensive use such as remote aboriginal communities – the same population where treatment of chronic suppurative otitis media (CSOM) is common.2 Of specific relevance to the use of ciprofloxacin ear drops, and counter to that stated by Coates, resistance has been selected during the treatment of CSOM. Jang and Park3 reported on failed ciprofloxacin therapy of CSOM in 88 patients. All patients were shown to have ciprofloxacin resistant Pseudomonas aeruginosa when recultured. Furthermore, topical use can provide the opportunity for local colonization with resistant organisms. Clinical trials of ciprofloxacin ear drops have not given rise to resistance selection in commensals in the short term. However, we cannot assume that it will not emerge in the long term, and indeed microbiological experience shows that eventually it does, with only rare exceptions. Evidence for resistance selection at sites distant to application of topical antibiotics has been demonstrated for tetracycline when used for acne4 and for he treatment of trachoma.5 We are therefore concerned that fluoroquinolones are now gaining currency as topical agents. While this can be justified for such sight-threatening conditions as bacterial keratitis, it is less than prudent to use them for minor conditions. Although CSOM is not a trivial infection, evidence has yet to emerge that topical antibiotics of any kind preserve hearing in the long term or prevent serious complications.6,7 Instead, the benefits are short-term and re-treatment is frequently required, leading to repeated antibiotic exposure and greater selection pressure for resistance. Hence, selecting for resistance with topical fluoroquinolone therapy for non-serious infections or for ‘prophylaxis’ puts the value of this class for truly serious infections under threat. To make our position clear, we are not strong supporters either of the use of topical aminoglycosides on resistance grounds. Aminoglycosides are also valuable antibiotics for serious infections. However, we know that neomycin, which is not used systemically, does not necessarily select for resistance to the standard systemic aminoglycosides such as gentamicin, and therefore not all topical aminoglycosides are alike. Unfortunately, using one fluoroquinolone topically will necessarily select for resistance to a range of valuable fluoroquinolones. The clinical problem addressed by Coates is a troublesome one, but rather than following his suggestions to lobby a wide range of groups to get ready access to ciprofloxacin ear drops, we believe that prescribers should urgently lobby the pharmaceutical industry to produce a safe and effective antimicrobial ear drop that does not have resistance implications for valuable reserve systemic antimicrobials. Australia has been very effective at controlling resistance to fluoroquinolones. We would like to keep it that way.

Regression of pustulosis palmaris et plantaris by periodontal treatment in a subject with severe periodontitis

A 38-year-old woman with pustulosis palmaris et plantaris (PPP) was referred to the dentist with a suspicion of metal allergy (Fig. 1A). She had noticed the symptoms 3 years previously on her palms and legs. As her symptoms did not improve spontaneously, she visited her dermatologist. There, she was recorded as follows: (1) pustules localized only on the palms and soles; (2) no lesions with eczema or psoriasis on any other parts of the body; and (3) no family history of pustulosis. On the basis of these observations, she was diagnosed with PPP. She was treated with topical application of corticosteroids; however, her symptoms did not improve. Her dermatologist then suspected metal allergy and performed a patch test. Her skin reacted positively with chromium sulfate, zinc chloride, and mercury bichloride. Therefore, she was referred to our dental hospital for the investigation of the possible association of PPP with oral restorative materials. Figure 1. Appearances of the palm at: (A) first visit; (B) after topical application of antibiotics for 1 month; (C) 1 week after surgery, including teeth extraction; (D) 1 month after surgery; (E) immediately after complete remission; (F) 2 years after complete remission Download figure to PowerPoint Her medical records were unremarkable, except for extremely severe periodontitis, which she was aware of more than 10 years previously; however, she had not received any comprehensive periodontal treatment. On presentation, we first analyzed the restorative materials of the teeth to determine whether they contained any positive elements detected by the patch test. Unfortunately, we could not detect any positive elements in the oral cavity. Her blood chemistry did not indicate any abnormalities. The white blood cell count was 5600/mm3, with differential counts of 49.3% neutrophils, 28.0% lymphocytes, 7.6% monocytes, 14.4% eosinophils, and 0.7% basophils. C-reactive protein was not elevated, as measured by the conventional method. Immunoglobulin G (IgG), IgA, and IgM concentrations were all within the normal range (IgG, 1243.0 mg/dL; IgA, 131.6 mg/dL; IgM, 149.6 mg/dL). Because of these laboratory data, we first treated her periodontitis to determine whether there was a possible relationship of her PPP to severe periodontitis. Her periodontitis was very severe (Fig. 2A,B). The extraction of several teeth, as well as the correction of the shape of the inflamed alveolar bone (the bone supporting the teeth) to a physiological shape, was necessary. We first treated her periodontitis with the topical application of antibiotics (minocycline-HCl ointment) to determine the effect of anti-infectious periodontal treatment on PPP. We administered the antibiotics in every periodontal pocket once a week for a period of 1 month. One month after treatment, her symptoms appeared to improve slightly (Fig. 1B). We next performed surgical treatment, including teeth extractions and correction of the bone shape. Two to three days after surgery, her PPP symptoms worsened suddenly (Fig. 1C), and continued for up to 1 month (Fig. 1D). From 1 month after surgery, however, her symptoms gradually improved and complete remission was observed (Fig. 1E). In addition, although more than 2 years have passed since complete remission, no recurrence has been observed (Fig. 1F). Figure 2. Appearance of the patient's periodontal tissues at first visit (A) and periapical radiograph of the periodontium (B). Note the severe alveolar bone resorption and so-called “floating teeth” due to severe periodontitis (B), as well as severe gingival inflammation (A). In addition, many metal restorations are observed Download figure to PowerPoint

Basal Cell Carcinoma of the Lip and Mentum

Basal Cell Carcinoma of the Lip and Mentum

Pediatric sinusitis: when do we operate?

The role of surgery in the treatment of pediatric sinusitis is still in evolution. This review of recent literature highlights developments in the study of pediatric sinusitis, particularly as it pertains to surgical intervention. There is growing support in the literature for adenoidectomy as a first-line surgical intervention for chronic rhinosinusitis in children when maximal medical management fails. Maxillary sinus aspiration or middle meatal culture can be performed at the same sitting to facilitate directed antibiotic therapy. Intravenous antibiotics seem to be a promising alternative to functional endoscopic sinus surgery (FESS), especially in younger children. Current literature continues to support FESS as a safe and effective treatment for pediatric sinus disease. Previous notions that FESS may inhibit midfacial growth have been challenged by several recent studies. There is no clear consensus regarding timing of FESS for chronic rhinosinusitis. However, the current literature seems to support FESS when maximal medical therapy, adenoidectomy, and culture-directed systemic antibiotics have all failed with persistence of sinonasal disease, when anatomic abnormalities predispose to chronic rhinosinusitis by obstructing normal sinonasal drainage pathways, in sinonasal polyposis to facilitate application of topical steroids, as an adjunct to desensitization in aspirin-sensitive patients, when orbital or intracranial complications of sinonasal disease occur, and in selected cystic fibrosis patients to improve quality of life and facilitate application of topical antibiotics with activity against Pseudomonas aeruginosa. Although the current literature lends some additional clarity to the indications for surgical intervention in pediatric chronic rhinosinusitis, additional research is still needed to elucidate appropriate timing for surgery and more specific indications.

Antimicrobial periodontal treatment decreases serum C-reactive protein, tumor necrosis factor-alpha, but not adiponectin levels in patients with chronic periodontitis.

Elevated levels of C-reactive protein (CRP) and decreased plasma adiponectin are associated with increased risk of atherosclerosis. Furthermore, recent observations suggested that adiponectin and tumor necrosis factor-alpha (TNF-alpha) suppressed each other's production. Since periodontal disease has been suggested to act as a risk factor for atherosclerosis, we examined the effects of antimicrobial periodontal treatment on CRP, adiponectin, and TNF-alpha levels. Fifteen chronic periodontitis patients with various systemic conditions at high risk for atherosclerosis were enrolled in the study. Patients were non-surgically treated with topical application of antibiotics and mechanical debridement of calculus once a week for 1 month. Before and after therapy, CRP, adiponectin, and TNF-alpha levels were measured. Both CRP and TNF-alpha levels were significantly decreased after treatment (P<0.01 and P<0.03, respectively), while adiponectin levels did not change significantly. Periodontal treatment is effective in reducing CRP and TNF-alpha, while adiponectin does not appear to be influenced by periodontal treatment. Elevated levels of CRP and TNF-alpha may be associated with increased risk for future development of atherosclerosis in periodontitis patients.

When should old therapies be abandoned? A modern look at old studies on topical ampicillin.

We sought to determine whether topical ampicillin can reduce the rate of wound infections in clean contaminated surgical wounds (appendectomy, colorectal surgery). All randomized controlled trials examining the use of topical ampicillin in appendectomy and colorectal surgery published in English were identified via a Medline, Advanced Medline, and Cochraine Controlled Trials Register search and a meta-analysis performed.Results. Topical ampicillin vs. no antibiotic prophylaxis in clean contaminated wounds significantly reduced surgical wound infection rates (Odds Ratio (OR)=0.084, 95% CI, 0.04-0.16, P<0.0001). Topical ampicillin vs. no antibiotic prophylaxis in contaminated wounds also reduced surgical wound infection (OR=0.262, 95% CI, 0.14-0.51, P<0.0001). Topical ampicillin combined with systemic antibiotics vs. systemic antibiotics alone did not reduce surgical wound infection rate (OR=0.927, 95% CI, 0.27-1.72, P=0.90). Topical ampicillin significantly reduces the rate of surgical wound infections in clean contaminated surgery. A significant but smaller effect is seen in appendectomies where the appendix is gangrenous or perforated. Topical ampicillin did not confer any additional benefit when systemic antibiotics are used. While ampicillin may no longer be an effective agent, topical application of antibiotics is effective.Summary. A meta-analysis of studies using topical ampicillin for the prevention of infection in clean contaminated wound suggests that topical ampicillin is effective, but no incremental benefit is seen with systemic antibiotics.

Novel methods for the prevention of infection of intravascular devices

Infections are often caused by bacteria that gain access either from the skin during insertion, or by migration from the skin to the surrounding subcutaneous tissue, once in situ. It is necessary to understand the pathogenesis with regard to sessile or planktonic forms of microorganisms and why they adhere to catheter surfaces, and the interactions between host, catheter and bug. To prevent colonization two general approaches have been adopted: the first concerns the site of insertion—prophylactic antibiotics, tunnelling, intraluminal antibiotic locks, topical application of antibiotics or antiseptics, cutaneous disinfection pre- and post-insertion with various antiseptics, catheter site care and frequency of dressing changes. The second approach considers the physical nature of catheter material—silver impregnated subcutaneous collagen cuffs, antiseptic hubs, antimicrobial coating of catheters with antibiotics e.g., rifampicin and minocycline, silver ions, chlorhexidine, chlorhexidine and silver sulphadiazine. Only by understanding the pathogenesis of catheter devices whereby micro-organisms are embedded in biofilms and are resistant to the action of antibiotics, will effective preventative measures be possible. In addition a proper risk-management review will be necessary to determine the cost-benefit of such modified catheters.

Treating Corneal Abrasions and Lacerations

Corneal abrasions and lacerations injure the ‘windshield’ of the eye, exposing the patient to potential vision loss and infection. Diagnosis rests on physical examination of the eye, using fluorescein stain and a penlight; further examination with a slit lamp is often required for lacerations. Treatment of abrasions includes application of topical antibiotics and—except in contact-lens users and perhaps with certain small abrasions—patching the eye. Lacerations require urgent referral to an ophthalmologist, who usually repairs the defect surgically. Topical anesthetics, though effective pain relievers, should never be prescribed for home use.

Treatment of chlamydial conjunctivitis.

Adult inclusion conjunctivitis and neonatal inclusion conjunctivitis are important infectious diseases commonly encountered by ophthalmologists in industrialized countries. Patients with inclusion conjunctivitis are frequently complicated by genitourinary and epipharyngeal chlamydial infections. Thus, treatment of chlamydial conjunctivitis must include systemic antibiotics and not be limited to topical application of antibiotics to the eye. Effective anti-chlamydial agents include tetracyclines, macrolides, and some of the fluoroquinolones. Prolonged and frequent application of ophthalmic agents is vital for successful treatment of conjunctivitis.

Surgical treatment of active infective endocarditis with paravalvular involvement

Aortic root infection remains a challenging problem in the surgical treatment of both native and prosthetic valve endocarditis. Between 1980 and 1991, 73 patients with active aortic valve endocarditis and paravalvular infection underwent operation. Indications for operation included congestive heart failure and uncontrolled sepsis. Aortic root abscesses were located in the noncoronary anulus or in the aorticomitral junction in 45% of cases, followed by the subannular interventricular septum in 23%. Two patients had an aorticoatrial fistula, seven an interventricular septal defect. Total or partial left ventricular–aortic dehiscence was observed in 27 patients. All patients underwent aortic valve replacement, nine with simultaneous mitral valve operations. Two of the latter required patch reconstruction of the destroyed aorticomitral septum with double valve replacement. Reconstruction of the aortic base was possible in 16 patients, whereas in 12 total replacement of the aortic root was necessary. In one patient, supracoronary aortic valve replacement was used. Recently, topical application of antibiotics in fibrin sealant was used in 25 patients. The operative mortality rate was 21% and correlated to preoperative uncontrolled sepsis and the presence of extensive root destruction. Operation for active endocarditis of the aortic root requires radical, individualized techniques and results in an acceptable operative and long-term risk. The use of an antibiotic fibrin compound appears to be a useful prophylactic tool to prevent postoperative residual endocarditis. (J THORAC CARDIOVASC SURG 1994;107:171-7)

Critical Care Pharmacotherapy

During recent years, research in critical care medicine has focused on the role of the gastrointestinal tract in the pathogenesis of multiple organ failure and nosocomial infection, and on preventive measures. Gram-negative bacterial overgrowth of the oropharynx and stomach has been proved to be a cause of nosocomial pneumonia. Topical application of antibiotics into the oropharynx and stomach, and preservation of gastric acidity have been shown to be effective prophylaxis in ventilated patients. Recent studies have demonstrated that gastric alkalinisation is no longer necessary for the prevention of stress ulcer bleeding in critically ill patients. Tissue hypoxaemia, not gastric acidity, is the underlying pathomechanism of stress ulcer bleeding. In experimental investigations, pirenzepine and sucralfate improved gastric mucosal oxygen supply. Both compounds effectively prevent bleeding without increasing gastric pH. In mechanically ventilated patients, significantly lower rates of pneumonia occur with both of these drugs compared with antacids or histamine H2-receptor antagonists. Topical antibiotics (selective digestive decontamination) are most effective in patients with alkaline gastric juice, but of only marginal clinical relevance in those with acidic gastric contents. Isoflurane, propofol and clonidine have been recently investigated for sedation of ventilated patients. Isoflurane may lead to fluoride accumulation after more than 1 day. Propofol dosage has to be increased more often after 4 to 7 days, leading to fat overload and significantly increased costs. Clonidine was highly effective in patients with 'sympathetic overshoot', e.g. those experiencing alcohol or opioid withdrawal. Wound infections are an important problem in burn patients.(ABSTRACT TRUNCATED AT 250 WORDS)

Fibrin-enmeshed tobramycin liposomes: single application topical therapy of Pseudomonas keratitis.

Treatment of bacterial keratitis requires frequent application of topical antibiotics. We studied the efficacy of a single topical administration of tobramycin incorporated in large multivesicular liposomes and enmeshed in a fibrin sealant on rabbit corneas infected with Pseudomonas aeruginosa. One cornea each of 25 New Zealand albino rabbits was infected with P. aeruginosa. Twenty-four hours later, the animals were randomly divided into five groups of five. Group A received single hourly drops (50 microliters) of fortified tobramycin (14.5 mg/ml, total of 17.4 mg). Group B received a single topical application of 3.5 mg tobramycin, in 0.1 ml multivesicular liposomes, enmeshed in a fibrin sealant with an overlaying bandage contact lens. Group C was treated in the same manner as group B without the addition of fibrin sealant. Groups D and E served as nondrug-treated controls, with group D receiving topical fibrin-enmeshed liposomes devoid of tobramycin and group E receiving hourly topical balanced salt solution (BSS) drops. All animals were killed 24 h after initiation of therapy. Significantly fewer colonies of Pseudomonas were present in corneas of all three treated groups, as compared with the two nondrug-treated control groups (p less than 0.02). There were significantly fewer colonies of Pseudomonas in groups A and B as compared with group C (p less than 0.02). No significant difference was noted between a single administration of topical fibrinen-meshed tobramycin-encapsulated liposomes (group B) and 24 doses of hourly fortified topical tobramycin (group A, p greater than 0.05). Tobramycin-encapsulated megaliposomes may serve as a useful adjunct in treatment of Pseudomonas keratitis.

Gentamicin Resistance in Staphylococcal Corneal Ulcers

Gentamicin sulfate is frequently included in the initial treatment of suspected bacterial keratitis. We present the gentamicin susceptibility test results for coagulase-negative staphylococci and Staphylococcus aureus isolated from corneal ulcers in nine patients. All were resistant to gentamicin by standard disk diffusion techniques. The minimum inhibitory concentrations were very high, some exceeding the concentration normally achieved by frequent topical application of antibiotics. Clinicians should be aware that strains of staphylococci extremely resistant to gentamicin may cause keratitis. An antibiotic with broad activity against Gram-positive organisms, such as a cephalosporin or vancomycin, should be included as part of the initial therapy for corneal ulcers that might be caused by staphylococci.

Pharmacokinetics of Cefazolin Applied Topically to the Surgical Wound

Topical application of antibiotics is used in the prophylaxis of postoperative surgical infections. However, whether topically applied antibiotics remain primarily in the surgical wound fluid or are systemically absorbed is uncertain. The pharmacokinetics of topically applied cefazolin were studied in a canine model that allowed simultaneous determination of serum and wound fluid antibiotic concentrations. Topical administration of cefazolin resulted in high antibiotic concentrations in the wound fluid for prolonged periods and rapid systemic absorption. Bioavailability after topical administration was 95%. Within 1 hour, the serum concentrations after topical administration equaled the serum concentrations after intravenous administration, and the concentration time curves declined in parallel. In wound fluid, the mean time above the susceptibility break point minimum inhibitory concentration after topical administration of cefazolin was 5.76 hours compared with the estimated time above the minimum inhibitory concentration of 2.55 hours after intravenous administration.

Bakterielles Keim- und Resistenzspektrum der reizfreien Konjunktiva - 7845 präoperative Abstriche der Erlanger Univ.-Augenklinik

The type and extent of bacterial colonization of the conjunctiva at the time of an intraocular operation may have a major influence on the risk of post-operative endophthalmitis. Between March 1986 and December 1988 a total of 7845 preoperative conjunctival smears were therefore examined by culture of Erlangen University Eye Hospital, in cooperation with the Department of Clinical Microbiology. Altogether, 1221 (15.6%) of the smears were classified as positive. Surgery on the patients concerned was postponed and gentamicin eye drops were administered for decontamination of their conjunctivae until the smears were negative. From the positive smears 1393 strains were cultured, identified, and submitted to a resistance test. The present paper reports the frequency distribution of the species of bacteria isolated and their resistance behavior, and compares the results with those of other authors. The value of preoperative cultures in the prevention of postoperative endophthalmitis and the importance of resistance tests in the context of topical application of antibiotics are discussed.

Intraincisional antibiotic in addition to systemic antibiotic treatment fails to reduce wound infection rates in contaminated abdominal surgery

One hundred ninety patients with peritonitis at the time of abdominal surgery were allocated at random to systemic antibiotic treatment alone or systemic antibiotic treatment combined with topical application of antibiotics in the wound at the time of wound closure. The overall wound infection rate was 17 percent without significant difference between the two treatment groups (P greater than 0.80).

Effect of Topical Application of Glycine and Proline on Recalcitrant Leg Ulcers of Prolidase Deficiency

<h3>To the Editor.—</h3> The most important dermatologic problem of prolidase deficiency¹is chronic recurrent leg ulcers that are recalcitrant in healing. We used 5% glycine-5% proline ointment on leg ulcers with excellent results. <h3>Report of a Case.—</h3> The patient, now a 22-year-old woman, is the same patient as described previously.¹She had never been free of recalcitrant leg ulcers despite various conventional therapies including topical application of antibiotics (gentamicin sulfate, colistin sulfate, polymyxin B sulfate, etc), leg bath in 0.01% potassium permanganate solution, oral administration of zinc sulfate (180 mg daily), and/or ascorbic acid (3 g daily). Oral L-proline (1 g daily) was also tried for two months with no apparent benefit. We decided to use topical proline and then a combination of topical proline and glycine on her leg ulcers. L-proline and/or L-glycine ground with a mortar and pestle was mixed in liquid paraffin (30 g) and

Endophthalmitis following intraocular lens implantation: report of 30 cases and review of the literature.

Bacterial endophthalmitis is a postoperative complication of intraocular lens implantation. A review of 30 cases showed that 77% occurred within seven days of initial cataract surgery and that all cases occurred within 32 days. Common presenting symptoms and signs included pain localized to the involved eye, decreased visual acuity, conjunctival injection, anterior chamber inflammation, hypopyon, and absent or poor red reflex. Diagnosis of an infectious etiology was made by aqueous and vitreous sampling for gram stain and culture. The commonest bacterial agents were coagulase-negative staphylococci and Staphylococcus aureus, but a wide variety of gram-negative and -positive organisms were also isolated. On the basis of this series and review of the literature, optimal treatment includes prompt diagnosis by culture and gram stain of vitreous fluid, frequent application of topical antibiotics, and administration of intraocular antibiotics. Intravenous antibiotics and vitrectomy may improve the visual outcome. Removal of the intraocular lens is unnecessary for bacteriologic cure and is not associated with improved visual outcome.

Toxicity of Mediastinal Irrigation With Bacitracin

To the Editor.— It is well recognized that topical application of antibiotics can result in systemic absorption. We report a case in which mediastinal irrigation with bacitracin solution resulted in a potentially toxic serum level of the drug. Report of a Case.— A 70-year-old man with severe chronic obstructive lung disease underwent a coronary artery bypass. The postoperative course was complicated by respiratory difficulties, and, on the 14th postoperative day, sternal dehiscence occurred. He was taken back to surgery, where sternal closure was performed. Because of the possibility of mediastinitis, three irrigation tubes were inserted with three sump drains. Through each irrigation tube, normal saline containing bacitracin, 50,000 units/L, was infused at 100 mL/hr for a total of three days. On the second day, a serum bacitracin level was measured at 10.7 units/mL. During this period, the serum urea nitrogen level rose from 16 mg/dL to 30 mg/dL. The serum