Original article Arch Argent Pediatr 2012;110(4):298-303 / 298

Systematic Review: which topical agent is more efficacious in the prevention of infections in patients?

María Teresa Rosanovaa, M.D., Daniel Stamboulianb, M.D. and Roberto Ledec, M.D.

SUMMARY use is quite extended. Therefore, the Topical agents are widely used in the care of question arises regarding which topi- burn patients; however the efficacy to prevent local infections and/or sepsis has not been clearly cal agent is more efficacious to pre- established in studies with a high level of evi- vent infections in the management of dence. This systematic review was conducted to . assess the comparative efficacy among different To answer this question, a system- topical agents. Material and Methods. The literature search atic review of the biomedical litera- was performed using the Medline database. ture was conducted; however, since a Key MESH terms were: (burn* or scald*) AND wide variety of topical anti-infective (antibacterial or *) AND (topic*) AND agents are used in the clinical practice, (therap* or prophylax*). Only randomized or quasi-randomized clinical trials, with a primary we thought it would be of interest to endpoint of local infection and/or sepsis were start by providing a brief description included. Studies were scored and classified re- of them.1 garding methodological key issues according to their level of evidence. Results. The initial search identified 457 studies -containing compounds of which 14 were eligible for final evaluation, and 1% Silver : Its mecha- full text was available for 11 of them. nism of action remains unclear. It is Conclusions. The evidence found in our review bactericidal against Gram-negative does not support differences in efficacy of topi- cal agents to reduce sepsis and/or local infec- aerobic rods and may be efficacious tions in burn patients. against fungi. Key words: topical agents, burn, infections. It does not penetrate the eschar ad- equately and its main adverse effect is http://dx.doi.org/10.5546/aap.2012.eng.298 the potential for leukopenia and, rare- ly, skin reactions. Some studies have demonstrated that it hinders the heal- INTRODUCTION ing process due to its toxicity on kera- According to several authors, topi- tinocytes and fibroblasts.1 This agent a. Epidemiological cal agents, together with early surgical may be combined with cerium ni- Control and 4,6 Infectology debridement have succeeded in de- trate, but it should not be used, either Department. creasing morbidity and mortality in alone or in combination, in patients al- Hospital “Prof. Dr. patients with severe burns.1-3 lergic to sulfa medications. Since sulfas Juan P. Garrahan”. When burnt tissues are colonized, may cause jaundice of the basal gan- b. Fundación Centro de Estudios healing becomes slower because the glia (kernicterus), they should not be Infectológicos. presence of bacteria causes an exag- used in infants younger than 2 months c. Instituto Argentino gerated inflammatory and deleterious and in pregnant women. de Medicina Basada response, which hinders wound heal- Other silver preparations are avail- en la Evidencia. 2-4 Buenos Aires, ing and favors infection. able in various forms such as silver- Argentina. The objectives of using topical anti- impregnated dressings (Acticoat®, microbial agents are to avoid bacterial among others), that can be replaced E-mail address: colonization and favor early wound every 24 hours. María Teresa Rosanova: [email protected] healing by reducing the chance of in- Silver Nitrate: It is used as a 0.5% fection. Nonetheless, these agents ef- solution. It is bactericidal and does Conflict of interest: ficacy has not been clearly established not penetrate the eschar. It must be None. in controlled studies. Evidence regard- applied every 2-3 hours and if used Received: 2-2-2012 ing their usefulness to prevent burn on large surface areas it may cause in- Accepted: 4-11-2012 related infections is scarce, but their tense fluid and electrolyte imbalance, Systematic Review: which topical agent is more efficacious in the prevention of infections in burn patients? / 299 and hyperpigmentation in the treated area.7 Objective of this review In some large burn centers, the development To evaluate the comparative efficacy of topical of bacterial resistance secondary to silver formu- agents most frequently used to prevent infection lations has been observed after their protracted in burn patients. use.8 MATERIAL AND METHODS Antibiotic agents for local use 0.5% Acetate: This is a local bacterio- Inclusion criteria static antibiotic with a broad spectrum of action a. Randomized or quasi-randomized clinical tri- against Gram-negative and Gram-positive organ- als (evidence levels 1b or 2a, as per the Cen- isms, and certain strains of anaerobes, but with no ter for Evidence-Based Medicine classification effect on fungi. Its main adverse effect is the po- system, Oxford University, UK); tential for metabolic acidosis. Like sulfadiazine, it b. Studies including only burn patients without may delay healing through the same mechanism. an initial infection; It diffuses into and through eschar better. c. Studies comparing the efficacy among topical Its disadvantage is that it causes pain on treat- agents; ed areas. Due to its short half life it must be ap- d. Development of infection and/or sepsis and/ plied every 12 hours.1,9 or time of wound re-epithelization; and/or : It is a petroleum-based polypeptide graft loss as main endpoints; antibiotic for local use, effective against Gram- e. Summary available at least either in Spanish positive cocci and bacilli. Its mode of action is by or in English. inhibiting the bacterial cell wall synthesis.1,10 Framycetin: It is a sulfate salt of B, Exclusion criteria bactericidal and active against Gram-positive coc- a. More than 20% of participants lost to follow- ci and Gram-negative rods. up in any of the groups; Other agents that have been used for the local b. Descriptive studies of any type or opinion pa- treatment of burn wounds, with variable results, pers, cohort or case-control studies; are polymixin and neomycin, either alone or com- c. Topical agents used in conditions unrelated to bined with bacitracin, and . burns.

Disinfecting agents Literature search Electrolytic Cloroxidant in a hypertonic solution A literature search was undertaken on No- of sodium chloride (Amuchina®): Initially used vember 18th., 2011 using the computerized da- as a disinfectant for fruits, vegetables and baby tabase (PUBMED), and the following key words: bottles. (burn* or scald*) AND (antibacterial or antibiotic*) Currently it is used to prevent burn-associated AND (topic*) AND (therap* or prophylax*). infections; however, data available are scarce.11 It A total of 457 citations were obtained, of which covers a broad spectrum of Gram-negative rods, 400 were discarded because the topical agent was Gram-positive cocci, fungi and viruses. Reports unrelated to burns, they were case reports, the have been published about its use for the local topical agents were mentioned as part of the treat- treatment of burn wounds.11 ment but their role was not investigated, or they Clorhexidine and povidone-iodine are among were basic experimental or in vitro trials. Another this group. 4 papers were obtained from the reference list of the systematic reviews found. Natural topical agents Honey: This substance is non-toxic and non- RESULTS irritant. In developing countries it has been used Fourteen papers were selected for final eval- in superficial and intermediate partial-thickness uation and the text was available in 11 of them. burns; it reduces the bacterial inoculate and fa- vors granulation of burn wounds.12,13 Excluded studies Papaya carica: In cases published, a favorable Brown et al.9 studied mafenide and its impact outcome has been described regarding its use to in reducing mortality. Mian et al. compared the cover burn wounds.14 incidence of sepsis and wound granulation in a series of 20 cases, in which Amuchina® was com- 300 / Arch Argent Pediatr 2012;110(4):298-303 / Original article pared to silver sulfadiazine. None of them was Muangman et al8 compared, apparently in a a randomized or quasi-randomized study. One quasi-randomized manner, (evidence level 2b), study was excluded because the summary was the efficacy of a dressing containing silver sulfadi- neither in English nor in Spanish; two were ex- azine versus 1% silver sulfadiazine in superficial cluded because they were experimental studies and intermediate burns affecting less than 25% of in rats and 5 because they were reviews. the BSA. Patients were children and adults, and were divided into 2 groups of 25 patients each Included studies of similar age, percent of BSA affected and burn Studies comparing silver-containing topical agents etiology. Of those 50 patients, 7 (28%) developed versus other agents infection; no significant differences were detect- De Gracia et al.,4 in an open randomized study, ed between groups. This paper was also written assigned the treatment according to a pre-estab- in Chinese, therefore we only worked with the lished sequence (evidence level 2a) and com- summary. pared 30 patients treated with silver sulfadiazine plus to 30 patients treated only Studies comparing topical agents versus placebo with silver sulfadiazine. Patients included were Livingston et al.10 conducted a prospective ran- older than 4 months of age, had a burnt surface domized study (the patient allocation method was area >15% of their body surface area (BSA) and not clarified in the summary) in 52 child and adult were admitted within 24 hours from the injury. patients subjected to tangential escharectomy and The main pre-determined endpoints were wound grafting upon admission (evidence level 2b). They healing and progression or not to sepsis. The compared the efficacy of two topical agents: 0.5% study concluded that patients in the group treated silver nitrate and neomycin (1 g/L) plus bacitra- with silver sulfadiazine plus cerium nitrate healed cin (50,000 U/L) versus a Ringer’s lactate solution their wounds 8 days before patients treated only (used as placebo), in preventing graft loss due to with silver sulfadiazine, and additionally, were infection. Seventeen patients (33%) suffered graft hospitalized for a shorter time (23 vs. 30 days). loss, which was attributed to infection in 16. However, no significant differences were found In patients with burns involving less than 20% in burn related infection or sepsis rates. of the BSA, graft loss was not a relevant issue. Ahuja et al.6 conducted a quasi-randomized, The use of any of the two antimicrobial drugs re- double blind study in children and adults; treat- sulted in less graft loss in patients with burns af- ments were assigned by drawing balls at random fecting 20-40% of their BSA: 1 out of 14 patients (evidence level 2a). They compared the outcome versus 4 out of 6 with Ringer’s lactate (p <0.05). of 20 patients with burns affecting 15-40% of their In patients with > 40% burn of their BSA, 67% BSA in whom they used 1% framycetin cream, to suffered graft loss regardless of the topical agent 20 patients treated with silver sulfadiazine. The employed; moreover, 50% of the patients in the main endpoint was bacterial count in the burn Ringer’s lactate group and the whole group that wound on days 4 and 7 in both groups. The 2 received bacitracin plus neomycin had infections groups were similar in age and percent of burnt caused by resistant microorganisms or fungi. The BSA (26% and 28.5% for framycetin and silver authors conclude that in patients with burns in- sulfadiazine, respectively). No differences were volving 20-40% of their BSA, topical observed in colony counts of biopsies obtained are efficacious in reducing infection related graft on days 4 and 7. loss and that neomycin and bacitracin induce an- Liao et al7 conducted a multicenter random- tibiotic resistance, whereas silver nitrate does not. ized study in 120 children and adults (80 with su- Only the summary was available for this study. perficial or intermediate partial-thickness burns, groups A and AB, and 40 with deep burns, group Studies that used antibiotics for systemic use B or “full-thickness” group). Patients were ran- administered topically, compared with domized (the summary does not describe the silver-containing agents method employed) to receive silver nitrate or sil- Snelling et al.15 conducted a randomized study ver sulfadiazine (evidence level 2a). The group assigning treatments with a system of odds and treated with silver nitrate exhibited the same an- even numbers at admission (evidence level 2a). A tibacterial effect than the group on silver sulfa- total of 71 child and adult patients were included; diazine. Since the paper is written in Chinese, we the study compared 1% silver sulfadiazine ver- only had access to the summary. sus 0.1% to prevent bacterial coloni- Systematic Review: which topical agent is more efficacious in the prevention of infections in burn patients? / 301 zation and sepsis. Patients had to be at 24 hours this agent exerts a better effect than others in re- post burn injury and have more than 10% of their ducing wound infection or the development of BSA affected. Children and adults were divided sepsis.4,6-8 The chance to slow the healing process into 2 groups: one with burns involving 11-40% of and generate reversible leukopenia should be the BSA and another group with burns affecting > weighted at the time of prescription. 41% of the BSA. Patients with odd numbers were Another problem to consider regarding non- treated with silver sulfadiazine (38 patients) and absorbable antibiotics used locally is the selec- patients with even numbers received topical gen- tion of resistant microorganisms, as shown in the tamicin (33 patients). Pseudomona aeruginosa colo- study by Livingston et al.,10 that reported a high nized the burns of 37% of the 38 patients treated rate of resistance in patients treated with bacitra- with sulfadiazine and 30% of the 33 patients treat- cin plus neomycin. This cautions about the pos- ed with gentamicin; of the latter, 21% had their sibility of generating resistance with the use of wound colonized with the same bacterium, but non-absorbable topical antibiotics. it was resistant to gentamicin. The emergence of Systemic antibiotics have also been used local- resistance warranted discontinuation of gentami- ly, but Snelling15 demonstrated the emergence of cin prophylaxis. gentamicin resistant strains during the course of topical treatment. Since this antibiotic is admin- Studies with natural topical agents compared to istered systemically, this finding points out the silver-containing agents need to avoid the use of systemic antibiotics as Malik et al.12 evaluated the usefulness of un- topical agents, since they could later be ineffective processed honey for the topical treatment of in patients requiring their systemic administration burns affecting less than 40% of the BSA. The because of a severe condition. method of randomization employed is not de- In developing countries, honey has been uti- scribed (evidence level 2a). Patients had to have lized for the treatment of burns.12-14 Malik12 and two contralateral similar body segments (for ex- Subrahmanyam13 demonstrated faster re-epithe- ample: both hands, right and left sides of the ab- lization in sites were honey was used compared domen) affected by second degree burns and one to those where silver sulfadiazine was used, but intervention was assigned to each segment. They burns treated were superficial or intermediate; included 150 patients of all ages. Interventions it was not used in deep burns. Additionally, the consisted in the topical application of honey or sil- small sample size in both studies would limit the ver sulfadiazine. Re-epithelization occurred faster validity of these conclusions. in sites where honey was used (13.47 days ± 4.06 We underscore that this review questions the versus 15.62 ± 4.4 days; p <0.0001). selection criteria used for one or another anti-in- Subrahmanyam,13 in a randomized study (the fectious agent in the preventive topical treatment randomization method was not specified; evi- of burn wound infections, since the evidence pre- dence level 2a), evaluated 50 patients, aged 3 to sented here is insufficient to define clear advan- 60 years, with burns affecting less than 40% of the tages of one agent over another. The conclusions BSA. Honey was used in 25 patients and silver of available studies are weakened by the following sulfadiazine in the remaining 25. After 21 days of caveats: marked heterogeneity of criteria regarding treatment, 100% of patients treated with honey the use of different kinds of topical agents, differ- had healed, compared to 84% of the other group; ent ages of patients included (all included children these data were confirmed by lesion biopsy. and adults), variable surface and extent of burns, and the various definitions of predetermined end- DISCUSSION points. Moreover, the lack of studies assessing the Numerous topical agents have been proposed efficacy of those agents against placebo must be and utilized in the management of burn patients. noted. In this review it was not possible to perform However, in spite of their massive use, evidence a metaanalysis given these limitations. regarding their usefulness in preventing the risk The potential for certain agents to cause a de- of infection is scarce.1-3 lay in healing and to generate bacterial resistance It has been postulated that silver sulfadiazine is an important reason for caution. reduces the bacterial inoculate, but data showing More prospective, randomized and controlled that it decreases the incidence of infection or sep- studies are required, with an adequate number of sis in burn patients are, in fact, scarce. The few patients, in order to reliably assess the benefits of studies cited in this review do not suggest that topical treatment in burn patients. n 302 / Arch Argent Pediatr 2012;110(4):298-303 / Original article

Table 1. Comparison of studies assessing the efficacy of topical treatment in burn patients

Study Objectives Design Patients Conclusions Evidence (n) Level Gracia et al. To compare the effects of silver Open, Age: older than Patients on cerium nitrate were 2a 20014 sulfadiazine with cerium nitrate randomized 4 months. Thirty hospitalized 17.2 days vs. 25.1 plus silver sulfadiazine on patients treated days; p: 0.03, and had a faster wound healing and with silver wound healing (9.5 days development of sepsis. sulfadiazine versus 14 days, p: 0.007). versus 30 There was no significant difference treated in sepsis rate (1/30 in the cerium with silver group vs. 4/30) or mortality rate sulfadiazine plus (1/30 in the cerium group cerium nitrate. vs. 4/30).

Ahuja et al. To compare bacterial count Randomized, Twenty child There was no difference 2a 20096 at day 4 and 7 from the initial double blind and adult in bacterial count. lesions in patients who patients received received framycetin vs. local treatment silver sulfadiazine. with 1% framycetin vs. 20 patients who received silver sulfadiazine.

Liao et al. To compare the time of healing Randomized, One hundred Time of healing of superficial 2a 20067 and the antibacterial effect of multicenter and twenty burns in the silver nitrate group silver sulfadiazine vs. silver child and adult was 9.5 ± 2.7 days vs. nitrate in superficial and patients: 10.8 ± 3.4 days in the intermediate burns. 60 treated with sulfadiazine group. sulfadiazine vs. Percent of healed tissue at 60 treated with day 7 was 77.9 ± 20.5% with silver nitrate. silver nitrate vs. 67.3 ± 22.6%, with sulfadiazine (p ≤0.01). In intermediate burns, time of healing was 21.5 ± 4.8 days with silver nitrate vs. 23.3 ± 6.4 days with sulfadiazine (p ≤0.01). Percent of healing at day 7 was 86.6 ± 15.9% with nitrate vs. 78.5 ± 17.7% with sulfadiazine (p ≤0.01).

Muangman P, To compare the efficacy of silver Quasi- Fifty child There was no difference 2b 20068 sulfadiazine vs. Acticoat® in randomized and patients between groups in the superficial and intermediate study incidence of infection burns affecting less than 25% of or hospital stay. the BSA.

Livingston To compare the effects of Randomized Fifty two Among patients with burns 2b et al. silver nitrate vs. neomycin children affecting 20-40% or more of their 199010 plus bacitracin vs. and adults BSA, graft loss was: 1/14 with Ringer’s lactate antibiotics vs. 4/6 with on graft loss. Ringer’s lactate (p <0.05). Among patients with burns affecting >40% of the BSA, graft loss was 67%, regardless of the topical agent used. Systematic Review: which topical agent is more efficacious in the prevention of infections in burn patients? / 303

Snelling et To compare silver sulfadiazine Randomized Seventy one Tissue colonization by 2a al. vs. gentamicin in preventing child and Pseudomonas spp was found in 197815 colonization and sepsis in burns adult patients 37% of the 38 patients treated with affecting more than 10% sulfadiazine and in 30% of the 33 of the BSA, at 24 hours patients treated with gentamicin; from the initial injury. 21% of Pseudomonas spp were resistant to gentamicin.

Malik et To compare the efficacy of Randomized One hundred Reepithelization time was shorter 2a al. honey vs. silver sulfadiazine and fifty child in the honey group than in the 201012 in superficial burns affecting and adult sulfadiazine group (13.47 ± 4.06 <40% of the BSA. patients vs. 15.62 ± 4.4 days; p: 0.0001).

Subrahman- To compare the time of Randomization Fifty A hundred percent of patients 2a yam M. healing with silver sulfadiazine method children and treated with honey healed by 199813 vs. honey. not adults day 21, vs. 84% specified in the other group (p: 0.01).

REFERENCES 9. Brown TP, Cancio LC, McManus AT, Mason AD Jr. Sur- 1. Palmieri T, Greenhalgh D. Topical treatment of pediatric vival benefit conferred by topical antimicrobial prepara- patients with burns: a practical guide. Am J Clin Dermatol tions in burn patients: a historical perspective. J Trauma 2002;3(8):529-34. 2004;56(4):863-6. 2. Atiyeh B, Costagliola M, Hayeck S, Dibo S. Effect of silver 10. Livingston DH, Cryer HG, Miller FB, Malangoni MA, et on burn wound infection control and healing: review of al. A randomized prospective study of topical antimicro- the literature. Burns 2007;33(2):139-48. bial agents on skin grafts after thermal injury. Plast Recon- 3. Mackway-Jones K. Best evidence topic report. Towards str Surg 1990;86(6):1059-64. evidence-based emergency medicine: best BETs from the 11. Mian EU, Gianfaldoni R, Mian M. Topical treatment of Manchester Royal Infrmary. Emerg Med J 2006;23(4):300. burn wounds with chloroxidating solution and silver 4. de Gracia CG. An open study comparing topical silver sulfadiazine: a comparative study. Drugs Exp Clin Res sulfadiazine and topical silver sulfadyazine cerium ni- 1991;17(4):243-52. trate in the treatment of moderate and severe burns. Burns 12. Malik KI, Malik MA, Aslam A. Honey compared with sil- 2001;(27):67-74. ver sulphadiazine in the treatment of superficial partial 5. Garner JP, Heppell PS. Cerium nitrate in the management thickness burns. Int Wound J 2010;7(5):413-7. of burns. Burns 2005;31(5):539-47. 13. Subrahmanyam M. A prospective randomised clinical and 6. Ahuja RB, Gupta A, Gur R. A prospective double blinded histological study of superficial burn wound healing with comparative analysis of framycetin and silver sulphadi- honey and silver sulphadiazine. Burns 1998;24(2):157-61. azine as topical agents for burns: a pilot study. Burns 14. Starley IF, Mohammed P, Schneider G, Bickler S. The 2009;35(5):672-6. treatment of paediatric burns using topical papaya. Burns 7. Liao ZJ, Huan JN, Lv GZ, Shou YM, Wang ZY. Multicenter 1999;25(7):636-9. clinical study of the effect of silver nitrate ointment on the 15. Snelling CF, Ronald AR, Waters WR, Yaworski DS, et al. partial thickness burn wounds. Zhongua Shao Shang Za Zhi Comparison of silver sulfadiazine and gentamicin for topi- 2006;22(5):359-61. cal prophylaxis against burn wound sepsis. Can Med Assoc 8. Muangman P, Chuntrasakul C, Silthram S, Suvanchote S, J 1978;119(5):466-70. et al. Comparison of efficacy of 1% silver sulfadiazine and Acticoat for treatment of partial thickness burn wounds. J Med Assoc Thai 2006;89(7):953-8.