A Prospective, Descriptive Study to Identify the Microbiological Profile of Chronic Wounds in Outpatients
Indiscriminate use of antibiotics for infected chronic wounds is a global problem that may contribute to delayed healing and the development of drug-resistant micro-organisms. A prospective, descriptive cohort study of 50 male and female outpatients (mean age 52.50 [± 14.84] years, range 18–90) with 52 chronic wounds was conducted to investigate the microbiological profile and prevalence of drug-resistant strains in chronic nonhealing wounds to develop an evidence-based approach to antibiotic therapy until drug sensitivity reports are available.
Mean wound duration was 8.23 (± 12.35) months (range 1.5–72), average wound size was 29.70 (± 37.83) cm, and most patients had a lower extremity wound and diabetes mellitus (n = 20). Pus and tissue samples were cultured and tested. Most (45) wounds contained a single organism and nine different genera were isolated. Of those, 39 were Gram-negative and 11 were Gram-positive (z = 5.50, P =<0.001). The most common organisms were Pseudomonas (21 wounds) and Escherichia coli (eight wounds). Pseudomonas aeruginosa was more common in patients with diabetes mellitus, in lower extremity ulcers, and in ulcers >20 cm2 (z-test, P <0.05). The presence of two organisms was more commonly observed in postsurgical/traumatic wounds. Ten (10) out of 55 pathogens (18.18%) isolated were drug-resistant, including Pseudomonas (seven), methicillin-resistant Staphylococcus aureus (one), and extended-spectrum beta lactamase (two — E. coli and Citrobacter). Most (70%) drug-resistant pathogens were obtained from persons with diabetes mellitus. Overall sensitivity to piperacillin and tazobactum combination was high. Because the prevalence of monomicrobial flora in chronic wounds is high, if a wound infection is suspected, empiric therapy should target the most prevalent flora. The high rate of drug-resistant Pseudomonas and MRSA strains should discourage antibiotic use in chronic ulcers before obtaining culture results.
When progress stalls during any stage of the healing process, a chronic wound (ie, a wound of >6 weeks’ duration) may develop.1 Such wounds are associated with great physical and psychological patient burden in terms of pain, wound discharge, infection, and resources, 2 demanding extensive therapy and increased dependence on nursing services. 3 Literature on microbial populations that colonize chronic wounds reflects the diversity of patient demographics, underlying wound etiology (eg, venous ulcers, diabetic foot ulcers), duration of treatment, technique used to collect specimens, and species. 4,5 The microflora is usually polymicrobial and complex and has been found in vivo to range from 1.6 to 4.4 species per ulcer, 6 with the potential to change over time. 7 In various prospective and retrospective studies, Staphylococcus aureus and coagulase-negative Staphylococcus are the species most commonly isolated, occurring in frequencies ranging from 43% of infected leg ulcers8 to 88% of nonsymptomatic ulcers. 9 Pseudomonas aeruginosa has been identified in 7% to 33% of ulcers. 9 Other aerobic species including Klebsiella, Escherichia coli, Proteus species, Enterobacter, and Enterococci8-10 make the list extensive and complex. Anaerobic organisms are also frequently identified, ranging in prevalence in chronic wounds from as low as 6% to as high as 88%8,11 and include Peptostreptococcus, Prevotella, Finegoldia magna, and Bacteroides. 11
Although not exhaustive, this list demonstrates the wide range of microbial flora that may colonize a chronic wound. Despite advances in wound care, the clinician still is faced with the dilemma of clinically differentiating between colonization and infection and in reckoning the ideal time to initiate the appropriate therapy. It is tempting to start antibiotic treatment in the absence of evidence of the nature of microbial flora to allay apprehension of “doing less”. This practice has led to the rise of multi drug-resistant bacterial strains such as methicillin-resistant S. aureus (MRSA), vancomycin-resistant S. aureus (VRSA), and extended-spectrum beta lactamase- (ESBL) producing bacteria. 12 Moreover, because bacterial profiles vary among wounds of different etiology and location, empiric therapy may not be the same for all chronic wounds and the antibiogram may change depending on a specific situation. 4 Further, guidance regarding antibiotic therapy may vary in different regions of the globe. In the developing world, where healthcare is subsidized or free, the prime concern probably is infrequent patient contact and follow-up, which may motivate the clinician to overprescribe; whereas, the scenario may be guided by the cost-effectiveness of treatment in developed nations where the rising cost of treatment is a major concern for both insured and uninsured individuals as well as state-run health facilities. 13 A review by O’Meara et al13 concluded that evidence for use of antibiotics in wound healing is insufficient and that until data are available, other criteria such as cost minimization may be used as guidelines for antibiotic use.
The present study was conducted with the aim to 1) document the prevalence of pathogens in chronic wounds, 2) analyze pathogen distribution according to wound type, site, size, and duration and eventually provide the clinician with clues to initiate best-guess treatment, and 3) determine the pattern of drug resistance in outpatients. The second goal — ie, analysis of distribution by wound type — was thought to be crucial in conditions where prevalence of drug resistance is unknown and microbiological reports take time to reach the clinician.
The present study was conducted from June to September 2007 at the Wound Clinic of the University Hospital of Banaras Hindu University, Varanasi, a tertiary teaching hospital in Northern India, and was approved by the Hospital Ethics Committee.
Sample population. All consecutive outpatients with a wound >6 weeks’ duration were recruited for the study irrespective of wound type, therapy, or comorbid disease. Written informed consent was secured from all participants. Only persons with malignant ulcers as determined by histopathological diagnosis from biopsy specimen were excluded from the study.
Procedure. Wound site and type, a brief clinical history, and examination findings were recorded. Wound duration was obtained via patient recall and two-dimensional size was measured by acetate tracing; area then was calculated by transferring the tracing to standard graph paper. Before formal cleansing and dressing, a pus/exudate sample was taken by sweeping a sterile swab in the center of the ulcer; a tissue sample was taken by scraping the slough. The material was transferred immediately to the Microbiology Laboratory along with case notes. Care was taken to avoid contamination with normal skin flora. In the laboratory, swabs were initially plated on blood agar and McConkey’s media and incubated overnight at 37° C. Various colonies were identified. Organisms were isolated using gram stain and standard biochemical tests and assessed for antibiotic sensitivity by disc diffusion technique. Culture for anaerobes could not be performed owing to the unavailability of anaerobic bacterial culture techniques for lack of funds. All variables were entered into a Microsoft Excel database. For analysis purposes, ulcer size and duration were divided into three groups. Statistical evaluation was done using the SPSS software for Windows (version 11, SPSS Inc, Chicago, IL). Tests for significance, including Z-test and Fischer’s exact test, were applied where necessary.
Fifty (50) patients with 52 wounds participated in the study, average age 52.5 (± 14.84) years; the majority (35) were men (see Table 1). The most common ulcers were foot ulcers (29; 55.76%). Of the 52 wounds evaluated, two were sterile, 45 contained one organism, and five contained two organisms. A total of 55 organisms were cultured and nine different genera were isolated (see Table 2).
The presence of two organisms was more commonly observed in postsurgical wounds (included as traumatic ulcers). Gram-negative bacteria were cultured more often than Gram-positive bacteria (75% versus 21.2% , z = 5.50, P <0.001) and were significantly more common in lower extremity and larger wounds (P <0.001) than in trunk wounds or wounds <10 cm2 (see Table 3). P. aeruginosa was the most common organism in all wounds with the exception of pressure ulcers and was significantly more common in lower extremity wounds and in ulcers >20 cm2 (P <0.05) (see Table 4).
During the study period, 70% of patients received some form of treatment including systemic antibiotics, topical antibiotic creams, emollients, and/or treatment for comorbid disease such as diabetes and venous insufficiency. Among these patients, 90% of wounds were positive for bacteria. The results of the antibiotic sensitivity tests showed that 10 out of 55 pathogens (18.18%) isolated were drug-resistant, including Pseudomonas (seven), MRSA (one) and extended-spectrum beta lactamase ([ESBL] two — E. coli and Citrobacter) (see Table 5). Most (70%) drug-resistant strains were isolated from patients with diabetes during multiple follow-up visits to the clinic. An overall sensitivity to piperacillin and tazobactum combination also was observed.
Healing stalls in wounds exhibiting ongoing inflammation with poor remodeling and re-epithelialization failure.14 One important cause of a prolonged inflammatory response is the constant presence of virulent bacteria in chronic wounds. 15 However, the mere presence of bacteria in a wound does not indicate infection or that healing will be impaired. 16 Although studies15,17 show a delay in healing due to the presence of multiple bacteria, it also has been proposed that a low number of certain bacteria may actually promote healing.18 The exact role of micro-organisms present in chronic wounds is not fully known, but it seems bacteria represent a critical bioburden that decides the wound’s fate. 19
It has been shown that any open wound is inevitably contaminated by normal skin flora.20 In some instances, wound flora may change and stabilize over time due to the bacterial synergy that develops as a result of specific species interaction,8 in which circumstance the wound still may progress toward healing. However, at a subtle point when the host defense is overwhelmed or a critical bacterial load is reached, critical colonization may occur, tilting the balance in favor of the microbe.21 Progressive bacterial invasion of the wound causes infection, the hallmarks of which are increased discharge and ulcer size, erythema, edema of the surrounding skin, and pain in a previously painless wound.22,23 When all of these characteristics are present, therapy is straightforward: systemic antibiotics usually are prescribed. Clinically, symptoms may be misleading — pain may be absent in neuropathic ulcers, erythema may be related to wound colonization with Morganella morganii,24 or signs of inflammation may be difficult to assess and/or recognize in immunocompromised states.25 When facing this dilemma, the clinician tends to overprescribe antibiotics to guard against undertreatment. Prolonged or indiscriminate antibiotic use may lead to development of drug resistance, associated drug-related complications, and/or immense financial stress on the patient and the health system.
Several options are available for preventing the development of resistant strains. Appropriate antibiotics should be prescribed in shorter courses instead of longer duration26,27 or the antibiotic therapy may be changed depending on the acute care setting’s antibiogram report.8 Microbial analysis is of benefit when considered in concert with clinical observation; sensitivity studies can refine the antibiotic selection process.28 However, the value of the microbiological report is diminished by the sensitivity accuracy of common techniques used to collect samples — ie, curettage (75%), needle aspirate (69%), and culture swabs (60%).6 Further, it may not be practically feasible to obtain the culture/sensitivity report before starting treatment. Organisms require 2 to 3 days to grow on a selective media and another 2 to 3 days to gauge susceptibility. During this time, in the absence of antibiotic therapy, existing infections will continue and may worsen. Thus, prior evidence-based knowledge of the bacterial flora and the sensitivity pattern in ulcers may guide the appropriate empiric use of antibiotic before definitive reports are available.
In the present study, most of the patient participants were breadwinners — in particular, daily wage bearers and laborers. They attended the authors’ clinic from great distances; money spent on conveyance was high when compared to their daily earning. Multiple visits meant lost wages above the cost of travel and medicine. Likely, these patients are reluctant to visit the hospital repeatedly for the same health issue. Hence, an evidence-based approach for selection of antibiotic therapy in nonhealing infected wounds was required until the drug sensitivity reports were available. The purpose of this study was to obtain data to provide clues for the initiation of best-guess treatments.
Most of the chronic wounds in the present study were found to be monomicrobial. Previous prospective studies29,30 report most leg and foot ulcers are polymicrobial. In prospective and retrospective studies, 31 S. aureus and coagulase-negative Staphylococci have been found to be the most common pathogens, with frequencies varying from 43% of infected leg ulcers to as high as 88% of non-long-standing chronic ulcers. In the present study, Staphylococcus was more common in diabetic wounds and, in concordance with previous studies,8 in chronic wounds that had been present for <3 months. Although Staphylococcus was found more commonly in diabetic ulcers, among all pathogens, P. aeruginosa was the most common organism isolated in patients with diabetes, followed by E. coli. Previous reports32 have cited a very high prevalence of MRSA (ie, 15% in all diabetic foot ulcers), which, during a 3 year follow up, had doubled to 30%.33 Other studies11,34 report a MRSA prevalence rate of approximately 12%. Consistent with reports of previous research, the current study found an 18.18% overall prevalence of drug-resistant microbes, with a high rate of resistant Pseudomonas (33.33%) and MRSA (25%). Most of these resistant organisms were isolated from diabetic ulcers.
Larger chronic ulcers of longer duration (>3-month duration) tended to accumulate Gram-negative bacteria, most likely Pseudomonas species. In the current study, the Klebsiella species occurred more commonly in the early wounds, in traumatic and postsurgical wounds, and in lower extremity wounds, findings in concordance with previous studies. 10,11 It colonized both lower extremity and trunk wounds. Overwhelming involvement of the foot (55.76% of the study ulcers), an expected statistic in rural India where people walk barefooted, calls for the provision of proper foot care to prevent chronic ulcers, especially in persons with diabetes and neuropathy.
The main point of concern remains the high prevalence of drug-resistant strains in the authors’ patients who are nutritionally compromised and the majority of whom have diabetes mellitus. Most had received treatment of some form before coming to the authors’ hospital, exacerbating overuse of broad-spectrum antibiotics. Most of the resistant strains were found in the group of patients with diabetes. These factors imply that more attention needs to be given to diabetes control and teaching self-care techniques of the foot.
Evaluation of sensitivity patterns showed an overall sensitivity to a combination of piperacillin and tazobactum. Pseudomonas species were highly sensitive to carbenecillin, ofloxacin, and amikacin. Sensitivity of many pathogens to multiple other drugs also was noted; Staphylococci were highly sensitive to gentamicin.
Anaerobic flora could not be assessed, forcing the authors to apply the results of this study to aerobic flora only. Further, all culture techniques have inherent deficiencies — eg, the poor sensitivity of culture swabs. A larger sample size would have made the study more robust and representative. The practice guidelines that emerged for this facility as a result of this study are based on the inadequate contact between patients and their clinicians, which is compromised by distance as much as by economics. Further studies that allow manipulation of the biofilm in a chronic wound, as well as development of clinical and biochemical parameters that can predict invasion of colonizing bacteria in nonhealing ulcers, are needed.
Although infection is a continuous problem in chronic wounds, the nature of the wound has been found to be a determining factor in therapy decisions — ie, healing might be hampered unless underlying comorbid factors are addressed.35 Thus, clinicians should not overlook the underlying cause and not remain focused only on the bacterial population. Wound healing is affected by many local as well as significant systemic factors, including nutrition,36 immunocompetence,25 and comorbid disease.1
A prospective study of 52 chronic wounds (the majority lower extremity ulcers) of various etiologies demonstrated that current practices in India may perpetuate inappropriate (overuse and/or mismatched to organism cultured) antibiotic treatment, strengthening the potential for antibiotic resistance. Wound etiology, duration, and location are key factors in determining treatment. In the majority of ulcers studied (primarily in the lower extremities), Pseudomonas spp. was found to be most prevalent. Although many drug-resistant micro-organisms were found in wounds, an overall high sensitivity to piperacillin and tazobactum combination was observed. Further study that addresses the clinical and biochemical parameters that can predict bacterial invasion and provide guidelines for treatment, particularly in resource-challenged countries, is warranted. Until then, all local and systemic factors relevant to healing should be considered before antibiotics are prescribed for the languishing wound.
Dr. Basu is a lecturer, Dr. Panray is an intern, and Prof. Shukla is a Professor, Department of General Surgery; Dr. Singh is a Reader, Department of Biostatistics; and Dr. Gulati is a Professor, Department of Microbiology, Institute of Medical Sciences, Banaras Hindu University, Varanasi, India. Please address correspondence to: Prof. Vijay K. Shukla, MS, MCh(Wales), FAMS, Department of General Surgery, Institute of Medical Sciences, Banaras Hindu University, Varanasi-221005, India; email:email@example.com.