Effectiveness of antiseptics in the bacterial load reduction after septic wound dressing at Bugando Medical Centre, Mwanza, Tanzania

¹Archbishop Anthony Mayalla School of Nursing, Catholic University of Health and Allied Sciences, Mwanza, Tanzania, ²School of Pharmacy, Catholic University of Health and Allied Sciences, Mwanza, Tanzania, ³Department of Microbiology and Immunology, Weill-Bugando School of Medicine, Catholic University of Health and Allied Sciences, Mwanza, Tanzania, ⁴Department of Surgery, Weill Bugando School of Medicine, Catholic University of Health and Allied Sciences, and Bugando Medical Centre, Mwanza, Tanzania, ⁵Department of Orthopaedic Surgery, Weill Bugando School of Medicine, Catholic University of Health and Allied Sciences, and Bugando Medical Centre, Mwanza, Tanzania

^&Corresponding author
Jeremiah Seni, Department of Microbiology and Immunology, Weill-Bugando School of Medicine, Catholic University of Health and Allied Sciences, Mwanza, Tanzania

Introduction    

Septic wounds are characterized by pus, excavation of the skin or underlying soft tissue, and are most often caused by bacteria or fungal pathogens [1,2]. Infected wounds are associated with profound discomfort, prolonged hospital stay due to delayed recovery, and enormous health care expenditures when implicated pathogens are multi-drug resistant bacteria (MDR) [3-5]. Globally, surgical site infections (SSIs) and non-healing or chronic wounds substantially contribute to healthcare-associated infections (HAI) and community-associated infections (CAI), respectively [6,7]. Although, SSIs are low in developed countries compared to developing countries [6], the overall prevalence of SSIs post-caesarian sections and among general surgical patients have remained relatively higher (10.0% and 26%, respectively) at Bugando Medical Center (BMC) tertiary hospital in Tanzania [8,9]. In this hospital, bacteria colonizing chronic lower limb ulcers are commonly MDR, with predominance of two phenotypes, namely Methicillin resistant Staphylococcus aureus (MRSA) in approximately 44.4% and Extended spectrum beta lactamases (ESBL) producing Gram negative bacteria in approximately 35.0% [10]. Therefore, searching for rational antimicrobial therapies and other management modalities for septic wounds remains an ongoing challenge among health care professionals. Antiseptics are chemical agents that kill or inhibit the multiplication of microorganisms, and are essential first-line chemical agents for wound management [11,12].

Antiseptics are prerequisites for the reduction of bacterial load in septic wounds prior to antibiotics administration. The most commonly used antiseptics in clinical practice include povidone?iodine, polyhexamethylene biguanide, chlorhexidine, alcohol, hydrogen peroxide, boric acid, silver nitrate, silver sulfadiazine, and sodium hypochlorite [11,12]. Their spectrum of activities varies depending on their chemical constituents, concentration of the active ingredients, inanimate or animate surfaces targeted, and other environmental conditions [12]. Despite extensive research on the causes of wound infections, the use of standardized surgical protocols/procedures and antibiotic therapeutic options for septic wound management, the role of antiseptics in the total bacteria load reduction has been sparsely studied [1,13]. Objective: this study had three objectives. Firstly, to evaluate the effectiveness of antiseptics in the total bacterial load reduction and MDR bacterial load reduction (i.e. ESBL producing Gram negative bacteria and MRSA) after septic wound to guide evidence-based infection prevention and control measures (IPC). Secondly, to determine bacteria species implicated in causing wound infections, and thirdly, to determine factors associated with bacterial load reduction after septic wound dressing at BMC Mwanza - Tanzania.

Methods     

Study design: this was a cross-sectional analytical study.

Settings: this study was conducted from May to August 2021. The BMC is a tertiary hospital in the northwestern part of Tanzania with approximately 1000 bed capacity. It attends around 300,000 patients each year and serves as a referral hospital for eight regions, which altogether constitute a catchment population of approximately 20 million people. Bugando Medical Centre is also a teaching hospital for the Catholic University of Health and Allied Sciences (CUHAS).

Participants: this study included patients (outpatients and inpatients) with septic wounds of all age groups attending surgical departments at BMC. Patients with septic wounds who were already enrolled in the study but were attending for subsequent hospital visits, and those who did not consent to the study were excluded.

Variables: outcome variables in this study included culture positivity after wound dressing, bacterial load reduction after wound dressing, and AMR bacterial phenotypes (MRSA and ESBL-producing Gram-negative bacteria) after wound dressing. Exposure variables in this study included sociodemographic and clinical characteristics of patients, and the type of antiseptics used for wound dressing.

Data sources/measurements: demographic data and clinical data from the study participants were collected using a pre-tested data collection tool. Two wound swabs (before and after wound dressing) were collected from each patient, placed in Stuart´s transport media, and taken to the CUHAS Microbiology laboratory for analysis within two hours. The swabs were placed into two different sterile bottles with 2 mL of normal saline and gently shaken. Thereafter, using a standard 1μl loop, each sample was quantitatively inoculated into plain Blood agar plate and plain MacConkey agar, and incubated at 35 - 37°C for 18 - 24 hours for subsequent isolation of Gram-positive bacteria and Gram-negative bacteria, respectively. Colony-forming units (CFU) were then quantified (before and after wound dressing) under the supervision of skilled laboratory scientists and medical microbiologists [14]. To enumerate MDR bacteria, a primary culture was also done onto MacConkey agar supplemented with 2μg/ml cefotaxime and on CHROMagarMRSA® for screening of ESBL-producing Gram-negative bacteria and MRSA, respectively. Confirmation of ESBL and MRSA was done using the combined disk method and zone of inhibition of ≤ 21mm to the cefoxitin disk (30 μg), respectively [15]. Bacterial identification and MDR bacteria susceptibility testing were limited only to samples post-wound dressing.

Bias: potential bias was largely reduced by using quality control measures. For example, all samples were transported from the wards/clinics and processed in the laboratory within two hours, and the used of Stuart´s transport media ensured viability while limiting overgrowth of bacteria. All culture media were prepared as per the manufacturer´s instructions. Performance and sterility tests were conducted. A 0.5 McFarland standard was used to standardize bacterial inoculum density for the antimicrobial susceptibility testing for ESBL and MRSA. The standard reference strains are Methicillin-sensitive S. aureus (ATCC-25923), MRSA (ATCC 29213), Non-ESBL E. coli (ATCC-25922), and ESBL-K. pneumoniae (ATCC -700603) were used to guide all laboratory procedures [15]. To ensure that the antiseptics used were not contaminated by microorganisms, antiseptics were collected in sterile containers prior to wound dressing weekly, and cultured for bacteria and fungi. In all cases, neither bacteria nor fungi grew.

Study size: a minimum sample size of 180 was estimated using Kish-Leslie formula (1965) with proportions of 10.9% from Mpogoro et al, 2014, and 67.7% from Moremi et al, 2014 in the same hospital using Z of 1.96 at 95% confidence interval and a sampling error of 0.05, a sample size between 149 and 336 would suffice [8,10,16]. A total of 223 patients with septic wounds were enrolled during the study period using a convenience sampling technique. Out of these, 20 were excluded (6 were already enrolled before, 2 had their wounds already dressed prior to sample collection, and 12 did not consent to be sampled). Therefore, the final sample size was 203.

Quantitative variables: anatomical sites where wound swabs were taken were categorized into three (i.e. upper limbs, lower limbs, and others). The other groups were a composite sub-category to cater for multiple sites which had small frequencies. Quantitative bacterial culture was done using a calibrated 1μL loop to get bacterial colony-forming units per ml of normal saline suspension.

Statistical methods: all data recorded in the questionnaires were transferred to Microsoft Excel® for coding and then transferred to STATA version 13.0 software for analysis. Results were presented in percentages or proportions for categorical variables and median (interquartile range) for continuous variables. Two-sample test of proportions was used to compare the proportions of MDR (i.e. ESBL and MRSA) before and after septic wound dressing. Significant reduction of bacterial CFU load was assessed using the Two-sample Wilcoxon rank-sum (Mann-Whitney) test. All variables with a p-value of ?0.05 on bivariate analysis were subjected to multivariate logistic regression analysis so as to assess the independent predictors for culture positivity after wound dressing using 95% CI and a cut-off value of ≤0.05.

Ethical considerations: ethical approval to conduct this study was sough and provided by the Joint CUHAS, BMC research Ethics and Review committee (CREC 1869/2021 and CREC 1873/2021). Permission to conduct this study was granted by the Director General of BMC and Heads of Surgical Departments. The importance of the study, study procedures and right to participate or withdrawal was explained to each participant in details. Then, a written informed consent was sought from each patient above 18 years, whereas for children, parents/guardians provided permission on their behalf (however, teenagers were also requested to sign assent forms). Confidentiality and privacy were strictly ensured. Results for laboratory investigations for samples post-wound dressing were promptly communicated to the attending doctors to guide specific patients´ management

Results     

Participants and descriptive data:the median age (IQR) of patients involved was 29 (26 - 37) years, with the youngest and oldest participants being 1 month and 80 years, respectively. Approximately two-thirds of participants were males. Out of 203 patients, half had wounds due to road traffic accidents (RTA), and 18.2% had a previous history of antibiotic use in the past month. The most commonly used antiseptics were povidone iodine in approximately 70.9% (Table 1).

Outcome data

The total bacterial load before and after wound dressing among patients with septic wounds: the majority of patients were culture positive before wound dressing (71.9%, 146/203), and the culture positivity significantly decreased to 39.4% (80/203) after wound dressing (p-value < 0.001). There was a significant total median (IQR) bacterial load reduction from 205.5 (112-330) before wound dressing to 128 (34.5-235) after wound dressing (p-value < 0.001), Figure 1. The reduction was also evident for povidone iodine [210 (112-347) to 0 (0-60) CFU/ml], Savlon [191.5 (124-295) to 96 (21-216) CFU/ml], and other antiseptic mixtures consisting of normal saline mixed with other agents such as hydrogen peroxide or metronidazole or spirit [234 (105-330) to 127 (27-205) CFU/ml]. A total of 107 bacteria were isolated from 80 positive culture swabs after wound dressing, with the most common bacteria species being Staphylococcus aureus (24.3%), Coagulase Negative Staphylococci (15.9%), and Escherichia coli (15%), Figure 2.

Multidrug-resistant bacteria load before and after wound dressing among patients: there was a significant decrease in the proportions of MDR bacteria after septic wound dressing among patients. Specifically, ESBL decreased from 34.0% (69/203) to 20.2% (41/203), p-value=0.0018, and MRSA decreased from 39.9% (81/203) to 11.3% (23/203), p-value<0.001. The median ESBL bacterial load reduction before and after wound dressing was 90 (40-147) CFU/ml and 40 (18-100) CFU/ml, respectively (p-value = 0.0140). On the other hand, the median (IQR) MRSA bacterial load reduction before and after wound dressing was 95 (50-167) CFU/ml and 30 (6-150) CFU/ml, respectively (p-value = 0.0169). The proportion of Staphylococcus aureus isolates which were confirmed to be MRSA was 88.5% (23/26), and the proportion of Gram-negative bacteria which were confirmed to produce ESBL was 68.3% (43/63).

Factors associated with culture positivity after wound dressing: out of 203 participants, 80 (39.4%) had aerobic culture-positive results after septic wound dressing. On bivariate analysis, the odds of culture positivity increased with education level (college and above) [OR: 95% CI = 2.83 (1.15-7.01),' p-value=0.024], and diluted antiseptics OR: 95% CI = 6.36 (3.21-12.61), p-value<0.001]. However, it decreased with povidone iodine [OR: 95% CI = 0.08 (0.02-0.24), p-value<0.001] and antibiotic use [OR: 95% CI = 0.43 (0.19-0.97), p-value=0.042]. On multivariate logistic regression analysis, the use of povidone iodine was independently associated with decreased odds of culture positivity after septic wound dressing [OR: 95% CI = 0.09 (0.01-0.63), p-value =0.016], Table 2.

Discussion     

The majority of patients in this study were males (66%) in the middle and productive ages (i.e., second and third decades of life). The predominance of males was similar to another study at the University of Benin Teaching Hospital (58%) [17]. This can be accounted for by the fact that most of septic wounds in this study were caused by RTA, and it is well known that males are mostly involved in outdoor activities, which put them at risk of being involved in RTA, in contrast to females. However, our findings contrast with another study in Nigeria, which showed that septic wounds associated with RTA ranked fourth after surgical wounds, diabetic sores, and cancer-related wounds [5]. This shows variations in the risk exposures for wound infections across countries, which calls for country-specific responsive measures. However, our findings are reiterating the significant contribution of RTA to septic wounds, calling for focused advocacy and preventive measures involving public transport operators [18,19]. The proportion of positive aerobic cultures was high before wound dressing (71%) and was halved after dressing. Staphylococcus aureus, CNS, Escherichia coli and Pseudomonas aeruginosa were predominant isolates after wound dressing. The high proportion before dressing was similar to other previous studies from Benin Teaching Hospital in Nigeria (89.4%), Mulago National Hospital in Uganda (68.8%), and Muhimbili National Hospital in Tanzania (90.0%) [17,20,21]. The findings are also similar to 67.7% from patients with chronic lower limb ulcers and 72% from women with SSIs post-caesarian section at BMC conducted 7 years ago [8,10]. However, there may be slight variations in these results due to differences in patient groups, culture techniques used, and history of previous use of antibiotics. This study underscores an increase in the prevalence of MDR strains among patients with skin and soft tissue infections in this tertiary hospital, notably, ESBL-producing Gram-negative bacteria (from 13.7% in 2014 to 34.0% in the current study), and MRSA from 2.7% in 2014 to 39.9% in the current study [10]. There is also a significant increase in the proportions of ESBL among Gram relative bacteria (from 35% in 2014 to 68.3%), and MRSA among Staphylococcus aureus (from 44.4% in 2014 to 88.5%) [10]. The predominance of MRSA and ESBL has also been previously reported in the same hospital among general surgical patients connoting predilection of MDR strains in the hospital settings [9]. However, the situation is different in community-associated skin and soft tissue infections in Morogoro, Tanzania, where the two MDR phenotypes are rarely found [22]. These findings highlight the fact that, although the proportions of culture positivity among patients with skin and soft tissue infections have remained relatively constant, the proportions of MDR strains in these patients are alarmingly high in the hospital settings, calling for strengthening of antimicrobial stewardship and IPC measures.

This study has shown a significant bacterial load reduction (both total bacteria, ESBL, and MRSA) after wound dressing, emphasizing on the the importance of this procedure in wound management prior to instituting antimicrobial therapies. However, the reduction was more with povidone iodine than salvon and normal saline mixed other antimicrobial agents. Povidone-iodine´s effectiveness was maintained even after controlling for potential confounders, whereby the OR and 95%CI remained protective [0.09 (0.01-0.63), p-value=0.016]. A study done in the USA showed that all antiseptics used (chlorhexidine, chlorhexidine in isopropyl alcohol, povidone-iodine, povacrylex-iodine in isopropyl alcohol, and isopropyl alcohol) were effective, with observed incidence of SSIs ranging from 4.0% to 6.0% [23]. In a systematic review involving 27 randomized controlled trials, iodine was shown to be safe and superior to silver sulfadiazine cream and non-antiseptic dressings. However, it was inferior to honey-based dressing and Rifamycin-based dressing in terms of time to complete bacterial clearance and wound healing [24]. A recent randomized controlled trial involving 54 hospitals in seven low-and-middle income countries showed similar results in preventing SSIs among patients with clean-contaminated or dirty surgical wounds when using 2% alcoholic chlorhexidine skin preparation compared with povidone-iodine, or with triclosan-coated sutures compared with non-coated sutures [13]. Therefore, these findings emphasize on the usefulness of the widely available and cost-effective povidone-iodine antiseptic in wound dressing. However, the fact that diluted antiseptics had little to no effect on bacterial load reduction in bivariate analysis calls for further studies to assess staff compliance with antiseptic dilution guidelines at BMC, review of shelf-life for antiseptics, review of sources/manufacturers of antiseptics as well as evaluation of antiseptics´ active ingredients so as to institute specific remedial IPC measures.

Limitations of the study: firstly, this study was cross-sectional in nature conducted between May and August 2021, and therefore could not assess the trend of septic wound infections at BMC over time. Secondly, anaerobic culture was not done due to the unavailability of diagnostic infrastructures, and therefore, this may underestimate the true burden of septic wound infections. Lastly, the active ingredients in various antiseptics were not evaluated, and can be an area of focus in future studies.

Conclusion     

This study showed that approximately 70% of septic wounds were microbiologically confirmed to be infected with bacteria at BMC tertiary hospital prior to wound dressing, and this significantly decreased to 39.4% after dressing. Half these wounds were due to RTA. There was a significant reduction of the total bacteria, ESBL and MRSA load, notably by povidone-iodine. Routine monitoring of antiseptics’ effectiveness is reiterated as adjunct to antibiotics during septic wound management. Assessment of the active ingredients in the antiseptics used will be of interest in future studies. Road traffic accident victims should be prioritized in the preventive measures for MDR attributable septic wounds.

Competing interests     

The authors declare no competing interest.

Authors' contributions     

Emmanuel Jagadi, Doris George Makweta andJeremiah Seni made substantial contributions to the conception, designing, and execution of the study; Emmanuel Jagadi, Doris George Makweta, Magwa Jisusi Kiyumbi, Helmut Nyawale, Zengo Kashinje, Vitus Silago, Francis Tegete, Inyasi Lawrence Akaro, Jeremiah Seni made substantial contributions to the acquisition of data and clinical samples; Francis Tegete and Inyasi Lawrence Akaro played critical roles in the patients’ management; Emmanuel Jagadi, Doris George Makweta, Magwa Jisusi Kiyumbi, Helmut Nyawale, Zengo Kashinje, Vitus Silago and Jeremiah Seni made substantial contributions to the analyses of data and clinical samples, and performed extensive literature searches; Emmanuel Jagadi, Doris George Makweta and Zengo Kashinje wrote the initial draft of the manuscript, which was critically revised by all authors. All authors have read and agreed on the final manuscript.

Acknowledgments     

The authors are thankful to all patients for their involvement in this study. Authors are also grateful to staff in the Surgical Departments, and Department of Microbiology and Immunology. Special thanks to Mr. Kenan Malindisa, Mr. Modest Benard and Mr. Boniface A. Remi for their technical support during sample analysis.

Tables and figures     

Table 1: socio demographic and clinical characteristics of patients with septic wound at BMC

Table 2: factors associated with bacteria culture positivity after wound dressing

Figure 1: comparison of median total bacterial load before and after wound dressing

Figure 2: bacterial species distribution among patients with septic wound after dressing

References