Bacteriological Profile and Antibiotic Susceptibility Pattern of Diabetic Foot Infection in a Tertiary Care Hospital in Lampung, Indonesia

Background Diabetic foot infection (DFI) is a serious complication of diabetes mellitus and identification of the causative bacteria is an essential step in selecting the appropriate antibiotic therapy. This study aimed to evaluate the bacterial pattern and antibiotic susceptibility of the bacteria causing DFI in Lampung Province in Indonesia. Methods This study is a retrospective study reviewing the medical records of DFI patients admitted to the Dr Hi Abdul Moeloek Regional General Hospital in 2017–2019. DFI patients with complete medical record data were included in this study. Demographic, clinical, laboratory, wound culture and antibiotic susceptibility data were collected from the medical records using a short structural chart. The data obtained then reviewed. Results In this study, 131 DFI patients met the study criteria and were included. Based on the wound culture results, Gram-negative bacteria were obtained in 112 (85.5%) subjects with Enterobacter spp. as the predominant bacteria. Gram-positive bacteria were found in 19 (14.5%) subjects with Staphylococcus spp. as the predominant bacteria. Gram-negative bacteria found in this study showed high susceptibility to amikacin, meropenem and sulbactam/cefoperazone. Meanwhile, the Gram-positive bacteria showed high susceptibility to meropenem, sulbactam/cefoperazone and amikacin. Conclusion The findings of the study revealed Enterobacter spp. as the most predominant bacteria causing DFI in the studied population. The highest antibiotic susceptibility was seen for amikacin, meropenem and sulbactam/cefoperazone.


Introduction
Diabetes mellitus is one of the leading causes of morbidity and mortality in the world. To this day, an estimated 463 million adults live with diabetes worldwide. This figure is projected to grow to up to 700 million by 2045 (1). The 2012 Indonesia Basic Health Research (RISKESDAS) reported that the prevalence of diabetes in Indonesia was 2.1% in the population aged 15 and above (2). and the prevalence of microorganisms in the local area and their antibiotic susceptibility. For DFI that occurs in diabetes patients in tropical/subtropical climates, IDSA recommends an empirical antibiotic regimen covering frequently isolated Gram-positive and Gramnegative pathogens as well as obligate anaerobes in moderate to severe DFI. The selection of subsequent antibiotic regimens must be carried out based on the clinical response, culture and antibiotic susceptibility results (5).
The variations in the bacteriological profile of DFI that depends on geographic, cultural and climatic factors could make the international clinical guidelines for the selection of appropriate empirical antibiotic therapy difficult to apply in developing countries (5,7). This study was conducted to evaluate the bacterial profile and antibiotic susceptibility of bacteria causing DFI in people with diabetes in Lampung Province, Indonesia, to aid in the appropriate selection of empirical antibiotics.

Research Design
This is a retrospective study reviewing the medical records of diabetic patients with DFI treated at the Dr Hi Abdul Moeloek Regional General Hospital in 2017-2019. This hospital is a provincial hospital and the main referral hospital of Lampung Province. Patients with complications of diabetes are usually referred to this hospital. The study population was defined as the total number of patients with diabetes treated at the Dr Hi Abdul Moeloek Regional General Hospital. Sample sizes were calculated using the single proportion formula. Using the global prevalence of diabetic ulcers among diabetics (6.3%) (8) and taking a precision of 0.05, the calculated minimum sample size was 91. The level of confidence was set at 95%. The sampling technique used in this study was nonprobability with total sampling, a purposive sampling method, due to the limited population size of this study. The inclusion criteria for this study included: i) patients with DFI admitted to the Dr Hi Abdul Moeloek Regional General Hospital of Lampung Province; ii) patients with complete medical records including demographic data (gender and age), clinical data (length of hospital stay, in-hospital mortality), laboratory data (complete blood count and blood chemistry), wound culture and antibiotic Diabetes mellitus is an endocrine disorder that can result in various complications, including diabetic foot infection (DFI). The lifetime risk of Type 1 and Type 2 diabetes patients with DFI is 34% (3). Circulatory disorders due to peripheral arterial disease and diabetic peripheral neuropathy predispose people with diabetes to develop DFI. Infection occurring in DFI varies from simple superficial cellulitis to chronic osteomyelitis (4).
The bacteriological profile of DFI varies depending on the severity of the disease. Staphylococcus aureus, Streptococcus pyogenes, Staphylococcus epidermidis, Escherichia coli, Pseudomonas aeruginosa, Klebsiella pneumonia, Acinetobacter spp., Proteus spp. and Enterococcus spp. are the most common pathogens. The typical bacteriological profile also differs according to geographic location. Studies in temperate regions (North America and Europe) have consistently shown that the most common pathogens in DFI are aerobic Gram-positive cocci, particularly Staphylococcus aureus and coagulase-negative Streptococcus and Staphylococcus. Recent studies of DFI from tropical/subtropical regions (especially Asia and northern Africa) have shown that the main cause of DFI is aerobic Gram-negative bacilli, either alone or in combination with Gram-positive cocci. Obligate anaerobic bacteria can also play a role in DFI, especially in ischaemic limbs and DFI with abscesses (5).
The type of infection in DFI, either monomicrobial or polymicrobial, also depends on the severity of the disease. The severe form of DFI is usually caused by polymicrobial infections, whereas mild infections are frequently monomicrobial. In severe forms of DFI, up to three or five organisms may be cultured (5).
DFI is a complex complication of diabetes and is costly to treat. Apart from being a major cause of morbidity, DFI contributes to many hospitalisations and hospital admissions for people with diabetes and is the most frequent cause of nontraumatic proximal amputation. DFI is also associated with a risk of death of up to 2.5 times that of diabetes patients without DFI (3,6).
Clinical guidelines from the Infectious Diseases Society of America (IDSA) recommend treating clinically infected diabetic foot ulcers with empirical antibiotics until microbiological culture results are available. Empirical antibiotics should be chosen in accordance with the severity of infection, clinical presentation of hospital stay. The standard identification of the causative microorganisms in our hospital was carried out using standard microbiological methods with samples taken from the patient's foot ulcer and antibiotic susceptibility testing was carried out using the disc diffusion method.

Statistical Analysis
The data obtained from this study are summarised and then presented in diagrams, tables and bar charts as appropriate. Data analysis was performed using Statistical Package for Social Science software (IBM version 21.0; SPSS Inc., Chicago, IL, USA). Continuous variables are reported as mean (standard deviation [SD]) or median (interquartile range [IQR]) and categorical variables are reported as proportions. Distribution of microorganisms isolated in wound culture was reported as frequency (%), while antibiotic resistance and susceptibility were reported as proportion (%). Comparisons were conducted via Pearson's chisquared test or Fisher's exact test for categorical variables and independent t-test or Mann-Whitney U test for continuous variables. The results were considered significant if the P-value was less than 0.05.

Results
From 2017 to 2019, 131 patients were treated with DFI and met the inclusion criteria of this study. Among them, 57 (43.5%) were males and 74 (56.5%) were females. The mean age of the study subjects was 53.9 (9.2) years old. The mean length of stay of the subjects was 10.7 (5.9) days, with a median of 10 (IQR 8) days and 18 (13.7%) subjects died during hospital stay. The characteristics of the research subjects are shown in Table 1.
The most widely used empirical antibiotic regimen for DFI treatment before antibiotic susceptibility data can be obtained in this study was a regimen consisting of ceftriaxone and metronidazole (56.5%), followed by ceftriaxone alone (15.3%), a regimen consisting of ciprofloxacin and metronidazole (14.5%), and other regimens (13.7%). susceptibility test results. Patients who had received systemic antibiotic therapy for more than 24 h in the 72 h prior to the collection of samples for microbiological and antibiotic susceptibility examinations were excluded from this study. Ethical clearance was obtained from the Medical Faculty of Universitas Lampung and the hospital authority before the commencement of this study.

DFI
is defined clinically as the manifestation of an inflammatory process due to the invasion and multiplication of microorganisms in any tissue in a diabetic patient's foot, the anatomical area under the malleoli. This condition usually results from disruption of the protective skin envelope at a site of trauma or ulceration in a patient with neuropathy or peripheral artery disease (5). We hospitalise DFI patients with moderate infection accompanied by limb ischaemia or a lack of response to outpatient treatment, severe infection, gangrenous tissue requiring amputation or the need for complex wound care. An endocrinologist or surgeon will take the lead as the physician in charge and treat the patients to increase the likelihood of limb salvage by combining medical treatment, revascularisation if needed and surgical management, which includes limited lower extremity amputation.

Data Collection
The collection of medical records of diabetic patients with DFI was carried out through the Dr Hi Abdul Moeloek Regional General Hospital medical record database according to the 10th International Classification of Disease (ICD-10) diagnosis. The diagnosis of diabetes mellitus was defined as defective code E11 on the diagnosis of ICD-10. The diagnosis of DFI was confirmed by the presence of a diagnosis in the form of a handwritten medical record: 'diabetic foot', 'diabetic foot ulcer', 'diabetic foot infection' or 'diabetic foot gangrene'. The data were then collected based on a questionnaire structured in several sections to collect various aspects of information from the medical record, including demographic data, history of diabetes, laboratory parameters during hospital stay, DFI bacteriological profile, antibiotic susceptibility profile, empirical antibiotic regimens given before the availability of microbiological culture results and susceptibility data, and observed outcomes in the form of mortality and length When the bacteria found in the wound culture of DFI patients were grouped based on Gram staining, it was found that Gram-negative bacteria were mostly susceptible to amikacin (95.5%), meropenem (92.6%) and sulbactam/ cefoperazone (90.6%). Meanwhile, Grampositive bacteria were mostly susceptible to meropenem (92.9%), sulbactam/cefoperazone (86.7%) and amikacin (80%). These results are shown in Tables 4 and 5.
The antibiotic susceptibility data revealed the most sensitive antibiotics for Enterobacter spp., Klebsiella spp. and Pseudomonas sp. as amikacin and for Proteus spp. as meropenem. Meanwhile, it was revealed that the most sensitive antibiotic for Staphylococcus spp. was sulbactam/cefoperazone (shown in Table 6). Based on the results of the wound culture, 13 pathogens were identified. Polymicrobial infection occurred among 10 (7.6%) subjects, while 111 (84.7%) others had a monomicrobial infection. The most common pathogens found were Gram-negative bacilli, with Enterobacter spp. (25.2%) as the predominant bacteria, followed by Klebsiella spp., Proteus spp., Pseudomonas spp., Alcaligenes spp., Morganella spp., Sphingomonas spp., Escherecia spp., Acinetobacter spp., Citrobacter spp. and Yersinia spp. Gram-positive cocci were also found with Staphylococcus spp. (10.7%) as the predmominant bacteria, followed by Streptococcus spp. These results are shown in Table 2.
There was no significant median length of stay difference between polymicrobial and monomicrobial infections among DFI (11.5 days versus 10 days; P = 0.204). Furthermore, mortality was revealed to not be associated with the nature of infection (P = 0.643) (shown in Table 3). The individual components of empirical antibiotic regimens used in the Dr Hi Abdul Moeloek Regional General Hospital intended to cover aerobes bacteria that showed low susceptibility against Gram-negative or Grampositive pathogens. Ceftriaxone showed only 34.1% and 50% susceptibility against Gramnegative and Gram-positive pathogens, respectively. Meanwhile, ciprofloxacin showed 52.2% and 37.5% susceptibility against Gramnegative and Gram-positive pathogens, respectively.
Apart from Klebsiella and Escherichia coli, other bacteria from Enterobacteriaceae families, such as Proteus, Enterobacter and Citrobacter, were also found in this study. The five of them live in the human digestive tract as normal flora (25). Environmental factors such as sanitation habits as well as the use of water to clean the perianal area after defecation, which causes hand contamination by the normal flora of faeces, are thought to affect the bacteriological profile of organisms causing DFI in developing countries, especially in rural areas (26).

Discussion
DFI is one of the main causes of morbidity in diabetic patients. It is more common in the older age group than in the younger groups. In this study, it was found that the mean age of patients with DFI was 53.9 years old. This is in accordance with the results of several other local studies that reported the average age of DFI patients to be over 50 years old (9-10). The majority of the patients with DFI in the present study were female. This contradicts previous studies that reported male predominance (9,(11)(12)(13). Poor glycaemic control, measured as HbA1c, fasting blood glucose or even single random blood glucose, has been shown to predict ulceration and subsequent amputation. In addition, it has also been reported to inhibit wound healing, making treatment even more difficult (14)(15). In the present study, the mean of random blood glucose was 265 (123.2) mg/ dL, above the acceptable limit for good glycaemic control.
Various types of bacteria can cause DFI and determining the specific bacteria causing this infection is an important step for a clinician to avoid using excessive and prolonged usage of broad-spectrum antibiotics. This aims to prevent various drug-related side effects, financial burdens, and antibiotic resistance (5).
Our study reported a predominance of monomicrobial infection in subjects of DFI, similar to several studies (12)(13)(16)(17)(18). However, other studies have reported polymicrobial infection predominance (19)(20). This discrepancy may be due to differences in clinical profiles or patient history (particularly the prior antibiotics scenario) of the study subjects (21). Another point to be noted here is showed high susceptibility to amikacin (100%) and meropenem (98.9%).
Meanwhile, most Gram-positive bacteria showed the highest susceptibility to meropenem (92.9%). Meropenem is an antibiotic from the carbapenem class with broad antibacterial activity against Gram-positive, Gram-negative and even anaerobic bacteria. Like other betalactam antibiotics, meropenem works by inhibiting the synthesis of bacterial cell walls, causing bacterial death. This antibiotic has excellent activity against Staphylococcus aureus and other methicillin-susceptible Staphylococci and most Streptococcus (32). Gram-positive bacteria also showed high susceptibility to sulbactam/cefoperazone (86.7%) and amikacin (80%).
In this study, the bacteria causing DFI showed resistance to commonly used antibiotics. Gram-negative bacteria showed the lowest susceptibility to chloramphenicol, tetracycline, cefixime, ampicillin, penicillin, erythromycin, amoxicillin and cefadroxil. Gram-positive bacteria showed the lowest susceptibility to chloramphenicol, amoxicillin, cefadroxil, cefotaxime, cefoperazone, tetracycline, cefixime, clindamycin, erythromycin, ampicillin and penicillin. Xie et al. (11) reported that Gram-negative bacteria, especially the Enterobacteriaceae family, exhibited a high degree of resistance to Ampicillin and antibiotics from the cephalosporin class. Xie et al. also reported that Gram-positive bacteria, especially Staphylococcus aureus, showed high resistance to commonly used antibiotics, such as penicillin and tetracyclines.
Individual components of empirical regimens given prior to the culture intended to cover aerobes in DFI were found to have low susceptibility against both Gram-negative and Gram-positive bacteria. The high rate of resistance to both the individual component of empirical antibiotic regimens given prior to the availability of causative pathogen identification and susceptibility data in the present study may be due to several factors, including hospitalisation history, recent use of broadspectrum antibiotics, history of surgery and chronic wounds, irrational use of antibiotics and horizontal transfer of antibiotic resistance genes (18).
In the clinical setting, empirical antibiotic regimens are normally commenced to treat patients with DFI before the availability of microbiological culture results. They are Our findings reported no association between the nature of infection and length of hospital stay or in-hospital mortality. This finding contrasts with a study by Lipsky et al. (27), which reported cultures yielding polymicrobial Pseudomonas aeruginosa or monomicrobial Gram-negative (other than Pseudomonas aeruginosa) as independent risk factors for increased length of hospital stay and in-hospital mortality. However, our study did not address other factors that may contribute to the length of hospital stay and in-hospital mortality, such as severity of illness, transfer from another acute care hospital, surgical site infection or prior antibiotic use.
Diabetic patients with DFI have several factors that cause a high risk of carrying multidrug-resistant microorganisms (MDR), such as inappropriate previous administration of antibiotics, chronic wounds and frequent hospitalisations. The causative pathogen and antimicrobial susceptibility profile should be considered when selecting an antibiotic regimen (28).
This study found that Gram-negative bacteria were mostly susceptible to amikacin (95.5%). Amikacin is a semisynthetic kanamycin derivative from an aminoglycoside class antibiotic. Like other aminoglycosides, amikacin works by binding to the 30S subunit ribosomal protein, causing mRNA reading error. This, in turn, inhibits protein synthesis or causes defective protein production, which leads to the death of microorganisms. Amikacin is unique among aminoglycosides due to its resistance to aminoglycoside-inactivating enzymes, enabling its usage for various microorganisms that are usually resistant to other aminoglycosides. Amikacin and other aminoglycosides have an antibacterial spectrum covering a wide range of bacteria, including Escherichia coli, Pseudomonas aeruginosa, Enterobacteria, Serratia, Proteus, Acinetobacter and Klebsiella. Amikacin also shows antibacterial activity against Staphylococcus (29)(30). The infrequent use of amikacin due to the high risk of ototoxicity and nephrotoxicity and the high price compared to other antibiotics may explain the fairly good susceptibility of various bacteria to this antibiotic (31). Apart from amikacin, Gram-negative bacteria also showed high susceptibility to meropenem (92.6%) and sulbactam/cefoperazone (90.6%). This is in line with a previous study by Kow et al. (12), which reported that Gram-negative bacteria in DFI intended to cover possible causative pathogens. The choice of empirical antibiotic regimens is often based on the patient's clinical presentation and the bacteriologic profile of the local setting (12). The findings of this study can be used as a consideration in designing an empirical antibiotic treatment protocol for the management of DFI in the local setting.
Limitations of this study include the inability to generalise the study to the whole of Indonesia, as it was mainly a single-centre study. Due to the retrospective study design, pathogenic bacteria's bacteriological profile differentiation based on the severity and grading of the DFI was not possible due to a lack of documented information in medical records. The lack of a specific culture for anaerobic bacteria is another limitation of this study. Further prospective studies are required to evaluate the clinical features and response to antibiotic treatment in DFI.

Conclusion
This study showed that the most common organisms found in DFI were aerobic Gramnegative bacteria, with Enterobacter spp. being the predominant pathogen in the studied population. The bacteria found in DFI have varying resistance to commonly used antibiotics. This study's findings provided valuable information concerning DFI in Lampung Province, Indonesia and might be used as a consideration for empirical antibiotic regimen selection to treat DFI in a local setting. Further studies with a larger sample size of the bacteria isolated from DFI in other areas in Indonesia are justified.