ABSTRACT

Objective

A cumulative antibiogram serves as a critical tool in guiding the selection of appropriate empirical therapy, facilitating de-escalation based on susceptibility results, and shaping institutional policies to combat antibiotic resistance effectively. This study aimed to evaluate changes in antimicrobial susceptibility over the years and provide guidance to clinicians in the selection of empirical therapies.

Methods

A retrospective analysis was conducted on the in vitro antimicrobial susceptibility test results of bacterial isolates obtained from clinical samples submitted to the Medical Microbiology Laboratory at Trakya University Hospital during the periods of 2015-2016 and 2022-2023. Cumulative antibiogram data were compiled in accordance with the guidelines outlined in the “Analysis and Presentation of Cumulative Antimicrobial Susceptibility Test Data” (CLSI 2014, M39-A4).

Results

A total of 7524 isolates (5009 Gram-negative, 2515 Gram-positive) were analyzed during the 2015-2016 period, while 5880 isolates (4202 Gram-negative, 1678 Gram-positive) were analyzed during the 2022-2023 period. In both timeframes, the most commonly isolated microorganisms were isolated from urine, blood/catheter, and wound/aspirate/tissue samples. The predominant isolates included Escherichia coli (E.coli), Klebsiella spp., and Enterococcus spp. Based on the findings, the following antimicrobials were identified as suitable for empirical treatment for E. coli infections: carbapenems and amikacin. For Klebsiella spp. infections: amikacin. For Enterococcus spp. infections: vancomycin, teicoplanin, linezolid, and tigecycline. For Acinetobacter spp. infections: combination therapy. Carbapenem susceptibility among Klebsiella spp., isolates decreased notably in 2022-2023, ranging between 55% and 62%, in contrast to higher rates observed during 2015-2016.

Conclusion

The regular evaluation of hospital-based antibiogram data and the revision of empirical treatment protocols based on these findings represent a crucial strategy for effectively combating antimicrobial resistance.

INTRODUCTION

According to the World Health Organization (WHO), antimicrobial resistance (AMR) has become one of the most urgent global public health challenges, significantly increasing morbidity and mortality worldwide.^{1, 2} In 2019, approximately 3.57 million deaths were attributed to antibiotic resistance. Projections by WHO estimate that this number could rise to 10 million annually by 2050.^{3, 4} The coronavirus disease-2019 pandemic has exacerbated the AMR crisis. During the pandemic, the inappropriate and excessive use of antibiotics, treatments that induce immunosuppression, and prolonged hospital stays have accelerated the development of antibiotic resistance.⁵ Additionally, financial constraints in healthcare systems and reductions in healthcare personnel have negatively impacted surveillance and control efforts aimed at combating AMR.⁶ The widespread use of hand sanitizers and surface disinfectants during the pandemic may further worsen resistance patterns in the coming years.⁷ AMR renders first-line antibiotics ineffective, leading to the replacement of these drugs with more expensive alternatives. This situation prolongs disease duration, increases healthcare costs, and imposes significant economic burdens on individuals and societies. According to the World Bank, resistant infections could trigger a global economic crisis. It is estimated that by 2050, AMR could push 28 million people into poverty each year and cost the global economy over $1 trillion annually.⁸ Under these circumstances, it is essential to develop effective global strategies to combat AMR, raise public awareness, and promote the judicious use of antibiotics. These measures are critical for mitigating the far-reaching impacts of AMR.

In community -or hospital-acquired infections, the rapid initiation of effective antibiotic therapy targeting the causative microorganisms is critical for improving survival rates.^9-12 However, the isolation and identification of the pathogen in culture, as well as the determination of its antimicrobial susceptibility profile, typically require 24-48 hours. Therefore, particularly in severe infection cases, ensuring an appropriate and broad-spectrum empirical antimicrobial therapy is of vital importance.^{13, 14}

One of the most essential tools guiding clinicians in the selection of empirical therapy are cumulative antibiogram reports. These reports were standardized for the first time in 2000 with the publication of the M39 guideline by the Clinical and Laboratory Standards Institute (CLSI). CLSI defines a cumulative antibiogram as an analysis and reporting method that reflects the percentage susceptibility of the first isolates per patient to tested antimicrobial agents, collected from a specific institution over a defined time period.¹⁵ Cumulative antibiograms provide clinicians with critical guidance for empirical therapy decisions before the antimicrobial susceptibility results of the patient’s isolated pathogen are available. Furthermore, these reports can be compiled at national and international levels, enabling the detection of regional antimicrobial susceptibility patterns and the emergence of new resistance trends.

The aim of this study is to compare the cumulative antibiogram data of bacterial isolates obtained from patient samples submitted to the Medical Microbiology Laboratory at Trakya University Hospital during the periods of 2015-2016 and 2022-2023. The study seeks to evaluate changes in antimicrobial susceptibility over the years and provide guidance to clinicians in the selection of empirical therapies.

MATERIAL AND METHODS

Ethics Approval and Consent to Participate

The study received ethical approval from the Non-interventional Scientific Research Ethics Committee of the Faculty of Medicine of Trakya University with the protocol code 2024/207 (approval no: 09/35, date: 06.05.2024).

In this study, the in vitro antimicrobial susceptibility test results of bacterial strains isolated from clinical samples submitted to the Medical Microbiology Laboratory at Trakya University Hospital during the periods of 2015-2016 and 2022-2023 were evaluated.

Identification of Bacterial Species and Susceptibility Testing

Bacterial identification and antimicrobial susceptibility testing were performed using both conventional methods and the automated VITEK-2 system (bioMérieux, France). Conventional methods included the evaluation of colony morphology, Gram staining, and basic biochemical tests such as catalase, oxidase, and coagulase when appropriate. Antimicrobial susceptibility test results were interpreted according to the current annual recommendations of the European Committee on Antimicrobial Susceptibility Testing (EUCAST).¹⁶

Data Collection and Processing

The data used in this study were retrospectively retrieved from the laboratory information system. The collected data were organized according to the guidelines outlined in the Analysis and Presentation of Cumulative Antimicrobial Susceptibility Test Data (CLSI 2014, M39-A4), based on criteria.¹⁵

Only verified test results from samples submitted for clinical diagnostic purposes were included in the study; samples submitted for surveillance or screening purposes were excluded.

Susceptibility rates (S%) were calculated exclusively for antimicrobial agents routinely tested in the laboratory.

When multiple isolates of the same bacterial species were identified from the same patient, only the first isolate was included in the study; subsequent isolates were excluded from the analysis.

While calculating S%, data categorized as “intermediate” (I) were included in the percentage of susceptible isolates (S%).

Antimicrobial Agents

The selection of which antimicrobial agents should be tested for each pathogen was determined based on the “Turkish Society of Microbiology Restricted Antimicrobial Notification Table”.¹⁷ This table aims to promote rational antimicrobial use and stewardship by recommending targeted antibiotic panels based on organism group, infection site, and clinical relevance. In line with these guidelines, agents were selected to include first-line, commonly used antibiotics, as well as critical agents for multidrug-resistant pathogens. The antibiotics tested were standardized across isolates of the same species and clinical significance. The susceptibility of the isolates included in the study was evaluated for the following antimicrobial agents:

Beta-lactams: ampicillin (AM), amoxicillin/clavulanic acid (AMC), piperacillin-tazobactam (TZP), cefuroxime (CXM), ceftazidime (CAZ), ceftriaxone (CRO), cefepime (FEP), ertapenem (ETP), imipenem (IPM), meropenem (MEM).

Aminoglycosides: gentamicin (GN), amikacin (AN).

Fluoroquinolones: ciprofloxacin (CIP), levofloxacin (LVX).

Other agents: trimethoprim-sulfamethoxazole (SXT), nitrofurantoin (F), tigecycline (TGC), erythromycin (E), clindamycin (CC), vancomycin (VA), teicoplanin (TEC), linezolid (LZD), tetracycline (TE).

The selection of antimicrobials should be guided by the S% rate based on the severity of the infection. For severe infections such as meningitis or sepsis, the WHO recommends a S% (percent susceptible) rate of ≥90% for penicillin in empirical treatment.¹⁸ In cases where the risk of mortality or severe morbidity is high, antimicrobials with a S% rate of at least 90-95% should be preferred. For milder infections, a S% rate of 80-85% may be acceptable. When no alternatives are available and the S% rate is below 80%, the agent with the highest S% rate or combination therapy should be considered. Local antimicrobial policies and guidelines must also be taken into account when making treatment decisions^19-21. In this study, we evaluated antimicrobials with a S% rate of ≥90% as suitable options for empirical treatment.

Statistical Analysis

The data were analyzed using IBM SPSS (Statistical Package for Social Sciences) Statistics 21.0. Descriptive statistics were presented as frequencies and percentages. The analyses were performed using descriptive statistics, the chi-square test, and Fisher’s Exact test. Results with P values < 0.05 were considered statistically significant.

RESULTS

Sample Types and Distribution of Bacteria

This study evaluated isolates from the periods of 2015-2016 and 2022-2023. A total of 7524 isolates (5009 Gram-negative, 2515 Gram-positive) were analyzed during 2015-2016, and 5880 isolates (4202 Gram-negative, 1678 Gram-positive) were analyzed during 2022-2023. In both periods, the most frequently isolated microorganisms were obtained from urine, blood, catheter, and wound, aspirate, and tissue samples. In 2015-2016, microorganisms were most commonly isolated from patients aged 18-65 years, whereas in 2022-2023, isolates were predominantly obtained from patients older than 65 years. For both time periods, the highest number of isolates came from samples submitted by the Internal Medicine department. The most frequently isolated microorganisms during 2015-2016 were Escherichia coli (E. coli), coagulase-negative Staphylococcus (CoNS), and Enterococcus spp. In 2022-2023, the most common isolates were E. coli, Klebsiella spp., and Enterococcus spp. (Table 1). E. coli was the most frequently isolated microorganism from urine samples, while CoNS was predominantly isolated from blood/catheter samples (Table 2). In intensive care units (ICUs), the most frequently isolated microorganisms during 2015-2016 were CoNS, Acinetobacter spp., and Klebsiella spp. In 2022-2023, the most common ICU isolates were Acinetobacter spp., Klebsiella spp., and Enterococcus spp. Among ICU samples, bacteria were most frequently isolated from blood/catheter samples (48.3%), respiratory samples (tracheal aspirate, sputum, bronchoalveolar lavage) (31.5%), and urine samples (12.8%). The highest bacterial isolation rate in ICUs was observed in Surgical ICUs, with Acinetobacter spp. being the most frequently isolated microorganism (Table 3).

Antimicrobial Susceptibility Findings

For E. coli isolates, S% exceeding 90% were observed for ETP, IMP, MEM, and AN in both urine and non-urine samples, as well as across outpatient, inpatient, and ICU settings, during both the 2015-2016 and 2022-2023 periods. Additionally, susceptibility to F in urine samples was found to be above 90%.

For Klebsiella spp. isolates, S% exceeding 90% were observed only for AN in both study periods. In 2022-2023, susceptibility to carbapenems ranged from 55% to 62%, representing a significant decline compared to the 2015-2016 period.

For Acinetobacter spp. isolates, no single antimicrobial agent demonstrated a susceptibility rate exceeding 90% in either study period. Given the limited susceptibility observed across the tested agents, empirical treatment of suspected Acinetobacter infections may require the use of combination therapy, especially in settings with high rates of multidrug resistance. For Pseudomonas spp. isolates, S% above 90% were observed for FEP and AN during 2015-2016, while in 2022-2023, only AN maintained a susceptibility rate above 90%.

For Staphylococcus aureus isolates, S% exceeding 90% were observed for LVX, VA, TEC, LZD, TGC, SXT, and GN. In methicillin-resistant S. aureus (MRSA) isolates, a significant decrease in susceptibility to E and CC was detected. Additionally, susceptibility to LVX, SXT, and GN fell below 90% in MRSA isolates. For CoNS, S% above 90% were observed for VA, LZD, and TGC. Similarly, for Enterococcus spp., VA, TEC, LZD, and TGC demonstrated S% exceeding 90%. Cumulative antimicrobial susceptibility percentages for Gram-negative and Gram-positive bacteria are presented in Table 4 and Table 5. Notably, the MRSA rate increased from 12.9% to 24.1%, while the vancomycin-resistant Enterococcus (VRE) rate rose from 2.5% to 4.2% (Table 6).

DISCUSSION

Cumulative antibiogram reports, regularly and systematically updated, serve as a critical guide for clinicians in selecting empirical antibiotic therapy. Monitoring the S% of commonly used antibiotics through cumulative antibiograms is an essential component of antibiotic stewardship programs.^{22, 23} This approach allows for the identification of regional resistance patterns and highlights areas requiring targeted interventions. Additionally, cumulative antibiograms facilitate the analysis of resistance trends over successive years, providing valuable insights into the evolution of AMR.²⁴ However, caution is warranted when interpreting these reports at the patient level. Patient-specific factors play a crucial role in antibiotic selection and in determining whether an isolated microorganism is a true pathogen or a colonizer. A notable limitation of cumulative antibiogram reports in the literature is their qualitative nature, as they typically lack minimum inhibitory concentration (MIC) values, which are critical for more nuanced decision-making, as noted in study.²⁴ In Türkiye, comprehensive cumulative antibiogram data are scarce. For this reason, the findings of this study have been compared with surveillance data from broad-scale programs. These include the ‘National Antimicrobial Resistance Surveillance System (NAMRSS)’ by the Ministry of Health Turkish Public Health Institute, the ‘Central Asian and European Surveillance of Antimicrobial Resistance (CEASER)’, and the ‘European Centre for Disease Prevention and Control’s ‘Antimicrobial Resistance Surveillance in Europe report (ECDC)’. These sources are considered valuable references for reflecting national and regional resistance patterns.

According to the CDC’s National Healthcare Safety Network surveillance, the most frequently isolated bacteria in healthcare-associated infections during 2015-2017 were E. coli (17.5%), Enterococcus spp. (14.8%), S. aureus (11.8%), Klebsiella spp. (8.8%), and Pseudomonas spp. (8.0%).²⁵ In Türkiye, the NAMRSS surveillance system, established in 2011, included 105 centers from 59 provinces by 2016, forming a nationwide surveillance network. Although the system continues to operate, no data have been published since 2016. The NAMRSS 2016 report, which included only blood and cerebrospinal fluid samples, indicated that the most frequently isolated bacteria in Türkiye were E. coli (23.8%), Enterococcus spp. (18.9%), Klebsiella spp. (17.6%), S. aureus (15.5%), and Acinetobacter spp. (15.1%).²⁶ In a study conducted in Saudi Arabia, the most frequently isolated Gram-negative bacteria were E. coli, Klebsiella spp., and Pseudomonas spp.²⁷ Similarly, a U.S.-based study focusing exclusively on Gram-negative bacteria found that Pseudomonas aeruginosa, E. coli, and Klebsiella pneumoniae were the three most frequently isolated microorganisms in both ICU and non-ICU settings. These three bacteria accounted for 56.7% of Gram-negative isolates in ICUs and 70% in non-ICU settings.²⁸ Our study also identified similar microorganisms, albeit with variations in their rankings across different periods. This underscores the critical importance of regularly and systematically monitoring surveillance data to track trends and inform infection control strategies.

A study conducted under the ‘International Network for Optimal Resistance Monitoring’ program involving 70 centers from the United States separately evaluated isolates from ICUs and non-ICU settings. In this study, the most frequently isolated bacterium from ICU samples was Pseudomonas spp., while E. coli was most commonly isolated from non-ICU samples. Furthermore, antimicrobial S% were found to be significantly lower in isolates from ICUs compared to those from non-ICU settings.²⁸ In a study conducted in Türkiye, the most frequently isolated bacteria from ICU samples were Acinetobacter spp. and Klebsiella spp.²⁹ Similarly, in our study, E. coli was the most frequently isolated microorganism from outpatient and inpatient ward samples, whereas Acinetobacter spp. and Klebsiella spp. were predominant in ICU samples. The observed differences in bacterial species between ICU and non-ICU isolates are likely due to the varying infection types common to these patient groups. In ICUs, bacterial growth was most frequently observed in blood/catheter and respiratory samples, whereas non-ICU infections were predominantly associated with urine samples.

Consistent with our findings, a large-scale study also reported that isolates from ICUs exhibited lower antimicrobial susceptibility compared to those from other units, such as outpatient clinics and inpatient wards.²⁸ In general, it is well-documented that isolates from ICUs tend to have lower S% than those from outpatient or inpatient ward settings. The higher rates of AMR observed in ICUs are attributed to several factors. These include the more intensive use of antimicrobial agents, prolonged hospital stays, and the increased risk of acquiring hospital-associated infections caused by resistant organisms.^{23, 25} This highlights the critical role of ICUs as hotspots for the development and proliferation of AMR, emphasizing the need for targeted infection control measures in these settings.

In a study conducted by Sader et al.²⁸ in the United States, S% for E. coli isolates from ICUs were reported as follows: TZP 92.1%, CAZ 81.3%, CRO 76.9%, MEM 99.8%, GN 87.6%, and AN 99.4%. For non-ICU E. coli isolates, S% were higher: TZP 97.4%, CAZ 90.2%, CRO 88.1%, MEM 99.8%, GN 90.2%, and AN 99.7%. According to the NAMRSS report, S% for E. coli in Türkiye were as follows: AM 21.5%, AMC 35.4%, TZP 72.3%, CRO 48.9%, CAZ 45.8%, GN 70.7%, AN 91.3%, CIP 45.5%, IMP/MEM 95%, and ETP 91.8%.²⁶ Similarly, the CEASER report for Türkiye reported the following rates: AMC 39%, TZP 78%, CAZ 53%, ETP 91%, IMP/MEM 97%, GN 74%, AN 98%, and CIP 48%.³⁰ The ECDC report indicated that in Türkiye, fluoroquinolone resistance in E. coli exceeded 25%, carbapenem resistance ranged between 1-5%, and third-generation cephalosporin resistance exceeded 50%.³¹ In another study conducted in Türkiye, E. coli isolates from urine samples demonstrated S% exceeding 90% for ETP, MEM and F. For non-urine samples, ETP and MEM also showed S% above 90% 29. In our study, antimicrobial S% were found to be higher than NAMRSS and CEASER reports. However, the S% reported by Sader et al.,²⁸ even for ICU isolates -particularly for beta-lactam antibiotics- were higher than our findings. While our results were consistent with the ECDC report, third-generation cephalosporin resistance in our study was below 50%. In conclusion, ETP, MEM, IMP, and AN were identified as effective empirical treatment options for E. coli-related infections in our study. However, it is important to note the significant decline in S% for ETP, IMP, and AN over time.

According to a study conducted by Sader et al.²⁸, Klebsiella spp. isolates from ICUs exhibited the following S%: TZP 89.1%, CAZ 82.7%, CRO 82.2%, MEM 97.6%, GN 91.5%, and AN 99.2%. For non-ICU Klebsiella spp. isolates, S% were higher: TZP 93.6%, CAZ 88.4%, CRO 87.6%, MEM 98.3%, GN 93.2%, and AN 99.3%. The NAMRSS report indicated that Klebsiella spp. isolates showed S% of 23.2% for AMC, 33.4% for TZP, 31.5% for CRO, 24.7% for CAZ, 50.8% for GN, 70% for AN, 37.3% for CIP, 59.9% for IMP/MEM, and 51.1% for ETP.²⁶ Similarly, the CEASER report for Türkiye: reported S% of 25% for AMC, 40% for TZP, 30% for CAZ, 49% for ETP, 61% for IMP/MEM, 55% for GN, 73% for AN, and 35% for CIP.³⁰ The ECDC report highlighted that in Türkiye, third-generation cephalosporin resistance in Klebsiella spp. exceeded 50%, while carbapenem resistance ranged between 10-25%.³¹ In our study, S% for carbapenems in Klebsiella spp. isolates were similar to those reported by NAMRSS and CEASER. However, higher S% were observed for other antibiotics. Sader et al.’s²⁸ findings demonstrated higher S% overall compared to our results. In our study, carbapenem resistance was found to be between 25% and 50%, which deviates from the range reported in the ECDC report. The high carbapenem resistance observed in Klebsiella spp. isolates is particularly concerning and underscores the critical need for stringent antibiotic stewardship strategies to mitigate resistance and improve treatment outcomes.

A study reported that antimicrobial S% for Klebsiella spp. isolates were lower than E. coli isolates. 27. Our findings align with these results. In our study, significant reductions in S% were observed for Klebsiella spp. isolates against TZP, CXM, CAZ, FEP, ETP, IMP, MEM, AN, CIP, and SXT. From an empirical therapy perspective, AN was identified as the only effective option for Klebsiella spp. infections. Notably, carbapenem S% were approximately 80% in outpatient settings but declined sharply to 30% in ICUs. This highlights the increasing resistance rates in ICUs and underscores the necessity of a more cautious approach when selecting empirical therapies in these settings.

In a study conducted by Sader et al.,²⁸ Acinetobacter spp. isolates from ICUs exhibited S% of 61.0% for TZP, 69.4% for CAZ, 70.1% for MEM, 70.8% for LEV, and 74.3% for GN. For non-ICU Acinetobacter spp. isolates, S% were slightly different: TZP 61.9%, CAZ 62.4%, MEM 72.2%, LEV 69.8%, and GN 79.4%.²⁸ According to the NAMRSS report, Acinetobacter spp. isolates showed S% of 22.7% for GN, 27.6% for AN, 8.8% for CIP, and 7.7% for IMP/MEM.²⁶ Similarly, the CEASER report indicated S% of 10% for IMP/MEM, 20% for GN, 30% for AN, and 9% for CIP in Türkiye.³⁰ The ECDC report highlighted carbapenem resistance exceeding 50% in Acinetobacter spp. isolates in Türkiye.³¹ The findings of our study were consistent with the NAMRSS, CEASER, and ECDC reports. However, the S% reported by Sader et al.²⁸ were notably higher than our results. Among Gram-negative bacteria, Acinetobacter spp. isolates exhibited the lowest antimicrobial S%. In our study, a significant decrease in S% was observed for GN, AN, and SXT over time. Notably, no single antimicrobial agent demonstrated sufficiently high susceptibility to be recommended alone for empirical treatment of Acinetobacter spp. infections. Given the overall low susceptibility profile, particularly in the context of multidrug resistance, the use of combination therapy may be considered a more appropriate empirical treatment strategy until definitive culture and susceptibility results are available.

In a study by Sader et al.²⁸, Pseudomonas spp. isolates from ICUs demonstrated S% of 76.9% for TZP, 81.7% for CAZ, 76.1% for MEM, 69.5% for LEV, and 88.1% for GN. For non-ICU isolates, S% were higher: TZP 83.4%, CAZ 87%, MEM 83.7%, LEV 68.4%, and GN 87.4%. According to the NAMRSS report, Pseudomonas spp. isolates showed S% of 69.9% for TZP, 76.5% for CAZ, 69.5% for FEP, 73.9% for GN, 62.3% for CIP, and 53.9% for IMP/MEM. Similarly: the CEASER report indicated S% of 66% for TZP, 72% for CAZ, 69% for FEP, 62% for IMP/MEM, 79% for GN, and 65% for CIP in Türkiye 30. The ECDC report noted that carbapenem resistance in Pseudomonas spp. isolates in Türkiye ranged between 10-25%.³¹ The findings of our study showed that S% were higher compared to the NAMRSS and CEASER reports and were consistent with the ECDC report. Among antimicrobials, AN was identified as the most effective option for empirical therapy in Pseudomonas spp. infections. While S% for other antimicrobials were below 90%, rates for CAZ, FEP, and CIP were above 80%. A significant decrease in susceptibility to FEP was observed, alongside a notable increase in susceptibility to IMP. However, it is crucial to consider that S% are even lower in ICU settings, which should be carefully accounted for when planning empirical therapy for critically ill patients.

According to the NAMRSS report, S. aureus isolates exhibited S% of 85.5% to CIP and 94% to LZD, with no resistance detected to VA or an unspecified agent, TEC.²⁶ Similarly, the CEASER report from Türkiye also reported no resistance to VA and LZD in S. aureus isolates.³⁰ Our study findings are consistent with these reports, indicating high S% for most antimicrobials in S. aureus isolates. However, a significant reduction in susceptibility was observed for E, CC, and TE. From an empirical therapy perspective, LVX, VA, TEC, LZD, TGC, SXT, and GN are viable options. Notably, susceptibility differences between methicillin-susceptible S. aureus (MSSA) and MRSA isolates are significant. For MSSA isolates, all antimicrobials except E and CC are appropriate for empirical therapy. In contrast, for MRSA isolates, empirical therapy should be limited to VA, TEC, LZD, and TGC. These findings underscore the importance of carefully tailored treatment strategies to ensure the effective management of S. aureus infections.

CoNS isolates were found to have lower S% than S. aureus. Significant reductions in susceptibility were observed for all antimicrobials except LZD and TE. From an empirical therapy perspective, only VA, LZD, and TGC were identified as effective treatment options. Methicillin-susceptible CoNS (isolates demonstrated higher S% than methicillin-resistant CoNS (MRCoNS) isolates. However, SXT, which was a viable empirical therapy option in 2015-2016, showed a significant decline in S% by 2022-2023, rendering it unsuitable for empirical use. In MRCoNS isolates, notable reductions in susceptibility were observed for E, CC, LVX, and TEC. Conversely, a significant increase in susceptibility to TE was recorded. Despite this, only VA, LZD, and TGC were deemed suitable for empirical therapy in MRCoNS infections. These findings highlight the importance of carefully assessing resistance patterns when planning treatment strategies for CoNS infections, ensuring that therapeutic choices are guided by the most current susceptibility data.

According to the NAMRSS report, Enterococcus faecalis isolates demonstrated S% of 94% for AM, 42.8% for GN, and 98.7% for VA. For Enterococcus faecium, S% were 8.4% for AM, 38.3% for GN, 84% for VA, and 99% for LZD 26. Similarly, the CEASER report indicated S% of 95% for AM, 66% for GN, 100% for VA, and 100% for LZD in E. faecalis isolates in Türkiye. For E. faecium isolates, S% were 11% for AM, 45% for GN, 87% for VA, and 100% for LZD. In our study, the findings were consistent with the NAMRSS and CEASER reports, except for VA susceptibility, which was slightly lower. In addition to VA, TEC, LZD, and TGC were identified as effective empirical therapy options. A significant decrease in susceptibility was observed for AM, LZD, and TGC in E. faecalis isolates. Despite this decline, AM, in addition to VA, TEC, LZD, and TGC, remains a viable option for empirical treatment of E. faecalis infections. However, AM susceptibility in E. faecium isolates is substantially lower than in E. faecalis. Furthermore, significant reductions in susceptibility to TEC, TGC, and GN were noted for E. faecium isolates. For E. faecium infections, only VA, TEC, LZD, and TGC were deemed suitable for empirical therapy. These findings underscore the critical importance of carefully evaluating resistance patterns in the management of Enterococcus infections to optimize therapeutic outcomes.

According to the NAMRSS report, the MRSA rate was reported as 23.6%. The CEASER report for Türkiye indicated a slightly higher rate of 31%,³⁰ while the ECDC report stated that MRSA rates in Türkiye range between 25 and 50%.³¹ In our study, MRSA rates were found to be consistent with the NAMRSS report but lower than the rates reported in the CEASER and ECDC reports. The observed increase in MRSA rates was primarily attributed to samples originating from inpatient wards. This highlights the need for enhanced infection control measures specifically targeting ward-based infections to mitigate the spread of MRSA.

According to the CEASER report, the VRE rate in Türkiye was reported as 1%. In contrast, the ECDC report indicated a higher VRE rate, ranging between 10-25%.³¹ In our study, the observed VRE rate was higher than the CEASER report, but lower than the range reported in the ECDC data. This increase in VRE rates was largely attributed to samples collected from outpatient clinics and ICU patients. These findings emphasize the critical importance of strengthening infection control measures, particularly in these settings, to effectively manage and reduce the spread of VRE.

According to ECDC reports, AMR rates are lower in Northern and Western Europe, while significantly higher rates are observed in Eastern and Southern Europe, as well as in Türkiye.³¹ In our study, the resistance rates at our hospital were found to be lower compared to the NAMRSS and CEASER reports, indicating the implementation of an effective antimicrobial stewardship program within our institution. However, the persistence of high resistance rates underscores that AMR cannot be fully mitigated through measures at the level of a single hospital or region alone. Addressing this global challenge requires comprehensive nationwide strategies and enhanced international collaboration and coordination to effectively combat AMR on a broader scale.

When comparing our findings with national and international surveillance data, we observed some differences in S%. These variations may be attributed to several factors, including differences in patient populations, hospital-specific antimicrobial stewardship strategies, local infection control practices, diagnostic methodologies, and sample selection criteria. As this study was conducted in a single tertiary care center, the observed resistance patterns may reflect the unique characteristics of the institution. These findings underscore the importance of cumulative antibiogram data tailored to specific regions or healthcare settings, because they provide more clinically relevant guidance for empirical therapy decisions than generalized national estimates.

In accordance with current WHO recommendations, antimicrobials with a susceptibility rate of ≥90% are considered appropriate for empirical treatment in severe infections.¹⁸ In this study, some antimicrobial agents demonstrated S% below this threshold. These findings suggest that such agents may be less reliable for empirical use in serious infections, such as bloodstream infections or meningitis, and should be used with caution. However, in cases of mild or moderate infections-especially when supported by local antibiogram data or when alternative agents are limited-agents with S% between 80-90% may still be considered, provided that close clinical monitoring and follow-up microbiological testing are ensured. In instances where no suitable alternatives are available, combination therapy or the agent with the highest S% rate may be considered, as reflected in local antimicrobial stewardship strategies. This underscores the importance of continually updating cumulative antibiograms and integrating them into institutional treatment guidelines.

Although empirical therapy suggestions in this study were based on antimicrobial agents with ≥90% S%, it is important to emphasize that such recommendations should not be generalized across all clinical contexts. The choice of empirical treatment must be guided not only by local susceptibility trends, but also by the specific clinical setting, infection site, severity of illness, and patient-related factors such as renal function, immune status, and comorbidities. Moreover, the pharmacokinetic and pharmacodynamic properties and toxicity profiles of antimicrobial agents should be carefully considered before empirical use. Therefore, the findings of this study should be viewed as a microbiological foundation to support, but not replace, individualized clinical decision-making and alignment with institutional treatment protocols.^{18, 32, 33}

One of the limitations of this study is the absence of MIC distribution data. Only categorical susceptibility data (i.e., susceptible or resistant) were presented. In addition, molecular characterization of resistance mechanisms -such as the detection of extended-spectrum β-lactamase, Klebsiella pneumoniae carbapenemase, or oxacillinase-48 type carbapenemase genes- was not routinely performed in our laboratory and was, therefore, not available for analysis in this study. Furthermore, interpretive breakpoints defined by EUCAST were not uniform across the two study periods. Specifically, isolates from 2015-2016 were interpreted using EUCAST versions 5.0 to 6.0, while data from 2022-2023 were evaluated based on versions 12.0 to 13.0. Updates in clinical breakpoints for certain antimicrobial agents -particularly aminoglycosides and fluoroquinolones- may have led to reclassification of isolates with MIC values near the susceptibility thresholds. Although overall trends remained consistent, this methodological variation should be taken into account when comparing the two time periods.

In conclusion, the cumulative antibiogram data from our region demonstrate higher S% compared to those of the national average, in Türkiye. However, the growing resistance problem, particularly among Gram-negative bacteria, is a significant concern. This emphasizes the critical importance of accurately selecting empirical treatment options to effectively manage infections.

The recommended antimicrobials for empirical therapy are as follows:

For E. coli infections: carbapenems, and AN.

For Klebsiella spp. and Pseudomonas spp. infections: AN.

For Acinetobacter spp. infections: combination therapy.

For S. aureus infections: LVX, VA, TEC, LZD, TGC, SXT, and GN.

For CoNS infections: VA, LZD, and TGC.

For Enterococcus spp. infections: VA, TEC, LZD, and TGC.

These findings highlight the necessity of regularly assessing regional antibiogram data and updating empirical therapy protocols based on these data. Such an approach is essential for improving the effectiveness of infection management and combating AMR.