Microbial isolates and resistance profiles in cerebrospinal fluid cultures: a five-year experience at a tertiary center

ABSTRACT

Purpose

Central nervous system infections are a significant global health concern that results in morbidity and mortality. In case of these infections, empirical antibiotic treatment should be initiated according to the microbiological findings of the cerebrospinal fluid (CSF), and the treatment should be rearranged according to the antibiogram results.The study aimed to determine the distribution of meningitis agents and antimicrobial resistance patterns in our hospital.

Methods

A total of 17,594 CSF culture samples collected from January 2020 to December 2024 were analyzed retrospectively. The study assessed demographic data, causative microorganisms, and antimicrobial resistance rates.

Results

Positive results were found in 7.3% of CSF culture samples, with 69.1% being gram-positive bacteria and 30.9% gram-negative bacteria. The highest positivity rate was detected in the palliative inpatient clinic (14.8%) despite most cultures being from the neonatal intensive care unit (23.5%). The most common microorganisms were coagulase-negative staphylococci(CoNS) (47.9%) and Klebsiella pneumoniae (K.pneumoniae) (9.1%). Methicillin resistance was present in 89.7% of CoNS, and the rate of extended-spectrum β-lactamase-producing strains in K.pneumoniae was 32.4%.

Conclusion

Analysis of five-year data revealed the types of microorganisms and their antimicrobial resistance in positive CSF cultures. These findings will guide clinicians in promptly choosing the most effective empirical treatment.

Plain Language Summary

Infections of the brain and spinal cord can be very serious and even life-threatening. To treat these infections effectively, doctors often begin treatment with broad antibiotics and then adjust the medications once test results are available. One of the most important tests involves analyzing the clear fluid that surrounds the brain and spinal cord, called cerebrospinal fluid (CSF).In this study, we looked at 17,594 CSF samples collected at our hospital between 2020 and 2024. We wanted to find out which bacteria were most commonly responsible for these infections and how resistant they were to antibiotics. Only 7.3% of the samples showed bacterial growth, but these results provided important insights.We found that most infections were caused by gram-positive bacteria, a group of bacteria that are identified by a special laboratory stain, and the most common type was coagulase-negative staphylococci (CoNS), which are bacteria often linked to infections in people with medical devices. Among gram-negative bacteria, the most common was Klebsiella pneumoniae, which is known for being resistant to many antibiotics.We also observed that infections were more common in certain hospital areas, especially among patients with serious or long-term illnesses. By understanding which bacteria are most likely to cause infection and how resistant they are to treatment, doctors can make better decisions when choosing antibiotics, especially before lab results are ready.

1. Introduction

Central nervous system (CNS) infections continue to represent a major global public health challenge, particularly in low- and middle-income countries. Contributing factors include socioeconomic inequalities, underdeveloped healthcare infrastructure, and limited access to vaccinations and early diagnostic tools, all of which exacerbate incidence rates and worsen clinical outcomes [1]. They are caused by various bacteria, viruses, fungi, and parasites and can lead to serious complications such as coma, seizures, hydrocephalus, and death. In addition to the immediate health risks, CNS infections can also result in long-term neurological issues, including motor problems, cognitive difficulties, behavioral changes, and sensory impairments, impacting both the individual’s quality of life and societal productivity [2,3].

Bacteria continue to be a major cause of CNS infections, with acute bacterial meningitis being a particularly severe and rapidly progressing form [4]. This condition occurs when bacteria invade the leptomeninges and subarachnoid space, leading to purulent inflammation and acute neurological decline. Symptoms of bacterial meningitis typically include sudden fever, headache, neck stiffness, sensitivity to light, vomiting, and changes in mental status. These symptoms are often accompanied by specific abnormalities in cerebrospinal fluid analysis, such as increased white blood cells, elevated protein levels, and decreased glucose levels [5]. Despite improvements in diagnosis and treatment, the global mortality rate for bacterial meningitis remains high, ranging from 10% to 30%, with the highest rates seen in neonates, immunocompromised individuals, and those in resource-limited areas [6].

Globally, bacterial meningitis affects around 1.2 million people each year, with varying incidence rates in different geographical and socioeconomic settings [7]. Sub-Saharan Africa, particularly the “meningitis belt,” continues to face frequent outbreaks, making it a hotspot for the disease [8,9]. In contrast, high-income countries like the United States have much lower incidence rates due to effective immunization programs targeting key pathogens [10]. However, survivors still experience significant long-term morbidity, with 30% to 50% suffering permanent sequelae such as sensorineural hearing loss, cognitive deficits, hydrocephalus, epilepsy, and motor impairments [5,11].

From an etiological perspective, acute bacterial meningitis can be broadly divided into two categories: community-acquired and healthcare-associated. Community-acquired meningitis is mainly caused by pathogens such as Streptococcus pneumoniae (S. pneumoniae), Neisseria meningitidis (N. meningitidis), and Haemophilus influenzae (H. influenzae) [12]. In specific patient groups – specifically neonates, older adults over 50, and immunocompromised individuals- Listeria monocytogenes (L. monocytogenes) is also a significant causative agent [13]. On the other hand, healthcare-associated meningitis typically occurs within 48 hours of hospitalization or within one year after the placement of a neurosurgical device [14,15]. These cases often follow invasive procedures such as craniotomy, lumbar puncture (LP), intrathecal drug administration, or the insertion of external or internal ventricular drainage systems [16,17]. The main pathogens in healthcare-associated meningitis include Staphylococcus species and various gram-negative bacilli [16,18].

Despite the increasing availability of molecular and antigen-based diagnostic methods, cerebrospinal fluid (CSF) culture remains the gold standard for definitively diagnosing bacterial meningitis. It not only confirms the presence of the pathogen but also allows for antibiotic susceptibility testing, which is crucial for guiding targeted antimicrobial treatment. However, various pre-analytical and procedural factors can impact culture sensitivity, such as prior antibiotic use before lumbar puncture, timing of sample collection, inadequate sample volume, and technical variations in laboratory processing [19,20].

Before the widespread use of vaccines, S. pneumoniae, N. meningitidis, and H. influenzae were commonly identified as the main causes of bacterial meningitis worldwide [21]. The introduction of vaccines targeting these bacteria has significantly changed the epidemiology of the disease, leading to a decrease in cases of vaccine-preventable meningitis [22]. However, this shift has also brought about the emergence of new pathogens and an alarming increase in antimicrobial resistance among CNS isolates [23]. Multidrug-resistant (MDR) organisms, such as extended-spectrum beta-lactamase (ESBL)-producing Enterobacterales and methicillin-resistant Staphylococci, have become increasingly prevalent, complicating therapeutic approaches and posing a serious threat to patient outcomes.

In CNS infections, delayed or incorrect empirical antibiotic therapy is closely linked to negative clinical outcomes. Therefore, it is crucial to quickly identify the causative agent and be aware of local antimicrobial resistance patterns to choose appropriate initial treatments. This is especially important in healthcare-associated infections, where pathogens may have intricate resistance mechanisms, requiring tailored empirical regimens based on local susceptibility data for optimal effectiveness [23,24]. Reliable local surveillance data is essential for informing clinical decisions, enhancing treatment protocols, and revising hospital antibiotic stewardship guidelines.

The aim of this study was to analyze the demographic and microbiological profiles of patients diagnosed with bacterial meningitis at Ankara Bilkent City Hospital, a leading tertiary referral center in Turkey, between January 2020 and December 2024. By investigating the results of CSF cultures and the corresponding antimicrobial resistance patterns, this research seeks to contribute valuable regional data to the growing body of knowledge on bacterial meningitis and support efforts to improve infection control measures and antimicrobial stewardship practices.

2. Materials and methods

2.1. Study design and period

This research was conducted as a retrospective and observational study at Ankara Bilkent City Hospital, a high-capacity tertiary care institution situated in the capital city of Turkey. Spanning a five-year interval from 1 January 2020, to 31 December 2024, the study encompassed 17,594 cerebrospinal fluid (CSF) culture samples obtained from 9,561 patients who were hospitalized with clinical suspicion of CNS infections.

Ankara Bilkent City Hospital, equipped with 4,050 beds, functions as a major referral center that provides comprehensive medical services to a heterogeneous population across both metropolitan and rural settings. The institution integrates a wide array of specialized departments and intensive care units (ICUs) and plays a pivotal role in the regional diagnosis and management of challenging infectious diseases, particularly those involving the CNS.

2.2. Inclusion and exclusion criteria

Patients of all age groups and sexes who underwent CSF sampling with an initial clinical impression of bacterial meningitis between January 2020 and December 2024 were considered eligible. CSF specimens collected either via lumbar puncture (LP) or external ventricular drain (EVD) and submitted in sterile tubes or blood culture bottles were included.

Patients were excluded from the study if their medical records were missing demographic data or if CSF culture and antimicrobial susceptibility results were incomplete. Additionally, in order to distinguish true infections from contamination, only CSF samples with both clinical and laboratory findings indicative of acute bacterial meningitis were deemed valid. Specifically, CSF samples that showed bacterial growth in the absence of microorganisms on Gram stain, had a leukocyte count below 30/mm³, and were collected from patients without clinical symptoms of meningitis were classified as “contaminated” and were not included in the analysis.

In the specific context of coagulase-negative staphylococci (CoNS), cultures were classified as “causative” if the patient had a history of neurosurgical intervention or the presence of a foreign device such as an EVD catheter. Conversely, CoNS isolates were considered “contaminants” when derived from CSF collected via LP in patients undergoing evaluation for noninfectious neurological conditions (e.g., multiple sclerosis, autoimmune encephalitis), in the absence of bacteria on Gram stain and without clinical evidence of meningitis. These cases were also excluded.

To prevent redundancy and excessive representation, only one isolate per patient was included within a seven-day period. However, CSF samples collected from the same patient more than seven days apart were treated as separate and included in the analysis if they indicated a separate infection episode.

2.3. Data collection

All relevant microbiological and clinical data were obtained from the hospital’s electronic information management system and analyzed and standardized. Extracted variables included the date of CSF sample collection, patient age and sex, isolated microorganisms’ identity, and antimicrobial susceptibility testing results. Data accuracy and completeness were ensured through verification with original laboratory records and structured consolidation using Microsoft Excel (version 2023, USA).

2.4. Study population

For age-based subgroup analysis, the patient cohort was stratified into three categories: newborns (≤29 days), pediatric patients (1 month to 17 years), and adults (≥18 years). This categorization helped in conducting age-specific evaluations of pathogen distribution and resistance profiles, consistent with epidemiological trends documented in the literature.

2.5. CSF collection and analysis

CSF samples were collected using aseptic techniques through either LP or EVD insertion and were immediately transported to the microbiology laboratory in sterile tubes or blood culture bottles. Specimens submitted in blood culture bottles were incubated using the BacT/ALERT® 3D (bioMérieux, France) automated microbial detection system. Samples that triggered positive growth signals, as well as those submitted in sterile containers, were inoculated onto standard culture media, including 5% sheep blood agar, chocolate agar, and Eosin Methylene Blue (EMB) agar. These media were incubated at 37°C 5% CO₂ for a period ranging from 48 to 72 hours. The identification of bacterial species was performed using the VITEK2 MS system (bioMérieux, France), an automated matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry platform.

2.6. Antibiotic susceptibility test

The automated VITEK system (bioMérieux, France) was used to determine minimum inhibitory concentrations (MICs) and susceptibility categories for isolated bacterial strains in antimicrobial susceptibility testing. This system also aids in identifying multidrug-resistant phenotypes, such as extended-spectrum beta-lactamase (ESBL)-producing Enterobacterales, methicillin-resistant Staphylococcus aureus (MRSA), and methicillin-resistant coagulase-negative staphylococci (MRCoNS). Colistin resistance was evaluated using the broth microdilution method. Results were interpreted based on the clinical breakpoints and methodological standards set forth annually by the European Committee on Antimicrobial Susceptibility Testing (EUCAST).

2.7. Statistical analysis

Data analysis was performed using SPSS Statistics (version 22.0). Descriptive statistics were used to summarize numerical and categorical variables. Continuous data were expressed as medians with minimum and maximum values, whereas categorical variables were evaluated using Pearson’s Chi-square or Fisher’s exact tests, where appropriate. A p-value of < 0.05 was considered statistically significant.

2.8. Ethical approval

This study was approved by the Ankara Bilkent City Hospital No. 1 Medical Research Scientific and Ethical Evaluation Board (TABED), with the protocol assigned decision number TABED 1/933/2025, dated 29 January 2025. All study procedures adhered to the principles outlined in the Declaration of Helsinki and were consistent with applicable national research ethics regulations.

3. Results

During the five-year study period from 1 January 2020, to 31 December 2024, a total of 17,594 CSF culture samples obtained from 9,561 individual patients were retrospectively analyzed at Ankara Bilkent City Hospital, a high-volume tertiary care institution. The patient population represented a wide age range: 51.7% of the patients were adults aged 18 years or older, 31.4% were pediatric patients aged between 1 month and 17 years, and 16.8% were neonates (≤28 days of life), reflecting the inclusion of a wide clinical spectrum of CNS infection cases (Table 1). Of the total samples collected, 15.1% were derived via external ventricular drain (EVD) procedures. This indicates that many cases involve neurosurgical intervention and potentially complex clinical scenarios.

Table 1.

Number of CSF culture samples and demographic characteristics of patients by year.

Year			2020	2021	2022	2023	2024	Total
Total number of CSF samples			2508	3518	3849	3995	3724	17.594
Patients	Gender	Female	609	869	1024	1276	1092	4870 (50.9%)
		Male	718	920	1015	1092	946	4691 (49.1%)
	Age groups	Neonates	298	364	437	336	171	1606 (16.8%)
		Children	522	672	608	655	550	3007 (31.4%)
		Adults	507	753	994	1377	1317	4948 (51.7%)
	Average age (Min-Max)		20,3 (1 day-89 years)	22,2 (5 days-102 years)	23,8 (9 days-93 years)	29,8 (6 days-94 years)	33,3 (8 days-100 years)	26,5 (6 days-102 years)
	Total		1327	1789	2039	2368	2038	9561
Number of Positive CSF Samples			284	327	285	193	191	1280
Patients	Gender	Female	53	74	73	50	47	297 (43.4%)
		Male	75	96	96	66	54	387 (56.6%)
	Age groups	Neonates	33	26	23	17	14	113 (16.5%)
		Children	50	61	65	35	20	231 (33.8%)
		Adults	45	83	81	64	67	340 (49.7%)
	Average age (Min-Max)		20,1 (9 days-86 years)	27,2 (11 days-89 years)	27,6 (1 month-91 years)	34,9 (25 days-93 years)	34,0 (20 days-77 years)	28,3 (9 days-93 years)
	Total		128	170	169	116	101	684

Neonates: 0–29 days, Children: 1 month-17 years, Adults: ≥ 18 years.

With respect to clinical service distribution, 50.5% of the CSF samples were collected from different intensive care units (ICUs), whereas 49.5% originated from inpatient clinics (ICs). It was observed that suspected CNS infections were frequently identified in both critically ill patients and the general patient population.

Microbiological evaluation revealed a total culture positivity rate of 7.3%, with 1,280 samples yielding identifiable bacterial growth. These culture-positive isolates belonged to 684 individual patients, of whom 56.6% were male and 43.4% were female. In the subset of patients with confirmed bacterial growth, 49.7% were adults, 33.8% were children, and 16.5% were neonates. The mean age of this culture-positive group was 28.3 years, with an age range extending from 9 days to 93 years (Table 1). This indicates a comprehensive epidemiological overview of CNS infection throughout all stages of life.

When annual trends were examined, the highest CSF culture positivity rate was observed in 2020 (11.3%), while the lowest was noted in 2023 (4.8%) (Table 2). This temporal variation may be indicative of changes in diagnostic approaches, infection control measures, or shifts in the patient demographic over time. The majority of CSF samples were submitted from the neonatal ICU (23.5%), neurosurgical ICU (15.6%), neurology inpatient clinic (14.8%), and pediatric inpatient clinic (12.8%). The highest positivity rates were recorded in the palliative inpatient clinic (14.8%), general ICU (12.4%), and pediatric ICU (10.4%) (Table 2). This could indicate a higher vulnerability to CNS infections in patients with severe underlying comorbidities or extended hospital stays.

Table 2.

Descriptive analysis for CSF culture-positive samples based on years 2020–2024.

			2020 (n = 284, 11.3%)	2021 (n = 327, 9.3%)	2022 (n = 285, 7.4%)	2023 (n = 193, 4.8%)	2024 (n = 191, 5.1%)	Total (n = 1280, 7.3%)
			n (%)	n (%)	n (%)	n (%)	n (%)	n (%)
CSF sample collection sites	EVD		110	105	92	48	49	404 (31.6%)
	LP		174	222	193	145	142	876 (68.4%)
Clinics	Neonatal ICU (n = 4139)		87	68	72	42	29	298 (7.2%)
	Neurosurgery ICU (n = 2744)		20	56	48	40	75	239 (8.7%)
	Neurology IC (n = 2610)		2	10	17	11	6	46 (1.8%)
	General Child IC (n = 2249)		57	75	48	13	16	209 (9.3%)
	Neurosurgery IC (n = 1139)		12	18	21	17	7	75 (6.6%)
	Pediatric ICU (n = 958)		29	23	16	19	13	100 (10.4%)
	Adult Emergency IC (n = 957)		8	19	9	18	12	66 (6.9%)
	Adult General ICU (n = 679)		18	19	17	13	17	84 (12.4%)
	Pediatric Emergency IC (n = 431)		10	4	3	2	3	22 (5.1%)
	Pediatric Infection IC (n = 388)		6	9	7	2	1	25 (6.4%)
	Adult Infection IC (n = 356)		5	4	11	4	1	25 (7%)
	Neurology ICU (n = 333)		6	1	3	4	4	18 (5.4%)
	Pediatric Hemato/Onco IC (n = 321)		3	15	8	4	0	30 (9.3%)
	Palliative IC (n = 290)		21	6	5	4	7	43 (14.8%)
Organism	Gram-positive	CNS	126	149	146	103	89	613 (47.9%)
		E.faecium	22	21	6	10	7	66 (5.2%)
		S.aureus	10	13	14	13	7	57 (4.5%)
		E.faecalis	23	13	8	2	1	47 (3.7%)
	Gram-negative	K.pneumoniae	29	24	24	20	19	116 (9.1%)
		A.baumannii	17	48	16	11	23	115 (9%)
		P.aeruginosa	11	22	29	16	11	89 (7%)
		E.coli	19	11	20	2	10	62 (4.8%)
	Others		27	26	22	16	24	115 (9%)

CSF: Cerebrospinal fluid, EVD: External Ventricular Drain, LP: Lumbar puncture, ICU: Intensive Care Unit, IC: Inpatient Clinic, CNS: Coagulase-negative Staphylococci, E.faecium: Enterococcus faecium, S.aureus: Staphylococcus aureus, E.faecalis: Enterococcus faecalis, K.pneumoniae: Klebsiella pneumoniae, A.baumannii: Acinetobacter baumannii, P.aeruginosa: Pseudomonas aeruginosa.

Of the culture-positive CSF samples, 31.6% were obtained via EVD. Statistical analysis confirmed a significantly higher rate of bacterial growth in EVD-derived specimens compared to those collected through LP (p < 0.001). This underscores the increased infection risk linked to invasive neurosurgical procedures and highlights the crucial role of thorough aseptic techniques and infection prevention protocols in managing EVD. Furthermore, the frequency of EVD use was significantly higher in the pediatric population compared to patients who did not undergo EVD placement (p < 0.001). This suggests either a higher clinical need or diagnostic preference for EVD in this age group.

Bacteriological classification of the isolates revealed a predominance of gram-positive organisms, which accounted for 69.1% of all positive cultures. Among these, coagulase-negative staphylococci (CoNS) were the most frequently isolated species, representing 47.9% of all bacterial isolates. Other significant gram-positive pathogens included Enterococcus faecium (E. faecium) (5.2%) and Staphylococcus aureus (S. aureus) (4.5%). In contrast, gram-negative organisms comprised 30.9% of the isolates, with Klebsiella pneumoniae (K. pneumoniae) (9.1%), Acinetobacter baumannii (A. baumannii) (9.0%), and Pseudomonas aeruginosa (P. aeruginosa) (7.0%) being the most prevalent among this group (Table 2). Notably, no fungal organisms were isolated during the entire study period.

A comparative evaluation of bacterial distribution based on sampling technique revealed that E. faecium, K. pneumoniae, and P. aeruginosa were significantly more frequently isolated from EVD-associated samples compared to those obtained via LP (p < 0.001, p = 0.049, and p = 0.001, respectively) (Table 3). This association highlights the increased risk of nosocomial CNS infections linked to the presence of foreign devices and supports the need for heightened clinical vigilance in such cases.

Table 3.

Types of microorganisms detected based on CSF sample collection sites.

Organism		CSF sample collection sites		p-value
		EVD (n = 404, 31.6%)	LP (n = 876, 68.4%)
Gram-positive bacteria	CNS	179	434	0.08
	E.faecium	35	31	<0.001
	S.aureus	22	35	0.24
	E.faecalis	11	36	0.22
Gram-negative bacteria	K.pneumoniae	46	70	0.049
	A.baumannii	37	78	0.88
	P.aeruginosa	42	47	0.001
	E.coli	22	40	0.49
Others		10	105	<0.001

EVD: External Ventricular Drain, LP: Lumbar puncture, CNS: Coagulase-negative Staphylococci, E.faecium: Enterococcus faecium, S.aureus: Staphylococcus aureus, E.faecalis: Enterococcus faecalis, K.pneumoniae: Klebsiella pneumoniae, A.baumannii: Acinetobacter baumannii, P.aeruginosa: Pseudomonas aeruginosa.

The antimicrobial susceptibility profiles of the major bacterial isolates were thoroughly analyzed across the five-year period. It has been determined that methicillin-resistant CoNS (MRCoNS) are quite common, with methicillin resistance rates ranging from 82.3% to 96.3% and a five-year average of 89.7%. This indicates that the resistant phenotype is consistently endemic in the population (Figure 1). Among E. faecium isolates, vancomycin resistance was observed in 37.3% of strains, with notable year-to-year fluctuations. Resistance to other agents, such as teicoplanin and levofloxacin, also reached concerning levels, averaging near 60% (Supplementary file). For S. aureus, methicillin resistance was found in 61.5% to 84.6% of isolates, with a five-year average of 71.9%. Nevertheless, there have been no instances of resistance to vancomycin, teicoplanin, or linezolid, suggesting that these drugs remain effective treatment choices. Inducible clindamycin resistance was present in 35.1% of S. aureus isolates, further complicating therapeutic decisions in empirical settings (Figure 2).

Figure 1.

Antimicrobial resistance rates of CoNS isolated from CSF, 2020–2024.

Figure 2.

Antimicrobial resistance rates of S. aureus isolated from CSF, 2020–2024.

Among gram-negative bacteria, K. pneumoniae exhibited alarmingly high resistance rates to third- and fourth-generation cephalosporins (e.g., ceftriaxone, ceftazidime, cefepime) and to β-lactam/β-lactamase inhibitor combinations (e.g., piperacillin-tazobactam), with resistance nearing 90%. Resistance to ertapenem demonstrated an upward trend, reaching nearly 80% by 2024. In addition, the rate of extended-spectrum β-lactamase (ESBL)-producing K. pneumoniae was detected at an average of 32.4%, indicating that management of multidrug-resistant strains is becoming increasingly difficult (Figure 3). A. baumannii isolates demonstrated extensive drug resistance, with resistance rates exceeding 90% for imipenem, amikacin, ciprofloxacin, and piperacillin/tazobactam. The mean colistin resistance rate among these isolates was 8.6%. This reflected the possibility of over-reliance on last resort agents in the setting of multidrug resistance (Figure 4). Lastly, P. aeruginosa showed the highest resistance to piperacillin-tazobactam (75%), followed by amikacin (50%). Interestingly, although the overall colistin resistance rate among P. aeruginosa isolates was 6.7%, no resistance was detected in samples collected during 2024 (Figure 5). This may show a positive response to institutional antimicrobial stewardship efforts and infection control strategies implemented in recent years.

Figure 3.

Antimicrobial resistance rates of K. pneumoniae isolated from CSF, 2020–2024.

Figure 4.

Antimicrobial resistance rates of A. baumannii isolated from CSF, 2020–2024.

Figure 5.

Antimicrobial resistance rates of P. aeruginosa isolated from CSF, 2020–2024.

4. Discussion

In this five-year retrospective analysis conducted at a tertiary care center, a total of 17,594 cerebrospinal fluid (CSF) samples obtained from 9,561 patients were evaluated. The overall culture positivity rate was 7.3%, with a disproportionately higher prevalence in male patients, pediatric populations, and specimens collected via EVD systems. The predominant pathogens isolated from CSF cultures included CoNS (47.9%), K. pneumoniae (9.1%), and A. baumannii (9.0%). A statistically significant association was observed between EVD use and increased isolation of E. faecium, K. pneumoniae, and P. aeruginosa (p < 0.05), underscoring the infection risk associated with indwelling neurosurgical devices. Alarmingly, antimicrobial susceptibility testing revealed markedly high resistance rates among key pathogens: methicillin resistance in CoNS averaged 89.7%, vancomycin resistance in E. faecium was 37.3%, and MRSA prevalence reached 71.9%. Among gram-negative isolates, resistance to extended-spectrum β-lactams, fluoroquinolones, and carbapenems approached or exceeded 90%, while colistin resistance, although variable, was present in both A. baumannii and P. aeruginosa. These results have shown important shifts in the type of microorganisms and their resistance patterns in CNS infections, which have significant implications for empirical treatment protocols and infection control practices, particularly in intensive care settings.

The epidemiological profile of bacterial meningitis has changed with regional health care inequalities, demographic variables such as age distribution, and the implementation and coverage of national immunization programs. More recently, the unprecedented global disruptions caused by the COVID-19 pandemic have also contributed to this change [25]. Furthermore, the fluctuations in CSF culture positivity rates documented across the five-year study period do not solely reflect changes in disease prevalence. They may also reflect improvements in hospital laboratory infrastructure, changes in diagnostic protocols, and improved clinician awareness of early recognition and investigation of suspected meningitis cases. Given the high morbidity and mortality rates associated with bacterial meningitis, particularly in vulnerable populations such as neonates, the elderly, and immunocompromised individuals, timely diagnosis and prompt initiation of appropriate antimicrobial therapy are of paramount importance [23,26]. Clinical delays or misdiagnosis can result in rapid neurological deterioration, long-term sequelae, or fatal outcomes. In this context, it is essential to have ongoing surveillance, and region-specific data are critical for optimizing empirical treatment strategies, guiding vaccination policies, and informing public health interventions.

When compared to international studies, our culture positivity rate of 7.3% is notably higher than that reported in studies from China (0.5%) [27] and slightly above that observed in another Turkish study (5.8%) [28], yet lower than the 26% reported in African cohorts [29]. The variations between studies are probably due to multiple factors and could be attributed to differences in how clinical samples were collected, variations in the capabilities of laboratory diagnostics, and the overall quality and availability of microbiological services in different regions. Additionally, differences in patient selection, such as the inclusion of only high-risk or severely ill individuals compared to a more general population, could impact the rates of positivity observed.

Moreover, our study revealed that male patients exhibited a higher rate of culture-positive CSF samples compared to female patients, a finding that is consistent with previous observations reported by Zhang et al. [27] in a large-scale retrospective cohort from China. Sex-based immunological differences, including hormonal modulation of inflammatory responses, have been proposed to underlie these discrepancies [30], though the precise immunological pathways underlying sex-related differences in the pathogenesis and clinical progression of bacterial meningitis remain incompletely understood. Therefore, further mechanistic studies incorporating molecular immunology and hormone profiling are warranted to elucidate the exact roles of sex hormones and genetic factors in shaping infection risk and host response in CNS infections.

Over the course of the five-year study period, a notable degree of interannual variability was observed in the incidence of culture-confirmed bacterial meningitis, as reflected by yearly fluctuations in CSF culture positivity rates. The highest positivity rate, 11.3%, was recorded in 2020, coinciding with the onset and peak of the global COVID-19 pandemic. This temporal alignment suggests that the pandemic may have exerted indirect effects on the epidemiology of bacterial meningitis. Potential contributing factors include changes in increased hospital admissions due to overlapping symptomatology with SARS-CoV-2 infection, modifications in diagnostic vigilance, and the implementation of stringent infection control measures such as universal masking, physical distancing, and enhanced personal protective equipment protocols. These public health interventions may have altered both the transmission dynamics and clinical thresholds for CSF sampling, thereby influencing the observed increase in culture positivity. Nonetheless, additional multicenter studies are necessary to investigate this connection more thoroughly due to the complex and likely multifactorial relationship between the COVID-19 pandemic and changes in bacterial meningitis incidence.

Approximately half of the confirmed meningitis cases in our study were observed in the pediatric population, with a disproportionate concentration among neonates and infants. This distribution is consistent with previously published studies that underscore the heightened susceptibility of this age group to invasive central nervous system infections [31,32]. The vulnerability of neonates and young infants can be attributed primarily to the immaturity of both innate and adaptive immune responses during the early stages of life. Moreover, the blood-brain barrier in neonates is not fully developed, potentially allowing pathogens to enter the cerebrospinal fluid through the bloodstream. Additional contributing factors include perinatal exposures, such as vertical transmission during childbirth, prematurity, low birth weight, and prolonged hospital stays in the neonatal ICU, where invasive procedures and device use further increase the risk of nosocomial infections. Together, these biological and clinical risk factors explain the increased incidence and severity of bacterial meningitis in this particularly fragile patient population. Thus, it emphasizes the critical importance of early diagnosis, empiric antimicrobial coverage, and close clinical monitoring.

While the majority of CSF culture requests in our study originated from critical care settings, including the neonatal ICU (23.5%), the neurosurgical ICU (15.6%), neurology inpatient clinic (14.8%), and the pediatric inpatient clinic (12.8%); the highest culture positivity rates were identified in the palliative inpatient clinic (14.8%), the general adult ICU (12.4%), and the pediatric ICU (10.4%). This difference in sample submission frequency and culture positivity rate highlights the significance of considering both diagnostic access and clinical context when interpreting CSF culture data. The elevated positivity rates in palliative and general ICUs may be multifactorial in origin. Patients in these units often present with advanced age, chronic debilitating illnesses, and a high degree of functional decline, all of which contribute to immunosenescence, an age-related deterioration in immune function that impairs host defenses against bacterial invasion [33,34]. Furthermore, such patients frequently require prolonged hospitalization, mechanical ventilation, urinary and central venous catheterization, and broad-spectrum antimicrobial exposure. Each of these is a well-known risk factor for healthcare-associated infections, including meningitis.

The incidence of microorganism growth was found to be significantly higher in CSF samples obtained via EVD placement compared to those collected through LP, with the difference reaching statistical significance (p < 0.001). This observation is in alignment with existing literature that consistently underscores the elevated risk of CNS infections associated with the presence of indwelling neurosurgical devices [35–37]. EVD systems, while essential for monitoring intracranial pressure and managing hydrocephalus or intracranial hemorrhage, create a direct communication pathway between the external environment and the sterile subarachnoid or ventricular spaces. As such, they serve as a potential conduit for microbial entry, particularly in settings with suboptimal aseptic technique or prolonged device use.

Studies have shown that the infection rate linked to EVD-associated meningitis ranges from 0% to 22%, depending on factors such as the patient population, duration of catheterization, institutional infection control practices, and surveillance definitions used [37]. This wide variability reflects the multifactorial nature of device-associated infections and highlights the challenges in establishing standardized diagnostic thresholds. Notably, EVD-related infections have been linked to longer hospital stays, increased ICU utilization, greater healthcare costs, and significantly worse clinical outcomes, including higher rates of neurological sequelae and mortality [37,38].

CoNS was the most frequently isolated gram-positive bacterium (47.9%), in line with national and international studies [27,39,40]. The predominance of CoNS, especially noted during the years 2021 and 2022 in our dataset, may be attributed to a relative increase in the use of external ventricular drains (EVDs) during that period. Although the comparison between EVD- and LP-derived CSF samples did not yield a statistically significant difference in CoNS isolation rates (p = 0.08), the trend toward increased prevalence in EVD-associated specimens remains clinically relevant. CoNS are well-known constituents of normal skin flora and are commonly implicated as contaminants in blood and CSF cultures. However, they are also recognized as opportunistic pathogens capable of causing serious infections in the presence of foreign materials, such as ventricular catheters or shunts. This dual role complicates the interpretation of positive culture results and often necessitates careful clinical correlation. Distinguishing true CoNS-related CNS infection from contamination is critically important but inherently challenging. Diagnostic accuracy requires integration of multiple parameters, including the patient’s clinical presentation (e.g., fever, altered mental status), CSF biochemical findings (e.g., pleocytosis, hypoglycorrhachia), Gram stain results, the presence of a compatible radiologic or neurosurgical context, and microbiological factors such as the number of positive cultures and time to positivity [41]. Misclassification may lead either to unnecessary antibiotic exposure or, conversely, to under-treatment of a potentially serious infection. Hence, it is crucial to have standardized diagnostic algorithms and interdisciplinary evaluation that includes microbiologists and treating clinicians to ensure accurate case interpretation and optimal patient management.

Among gram-negative bacteria, K. pneumoniae (9.1%) and A. baumannii (9%) were the most commonly isolated pathogens. Their prominence reflects nosocomial origins, particularly in patients subjected to long ICU stays, broad-spectrum antibiotics, and invasive neurosurgical interventions [42]. P. aeruginosa was detected in 7% of cases, contrasting with higher rates reported in Saudi Arabia (43%) [42] and lower rates in China (1.4%) [27], indicating strong geographical variation in pathogen distribution.

In studies, the most commonly identified microorganisms in community-acquired bacterial meningitis are S. pneumoniae, N. meningitidis, and H. influenzae [31,43,44]. However, our results indicate a shift in the predominant pathogens of meningitis, with these microorganisms being detected at very low rates of 1.1%, 0.2%, and 0.1%, respectively. When investigating whether this change in etiology is linked to the COVID-19 pandemic, it is revealed that S. pneumoniae, N. meningitidis, and H. influenzae were still the most frequently isolated bacteria from CSF in studies conducted in Iran [45] and Africa [29] prior to the pandemic. In studies covering the pandemic years, a study in China from 2012 to 2020 reported S. pneumoniae at a rate of 13.7%, N. meningitidis at 0.6%, and no detection of H. influenzae [4]. In the study carried out in Turkey between 2017 and 2022, S. pneumoniae was identified at a rate of 13.2%, N. meningitidis at a rate of 1.8%, and H. influenzae at a rate of 1.8% [46]. Particularly during the years of the COVID-19 pandemic, a noticeable decrease was observed in the detection rates of these pathogens. It is thought that the increased use of protective medical masks and the implementation of control measures such as social distancing not only limited the transmission of SARS-CoV-2 but also decreased infections that can be transmitted through droplets and close contact. However, since our study did not encompass the pre-pandemic years, a definitive conclusion cannot be drawn; further comprehensive research is needed to clarify this issue. Additionally, the variation in detection rates of these microorganisms, which are considered primary pathogens in certain studies, in other studies may also be impacted by regional socioeconomic conditions, age demographics in the study population, and the vaccination status in the respective countries.

Antimicrobial resistance analysis revealed alarming trends. Methicillin resistance in CoNS was 89.7%, aligning with studies from China (80–93.6%) [47,48], but higher than rates in India (57.7%) [49] and Ethiopia (37%) [50]. E. faecium was the fifth most commonly isolated microorganism from CSF samples, with 37.3% of these being VRE. Literature reviews indicate that vancomycin-resistant E. faecium meningitis is rare and typically associated with neurosurgical procedures or devices [51–53]. Enterococci, which are normal skin flora, can enter the body through neurosurgical procedures, traumatized tissue, or colonization of patient devices like EVDs [54]. Our study found a statistically significant increase in E. faecium growth in CSF samples collected via EVD compared to LP samples, supporting this observation (p < 0.001). Factors such as prolonged hospital stays, extensive antibiotic use, and close contact with VRE-infected patients contribute to the spread of VRE infections in healthcare settings. Identifying the reasons for the high prevalence of vancomycin-resistant E. faecium in our hospital and implementing necessary precautions and training are crucial. S. aureus exhibited a 71.9% MRSA rate, substantially exceeding that reported in Saudi Arabia (1.4%) [42] and China (29%) [47]. Risk factors such as extended hospitalization, the presence of urinary or central venous catheters, and the use of multiple antibiotics can increase the risk of developing MRSA meningitis, a type of hospital-acquired infection [55]. Our study highlights the need to control this high percentage to enhance local screening programs.

K. pneumoniae was the most commonly isolated gram-negative bacteria from CSF. 32.4% of K. pneumoniae isolates were identified as ESBL-producing, which is a lower percentage compared to studies conducted in Saudi Arabia and China [42,56]. While this finding is promising, it highlights the importance of maintaining this resistance pattern at current levels. ESBL-producing Klebsiella species present a significant threat in hospital settings due to their virulence, rapid spread, and high antibiotic resistance rates [57]. K. pneumoniae exhibited resistance rates of approximately 90% to piperacillin/tazobactam, cefepime, ceftriaxone, and ceftazidime, which are significantly higher than those reported in studies from China [47] and Pakistan [58]. The average colistin resistance rate in A. baumannii was determined to be 8.6%. In a recent multicenter study conducted in China, this rate was found to be 0% [27]. The high value observed in our study is attributed to our reliance on colistin as the primary treatment option against this microorganism, given the high resistance rates of 90% against imipenem, amikacin, ciprofloxacin, and piperacillin/tazobactam in A. baumannii. P. aeruginosa resistance to piperacillin-tazobactam reached 75%, nearly double that reported by Peng et al. [47]. This significant rise in resistance poses a serious public health concern that requires close monitoring and control. Furthermore, Bathija et al. [58] observed a colistin resistance rate of 22.2% in their study. While our study found an average colistin resistance rate of 6.7%, it is noteworthy that no resistance was determined in 2024.

The study aimed to analyze a five-year dataset from 2020 to 2024 to provide information on changing trends in meningitis epidemiology in Turkey. However, our study has limitations. Firstly, it is a single-center and retrospective study, even though our hospital is one of the largest in Turkey. The accuracy of our findings depends on the data in the hospital information management system. Secondly, certain clinical variables related to bacterial meningitis were not considered, such as patient history, previous culture results, clinical features, length of hospital stay, and molecular characterization of the microorganisms isolated.

Collaborating with clinical departments to carry out multicenter and prospective studies is essential for identifying bacterial meningitis agents and patterns of antimicrobial resistance in our country. Furthermore, molecular analysis of resistant strains and ongoing monitoring of epidemiological data changes will help inform public health strategies for preventive medicine and assist clinicians in managing bacterial meningitis effectively.

5. Conclusion

Our research aimed to contribute to the understanding of the epidemiology and antimicrobial resistance patterns of bacterial meningitis in a tertiary care hospital in Turkey. The results show that bacterial meningitis predominantly affects the pediatric population. Notably, the spectrum of microorganisms isolated from CSF cultures differed from the usual pathogen profiles, underscoring the necessity for country-specific epidemiological surveillance to inform national public health strategies. Furthermore, the frequent identification of antimicrobial-resistant bacteria, particularly ESBL-producing gram-negative bacilli, MRSA, and MRCoNS, highlights the increasing challenge in effectively managing bacterial meningitis and emphasizes the urgent need for updated empirical treatment guidelines and strong antimicrobial stewardship programs.

Funding Statement

This paper was not funded.

Article highlights

• Central nervous system infections are a prevalent health issue that leads to illness and death, particularly in less developed and developing nations.

• Microorganisms were isolated from 7.3% of cerebrospinal fluid (CSF) samples. Out of these patients, 49.7% were adults.

• Between January 2020 and December 2024, the samples sent had a peak positivity rate of 11.3% in 2020 and a low of 4.8% in 2023.

• Although CSF culture samples were most frequently sent from the neonatal intensive care unit (23.5%), the clinics with the highest positivity rates were the palliative intensive care unit (14.8%).

• Bacterial growth was significantly more common in CSF samples collected via external ventricular drain (EVD) compared to those collected via lumbar puncture (LP) (p < 0.001).

• The most frequently identified gram-positive bacteria in CSF samples were coagulase-negative staphylococci (CoNS) (47.9%), Enterococcus faecium (E. faecium) (5.2%), and Staphylococcus aureus (S. aureus) (4.5%), while the most common gram-negative bacteria were Klebsiella pneumoniae (K. pneumoniae) (9.1%), Acinetobacter baumannii (A. baumannii) (9%), and Pseudomonas aeruginosa (P. aeruginosa) (7%).

• Detection rates for MRCoNS have averaged 89.7% over the past five years, showing a growing trend in resistance detection.

• The average detection rate of vancomycin-resistant enterococci (VRE) was 37.3%. The resistance rates for teicoplanin and levofloxacin approached 60%.

• The rates of MRSA detection varied between 61.5% and 84.6%, with an average rate of 71.9%. The average rate of clindamycin resistance that can be induced was 35.1%.

• In 2024, the resistance to Ertapenem in K. pneumoniae was around 80%. The prevalence of ESBL-producing strains fluctuated over the years, averaging 32.4%.

• Colistin resistance in A. baumannii reached nearly 10%.

• Piperacillin/tazobactam had the highest resistance rate in P. aeruginosa at 75%, while the average resistance rate to colistin was 6.7%.

Disclosure statement

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.

Author contributions

Study conception and design: Merve Gürler., Füsun Kırca., Figen Türen., İrem Tümkaya Kılınç., Ferhat Gürkan Aslan., Emrah Salman., Nilay Çöplü., Bedia Dinç., Material preparation, data collection, and analysis: Merve Gürler., Füsun Kırca., Figen Türen., İrem Tümkaya Kılınç., Ferhat Gürkan Aslan., E.S., Writing-original draft: Merve Gürler., Writing – review and editing: Merve Gürler., Füsun Kırca., Figen Türen., İrem Tümkaya Kılınç., Ferhat Gürkan Aslan., Emrah Salman., Nilay Çöplü., Bedia Dinç. All authors have read and agreed to the published version of the manuscript.

Ethical declaration

This study was conducted in accordance with the principles of the Declaration of Helsinki. The ethical approval was obtained from the Ankara Bilkent City Hospital Medical Research Scientific and Ethical Review Board No.1 (29 January 2025, no: TABED 1/933/2025)

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/17460913.2025.2520666