Impact of Culture Status on Outcomes in Vertebral Osteomyelitis: A Retrospective Cohort Study of 206 Patients

Abstract

Study Design

Retrospective cohort study.

Objectives

To compare clinical features and outcomes between culture-positive and culture-negative vertebral osteomyelitis (VO), describe empirical antimicrobial regimens used for culture-negative VO, and identify independent predictors of 1-year recurrence and mortality.

Methods

We retrospectively reviewed 252 consecutive patients treated for VO at a tertiary center (2010–2023). Finally, 206 patients were analyzed (173 culture-positive, 33 culture-negative). Demographics, comorbidities, laboratory and imaging findings, microbiology, antimicrobial regimens, and surgical treatment were collected. Antimicrobials were generally administered for ≥6 weeks. Primary outcomes were initial treatment failure, recurrence, 1-year all-cause mortality, and infection-related mortality.

Results

Culture-negative VO showed lower temperature, C-reactive protein, and white blood cell count; higher albumin; and fewer epidural abscesses than culture-positive VO, consistent with a milder phenotype. Initial treatment failure and recurrence occurred in 24/206 (11.7%) and 3/206 (1.5%) patients, respectively. All recurrences occurred in the culture-positive group. One-year all-cause mortality was 17% (29/173) in culture-positive and 24% (8/33) in culture-negative patients (p=0.305); infection-related mortality was 13% and 18%, respectively (p=0.424). Among culture-negative patients, outcomes did not differ between cefazolin-based versus other empirical regimens. Multivariable analysis identified older age, higher Charlson Comorbidity Index, hypoalbuminemia, and absence of surgery as independent predictors of mortality, whereas culture status was not.

Conclusions

In this cohort with high microbiological yield (84%), culture-negative VO presented with milder clinical features and had recurrence and 1-year mortality comparable to culture-positive VO when treated with ≥6 weeks of antimicrobials. Prognosis was driven primarily by host factors rather than culture status.

Keywords

vertebral osteomyelitis culture mortality recurrence level 4

Introduction

Vertebral osteomyelitis (VO) is a rare but serious spinal infection predominantly affecting older adults and those with significant comorbidities.¹ Improvements in magnetic resonance imaging (MRI) and population aging have contributed to the increasing incidence of VO, which remains associated with prolonged hospitalization, neurological impairment, and substantial mortality.¹ Although establishing a definitive diagnosis of native vertebral osteomyelitis can be time-consuming in some patients,² it can often be achieved promptly with early use of advanced imaging and appropriate microbiological investigation. A microbiological diagnosis is central to VO management, as it allows targeted antimicrobial therapy and supports antimicrobial stewardship.^2,3 Guidelines from the Infectious Diseases Society of America recommend performing multiple sets of blood cultures, and whenever feasible, conducting image-guided or open spinal biopsy before initiating antimicrobial therapy.²

Despite these strategies, culture yield remains imperfect. In approximately 20–40% of VO cases, the causative organisms fail to be identified, often because of prior antibiotic exposure, low bacterial burden, or sampling limitations.^2-5 In culture-negative VO, clinicians must rely on empirical regimens directed mainly against Staphylococcus aureus and common Gram-negative bacilli, tailored to the local epidemiology and host risk factors.^2,5,6 Several observational studies have previously suggested that culture-negative VO presents with a less intense systemic inflammatory response and may have outcomes comparable to or even better than those of microbiologically confirmed infections when adequately treated.^5,7-9 Conversely, other reports and reviews have emphasized the uncertainty surrounding empirical choices, the risk of overly narrow or unnecessarily broad coverage, and the potential for relapse or treatment failure in culture-negative VO.^5,8,10 Consequently, there is currently no consensus regarding the optimal empirical strategy, particularly the balance between narrow-spectrum regimens targeting methicillin-sensitive S. aureus and streptococci versus broad-spectrum combinations with routine anti-MRSA coverage—and relatively few studies have examined how culture status relates to long-term outcomes, such as recurrence and mortality, or interacts with established prognostic factors.

To address these gaps, we conducted a single-center retrospective study of patients with VO to compare the clinical characteristics, imaging findings, antimicrobial use, and outcomes between culture-positive and culture-negative cases, focusing specifically on real-world empirical regimens for culture-negative VO (commonly based on first-generation cephalosporins in our center) and their association with recurrence and 1-year mortality. Therefore, this study addressed the following questions: (1) Do culture-negative and culture-positive VO differ in clinical presentation and outcomes, including recurrence and 1-year mortality? (2) What empirical antimicrobial regimens are used for culture-negative VO in routine clinical practice, and is empirical antimicrobial choice associated with prognosis? (3) Which factors independently predict 1-year all-cause and infection-related mortality in patients with VO?

Materials and Methods

Study Design and Setting

This study was a level IV retrospective cohort study conducted at a single tertiary medical center. We identified 252 consecutive patients who were diagnosed with and treated for VO between January 2010 and April 2023. We excluded patients with tuberculous spondylitis, postoperative spinal infection, isolated epidural abscesses without vertebral involvement, and those managed exclusively as outpatients.

Participants

Of the 252 patients initially diagnosed with VO during the study period, 46 were excluded because 1-year follow-up data were unavailable (those who died within 1 year were retained), leaving 206 patients for the analysis (173 culture-positive and 33 culture-negative). Thus, none of the patients included in the final cohort were lost to follow-up after the 1-year time point.

In this study, the term vertebral osteomyelitis (VO) was used broadly to include infectious involvement of the vertebral bodies, intervertebral discs, or both (i.e., spondylodiscitis/discitis and vertebral osteomyelitis). VO was suspected on the basis of new or worsening back or neck pain, fever, and laboratory evidence of inflammation. To facilitate early diagnosis, the initial evaluation included MRI, contrast-enhanced CT, and at least two sets of blood cultures. The diagnosis was confirmed by integrating the clinical, imaging, laboratory, and microbiological findings. MRI features consistent with VO include endplate destruction, increased T2-weighted signal intensity in the vertebral bodies and intervertebral discs, and epidural or iliopsoas abscesses. CT was used to evaluate the extent of endplate and vertebral body destruction. Imaging diagnoses were jointly confirmed by a senior spine surgeon with 30 years of experience in orthopedic and spine surgery and an experienced radiologist. In culture-negative cases, imaging was re-evaluated to exclude sacroiliitis and non-infectious conditions, such as ankylosing spondylitis or axial spondyloarthritis.

Overview of Treatment and Surgery

At the initial presentation, patients with suspected VO underwent at least two sets of blood cultures. When clinically indicated for microbiological diagnosis, percutaneous aspiration of the disc space or a vertebral body abscess was performed to identify the causative organism. CT-guided drainage was performed for psoas abscesses, and all specimens were cultured. Surgical intervention was indicated in patients with poor response to antimicrobial therapy, progressive neurological deficit, spinal instability or deformity, or extensive vertebral destruction. Poor response to antimicrobial therapy was defined as persistent or worsening pain, fever, inflammatory markers, or imaging findings despite appropriate antimicrobial treatment. Deformity was defined as progressive kyphotic or segmental alignment abnormality associated with infectious destruction. Meanwhile, vertebral destruction was defined as marked endplate or vertebral body destruction on CT or MRI judged to compromise spinal stability or risk progressive collapse. In such cases, debridement of the infection focus with decompression and/or spinal stabilization was conducted, and intraoperative specimens were obtained for culture. All microbiological results were reviewed by infectious disease specialists and any isolates judged to be affected by contamination were excluded from the analysis.

Antimicrobial regimens were selected by spine surgeons and infectious disease specialists based on each patient’s clinical presentation, culture results, and antimicrobial susceptibility testing. In principle, antibiotic therapy was administered for a total duration of at least 6 weeks, in accordance with the current guideline recommendations for VO, typically beginning with intravenous therapy.^2,11,12 Oral antibiotic therapy was subsequently introduced in selected patients under infectious disease specialist supervision, based on the clinical course, including trends in inflammatory markers, overall clinical condition, and the presence of implanted foreign material such as spinal instrumentation or a prosthetic valve.

Patients underwent regular follow-up in the outpatient clinic with clinical assessment, laboratory tests (including C-reactive protein [CRP] and white blood cell count [WBC]), and imaging conducted as needed. Follow-up MRI or CT was conducted when clinical or laboratory improvements were insufficient, or when recurrence was suspected. Patients with at least 1 year of follow-up after diagnosis were included in the analysis; deaths occurring within 1 year were also recorded and classified as either infection-related or non-infection-related.

Variables, Outcome Measures, Data Sources, and Bias

Clinical data, including age, sex, body mass index (BMI), and medical history, were collected from the electronic medical records. Comorbidity burden was assessed using the Charlson Comorbidity Index (CCI) according to the original, non-age-adjusted Charlson method.¹³ The maximum body temperature (°C) within 1 week of presentation was recorded as a clinical variable. Laboratory variables obtained within 1 week of presentation included white blood cell count (×10⁹/L), C-reactive protein (mg/dL), and minimum serum albumin level (g/dL). Imaging studies were reviewed for the presence of epidural and iliopsoas abscesses, the site of vertebral involvement, as well as the number of infected intervertebral discs. When infection was confined to a single vertebral body without any apparent disc involvement, the number of infected discs was recorded as one. The presence of infective endocarditis (IE) and antibiotic exposure prior to culture sampling were also documented. The presence of infective endocarditis (IE) was determined by cardiology specialists using transthoracic echocardiography and/or transesophageal echocardiography, as clinically appropriate. Motor function was graded using the American Spinal Injury Association Impairment Scale (AIS), with AIS grades A–C indicating severe paralysis.

All patients were evaluated by spine surgeons and infectious disease specialists and classified as culture-positive when a causative organism was identified from any collected specimen, including blood cultures, disc space aspiration, CT-guided drainage, or intraoperative specimens. Patients were classified as culture negative only when no causative organism was identified from any of these microbiological specimens. The primary outcomes included initial treatment failure, infection recurrence, and 1-year all-cause and infection-related mortality. Treatment success was defined as clinical improvement or stabilization after the initial treatment course, without death due to uncontrolled vertebral osteomyelitis during that course. Initial treatment failure was defined as death due to uncontrolled vertebral osteomyelitis with sepsis progressing to multiple organ failure during the initial treatment course. Recurrence was defined as clinical and radiological deterioration requiring renewed or prolonged antimicrobial therapy and/or repeat surgery after initial improvement. Infection-related death was defined as death attributable to uncontrolled vertebral osteomyelitis, persistent or recurrent bloodstream infection associated with vertebral osteomyelitis, or infectious complications occurring during treatment. Deaths primarily attributable to non-infectious causes, such as malignancy, cardiovascular events, or senility, were classified as non-infection related.

To mitigate misclassifications, all imaging and microbiological results were jointly reviewed by spine surgeons, radiologists, and infectious disease specialists.

Ethical Approval

This study was approved by our institutional review board (approval number: zn231213). Participants’ notifications were managed using a general opt-out approach. The study complied with the institutional ethical guidelines and the Declaration of Helsinki, and the need for individual informed consent was waived.

Statistical Analysis

All statistical analyses were conducted using JMP 17.0 (JMP, Cary, NC, USA). Continuous variables were tested for normality and compared using Student’s t-test or the Mann–Whitney U test, as appropriate. Categorical variables were compared using the chi-square test or Fisher–Freeman–Halton exact test. Univariate logistic regression analyses were applied to investigate the variables associated with 1-year all-cause and infection-related mortality. Variables that were statistically significant in the univariate analysis, and judged to be clinically relevant, were entered into the multivariate logistic regression models. Collinearity among candidate variables was assessed using correlation matrices and variance inflation factors (VIFs), together with clinical judgment. A VIF of greater than 5 was considered to indicate potentially important multicollinearity. Variables identified as highly collinear were not included in the same multivariable model. Odds ratios (ORs) with 95% confidence intervals (CIs) were calculated, and the level of significance was set at p < 0.05.

Results

Patient Characteristics and Antibiotic Treatment

A total of 206 patients with VO were included, comprising 173 culture-positive (84%) and 33 culture-negative (16%) patients (Table 1). The mean age was 68.6 ± 15.2 years, and sex distribution, BMI, CCI, and the prevalence of diabetes mellitus and alcoholism were similar between groups. As shown in Table 2, the distribution of individual CCI components was broadly similar between the culture-positive and culture-negative groups, with no statistically significant between-group differences for any component (Table 2). In the overall cohort, uncomplicated diabetes mellitus, localized malignant tumor, and chronic heart failure were the most prevalent comorbidities. Leukemia and AIDS were not present in this series. Surgical treatment was performed in 65 of 206 patients (31%), including 59 of 173 culture-positive patients (34%) and 6 of 33 culture-negative patients (18%), and the detailed surgical treatment is shown in Table 1. All patients initially received intravenous antibiotic therapy. Although a minimum treatment duration of 6 weeks was generally targeted, the actual duration of intravenous antibiotic therapy was individualized under the supervision of an infectious disease specialist according to the clinical course, including trends in inflammatory markers, overall clinical condition, and the presence of implanted foreign materials. The mean duration of intravenous antibiotic therapy was 7.0 weeks, indicating that extended treatment beyond 6 weeks was required in some patients. Oral antibiotic therapy was subsequently administered in 144 of 206 patients (70%), primarily as suppressive therapy when clinically indicated.

Table 1.

Demographic and Clinical Characteristics of Patients With Vertebral Osteomyelitis According to Culture Status

	Total	Culture positive group	Culture negative group	p value
No. of patients	206	173(84%)	33(16%)
Age (years)	68.6±15.2	67.9±15.4	72.2±13.9	0.221^*
Sex (M:F)	137:69	112:61	25:8	0.314^b
Temperature	38.8±1.0	38.9±1.0	38.3±1.0	0.004^*
CRP	19.1±12.4	20.1±12.0	14.0±13.6	0.003^*
WBC	14.2±6.8	15.0±7.0	10.4±4.1	<0.001^*
ALB	2.23±0.65	2.13±0.62	2.75±0.57	<0.001^a
BMI	23.1±4.9	22.9±4.7	24.1±5.6	0.114^*
Number of infected discs	1.4±1.0	1.5±1.0	1.2±0.5	0.295^*
CCI	2.42±2.43	2.42±2.46	2.45±2.32	0.819^*
IP abscess	88(42%)	78(45%)	10(30%)	0.128^b
Epidural abscess	74(35%)	71(41%)	3(9%)	<0.001^b
IE	27(13%)	26(15%)	1(3%)	0.088^b
Prior antibiotic exposure	35(17%)	22(12%)	13(39%)	0.001^b
Alcoholism	17(8%)	15(8%)	2(6%)	1.000^b
Diabetes mellitus	65(31%)	54(31%)	11(33%)	0.839^b
Surgical treatment	65(31%)	59(34%)	6(18%)	0.101^b
Debridement	7(11%)	7(12%)	0(0%)	0.585^b
Posterior decompression	6(9%)	5(9%)	1(17%)
Posterior fusion with instrumentation without decompression	10(16%)	10(17%)	0(0%)
Anterior and/or posterior fusion with instrumentation and bone graft	42(65%)	37(63%)	5(83%)
Use of instrumentation	52(80%)	47(80%)	5(83%)	1.000^b
AIS A-C	10(5%)	8(5%)	2(6%)	0.664^b
Follow up period (months)	56.4±45.3	58.2±45.2	46.7±45.0	0.128^*

For continuous variables, data are expressed as the mean±SD.

^*Mann–Whitney U test.

^aStudent’s t-test.

^bFisher’s exact test.

Values are expressed as mean ± SD unless otherwise indicated. CRP, C-reactive protein; WBC, white blood cell count; ALB, albumin; BMI, body mass index; CCI, Charlson Comorbidity Index; IP, iliopsoas; IE, infective endocarditis; AIS, American Spinal Injury Association Impairment Scale.

Table 2.

Distribution of Charlson Comorbidity Index Components by Culture Status

Charlson comorbidity index	Category	Total, n (%)	Culture-positive, n (%)	Culture-negative, n (%)	P Value
Myocardial infarction		17 (8.3%)	14 (8.1%)	3 (9.1%)	0.740^b
Chronic heart failure		33 (16.0%)	26 (15.0%)	7 (21.2%)	0.375^a
Peripheral vascular disease		7 (3.4%)	5 (2.9%)	2 (6.1%)	0.312^b
CVA or TIA		21 (10.2%)	19 (11.0%)	2 (6.1%)	0.539^b
Dementia		12 (5.8%)	9 (5.2%)	3 (9.1%)	0.413^b
Chronic pulmonary disease		12 (5.8%)	9 (5.2%)	3 (9.1%)	0.413^b
Connective tissue disease		10 (4.9%)	8 (4.6%)	2 (6.1%)	0.664^b
Peptic ulcer disease		10 (4.9%)	8 (4.6%)	2 (6.1%)	0.664^b
Liver disease	Mild	15 (7.3%)	12 (6.9%)	3 (9.1%)	0.713^b
Liver disease	Moderate to severe	12 (5.8%)	11 (6.4%)	1 (3.0%)	0.695^b
Diabetes mellitus	Uncomplicated	45 (21.8%)	37 (21.4%)	8 (24.2%)	0.716^a
Diabetes mellitus	End-organ damage	19 (9.2%)	16 (9.2%)	3 (9.1%)	1.000^b
Hemiplegia		4 (1.9%)	3 (1.7%)	1 (3.0%)	0.505^b
Moderate to severe CKD		18 (8.7%)	13 (7.5%)	5 (15.2%)	0.176^b
Malignant tumor	Localized	44 (21.4%)	38 (22.0%)	6 (18.2%)	0.627^a
Malignant tumor	Metastatic	4 (1.9%)	4 (2.3%)	0 (0.0%)	1.000^b
Leukemia		0 (0.0%)	0 (0.0%)	0 (0.0%)	—
Lymphoma		7 (3.4%)	6 (3.5%)	1 (3.0%)	1.000^b
AIDS		0 (0.0%)	0 (0.0%)	0 (0.0%)	—

^aChi-square test.

^bFisher’s exact test.

CVA, cerebrovascular accident; TIA, transient ischemic attack; CKD, chronic kidney disease; AIDS, acquired immune deficiency syndrome.

Differences in Clinical Presentations and Prior Antibiotic Exposure Between Culture-positive and Culture-negative Vertebral Osteomyelitis

At presentation, culture-negative patients showed significantly milder systemic inflammatory responses than culture-positive patients. Maximum body temperature (38.3 ± 1.0°C vs 38.9 ± 1.0°C, p = 0.004), CRP (14.0 ± 13.6 vs 20.1 ± 12.0 mg/dL, p = 0.003), and WBC count (10.4 ± 4.1 vs 15.0 ± 7.0 ×10⁹/L, p < 0.001) were all lower in the culture-negative group, whereas serum ALB was higher (2.75 ± 0.57 vs 2.13 ± 0.62 g/dL, p < 0.001) (Table 1). The number of infected discs and distribution of the affected spinal levels did not differ between groups. Epidural abscesses were identified in 41% of culture-positive patients (71/173), but in only 9% of culture-negative patients (3/33; p < 0.001). The frequencies of iliopsoas abscesses and infective endocarditis did not differ significantly between the groups. Antibiotic exposure before culture sampling was more frequent in culture-negative patients than in culture-positive patients (39% vs 12%; p = 0.001). To further characterize this finding, additional analysis of prior antibiotic exposure showed that antibiotic administration before blood culture was more common in the culture-negative group than in the culture-positive group (18/50 vs 17/155, p < 0.001), and its duration was also significantly longer (22.2 ± 23.3 vs 5.1 ± 5.9 days, p = 0.002). Among patients who underwent biopsy, prior antibiotic use before biopsy was common in both groups and did not differ significantly (100% vs 87.9%, p = 0.114), although the duration of antibiotic administration before biopsy was significantly longer in the culture-negative group (42.8 ± 29.0 vs 19.1 ± 21.7 days, p < 0.001). Antibiotic interruption before biopsy was infrequent and similar between the groups (12.9% vs 13.8%, p = 1.000), as was the duration of interruption (4.5 ± 2.9 vs 6.3 ± 4.4 days, p = 0.530) (Table 3).

Table 3.

Impact of Prior Antibiotic Exposure on Culture Yield of Blood Culture and Biopsy, including Fluoroscopic Disc Biopsy, CT-Guided Drainage, and Intraoperative Specimen

			Culture-negative	Culture-positive	p-value
Blood culture	Antibiotics before blood culture	yes	18	17	<0.001^a
	Antibiotics before blood culture	no	32	138
	Antibiotics duration before blood culture(days)		22.2±23.3	5.1±5.9	0.002^*
	Form of Antibiotics before blood culture	Intravenous	11(91.7%)	18(85.7%)	1.000^a
	Form of Antibiotics before blood culture	Oral	1(8.3%)	3(14.3%)
Biopsy	Antibiotics before biopsy	yes	31(100%)	29(87.9%)	0.114^a
	Antibiotics before biopsy	no	0	4(12.1%)
	Antibiotics duration before biopsy(days)		42.8±29.0	19.1±21.7	<0.001^*
	Antibiotics interruption	yes	4(12.9%)	4(13.8%)	1.000^a
	Antibiotics interruption	no	27(87.1%)	25(86.2%)
	Antibiotics interruption period(days)		4.5±2.9	6.3±4.4	0.530^*

For continuous variables, data are expressed as the mean±SD.

^*Mann–Whitney U test.

^aFisher’s exact test.

What empirical antimicrobial regimens are routinely used for culture-negative VO in clinical practice, and is the treatment choice associated with prognosis?

S. aureus was the most common pathogen in culture-positive VO, including 60 cases of methicillin-sensitive S. aureus (MSSA), 18 cases of methicillin-resistant S. aureus (MRSA); streptococcal species accounted for 47 cases, while Gram-negative bacilli accounted for 33 cases (Table 4). Regarding the source of microbiological diagnosis, blood culture yielded positive results in 155 of 205 patients (75.6%), CT-guided drainage in 32 of 44 (72.7%), fluoroscopic disc biopsy in 9 of 28 (32.1%), and intraoperative specimens in 28 of 56 (50.0%). Overall, causative microorganisms were identified in 173 of 206 patients (84.0%). Among these methods, blood culture showed the highest diagnostic yield in this cohort, closely followed by CT-guided drainage (Table 5). Cefazolin was the most frequently used agent for MSSA infections (44/60), whereas MRSA infections were treated exclusively with glycopeptides (Table 4). Streptococcal infections were predominantly treated with third-generation cephalosporins or other β-lactams, and Gram-negative infections with cephalosporins, broad-spectrum β-lactams, carbapenems, or quinolones according to susceptibility. In our cohort, antimicrobial therapy was administered for a total duration of at least 6 weeks, in accordance with the current guidelines for VO.^2,11,14 In culture-negative cases, empirical therapy was predominantly based on narrow-spectrum β-lactams: cefazolin was used in 18 of 33 patients (55%), followed by third-generation cephalosporins in six and other β-lactams in five. Glycopeptides and antifungal agents were used in two patients, while carbapenems or quinolones were not routinely employed. Among the 33 culture-negative patients, 4 received combination antibiotic therapy. In these patients, the primary regimen was categorized as either cefazolin or a third-generation cephalosporin, with temporary concomitant administration of a glycopeptide. Among the 33 culture-negative patients, the outcomes did not differ between those treated with cefazolin and those treated with other regimens (recurrence: 0/18 vs 0/15; infection-related death: 2/18 vs 4/15, p = 0.660; all-cause death: 4/18 vs 4/15, p = 1.00).

Table 4.

Primary Intravenous Antibiotics Used According to the Causative Organism and Culture Status

Used antibiotics	No.of patients	Cefazolin	2nd gen cefalosporin	3rd gen cefalosporin	Beta lactam	Glycopeptide	Carbapenem	Quinolone	Antifungal	Others
Culture positive group	173
Staphylococcus aureus
Methicillin-sensitive S. aureus	60	44		5	7	3				1
Methicillin-resistant S. aureus	18					18
Streptococcus species
Streptococcus agalactiae	10			1	9
Streptococcus pneumoniae	9			3	5	1
Other streptococci	28	3		1	21	2	1
Gram-negative bacilli
Escherichia coli	23	9	1	4	3		3	1		2
Other Gram-negative bacilli	10	3		3	2		1	1
Coagulase-negative staphylococci	2	1								1
Anaerobes	5				5
Fungi	2								2
Others	6			1	5
Culture negative group	33	18		6	5	2				2

Table 5.

Details of Culture Methods and Identified Pathogens

Method	Yield	%
Blood culture	155 of 205	75.6
CT-guided drainage	32 of 44	72.7
Fluoroscopic disc biopsy	9 of 28	32.1
Intraoperative specimens	28 of 56	50.0
Overall	173 of 206	84.0

Clinical Outcomes and Predictors of 1-year Mortality in Culture-positive and Culture-negative VO

Initial treatment failure occurred in 24 of 206 patients (11.7%), including 21 of 173 culture-positive patients (12.1%) and 3 of 33 culture-negative patients (9.1%; p = 0.773) (Table 6). Infection recurrence after initial improvement was rare, occurring in only three patients, corresponding to 1.5% of the entire cohort. All recurrences occurred in the culture-positive group (3/173, 1.7%), and no recurrence was observed in the culture-negative group (0/33). The overall 1-year mortality rate was 18% (37/206): 17% in the culture-positive group (29/173) and 24% in the culture-negative group (8/33; p = 0.305). Infection-related deaths occurred in 29 patients (14% overall), including 23 culture-positive (13%) and 6 culture-negative patients (18%, p = 0.424). Among the 29 infection-related deaths, 24 were classified as initial treatment failure due to uncontrolled vertebral osteomyelitis with sepsis progressing to multiple organ failure. The remaining infection-related deaths were attributed to pneumonia including aspiration pneumonia during treatment (4 patients) and sepsis-associated circulatory failure/heart failure (1 patient). In univariate analysis (Table 7), non-survivors were significantly older (77.9 ± 12.0 vs 66.5 ± 15.1 years, p < 0.001) and had lower serum albumin (1.86 ± 0.59 vs 2.31 ± 0.64 g/dL, p < 0.001), higher CRP (23.0 ± 12.5 vs 18.3 ± 12.3 mg/dL, p = 0.043) and CCI (3.30 ± 2.63 vs 2.24 ± 2.35, p = 0.020) than those of survivors. Surgical treatment was less frequent among non-survivors and was associated with lower all-cause mortality (p = 0.010). For infection-related deaths, low ALB level, high CRP level and high CCI were significantly associated with mortality, whereas culture status and prior antibiotic exposure were not. In the multivariate logistic regression analysis (Table 8), advanced age, low serum albumin levels, higher CCI, and the absence of surgical treatment were identified as independent predictors of all-cause mortality. Low serum albumin levels, high CCI, and non-surgical management remained independent risk factors for infection-related deaths. Advanced age was independently associated with all-cause mortality; for infection-related mortality, it showed only a borderline association (OR 1.034, 95% CI 0.999–1.075, p = 0.055). A culture-negative status was not identified as an independent predictor of all-cause or infection-related mortality.

Table 6.

Initial Treatment Failure, Recurrence, and 1-Year Mortality According to Culture Status

	Total	Culture positive group	Culture negative group	p value
No. of patients	206	173	33
Initial treatment failure	24 (11.7%)	21 (12.1%)	3 (9.1%)	0.773^b
Recurrence after initial improvement	3 (1.5%)	3 (1.7%)	0	1.000^b
Mortality (1 year)
All-cause deaths	37 (18%)	29 (17%)	8 (24%)	0.305^a
Infection-related deaths	29 (14%)	23 (13%)	6 (18%)	0.424^b

^aChi-square test.

^bFisher’s exact test.

Table 7.

Results of Univariate Analysis of Risk Factors for 1-Year All-Cause and Infection-Related Mortality

Variable	All-cause deaths	Survivors	Odds ratio	95% Cl	P Value	Infection-relateddeaths	Others	Odds ratio	95%Cl	P Value
No of patients	37	169				29	177
Age	77.9±12.0	66.5±15.1	1.080	1.04–1.12	<0.001	76.0±12.6	67.4±15.3	1.051	1.02–1.09	0.005
Sex M:F	28:9	109:60	1.710	0.76–3.87	0.249	22:7	115:62	1.694	0.69–4.2	0.253
BMI	22.5±3.9	23.3±5.1	0.967	0.89–1.05	0.866	22.9±3.9	23.2±5.0	0.991	0.91–1.08	0.830
ALB	1.86±0.59	2.31±0.64	0.283	0.14–0.56	<0.001	1.79±0.6	2.32±0.6	0.220	0.10–0.48	<0.001
CRP	23.0±12.5	18.3±12.3	1.030	1.00–1.06	0.043	24.9±12.8	18.2±12.2	1.042	1.01–1.08	0.010
WBC	15.7±6.9	13.9±6.7	1.040	0.99–1.09	0.115	16.0±6.9	14.0±6.7	1.041	0.99–1.10	0.146
CCI	3.30±2.63	2.24±2.35	1.178	1.03-1.35	0.020	3.40±2.5	2.27±2.4	1.182	1.02–1.37	0.026
DM	16	49	1.870	0.90–3.87	0.118	13	52	1.953	0.88–4.3	0.101
Alcohol	2	14	0.995	0.27–3.67	1.000	3	14	1.343	0.36–5.00	0.660
Epidural abscess	9	65	0.514	0.23–1.16	0.131	8	66	0.641	0.27–1.53	0.316
IP abscess	21	67	2.000	0.97–4.1	0.067	16	72	1.795	0.81–3.96	0.147
IE	4	23	0.769	0.25–2.37	0.792	3	24	0.736	0.21–2.62	0.636
Number of infected discs	1.3±0.6	1.5±1.0	0.805	0.51–1.27	0.625	1.3±0.7	1.4±1.0	0.883	0.56–1.41	0.600
Culture Positive	29	144	0.629	0.26–1.53	0.325	23	150	0.690	0.26–1.85	0.461
Prior antibiotic exposure	7	28	1.180	0.47–2.94	0.809	6	29	1.331	0.500–3.56	0.568
MRSA	4	14	0.969	0.33–2.62	0.952	3	15	1.246	0.34–4.60	0.741
Surgical treatment	5	60	0.284	0.105-0.767	0.010	3	62	0.214	0.06-0.74	0.009
AIS A-C	10	21	1.500	0.67–3.25	0.317	10	21	3.910	1.60–9.53	0.003

CI, confidence interval; BMI, body mass index; ALB, albumin; CRP, C-reactive protein; WBC, white blood cell count. CCI, Charlson Comorbidity Index; DM, diabetes mellitus; IP, iliopsoas; IE, infective endocarditis; AIS, American Spinal Injury Association.

Table 8.

Results of Multivariate Analysis of Risk Factors for 1-Year All-Cause and Infection-Related Mortality

	All-cause deaths			VIF	Infection-related deaths			VIF
	OR	95% CI	p value		OR	95% CI	p value
Age	1.063	1.027-1.107	<0.01	1.10	1.034	0.999-1.075	0.055	1.10
ALB	0.231	0.099-0.490	<0.01	1.08	0.149	0.055-0.358	<0.01	1.08
CCI	1.179	1.010-1.380	0.038	1.04	1.205	1.019-1.429	0.029	1.04
Surgical treatment	0.235	0.071-0.653	<0.01	1.02	0.152	0.033-0.502	<0.01	1.02

CI, confidence interval; OR, odds ratio; ALB, albumin; CCI, Charlson Comorbidity Index; VIF, Variance inflation factors.

Clinical Characteristics and Outcomes According to Surgical Treatment

Since multivariate analysis identified surgical treatment as an independent factor associated with better survival, baseline characteristics were further compared between surgically and non-surgically treated patients. No significant differences were observed in age, CCI, temperature, CRP, WBC count, albumin, or BMI. By contrast, epidural abscesses and severe neurological impairment were significantly more frequent in the surgically treated group (35/65 vs 39/141, p < 0.001, and 15/65 vs 16/141, p = 0.036, respectively), consistent with the clinical indications for surgery. Of the 65 surgically treated patients, 52 had instrumentation and 13 did not. In the surgically treated subgroup, comparison according to the use of instrumentation showed no significant differences in the duration of intravenous antibiotic therapy, the proportion of patients receiving intravenous therapy for more than 6 weeks, the use of oral suppressive antibiotic therapy, or recurrence risk. The mean duration of intravenous antibiotic therapy was 8.1 ± 4.2 weeks in the instrumentation group and 9.7 ± 4.6 weeks in the non-instrumentation group (p = 0.130). Intravenous therapy for more than 6 weeks was administered in 40 of 52 patients (76.9%) in the instrumentation group and 12 of 13 patients (92.3%) in the non-instrumentation group (p = 0.438). Oral suppressive antibiotic therapy was used in 46 of 52 patients (88.5%) and 13 of 13 patients (100%), respectively (p = 0.335). Recurrence was rare in both groups, occurring in 1 of 52 patients (1.9%) in the instrumentation group and in none of the 13 patients in the non-instrumentation group (0%, p = 1.000).

Discussion

Overall, we found that patients with culture-negative VO had milder inflammatory profiles but similar recurrence and mortality rates, while mortality was driven by age, comorbidity, albumin level, and surgery rather than culture status.

Clinical Significance of Culture-negative Vertebral Osteomyelitis

Several studies have examined the clinical significance of culture-negative VO and reported outcomes comparable to those of microbiologically confirmed diseases.^5,7,9,15 Kim et al. found that culture-negative patients presented with lower inflammatory markers and achieved similar outcomes to culture-positive patients when treated with prolonged antimicrobial therapy,⁷ while Yu et al. reported no significant differences in mortality or relapse between the two groups.⁸ Our results are consistent with these observations; although culture-negative cases in our cohort represented a clinically milder phenotype, their recurrence rate was 0%, and their 1-year mortality was not higher than that of culture-positive patients. These data support the notion that host factors and infection severity are more important determinants of outcomes than culture status alone in cases of VO.

The more favorable inflammatory and radiologic profiles observed in our culture-negative group provide a plausible explanation for the non-inferior outcomes. Culture-negative patients presented with lower fever, CRP, and WBC counts, and higher serum albumin levels, suggesting a lower systemic inflammatory burden, which may also have contributed to relatively preserved serum albumin levels. This interpretation is also consistent with the report by Ghassibi et al., who showed that the responsiveness of routine diagnostic laboratory tests in VO may vary according to the infecting organism, with lower inflammatory marker values observed in culture-negative and other lower-virulence infections than in Staphylococcus aureus or antibiotic-resistant infections.¹⁶ They were also less likely to have an epidural abscess, which is typically associated with a more extensive infection, as well as a higher risk of neurological compromise.⁵ Antibiotic exposure prior to culture sampling was more common among culture-negative patients, which may have partially sterilized the bloodstream and reduced the culture yield.^3,17,18 Additional analysis showed that antibiotic administration before blood culture was both more frequent and longer in culture-negative cases. Among patients who underwent biopsy, the duration of antecedent antibiotic administration before biopsy was also significantly longer in culture-negative cases, whereas biopsy interruption practices did not differ between groups. The low frequency of antibiotic interruption before biopsy may reflect the fact that, in many cases, microbiological diagnosis had already been established by blood culture and treatment was therefore continued without interruption. These findings suggest that prolonged antecedent antimicrobial exposure, rather than biopsy interruption practice itself, may have contributed to culture-negative results in some patients. Thus, in many culture-negative cases, failure to isolate a pathogen probably reflects a combination of earlier or partial treatment and an intrinsically milder disease, rather than an uncontrolled or atypical infection.

Our culture yield of 84% lies at the upper end of the contemporary VO series.^3-5 This finding is consistent with our previous institutional study, in which higher inflammatory activity, lower serum albumin levels, the presence of an epidural abscess, and the absence of previous antibiotic exposure were associated with successful microbiological identification. Conversely, previous antibiotic exposure was the strongest factor associated with failure to identify the causative organism.¹⁷ This interpretation is also supported by the recent study of Naganuma et al., who reported that pre-culture antimicrobial therapy was associated with a significantly lower proportion of positive blood and abscess cultures, although it did not significantly reduce bone culture positivity.¹⁸ Kim et al. also demonstrated that, in patients undergoing needle biopsy, soft tissue specimens, including disc and paraspinal or psoas abscess samples, had a significantly higher microbiological yield than vertebral body specimens.¹⁹ Achieving a microbiological diagnosis has been identified as an important predictor of successful treatment, as it allows for de-escalation from broad-spectrum empirical regimens to organism-directed therapies.^4,10,20 In our study, the high culture yield may have facilitated timely targeted treatment in most patients, thus contributing to a low recurrence rate of 1.5%. Under such circumstances, culture status becomes a less dominant determinant of outcomes than host factors and the adequacy of medical and surgical management.

What empirical antimicrobial regimens are routinely used for culture-negative VO in clinical practice, and is the treatment choice associated with prognosis?

The antimicrobial strategy used at our center merits attention. Current guidelines recommend providing empirical regimens that combine anti-staphylococcal coverage, including MRSA, with activity against Gram-negative bacilli, such as vancomycin plus third- or fourth-generation cephalosporins, particularly in patients with severe sepsis or healthcare-associated infections.^2,10 In contrast, most culture-negative patients in our cohort were treated empirically with cefazolin-based regimens, without the routine use of glycopeptides, carbapenems, or fluoroquinolones; however, they did not experience higher rates of recurrence or mortality than culture-positive patients. In our cohort, glycopeptide co-administration was used in a small subset of culture-negative patients, indicating that empirical therapy was not uniformly narrow-spectrum when resistant pathogens were a clinical concern. This selective escalation approach is also consistent with a pediatric osteomyelitis cohort in which MRSA-active coverage in culture-negative cases increased during treatment adjustment, although extrapolation to adult vertebral osteomyelitis should be made cautiously.²¹ This likely reflects the relatively low prevalence of MRSA and multidrug-resistant Gram-negative bacilli in our setting and the milder profile of culture-negative cases. Our data extend the findings of Kim, Yu, Lee, Dai, and colleagues—who reported similar or non-inferior outcomes in culture-negative compared with culture-positive pyogenic vertebral osteomyelitis—to a setting with high culture yield and predominant use of cefazolin-based empirical therapy.^7-9,22 Our findings suggest that cefazolin-based regimens represent a feasible empirical option for culture-negative VO in centers with similar epidemiological conditions and in clinically stable patients without risk factors for resistant organisms. However, these observations are hypothesis-generating rather than conclusive and should be interpreted cautiously, as they may not be generalizable to regions with high rates of MRSA or multidrug-resistant Gram-negative organisms or to settings with substantially different pathogen epidemiologies.

Which factors independently predict 1-year all-cause and infection-related mortality in patients with VO?

Importantly, our analysis indicates that 1-year mortality in VO is primarily driven by host-related factors rather than by culture status. A higher CCI score and hypoalbuminemia emerged as independent predictors of all-cause and infection-related mortality, whereas culture-negative status did not. These findings mirror those of previous studies demonstrating that older age, comorbidity burden, poor general condition, and neurological compromise are major determinants of VO outcomes.^1,20,23-25 This is also consistent with the report by Vettivel et al., who found that increasing age was independently associated with 1-year mortality and that epidural abscess and spondylodiscitis were more common in surgically treated patients.²⁶ An apparent difference between all-cause and infection-related mortality was the role of age. In the multivariable analysis, advanced age remained independently associated with all-cause mortality, whereas for infection-related mortality it showed only a borderline association (OR 1.034, 95% CI 0.999–1.075, p = 0.055). This may reflect the fact that all-cause mortality is influenced not only by infection itself but also by age-related frailty, reduced physiologic reserve, and competing non-infectious causes of death. By contrast, infection-related mortality may be more directly determined by sepsis-related deterioration, comorbidity burden and nutritional status. We further observed that surgical intervention was independently associated with a lower risk of death, thus supporting earlier reports that timely decompression and stabilization can improve infection control and functional outcomes in appropriately selected patients, although a selection bias cannot be excluded.^10,23,27,28 In the subgroup analysis of surgically treated patients, the presence of instrumentation was not associated with a longer intravenous antibiotic course, more frequent oral suppressive therapy, or a higher recurrence rate. These findings suggest that, in this cohort, surgical instrumentation itself did not clearly increase the risk of recurrence or mandate prolonged antimicrobial treatment. Recent systematic reviews have indicated that early surgery yields better overall clinical and neurological outcomes than conservative treatment in many patients with pyogenic spondylodiscitis.²⁹ In our cohort, surgical indications were determined primarily based on neurological impairment and structural factors, such as spinal instability or deformity, rather than age or general condition alone. Comparison between the surgically and non-surgically treated groups showed no significant differences in age, CCI, or systemic inflammatory markers. Conversely, epidural abscesses and severe neurological impairment were significantly more frequent in the surgically treated group, consistent with the established indications for operative management. Accordingly, these findings should not be interpreted as evidence of surgical superiority; rather, they suggest that surgical intervention may have contributed to improved outcomes in appropriately selected patients, although residual selection bias cannot be excluded.

Practical Approach in Suspected Culture-negative VO

Based on these findings, the following practical approach may help guide clinical decision-making in suspected culture-negative VO.

1. Obtain blood cultures before antibiotics whenever feasible. Pursue microbiological sampling when indicated, including image-guided biopsy, CT-guided drainage, or aspiration of the disc space or a vertebral body abscess.

2. In patients with sepsis, severe pain, neurological impairment, or structural instability, the risks of delaying treatment for an antibiotic holiday should be carefully weighed against the potential benefit of improving culture yield.

3. If blood cultures yielded negative results and the patient is systemically well, not in severe pain, and neurologically intact, further microbiological sampling should be pursued whenever feasible. A short antibiotic holiday before biopsy or drainage may be considered on an individual basis.

4. If no organism is identified, empirical therapy should be selected according to local epidemiology, disease severity, host factors, and the most likely pathogens, rather than using a single uniform regimen.

5. Clinical response should be reassessed continuously. Surgery should be considered when neurological deterioration, instability, deformity, or poor response to antimicrobial therapy is present.

Limitations

This study had several limitations. First, its retrospective, single-center design may have introduced selection and information biases; therefore, our findings may not be generalizable to settings with different pathogen distributions or antimicrobial resistance patterns. Second, antimicrobial regimens, treatment duration, and surgical indications were not fully standardized and were determined by the treating physicians. Although this approach reflects real-world practice, it may have introduced treatment-selection bias. The number of culture-negative cases was relatively small, thus limiting the statistical power to detect modest differences between the culture-positive and culture-negative groups. Third, molecular diagnostic techniques, such as broad-range PCR or metagenomic sequencing, are not routinely used and might have further increased the microbiological yield in culture-negative VO.^3,5 In addition, no patients in this cohort had documented intravenous drug use. Although this likely reflects the characteristics of the local patient population, it may limit the generalizability of our findings to patients with vertebral osteomyelitis associated with intravenous drug use, which may differ in microbiology and clinical course. Prospective multicenter studies with standardized treatment algorithms and contemporary microbiological methods are required to refine empirical strategies for culture-negative disease and to clarify the relative contributions of microbiological, host-related, and surgical factors to prognosis.

Conclusion

Culture-negative cases (16%) presented with milder inflammation, higher serum ALB levels, and fewer epidural abscesses than those in culture-positive cases; however, infection recurrence was rare, and 1-year mortality did not differ according to culture status. Advanced age, greater comorbidity burden, hypoalbuminemia, and absence of surgical intervention were all identified independent predictors of mortality, whereas culture negativity was not. In a setting with a low prevalence of multidrug-resistant organisms, our findings support the feasibility of cefazolin-based empirical therapy for clinically stable culture-negative vertebral osteomyelitis and indicate that prognosis is determined primarily by host factors and the adequacy of surgical and medical management, rather than by culture status alone. These results will help inform other clinicians regarding the optimal treatment strategies for culture-negative VO.

Footnotes

Acknowledgments

The authors thank Editage for English language editing and proofreading.

ORCID iDs

Sadaki Mitsuzawa

Eijiro Onishi

Ethical Considerations

Approval for this study was received from the Ethics Committee of Kobe City Medical Center General Hospital (Approval Date: Dec 14, 2023; Approval Number: zn231213).

Consent to Participate

The need for informed consent was waived owing to the retrospective study design.

Author Contributions

D.S., E.O., and Y.S. designed the study, collected and analyzed the data, and drafted the manuscript. M.K. and M.T. collected the data. T.Y. supervised the study. All authors have read and approved the submitted version of the manuscript.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Device Status/Drug Statement

The device in this manuscript is FDA-approved or approved by a corresponding national agency for this indication.

References

Mylona

Samarkos

Kakalou

Fanourgiakis

Skoutelis

. Pyogenic Vertebral Osteomyelitis: A Systematic Review of Clinical Characteristics. Semin Arthritis Rheum. 2009;39(1):10-17. doi:10.1016/j.semarthrit.2008.03.002.

Berbari

Kanj

Kowalski

, et al. Infectious Diseases Society of America (IDSA) Clinical Practice Guidelines for the Diagnosis and Treatment of Native Vertebral Osteomyelitis in Adults. Clinical Infectious Diseases. 2015;61(6):e26-e46. doi:10.1093/cid/civ482.

Kim

. Microbiologic diagnosis of pyogenic spondylitis. Infect Chemother. Korean Society of Infectious Diseases, Korean Society for Antimicrobial Therapy, Korean Society for AIDS. 2021;53:238-246. doi:10.3947/IC.2021.0054.

Maamari

Tande

Tai

DBG

, et al. Factors Impacting the Yield of Image-Guided Biopsy in Native Vertebral Osteomyelitis: A 10-Year Retrospective Study. Open Forum Infect Dis. 2022;9(12):ofac616. doi:10.1093/ofid/ofac616.

Alavi

SMA

Petri

Mahmoud

, et al. Culture-Negative Native Vertebral Osteomyelitis: A Narrative Review of an Underdescribed Condition. Multidisciplinary Digital Publishing Institute (MDPI). 2024;13. doi:10.3390/jcm13195802.

Lee

Kim

. Treatment Guideline for Patients with Native Culture-negative Pyogenic Vertebral Osteomyelitis. Clin Orthop Relat Res. 2022;480(1):124-136. doi:10.1097/CORR.0000000000001866.

Kim

Peck

, et al. Outcome of culture-negative pyogenic vertebral osteomyelitis: Comparison with microbiologically confirmed pyogenic vertebral osteomyelitis. Semin Arthritis Rheum. 2014;44(2):246-252. doi:10.1016/j.semarthrit.2014.04.008.

Kim

Jeon

. Antimicrobial therapy and assessing therapeutic response in culture-negative pyogenic vertebral osteomyelitis: a retrospective comparative study with culture-positive pyogenic vertebral osteomyelitis. BMC Infect Dis. 2020;20(1):939. doi:10.1186/s12879-020-05669-1.

Dai

Yin

, et al. Culture-negative versus culture-positive in pyogenic spondylitis and analysis of risk factors for relapse. Br J Neurosurg. 2024;38(2):527-531. doi:10.1080/02688697.2021.1896677.

10.

Lener

Hartmann

Barbagallo

GMV

Certo

Thomé

Tschugg

. Management of spinal infection: a review of the literature. Acta Neurochir (Wien). 2018;160(3):487-496. doi:10.1007/s00701-018-3467-2.

11.

Bernard

Dinh

Ghout

, et al. Antibiotic treatment for 6 weeks versus 12 weeks in patients with pyogenic vertebral osteomyelitis: an open-label, non-inferiority, randomised, controlled trial. The Lancet. 2015;385(9971):875-882. doi:10.1016/S0140-6736(14)61233-2.

12.

Rutges

JPHJ

Kempen

van Dijk

Oner

. Outcome of conservative and surgical treatment of pyogenic spondylodiscitis: a systematic literature review. European Spine Journal. Springer Verlag. 2016;25(4):983-999. doi:10.1007/s00586-015-4318-y.

13.

Charlson

Pompei

Ales

MacKenzie

. A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. J Chronic Dis. 1987;40(5):373-383. doi:10.1016/0021-9681(87)90171-8.

14.

Park

Cho

Lee

, et al. Optimal Duration of Antibiotic Therapy in Patients With Hematogenous Vertebral Osteomyelitis at Low Risk and High Risk of Recurrence. Clin Infect Dis. 2016;62(10):1262-1269. doi:10.1093/CID/CIW098.

15.

Lee

Jeon

Kim

, et al. Clinical characteristics and outcomes of patients with culture-negative pyogenic spondylitis according to empiric glycopeptide use. Infect Chemother. 2019;51(3):274-283. doi:10.3947/ic.2019.51.3.274.

16.

Ghassibi

Yen

Harris

Leary

Choma

. Responsiveness of routine diagnostic tests for vertebral osteomyelitis may be influenced by the infecting organism. Spine Journal. 2021;21(9):1479-1488. doi:10.1016/j.spinee.2021.04.001.

17.

Sakamoto

Onishi

Mitsuzawa

, et al. Risk Factors Preventing Identification of the Microorganism Causing Vertebral Osteomyelitis. Global Spine J. 2025;15(7):3055-3064. doi:10.1177/21925682251317136.

18.

Naganuma

Takahashi

Aita

, et al. Association of pre-culture antimicrobial therapy with culture positivity in patients with vertebral osteomyelitis: a 16-year cohort study. BMC Infect Dis. 2026;26:290. doi:10.1186/s12879-026-12549-7.

19.

Kim

Kang

Choe

, et al. Which tissues are best for microbiological diagnosis in patients with pyogenic vertebral osteomyelitis undergoing needle biopsy. Clinical Microbiology and Infection. 2015;21(10):931-935. doi:10.1016/j.cmi.2015.06.021.

20.

McHenry

Easley

Locker

. Vertebral osteomyelitis: long-term outcome for 253 patients from 7 Cleveland-area hospitals. Clin Infect Dis. 2002;34(10):1342-1350. doi:10.1086/340102.

21.

Chen

Lin

Wei

, et al. Clinical characteristics and outcomes of culture-negative versus culture-positive osteomyelitis in children treated at a tertiary hospital in central Taiwan. Journal of Microbiology, Immunology and Infection. 2021;54(6):1061-1069. doi:10.1016/j.jmii.2020.08.005.

22.

Lee

Song

Son

Choi

. Pyogenic Spondyliodiscitis with Negative Culture Result: A Comparative Study of Clinical Course Versus Positive Culture Result. 2008;5.

23.

Lee

Kim

. Clinical outcomes in older patients aged over 75 years who underwent early surgical treatment for pyogenic vertebral osteomyelitis. J Clin Med. 2021;10(22):5451. doi:10.3390/jcm10225451.

24.

Yamada

Ieguchi

Takahashi

Nakamura

. Life Expectancy Is Poor in Patients with Diffuse Idiopathic Skeletal Hyperostosis-Related Pyogenic Vertebral Osteomyelitis. Spine Surg Relat Res. 2022;6(6):654-663. doi:10.22603/ssrr.2022-0021.

25.

Schindler

Walter

Reinhard

, et al. Midterm survival and risk factor analysis in patients with pyogenic vertebral osteomyelitis: a retrospective study of 155 cases. Front Surg. 2024;11:1357318. doi:10.3389/fsurg.2024.1357318.

26.

Vettivel

Bortz

Passias

Baker

. Pyogenic vertebral column osteomyelitis in adults: Analysis of risk factors for 30-day and 1-year mortality in a single center cohort study. Asian Spine J. 2019;13(4):608-614. doi:10.31616/asj.2018.0295.

27.

Chen

Jiang

Dai

. Surgical treatment of pyogenic vertebral osteomyelitis with spinal instrumentation. European Spine Journal. 2007;16(9):1307-1316. doi:10.1007/s00586-006-0251-4.

28.

Yagdiran

Otto-Lambertz

Lingscheid

, et al. Quality of life and mortality after surgical treatment for vertebral osteomyelitis (VO): a prospective study. European Spine Journal. 2021;30(6):1721-1731. doi:10.1007/s00586-020-06519-z.

29.

Thavarajasingam

Vemulapalli

Vishnu

, et al. Conservative versus early surgical treatment in the management of pyogenic spondylodiscitis: a systematic review and meta-analysis. Sci Rep. 2023;13(1):15647. doi:10.1038/s41598-023-41381-1.