Abstract
Study Design
Registry-based retrospective cohort study.
Objective
Identify risk factors for in-hospital mortality in patients with vertebral osteomyelitis (VO) treated conservatively or surgically using a national dataset.
Methods
Hospital data from 2019 to 2023 were analyzed using the database of the German Institute for the Hospital Remuneration System (InEK). Cases were identified by ICD-10 codes M46.20–M46.59. Diagnosis-related groups (DRG) were used to analyze demographics and treatment.
Results
A total of 49,951 VO cases were identified, with an overall in-hospital mortality of 6.7%. The incidence of VO in Germany from 2020 to 2023 was 46.2 cases per 100,000 population. VO in the cervicothoracic (OR 2.31), thoracolumbar (OR 2.88), lumbosacral (OR 2.60) and multifocal (OR 3.93) regions was associated with increased in-hospital mortality. Conservative treatment was linked to higher mortality in intraspinal abscesses (OR 1.9) and multifocal disease (OR 2.29). In unadjusted, treatment-stratified analyses, surgical treatment was associated with higher in-hospital mortality in patients aged ≥65 years (OR 1.47) and ≥80 years (OR 1.56). Regarding pathogens, the highest mortality ORs were observed in patients infected with Bacteroides fragilis and other anaerobic gram-negative bacteria (OR 2.81, conservative treatment) as well as Haemophilus influenzae (OR 2.00, surgical treatment). In the pooled risk-factor analysis, Pseudomonas aeruginosa carried the highest pathogen-associated in-hospital death (OR 2.01).
Conclusions
This nationwide analysis identifies factors associated for in-hospital mortality in VO. Conservative management, though most common, carries increased risk in high-risk settings such as epidural abscesses or multifocal disease. In unadjusted analyses, surgical treatment in elderly patients was associated with higher in-hospital mortality. Because treatment was not randomly allocated and surgical patients differ systematically in disease severity, this association is likely confounded by indication and cannot be read as a treatment effect. Findings underscore the need for individualized, risk-adapted strategies and for adjusted, longitudinal studies.
Keywords: vertebral osteomyelitis, in-hospital mortality, treatment options
Introduction
Pyogenic vertebral osteomyelitis (VO), including spondylitis, discitis, and vertebral osteomyelitis, represents a clinical challenge.1,2 This diagnosis requires hospitalization and is associated with a high risk of neurological and septic complications.3,4 The most common causative agent is Staphylococcus aureus, although other pathogens can play a role. 5
Despite being relatively rare, VO accounts 3% to 5% of all infectious bone diseases. 6 Epidemiological studies indicate a steadily increasing incidence of VO in Europe, posing a growing burden on healthcare systems.6,7 In Germany, VO prevalence is 14.4 per 100,000, with 60% of cases occurring in patients aged ≥70 years. 8
Delayed diagnosis is common, particularly in infections caused by low-virulence pathogens, and contributes to substantial morbidity and mortality.8,9 In Germany, VO is associated with an in-hospital mortality rate of 64.7 per 1,000 patients. 8 The condition is associated with high morbidity and an in-hospital mortality rate of 17.2% within the first year after diagnosis. 9 VO has a comparatively lower mortality rate than conditions such as major amputations.10-12
Effective treatment requires a careful balance between conservative and surgical approaches to minimize severe complications. The management of elderly patients is particularly challenging due to age-related frailty and comorbidities. 13 Secondary diagnoses such as liver cirrhosis, congestive heart failure, and kidney failure elevate the risk of poor outcomes in this population. 14 VO has long-term consequences, with patients reporting reduced quality of life even an average of seven years after treatment compared to the general population. 15
This study aims to identify risk factors for in-hospital mortality and compare risks between conservative and surgical treatment.
Material and Methods
In this registry-based retrospective cohort study, data from the German Diagnosis Related Groups (DRG) system were analyzed, focusing on the International Classification of Diseases, 10th Revision (ICD-10). 16 Data were provided by the Institute for the Hospital Remuneration System (InEK GmbH; Germany) and included in-hospital cases in Germany from January 1, 2019, to December 31, 2023.
Cases of VO were identified using ICD-10 codes M46.20–M46.59. Extracted data included total case numbers, secondary diagnoses for comorbidities and complications, and in-hospital mortality. Conservative and surgical therapies were identified using Operation and Procedure Classification (OPS) procedure codes 5-838.g to 5-838.y and 5-83b.00 to 5-83b.y. The PCCL (Patient Clinical Complexity Level) reflects overall clinical complexity based on comorbidities and complications, ranging from 0 (none) to 6 (very severe).
Since January 1, 2019, all German hospitals have been required to submit DRG data to the InEK system, which ensures comprehensive nationwide coverage. 17 Data extraction was performed using the InEK data browser (last accessed November 12, 2024). All data were anonymized.
Statistical Analysis: Secondary diagnoses are reported as absolute numbers and proportions. Univariate analysis was conducted to evaluate factors associated with in-hospital mortality. Because treatment allocation in routine care is driven by disease severity, no multivariable or propensity-based adjustment could be performed on the aggregated InEK export. All treatment-stratified comparisons are therefore unadjusted and susceptible to confounding by indication. Odds ratios (OR) with 95% confidence intervals (CI) were calculated, and associations were tested using the chi-square test. Statistical significance was defined as p < 0.05. The overall incidence of VO was calculated based on population figures from 2019-2023. The reported figure represents a cumulative multi-year rate over the observation period rather than an annualised incidence, and the broader ICD-10 range applied here (M46.20–M46.59) captures a wider case spectrum than some prior studies, which should be considered when comparing with earlier German estimates. To determine the age-specific incidence rates of VO in Germany for the year 2020-2023, we utilized population data from Destatis (Federal Statistical Office of Germany) alongside reported case numbers of VO across different age groups. 18
Ethical Considerations: Ethics approval and informed consent were not required due to the use of anonymized administrative data.
Results
In our study, 49,951 hospital admissions for VO were recorded, corresponding to an incidence of 46,2 cases per 100,000 population from 2020 to 2023. This represents a cumulative rate across the multi-year observation window. The corresponding mean annual incidence is substantially lower and more directly comparable with prior German estimates (e.g., 14,4 per 100,000). Because DRG records represent hospital cases rather than unique patients, transfers, staged procedures and readmissions may inflate this figure. Most patients were aged ≥65 years (72.3%), including 30.6% aged ≥80 years. The cohort was predominantly male (n=30,444; 61.0%). Lumbar VO was most common (56.4%), followed by thoracic involvement (18.3%) (Figure 1). Overall, 62.8% of cases were managed conservatively, while 37.1% underwent surgery (Figure 2).
Figure 1.
Height localization of VO
Figure 2.
Treatment distribution and age distribution with in-hospital mortality
The average length of hospital stay was 22.4 ± 19 days. According to PCCL classification, higher clinical complexity was significantly associated with increased mortality, with marked differences between conservative and surgical groups (Table 1).
Table 1.
PCCL Distribution by Treatment Type (N=47,757)
| PCCL | Conservative (n=29,210; 61.2%) | In-hospital mortalities (n=2,038; 61.3%) | Surgical (n=18,547; 38.8%) | In-hospital mortalities (n=1,288; 38.7%) |
|---|---|---|---|---|
| 0 | 8,449 (17.7%) | 108 (0.2) | 2,080 (4.4%) | 10 (0.1%) |
| 1 | 3,968 (8.3%) | 94 (3.6%) | 1,230 (2.6%) | 8 (4.1%) |
| 2 | 3,832 (8.0%) | 196 (4.1%) | 1,707 (3.6%) | 26 (0.1%) |
| 3 | 7,148 (15.0%) | 590 (1.2%) | 4,006 (8.4%) | 136 (0.3%) |
| 4 | 4,612 (9.7%) | 727 (1.5%) | 5,919 (12.4%) | 497 (1.0%) |
| 5 | 1,055 (2.2%) | 260 (0.5%) | 3,245 (6.8%) | 516 (1.1%) |
| 6 | 146(0.3%) | 63 (0.1%) | 360 (7.5%) | 95 (0.2%) |
In total, 48.8% (n=24,384) cases were treated in a public, 32.1% (n=16,030) in a non-profit hospital and 18.1% (n=9,036) in a private hospital sponsor. The majority of surgically treated cases (38%) were managed in a university hospital (≥800 beds). In primary care facilities (<300 beds) and general hospitals (≥300 beds), the majority of patients (70.2%) received conservative treatment (Figure 3).
Figure 3.
Proportion of conservative and surgical treatments across different hospital sizes (<300, ≥300, ≥600, and ≥800 beds)
In general Staphylococci were the most commonly detected bacteria, accounting for 36.5% of cases, followed by Streptococci and Enterococci (27.1%) and gram-negative bacteria (21.8%) (Figure 4).
Figure 4.
Distribution of detected bacterial groups in clinical samples
General Risk Factors for In-Hospital Mortality of VO Patients
The observed in-hospital mortality rate was 6.7% (n=3,326). VO in the cervicothoracic (OR 2.31), thoracolumbar (OR 2.88), lumbosacral (OR 2.60) and multifocal (OR 3.93) regions was associated with increased in-hospital mortality (Figure 5).
Figure 5.
In-hospital mortality OR of spinal region distribution
When analyzing the influence of pathogens, documented infection with Pseudomonas aeruginosa had the highest risk of in-hospital mortality (OR = 2.01), followed by Streptococci and Enterococci infections. (OR = 1.97).
Numerous comorbidities including pneumonia, COVID-19 infection, SIRS, septic shock, renal failure, respiratory insufficiency, atrial fibrillation, coronary artery disease, diabetes mellitus, and acute anemia were significantly associated with increased in-hospital mortality, whereas arterial hypertension and rheumatoid arthritis were not (Figure 6).
Figure 6.
In-hospital mortality OR for comorbidities and complications
Conservative Treatment
The majority of patients underwent conservative therapy (n=31,404; 62.8%). In this cohort the in-hospital mortality rate was 6.5% (n=2,038). (Table 2).
Table 2.
Number Cases and In-Hospital Mortality; N=49,951; In-Hospital Mortality n= 3,326
| | Conservative (n=31,404; 62.8%) | In-hospital mortalities (n=2,038; 61.3%) | Surgical (n=18,547; 37.2%) | In-hospital mortalities (n=1,288; 38.7%) |
|---|---|---|---|---|
| Age <65 | 9,010 (18.0%) | 119 (3.6%) | 4,823 (9.7%) | 138 (4.1%) |
| Age ≥65 | 25,874 (51.8%) | 1,450 (43.6%) | 10,244 (20.5%) | 1,185 (35.6%) |
| Age ≥ 80 | 11,685 (23.4%) | 897 (27.0%) | 3,614 (7.2%) | 622 (18.7%) |
| Cervical VO | 1,885 (3.8%) | 82 (2.5%) | 1,451 (2.9%) | 98 (2.9%) |
| Cervicothoracic VO | 229 (0.5%) | 7 (0.2%) | 117 (0.2%) | 7 (0.2%) |
| Thoracic VO | 5,518 (11.0%) | 364 (10.9%) | 3,606 (7.2%) | 210 (6.3%) |
| Thoracolumbar VO | 1,716 (3.4%) | 132 (4.0%) | 1,070 (2.1%) | 74 (2.2%) |
| Lumbar VO | 17,630 (35.3%) | 1,124 (34.0%) | 10,562 (21.1%) | 720 (21.6%) |
| Lumbosacral VO | 1,992 (4.0%) | 87 (2.6%) | 1,111 (2.2%) | 31 (0.9%) |
| Multifocal VO | 861 (1.7%) | 118 (3.5%) | 204 (0.4%) | 6 (0.2%) |
| Community-acquired pneumonia | 513 (1.0%) | 102 (3.1%) | 301 (0.6%) | 64 (1.9%) |
| Hospital-acquired pneumonia | 953 (1.9%) | 291 (8.7%) | 1,040 (2.1%) | 303 (9.1%) |
| Endocarditis | 715 (1.4%) | 90 (2.7%) | 457 (0.9%) | 97 (2.8%) |
| Coronary artery disease | 5,118 (10.2%) | 569 (17.1%) | 2,829 (5.7%) | 293 (8.8%) |
| Intraspinal abscess | 2,465 (4.9%) | 287 (8.6%) | 3,927 (7.9%) | 277 (8.3%) |
Conservative treatment at PCCL level 3 was associated with a significantly higher in-hospital mortality compared to surgical treatment (Table 1). Up to Level 2, all PCCL levels were associated with lower odds of in-hospital mortality, whereas from Level 4 onwards, higher PCCL levels served as risk factors for increased in-hospital mortality.
Intraspinal abscesses (OR 1.9) and multifocal disease (OR 2.29) were associated with significantly increased in-hospital mortality.
Surgical Treatment
In the surgical treatment group, the in-hospital mortality rate was 6.9% (n=1,288).
The mean length of hospital stay for the entire cohort was 28.8 ± 21 days. Patients in the surgical treatment group had a mean hospital stay of approximately 6 days longer than the overall cohort average.
A total of 7,191 (21.7%) procedures were performed using a microsurgical technique. The posterior approach was the most commonly employed access route (n=17,949), followed by the lateral and anterior approaches. Laminectomy accounted for 7,760 procedures. Among specific interventions, discectomy, excision of soft or bony tissue, and osteosynthesis using various techniques were frequently performed. Osteosynthesis with a screw-rod system, cages and Osteosynthesis with dorsal screws were among the most commonly used spinal stabilization methods (Table 3).
Table 3.
Surgical Techniques, N=18,547
| Surgery | OPS Code | n |
|---|---|---|
| Microsurgical technique | 5-984 | 7,191 |
| Anterior approach | 5-030.70, 5-030.71, 5-030.72 | 2,369 |
| Lateral approach | 5-031.5, 5-031.6, 5-031.7, 5-031.8, 5-031.9, 5-032.5, 5-032.6, 5-032.7, 5-032.8, 5-032.9, 5-032.a, 5-032.b | 3,910 |
| Posterior approach | 5-030.30, 5-030.31, 5-030.32 | 17,949 |
| Laminectomy | 5-030.60, 5-030.61, 5-030.62 | 7,760 |
| Decompression | 5-033.0 | 2,072 |
| Bone Decompression | 5-839.60, 5-839.61, 5-839.62, 5-839.63 | 8,423 |
| Discectomy | 5-831.0, 5-831.2 | 10,901 |
| Excision of soft/bone tissue | 5-832.0, 5-832.1, 5-832.2, 5-832.4, 5-832.6 | 10,829 |
| Osteosynthesis with ventral screws | 5-836.50, 5-836.51, 5-836.55 | 3,887 |
| Osteosynthesis with dorsal screws | 5-836.30, 5-836.31, 5-836.34, 5-836.35, 5-836.36, 5-836.37 | 4,681 |
| Osteosynthesis with a ventral screw- plate system | 5-83b.30, 5-83b.31, 5-83b.32 | 1,359 |
| Osteosynthesis with a dorsal screw- plate system | 5-83b.40, 5-83b.42, 5-83b.44 | 161 |
| Osteosynthesis dorsal and ventral | 5-836.40, 5-836.41, 5-836.42, 5-836.43, 5-836.44, 5-836.45, 5-836.46, | 2,746 |
| Osteosynthesis with a screw-rod system | 5-83b.50, 5-83b.51, 5-83b.52, 5-83b.53, 5-83b.54, 5-83b.55, 5-83b.56, 5-83b.57 | 16,983 |
| Osteosynthesis with a cage | 5-83b.70, 5-83b.71, 5-83b.72 | 8,367 |
| Vertebral body replacement | 5-837.00, 5-837.01, 5-837.02, 5-837.a0, 5-837.a1 | 2,688 |
| Revision | 5-839.5 | 2,345 |
| Reoperation | 5-983 | 948 |
PCCL levels 0,1 and 2 were identified as a theoretical protective factor against in-hospital mortality. In contrast, higher PCCL levels, starting from Level 3, were associated with a statistically significant increase in-hospital mortality (Table 1).
Age-specific analysis revealed a significantly increased in-hospital mortality rate for surgical treatment compared to conservative management in patients aged 65 years and older and in those aged 80 years and older (Table 1; Figure 7).
Figure 7.
Age-Related cases of conservative and surgical therapy. Percentages show the share of alive (black) and deceased (blue) patients within each treatment group
The highest in-hospital mortality OR were observed in patients with Bacteroides fragilis and other anaerobic gram-negative bacteria (conservative treatment) and Haemophilus influenzae (surgery group). Overall, Staphylococci and Streptococci also showed elevated ORs across both treatment types. Gram-negative pathogens generally indicated higher in-hospital mortality risk in the conservatively treated group (Figure 8).
Figure 8.
Mortality OR by pathogens
Discussion
This nationwide analysis of nearly 50,000 VO cases represents the largest population-based study to date comparing conservative and surgical treatment strategies. By integrating age, anatomical localization and pathogens, the study provides new insights into in-hospital mortality risk stratification.
The findings confirm that VO predominantly affects elderly patients and is associated with considerable in-hospital mortality. Transitional spinal regions were linked to increased in-hospital mortality risk. Conservative treatment was most common, subgroups particularly those with intraspinal abscesses and multifocal VO showed higher in-hospital mortality. Surgical treatment was associated with increased in-hospital mortality in patients aged ≥65 years, underscoring the need for careful risk-benefit evaluation in frail populations.
General Risk Factors for In-Hospital Mortality of VO Patients
In-hospital mortality rates for VO showed considerable variation across studies, with values between 0.9% and 14.2%.8,9,18-21 Our analysis revealed an in-hospital mortality rate of 6.7%, which is consistent with national data from Germany ranging from 6.5% to 14.2%.2,8,9,13,19 Previous studies may overestimate in-hospital mortality due to selection bias toward tertiary centers managing advanced disease. These centers are more likely to treat critically ill individuals with complex comorbidities or neurologic deficits, whereas smaller hospitals may care for patients with milder disease who can be managed conservatively. As a result, previously reported mortality rates may reflect a selection bias toward more severe cases and may not be generalizable to the full clinical spectrum of VO. Notably, a substantial proportion of patients in our cohort were treated in hospitals with fewer than 600 beds, supporting the external validity of our findings.
Length of hospital stay was comparable to prior German reports (22.4 vs. 22.3 days). 8
The distribution of affected spinal levels aligned with existing literature, with lumbar involvement most common. 20 A U.S.-based study further highlighted that the anterior/posterior surgical approach and thoracolumbar fusion were among the strongest predictors of negative outcomes. 22 Importantly, our study is the first to demonstrate significantly increased in-hospital mortality associated with transitional zones. We hypothesise that biomechanical factors in these transitional regions may contribute to worse outcomes. However, the dataset contains no information on instability, deformity, vertebral destruction, neurological status or abscess extent, so this remains a hypothesis the present data cannot test. This concept is reinforced by Spinal Instability Spondylodiscitis Score (SISS), which assigns additional weight to transitional zones. 23 Our findings underscore segmental stability as a potential determinant of survival, rather than solely mechanical alignment.
Pathogen-specific analysis revealed the highest in-hospital mortaliy in infections with Pseudomonas aeruginosa and Enterococcus, followerd by other anaerobic gram-negative bacteria. These findings align with Ziarko et al, who identified Pseudomonas aeruginosa as the most lethal pathogen, followed by Streptococci, Escherichia coli, and Staphylococcus aureus. 14
Numerous risk factors have been associated with the development and poor prognosis of VO. In general, these include advanced age, obesity, diabetes mellitus, substance abuse and immunodeficiency.19,24 In our study, several comorbid conditions present at admission were associated with an increased risk of in-hospital mortality, including pneumonia, COVID-19 infection, septic shock, acute renal failure, acute respiratory insufficiency, pleural effusion, atrial fibrillation, coronary artery disease, hypertension, diabetes mellitus, and acute hemorrhagic anemia. 25 These findings align with prior German and international studies emphasizing advanced age, systemic inflammation, and chronic disease burden as key prognostic factors.19,21,26
VO is managed conservatively or surgically, depending on clinical factors, with goals of infection control, neurological preservation, spinal stabilization, and adequate decompression. 24
Conservative Treatment
Historically, conservative management with intravenous antibiotic therapy has been the first-line treatment for VO in patients without neurological deficits, spinal instability, deformity, sepsis, or those with a high surgical risk.24,26,27 The standard antibiotic regimen typically extends over a period of six weeks. 28
Our analysis showed an OR of 0.7 for in-hospital mortality in patients aged ≥65 and ≥80 years treated conservatively. In these unadjusted data, advanced age was associated with lower OR of in-hospital mortality within the conservative group. This association is highly susceptible to selection bias, because elderly patients selected for conservative care may have less severe local disease, no neurological deficit or instability, or may be considered unfit for surgery.
However, intraspinal abscesses and multifocal disease were associated with increased in-hospital mortality. Epidural abscesses are recognized in the literature as a indication for surgical intervention. 29 A retrospective study from a tertiary referral center identified diabetes mellitus, women, and higher ASA scores as predictors of neurological deficits, suggesting that early surgical intervention may be beneficial in these subgroups. 30 This is supported by a survey among German spine surgeons, in which 38% agreed that intraspinal empyemas should always be surgically evacuated, regardless of spinal cord compression. 31 A U.S. study from Ohio found that epidural empyema, neurological deficits, in-hospital acquisition, and delayed diagnosis were associated with worse outcomes. 32 Similarly, a systematic review by Rutges et al emphasized that surgical treatment is often warranted in the presence of high-risk features. 33 A German–UK comparative study further reported higher mortality rates in conservatively treated patients, while those receiving surgery had shorter hospital stays. Notably, 38% of conservatively treated cases presented with epidural abscesses, and many had additional risk factors such as thoracic involvement and neurological symptoms. 34 In contrast, a Danish study found no significant mortality difference between surgical and conservative management when excluding patients with postoperative VO and normal CRP levels. Importantly, all patients with epidural abscesses in this cohort underwent surgical treatment. A Chinese study also found no clear superiority of surgical over conservative therapy in 112 non-surgical VO patients, reporting comparable outcomes in terms of hospitalization, complications, recurrence, and pathogen detection. 35
Taken together, these findings suggest that while surgery is often necessary in cases involving instability, neurological compromise, or abscess formation, conservative therapy may be appropriate in selected patients without high-risk features. Treatment decisions should therefore be individualized based on clinical presentation and risk stratification. Overall, the existing literature consistently indicates that the presence of an epidural abscess significantly increases the risk of in-hospital mortality when managed conservatively. These findings highlight the need for thorough risk assessment and suggest early surgical intervention should be considered in the presence of high-risk features such as epidural abscesses.
Surgical Treatment
Surgery is indicated for instability, neurological compromise, abscess formation, sepsis, or failure of conservative therapy.36-38 Following surgical intervention, patients receive intravenous antibiotic therapy equivalent in duration and intensity to that of conservatively treated individuals. 38 The data showed that surgical treatments were predominantly performed in university hospitals, while conservative management prevailed in smaller facilities. This suggests a potential link between hospital size and treatment strategy. Larger hospitals may perform more surgeries due to better resources and infrastructure, as well as a higher likelihood of receiving complex cases through referral.
To support objective treatment decisions, SISS was developed as a classification tool to evaluate spinal stability in patients with spontaneous spondylodiscitis, aiming to better guide indications for surgery. 23 Additionally, more recent scoring systems such as the Mortality of Spinal Infection- 20 (MSI-20) and Spinal Infection Treatment Evaluation (SITE) score have emerged to help differentiate between patients who might benefit from early surgical intervention and those who may be “too sick to operate”.39,40 An MSI-20 score of 11 or more, for instance, has been suggested to identify patients with prohibitively high surgical risk, whereas those with scores of 10 or lower might benefit from timely surgery. The SITE-Score incorporates radiological findings and patient-related comorbidities to aid treatment decisions and has shown promise in early evaluations.
VO located in transitional zones was associated with markedly increased mortality, supporting the hypothesis that stabilization may contribute to infection control and survival, beyond mechanical considerations. This highlights the critical role of spinal stability not only in mechanical support but potentially also in infection control-a concept supported by scoring systems like SISS, where instability strongly drives surgical indication. Given that mortality was the primary outcome in our study, this finding is hypothesis-generating: whether surgical stabilisation improves infection control or survival in anatomically unstable regions cannot be determined from these unadjusted administrative data and requires prospective, adjusted evaluation.
Despite these tools, surgical decision-making remains complex, particularly in elderly and frail patients. In our nationwide study, increased age was associated with higher in-hospital mortality in the surgical group. Patients aged ≥65 years had significantly higher odds of in-hospital mortality compared to their younger counterparts, while those aged ≥80 years showed an even more pronounced risk. This stands in contrast to the conservative treatment group, where advanced age did not correlate with increased in-hospital mortality. However, these findings must be interpreted with caution. There are no data about the long-term outcomes and radiologic results of conservative therapy, as well as on conversion rates to surgery. Without this information, it remains unclear how durable or effective conservative treatment truly is in the long run.
Similarly, another German study including 180 patients with spinal epidural abscesses found that patients ≥80 years had significantly higher in-hospital mortality compared to younger groups, although no significant differences in 30-day readmission or 90-day mortality were observed. 41 This emphasized that surgery for VO is often major and complex. Candidates for surgical treatment frequently present with more advanced disease, which in itself contributes to increased risk. This raises the question of whether frail elderly patients truly benefit from surgery, especially given the already high mortality associated with VO. 42
Nonetheless, growing evidence suggests that delayed surgical intervention can lead to worse outcomes, including increased mortality. 43 Therefore, early diagnosis, rapid initiation of targeted antibiotic therapy, and timely surgical decision-making are essential components of effective VO management. The timing of surgery plays a crucial role, with 10–14 days post-admission considered ideal in some studies; however, this aspect cannot be evaluated based on our dataset. 44 While multimorbid patients might benefit from surgical debridement and stabilization due to impaired immune defenses, they simultaneously carry higher perioperative risks. This therapeutic dilemma highlights the need for nuanced, case-by-case, multidisciplinary decisions rather than binary treatment strategies.
Strengths and Limitations
The main strength of this nationwide registry study is its comprehensive coverage of VO cases across Germany, including both tertiary referral centers and smaller hospitals. By capturing severe surgically treated cases as well as less severe, conservatively managed patients who are often underrepresented in clinical cohorts and trials, this study provides a broad and realistic overview of current treatment patterns and in-hospital outcomes.
Several limitations must be acknowledged. First, the analysis is based on aggregated administrative DRG data using ICD-10 and OPS codes. Therefore, the unit of observation is the hospital case rather than the individual patient, meaning that transfers, staged procedures, revisions, and readmissions may have been counted more than once and may have inflated case numbers and incidence estimates. In addition, administrative data do not provide detailed clinical information such as disease severity, radiological findings, laboratory parameters, microbiological confirmation, surgical timing, procedural invasiveness, or long-term follow-up.
Second, no multivariable or propensity-score adjustment was possible using the aggregated dataset. Consequently, all treatment-stratified mortality comparisons are unadjusted and should be interpreted as associations rather than treatment effects. In particular, confounding by indication is likely, as surgically treated patients may differ substantially from conservatively treated patients with regard to disease severity, comorbidity burden, neurological status, or spinal instability.
Third, German DRG data do not include a present-on-admission flag. As a result, coded severity markers such as sepsis, SIRS, septic shock, acute renal failure, respiratory insufficiency, or acute anaemia cannot be reliably distinguished as baseline comorbidities versus in-hospital complications. Accordingly, the term “risk factor” is used descriptively and should not be interpreted as indicating admission-level predictors.
Fourth, PCCL is a reimbursement-derived complexity measure based on coded diagnoses and complications during hospitalization. It may therefore partly reflect the clinical course itself rather than baseline prognosis. Similarly, pathogen attribution relies on ICD coding without access to culture results, timing of microbiological sampling, or differentiation of polymicrobial infections. Pathogen-specific findings should therefore be considered hypothesis-generating.
Fifth, the OPS-based surgical category combines heterogeneous procedures, ranging from biopsy or limited decompression to extensive instrumented reconstruction. It also does not capture surgical timing or whether initially conservative treatment was later converted to surgery. This limits the ability to draw conclusions about the relative effects of specific surgical strategies.
Finally, the endpoint was restricted to in-hospital mortality. Because surgically treated patients generally have longer hospital stays and therefore a longer in-hospital observation window, comparisons of in-hospital mortality between treatment groups may be biased. Moreover, vertebral osteomyelitis-related mortality, recurrence, and treatment failure frequently occur after discharge. Therefore, the present results should not be extrapolated to 30-day, 90-day, or one-year survival.
Overall, while this nationwide analysis offers a valuable population-level perspective on VO care in Germany, its findings should be interpreted as descriptive and hypothesis-generating rather than causal.
Conclusion
This large analysis provides nationwide insights into treatment strategies and mortality risks in VO. As the majority of patients received conservative therapy, early identification of high-risk constellations such as epidural abscesses or multifocal involvement is essential to avoid adverse outcomes. Higher in-hospital mortality was observed among surgically treated elderly and frail patients, but because these are unadjusted associations subject to confounding by indication, they cannot establish that surgery increases mortality; surgical indications in this group warrant careful, individualized evaluation. Elevated in-hospital mortality in transitional spinal zones is hypothesis-generating and may reflect unmeasured disease severity rather than instability per se. Integrating clinical, anatomical and microbiological factors into risk-adapted treatment decisions is crucial and highlights the need for prospective studies to optimize care pathways for vulnerable patient populations.
Supplemental Material
Supplemental material for Comparison of Hospital Mortality and Risk Factors in Conservative and Surgical Treatment of Pyogenic Vertebral Osteomyelitis: A Nationwide Analysis of 49,951 Cases in Germany by Melanie Ardelt, Josina Straub, Jonas Krückel, Adriana Palacio Giraldo, Markus Rupp, Markus Loibl, Volker Alt, Dietmar Dammerer and Siegmund Lang in Global Spine Journal.
Acknowledgments
The authors want to appreciate the contribution of NÖ Landesgesundheitsagentur, legal entity of University Hospitals in Lower Austria, for providing the organizational framework to conduct this research. The authors also would like to acknowledge support by Open Access Publishing Fund of Karl Landsteiner University of Health Sciences, Krems, Austria.
Author Contributions: Melanie Ardelt: Writing − original draft, Visualization, Software, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Josina Straub: Writing − original draft, Visualization, Methodology, Formal analysis. Jonas Krueckel: Writing − original draft, Software, Methodology, Formal analysis, Data curation. Adriana Palacio Giraldo: Writing − review & editing, Data curation. Markus Rupp: Writing − original draft, Data curation. Sebastian Siller: Writing − review & editing, Data curation. Markus Loibl: Writing − review & editing, Formal analy sis. Dietmar Dammerer: Writing − review & editing, For mal analysis. Volker Alt: Writing − review & editing, Formal analysis. Siegmund Lang: Writing − original draft, Formal analysis, Conceptualization.
Funding: The authors received no financial support for the research, authorship, and/or publication of this article.
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Supplemental Material: Supplemental material for this article is available online.
ORCID iDs
Melanie Ardelt https://orcid.org/0000-0002-5050-0316
Jonas Krückel https://orcid.org/0009-0002-3057-2600
Siegmund Lang https://orcid.org/0000-0003-0459-9092
References
- 1.Gouliouris T, Aliyu SH, Brown NM. Spondylodiscitis: update on diagnosis and management. Journal of Antimicrobial Chemotherapy. 2010;65(Supplement 3):11-24. doi: 10.1093/jac/dkq303. [DOI] [PubMed] [Google Scholar]
- 2.Rupp M, Walter N, Baertl S, Lang S, Lowenberg DW, Alt V. Terminology of bone and joint infection. Bone & Joint Research. 2021;10(11):742-743. doi: 10.1302/2046-3758.1011.BJR-2021-0371. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Zimmerli W. Vertebral Osteomyelitis. N Engl J Med. 2010;362(11):1022-1029. doi: 10.1056/NEJMcp0910753. [DOI] [PubMed] [Google Scholar]
- 4.Lang S, Rupp M, Hanses F, Neumann C, Loibl M, Alt V. Infektionen der Wirbelsäule: Pyogene Spondylodiszitis und implantatassoziierte vertebrale Osteomyelitis. Unfallchirurg. 2021;124(6):489-504. doi: 10.1007/s00113-021-01002-w. [DOI] [PubMed] [Google Scholar]
- 5.Bonfiglio M, Lange TA, Kim YM. THE CLASSIC: Pyogenic Vertebral Osteomyelitis: Disk Space Infections. Clinical Orthopaedics & Related Research. 2006;444:4-8. doi: 10.1097/01.blo.0000201172.64764.bb. [DOI] [PubMed] [Google Scholar]
- 6.Issa K, Diebo BG, Faloon M, et al. The Epidemiology of Vertebral Osteomyelitis in the United States From 1998 to 2013. Clinical Spine Surgery: A Spine Publication. 2018;31(2):E102-E108. doi: 10.1097/BSD.0000000000000597. [DOI] [PubMed] [Google Scholar]
- 7.Kramer A, Thavarajasingam SG, Neuhoff J, et al. Epidemiological trends of pyogenic spondylodiscitis in Germany: an EANS Spine Section Study. Sci Rep. 2023;13(1):20225. doi: 10.1038/s41598-023-47341-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Lang S, Walter N, Schindler M, et al. The Epidemiology of Spondylodiscitis in Germany: A Descriptive Report of Incidence Rates, Pathogens, In-Hospital Mortality, and Hospital Stays between 2010 and 2020. JCM. 2023;12(10):3373. doi: 10.3390/jcm12103373. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Schindler M, Walter N, Reinhard J, 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. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Szymski D, Walter N, Krull P, et al. Comparison of mortality rate and septic and aseptic revisions in total hip arthroplasties for osteoarthritis and femoral neck fracture: an analysis of the German Arthroplasty Registry. J Orthop Traumatol. 2023;24(1):29. doi: 10.1186/s10195-023-00711-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Schindler M, Baertl S, Walter N, et al. Retrospective analysis of mortality and quality of life after hip disarticulation or hemipelvectomy: a report on 15 patients. Arch Orthop Trauma Surg. 2023;143(8):4943-4949. doi: 10.1007/s00402-023-04783-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Schindler M, Huber L, Walter N, et al. Survival and risk factor analysis in patients with septic arthritis: a retrospective study of 192 cases. BMC Infect Dis. 2025;25(1):374. doi: 10.1186/s12879-024-10316-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Herren C, Von Der Hoeh NH, Zwingenberger S, et al. Spondylodiscitis in Geriatric Patients: What Are the Issues? Global Spine Journal. 2023;13(1_suppl):73S-84S. doi: 10.1177/21925682221121300. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Ziarko T, Walter N, Schindler M, Alt V, Rupp M, Lang S. Risk Factors for the In-Hospital Mortality in Pyogenic Vertebral Osteomyelitis: A Cross-Sectional Study on 9753 Patients. JCM. 2023;12(14):4805. doi: 10.3390/jcm12144805 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Lang S, Walter N, Froemming A, et al. Long-term patient-related quality of life outcomes and ICD-10 symptom rating (ISR) of patients with pyogenic vertebral osteomyelitis: What is the psychological impact of this life-threatening disease? Eur Spine J. 2023;32(5):1810-1817. doi: 10.1007/s00586-023-07616-5. [DOI] [PubMed] [Google Scholar]
- 16.International Statistical Classification of Diseases and Related Health Problems, 10th Revision, German Modification, Version; 2024. [Google Scholar]
- 17.The Institute for the Remuneration System in Hospitals (InEK) data browser.
- 18.Conan Y, Laurent E, Belin Y, et al. Large increase of vertebral osteomyelitis in France: a 2010–2019 cross-sectional study. Epidemiol Infect. 2021;149:e227. doi: 10.1017/S0950268821002181. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Gerstmeyer J, Avantaggio A, Pierre C, et al. In-Hospital mortality in Spondylodiscitis: Risk factors assessed through the National Inpatient Sample analysis. Journal of Clinical Neuroscience. 2025;135:111183. doi: 10.1016/j.jocn.2025.111183. [DOI] [PubMed] [Google Scholar]
- 20.Joerger AK, Albrecht C, Lange N, Meyer B, Wostrack M. In-Hospital Mortality from Spondylodiscitis: Insights from a Single-Center Retrospective Study. JCM. 2023;12(23):7228. doi: 10.3390/jcm12237228. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Heuer A, Strahl A, Viezens L, Koepke LG, Stangenberg M, Dreimann M. The Hamburg Spondylodiscitis Assessment Score (HSAS) for Immediate Evaluation of Mortality Risk on Hospital Admission. J Clin Med. 2022;11(3):660. doi: 10.3390/jcm11030660. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Kurowicki J, Changoor S, Coban D, et al. The Impact of Patient Characteristics on Outcomes of Surgically Managed Vertebral Osteomyelitis in the United States: Insights from a National Database Study. J Long Term Eff Med Implants. 2024;34(3):83-94. doi: 10.1615/JLongTermEffMedImplants.2023049402. [DOI] [PubMed] [Google Scholar]
- 23.Schömig F, Li Z, Perka L, et al. Georg schmorl prize of the German spine society (DWG) 2021: Spinal Instability Spondylodiscitis Score (SISS)—a novel classification system for spinal instability in spontaneous spondylodiscitis. Eur Spine J. 2022;31(5):1099-1106. doi: 10.1007/s00586-022-07157-3. [DOI] [PubMed] [Google Scholar]
- 24.Lener S, Hartmann S, Barbagallo GMV, Certo F, Thomé C, Tschugg A. Management of spinal infection: a review of the literature. Acta Neurochir. 2018;160(3):487-496. doi: 10.1007/s00701-018-3467-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Schoof B, Stangenberg M, Mende KC, Thiesen DM, Ntalos D, Dreimann M. Obesity in spontaneous spondylodiscitis: a relevant risk factor for severe disease courses. Sci Rep. 2020;10(1):21919. doi: 10.1038/s41598-020-79012-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Mavrogenis AF, Megaloikonomos PD, Igoumenou VG, et al. Spondylodiscitis revisited. EFORT Open Reviews. 2017;2(11):447-461. doi: 10.1302/2058-5241.2.160062. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Duarte RM, Vaccaro AR. Spinal infection: state of the art and management algorithm. Eur Spine J. 2013;22(12):2787-2799. doi: 10.1007/s00586-013-2850-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Bernard L, Dinh A, Ghout I, 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. [DOI] [PubMed] [Google Scholar]
- 29.Park J, Han S, Jeon Y, Hong JY. Spinal epidural abscess as predicting factor for the necessity of early surgical intervention in patients with pyogenic spondylitis. BMC Musculoskelet Disord. 2023;24(1):586. doi: 10.1186/s12891-023-06703-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Sircar K, Jung N, Kernich N, et al. Risk Factors for Neurologic Deficits in Patients With Spinal Epidural Abscess: An Analysis of One-Hundred-Forty Cases. Global Spine J. 2025;15(2):474-481. doi: 10.1177/21925682231194467. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Lawson McLean A, Senft C, Schwarz F. Management of Lumbar Pyogenic Spondylodiscitis in Germany: A Cross-Sectional Analysis of Spine Specialists. World Neurosurgery. 2023;173:e663-e668. doi: 10.1016/j.wneu.2023.02.128. [DOI] [PubMed] [Google Scholar]
- 32.McHenry MC, Easley KA, Locker GA. 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. [DOI] [PubMed] [Google Scholar]
- 33.Rutges JPHJ, Kempen DH, Van Dijk M, Oner FC. Outcome of conservative and surgical treatment of pyogenic spondylodiscitis: a systematic literature review. Eur Spine J. 2016;25(4):983-999. doi: 10.1007/s00586-015-4318-y. [DOI] [PubMed] [Google Scholar]
- 34.Neuhoff J, Kramer A, Thavarajasingam SG, et al. Comparing Conservative and Early Surgical Treatments for Pyogenic Spondylodiskitis: An International Propensity Score–Matched Retrospective Outcome Analysis. Neurosurgery. 2025;96(5):1008-1022. doi: 10.1227/neu.0000000000003223. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Xie B, du WJ, Xiong J, Zhong B, Ai Q, Yang D. Comparative analysis of different treatment strategies for septic spondylitis: a retrospective study of one hundred and twelve patients. International Orthopaedics (SICOT). 2024;48(9):2445-2454. doi: 10.1007/s00264-024-06247-9. [DOI] [PubMed] [Google Scholar]
- 36.OʼDaly BJ, Morris SF, OʼRourke SK. Long-term Functional Outcome in Pyogenic Spinal Infection. Spine. 2008;33(8):E246-E253. doi: 10.1097/BRS.0b013e31816b8872. [DOI] [PubMed] [Google Scholar]
- 37.Nickerson EK, Sinha R. Vertebral osteomyelitis in adults: an update. Br Med Bull. 2016;117(1):121-138. doi: 10.1093/bmb/ldw003. [DOI] [PubMed] [Google Scholar]
- 38.Gregori F, Grasso G, Iaiani G, Marotta N, Torregrossa F, Landi A. Treatment algorithm for spontaneous spinal infections: A review of the literature. J Craniovert Jun Spine. 2019;10(1):3-9. doi: 10.4103/jcvjs.JCVJS_115_18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Appalanaidu N, Shafafy R, Gee C, et al. Predicting the need for surgical intervention in patients with spondylodiscitis: the Brighton Spondylodiscitis Score (BSDS). Eur Spine J. 2019;28(4):751-761. doi: 10.1007/s00586-018-5775-x. [DOI] [PubMed] [Google Scholar]
- 40.Lener S, Wipplinger C, Lang A, Hartmann S, Abramovic A, Thomé C. A scoring system for the preoperative evaluation of prognosis in spinal infection: the MSI-20 score. The Spine Journal. 2022;22(5):827-834. doi: 10.1016/j.spinee.2021.12.015. [DOI] [PubMed] [Google Scholar]
- 41.Lenga P, Gülec G, Kiening K, Unterberg AW, Ishak B. Mortality, complication risks, and clinical outcomes after surgical treatment of spinal epidural abscess: a comparative analysis of patients aged 18-64 years, 65-79 years, and ≥ 80 years, with a 3-year follow-up. Neurosurg Rev. 2023;46(1):96. doi: 10.1007/s10143-023-02003-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Zadran S, Pedersen PH, Eiskjær S. Vertebral Osteomyelitis: A Mortality Analysis Comparing Surgical and Conservative Management. Global Spine J. 2020;10(4):456-463. doi: 10.1177/2192568219862213. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Thavarajasingam SG, Vemulapalli KV, Vishnu KS, 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. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Kramer A, Thavarajasingam SG, Neuhoff J, et al. Management of severe pyogenic spinal infections: the 2SICK study by the EANS spine section. The Spine Journal. 2025;25(5):876-885. doi: 10.1016/j.spinee.2024.12.018. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplemental material for Comparison of Hospital Mortality and Risk Factors in Conservative and Surgical Treatment of Pyogenic Vertebral Osteomyelitis: A Nationwide Analysis of 49,951 Cases in Germany by Melanie Ardelt, Josina Straub, Jonas Krückel, Adriana Palacio Giraldo, Markus Rupp, Markus Loibl, Volker Alt, Dietmar Dammerer and Siegmund Lang in Global Spine Journal.