Clinical characteristics, microbiological landscape, and severity-stratified management of pyogenic spondylitis: a retrospective cohort study.

OBJECTIVE

To characterize the clinical, radiological, and microbiological features of pyogenic spondylitis (PS) and to describe outcomes observed within a severity-stratified real-world management pathway integrating antimicrobial therapy, stabilization, debridement, and fusion according to structural instability and neurological risk.

METHODS

We conducted a retrospective observational cohort study of 149 patients with PS treated at Ningxia Medical University General Hospital between January 2015 and December 2023. Patients were retrospectively grouped according to baseline structural severity, neurological status, and the management strategy actually used in routine care: Group A, antimicrobial therapy plus external immobilization (n = 48); Group B, posterior fixation and fusion plus antimicrobial therapy (n = 57); and Group C, posterior fixation with debridement and fusion via posterior and/or anterior approaches plus antimicrobial therapy (n = 44). Clinical characteristics, microbiological findings, inflammatory markers, VAS, ODI, Cobb angle, perioperative variables, complications, and endpoint-defined recovery were analyzed. Principal-component analysis and adjusted response models were prespecified as exploratory analyses.

RESULTS

The three groups represented distinct severity strata, ranging from structurally stable disease to destructive disease with neurological compromise. Final diagnostic adjudication identified 73 microbiologically confirmed cases, whereas the prespecified patient-level yield analysis was limited to a 66-case interpretable pretreatment subset, within which positivity was lower in Group C than in Groups A and B (40.0% vs 82.4% and 84.2%, respectively; P = 0.001). At the specimen level, 95 bacterial isolate records were summarized, and S. aureus was the predominant isolate (48/95). All groups showed improvement in inflammatory, pain, functional, and radiological measures over follow-up, although the tempo of recovery differed across severity-matched management strata. Exploratory adjusted analyses were directionally consistent with these descriptive patterns but should be interpreted cautiously.

CONCLUSIONS

In this retrospective real-world cohort, a severity-stratified management pathway for PS was feasible and showed coherent microbiological, inflammatory, functional, and radiological recovery patterns within each clinical stratum. These findings support pragmatic escalation according to structural instability and neurological risk, but they should not be interpreted as comparative evidence of treatment superiority.