Network meta-analysis of antibiotic exposure and holiday duration, biopsy site, imaging guidance, and sampling approach on culture yield in vertebral osteomyelitis.

BACKGROUND CONTEXT: Microbiological confirmation in suspected native vertebral osteomyelitis/spondylodiscitis (NVO/SD) guides pathogen-directed therapy, yet culture yields vary and the impact of antecedent antibiotics is disputed.

PURPOSE: The primary aim of this study was to determine how different factors-such as prior antibiotic use, the length of any antibiotic-free interval before biopsy, biopsy site selection, sampling method, and imaging guidance-affect the chances of obtaining a positive culture in adults suspected of having native vertebral osteomyelitis or spondylodiscitis. The study also explored whether planning a second biopsy after an initial negative attempt could help improve diagnostic confirmation.

STUDY

DESIGN

Systematic review and meta-analysis.

PATIENT SAMPLE: 27 adult cohort studies were included in this meta-analysis. Each study involved patients being investigated for suspected native vertebral osteomyelitis/spondylodiscitis. A comparative analysis was done. Included studies were both retrospective and prospective and mostly focused on pyogenic cases and some with mycobacterial infections. We excluded pediatric and nonspinal cases.

OUTCOME MEASURES: The primary outcome assessed was the rate at which bacterial cultures yielded positive results from sampled material (Biopsy or Aspirate). This was reported across studies and based on laboratory microbiology results. No self- report or functional measures were analyzed.

METHODS

We synthesized comparative evidence from an investigator-assembled corpus supplemented by a reproducible MEDLINE (via PubMed), Embase (via Ovid), and Cochrane CENTRAL search (2000-21 Sept 2025) and reference checking. Eligible adult cohort studies reported arm-level culture positivity for: prior antibiotics exposure, timing of an "antibiotic holiday," biopsy target, guidance modality, open/intraoperative versus percutaneous sampling, and a planned repeat-biopsy strategy. Analyses used MetaInsight (random-effects pairwise/network metanalysis). Effect size was risk ratio (RR) for positive culture (RR>1 favors the listed strategy). Overall, 27 unique studies contributed to ≥1 comparison.

RESULTS

Across 12 cohorts (n=967 biopsy procedures), prior antibiotic exposure was not associated with a statistically significant difference in culture yield (pooled RR 0.86, 95% CI 0.72-1.03). In the 3-arm network (biopsy performed while on antibiotics or within 48 hours of the last dose, 72 hours to 9 days after the last dose, and ≥10 days after the last dose) estimated yields were higher with longer antibiotic-free intervals; however, neither comparison reached statistical significance (72 h-9 days vs ≤48 h: RR 1.37, 95% CI 0.58-3.24; ≥10 days vs ≤48 hours: RR 2.53, 95% CI 0.81-7.92). For biopsy target (reference = paravertebral/epidural soft tissue collection/abscess), endplate+disc (RR 0.52, 95% CI 0.37-0.74) and bone/cortex (RR 0.34, 95% CI 0.13-0.87) were inferior; disc-only was similar (RR 0.92, 95% CI 0.66-1.27). Guidance modality showed no statistically significant difference (fluoroscopy vs CT RR 1.12, 95% CI 0.87-1.43). Open/intraoperative sampling was associated with statistically significant higher culture yield than percutaneous sampling (RR 1.65, 95% CI 1.34-2.03). A planned repeat biopsy after a negative first attempt showed suggested benefit however did not reach statistical significance (RR 1.37, 95% CI 0.81-2.34).

CONCLUSIONS

In suspected native vertebral osteomyelitis/spondylodiscitis, culture yield is generally lower after antecedent antibiotics and may improve with longer prebiopsy pauses. Targeting soft-tissue collections (or disc when collections are absent) maximizes yield; guidance modality is unlikely to be decisive; open sampling yields more positives in surgical candidates; and repeating biopsy after an initial negative may increase confirmation. These findings support practical, testable refinements to diagnostic pathways.