The enigma of culture-negative spondylodiscitis: comparable outcomes despite diagnostic uncertainty - a systematic...

Systematic review.

Culture-negative(CN) spondylodiscitis presents a diagnostic dilemma, as empirical antimicrobial therapy is often initiated without microbiological confirmation. This systematic review and meta-analysis compared the prevalence, risk factors, empirical antibiotic regimens, and outcomes of CN versus culture-positive(CP) spondylodiscitis.

A comprehensive literature search identified studies reporting comparative outcomes between CN and CP cohorts. Data on prevalence, risk factors, antibiotic therapy, surgical rates, complications, and mortality were extracted and synthesised according to PRISMA guidelines.

Across 25 studies, the pooled CN prevalence was 43.2%(95%CI:33.8-52.6). CN status was consistently associated with prior antibiotic exposure, previous spinal/invasive procedures, bacteremia, diabetes, chronic kidney or liver disease, and intravenous drug use. CN cohorts had lower inflammatory markers(WBC, ESR, CRP) than CP. The mean duration of antibiotic therapy was similar between groups (CN:49-105 days; CP:56-112 days), with glycopeptides, β-lactams, fluoroquinolones, clindamycin, and tigecycline most frequently employed empirically. CN cohorts had significantly shorter hospital stays (SMD=-0.36;p = 0.002). Time to clinical recovery, cure rates, and recurrence(6-17%) showed no statistically significant difference compared to CP. Surgery was required in 20-40% of CN cases, with generally favourable outcomes and residual disability in 20-30%. Complication rates were not higher than CP. Mortality in CN cohorts was 1.8%, rising with comorbidity burden.

CN spondylodiscitis accounts for nearly half of all cases. Despite diagnostic uncertainty, no statistically significant difference in outcomes was detected compared to CP, with shorter hospitalisation and relatively low mortality, supporting the effectiveness of empiric antibiotic regimens guided by clinical and epidemiological context.