Bimodal mortality pattern in pyogenic spine infection identified from a nationwide population based cohort study

Previous studies on mortality in pyogenic spine infection have provided limited insight into temporal trends. This study aimed to delineate the sequential mortality pattern at fine time intervals and, based on these findings, to propose an optimal follow-up period for clinical care. This retrospective cohort study analysed data from the Korean Health Insurance Review and Assessment Service (HIRA) database. Adults (≥ 19 years) hospitalised for pyogenic spine infection between 2014 and 2018 were included. Sequential mortality was evaluated in 2-week intervals during one year after diagnosis, and the six-month period-where mortality remained elevated-was designated as the target interval. Independent predictors of mortality were identified via multivariable logistic regression with bootstrap validation. Among 12,538 eligible patients (mean age 64.7 years; 52.1% male), 339 (2.7%) died within six months. Independent risk factors included advanced age, male sex, rheumatologic disease, moderate-to-severe liver disease, end-stage renal disease, metastatic malignancy, sepsis, and the use of antibiotics targeting resistant organisms. Sequential analysis revealed a bimodal mortality pattern, with early deaths (≤ 8 weeks) related to infection severity and a late phase between 12 and 24 weeks driven by comorbidity burden-particularly liver and renal dysfunction. Mortality after pyogenic spinal infection does not decline uniformly over time. In high-risk groups, a secondary rise in mortality is observed during the 12-24-week phase, a critical period that may have been under-recognised in the literature. These findings emphasise the need for extended six-month follow-up and phase-specific management: intensive infection control early, followed by multidisciplinary support for comorbidity and systemic recovery thereafter.