Abstract
Aim
Spinal infection is the most frequent complication of spine surgery. Its incidence varies between 1% and 14% in the literature, depending on various studied populations and surgical procedures. The aim of this study was to describe a consecutive 2706 case series.
Method
We analyzed a prospective cohort of 2706 patients operated for spine disease between 2013 and 2016 in a University Hospital. The infection rates, germs, time between surgery and infection and outcomes after surgical revision were assessed with a minimum follow-up of 7 months. We developed a mathematical model to analyze risk factors in this difficult-to-treat population.
Results
Among 2706 patient who underwent spinal surgery during the three-year study period, 106 developed a postoperative spine infection. Clinical indicators for infection were the sudden onset of local pain and swelling without fever after an initial pain-free interval. We observed a masculine predominance (68%); the median age was 56 years. The rate of infection was comprised between 0,3% (discal herniation surgery) to over 20% in posterior cervical instrumented surgery (acute cervical fractures), with a global rate of 4%. Polymicrobial infections with more than 3 germs were found in only 2 case, with 3 germs in 8 cases, 2 germs in 27 cases and 1 germ in 69 cases. Staphylococcus aureus, Propionibacterium acnes and Staphylococcus epidermidis were the three main germs identified (53, 36 and 22 respectively). Propionibacterium acnes was involved with a higher rate in instrumented surgery but also in 8% of conventional non-instrumented surgery, with a median relapse time of 24 days (12 days to 4 years). Staphylococcus aureus was involved at a higher rate in posterior non-instrumented surgery with a median relapse time of 18 days (8–66 days). The rate of infection per month was globally stable along the year except an increased rate in February-March. All patients with a suspicion of post-op infection were initially treated with wound/deep tissues revision within the first month after surgery and associated with implant removal after one-month post-op. Pejorative outcomes were associated with incomplete revision surgery, several surgeries and polymicrobial infection.
Conclusions
In this study, the rate of postoperative infection is comparable to the literature. In contrast, Propionibacterium incidence is high, especially for acute infections. This unexpected rate can be linked to technical improvements in culture detection but this should also lead us to further discuss the natural process of spine/disk colonization of this germ.