Outcomes Associated with Skull-Base Fractures: A Retrospective Nationwide Analysis

Background Although much attention has been devoted to head trauma, little is known regarding the hospitalization burden of skull-base fractures (SBF) that occur in association with head-trauma.

Methods Using ICD-9 codes, we identified all trauma-related SBF in the 2013 Nationwide Readmissions Database and assessed trends/outcomes. A retrospective analysis was performed using standard descriptive techniques and multivariate logistic regression explored the odds-ratios of the likelihood of readmissions. All data were weighted to obtain US population estimates.

Results In total, 12,317 SBF cases were identified. The mean age of patients was 42.77 years (Standard Error (SE) ± 0.49). SBF was categorized as closed-SBF (91.10%) or as open-SBF (8.90%). Generally, SBF was more common among males (67.34%), who compared with females were more likely to have open-SBF (7.62 vs. 3.44%, p < 0.001). There were associated cervical spine fractures in 2.51%. Injury mechanisms implicated were falls (58.79%), struck-by/hit (10.03%), and motor-vehicle accidents (5.51%), p < 0.001; however, struck-by/hit versus falls were more likely to result in open-SBF (6.86 vs. 2.14%, p < 0.001). Cerebrospinal fluid (CSF) leaks occurred in 2.14%, among whom 10.37% underwent surgical repair—more commonly in patients who also had open-SBF versus closed-SBF (12.48 vs. 2.11%, p < 0.001)—while placements of lumbar drains occurred in 15.99%. Significant intracranial hemorrhage (ICH) was also common (62.81%, p < 0.001), and similar to open-SBF, more likely to undergo craniotomy/cranioplasty (p < 0.001). Complications of bacterial meningitis (BM) occurred in 0.37% overall and were commoner among patients presenting with CSF-leaks versus no CSF-leaks (5.40 vs. 0.26%, p < 0.001). There were no significant differences in the incidence of BM between patients who underwent surgical versus nonsurgical repair of CSF-leaks (<0.01 vs. 6.02%, p = 0.48), nor between patients who underwent placement of lumbar drains versus no lumbar drains for management of CSF-leaks (12.48 vs. 4.05%, p = 0.27). Overall mortality associated with SBF was 10.16% and was higher in open SBF versus closed SBF (41.12 vs. 8.09%, p < 0.001). There were no significant differences in mortality rates between patients presenting with versus without BM (<0.01 vs. 10.20%, p = 0.21). The mean individual charges were $74,578.28 (SE ± $2,163.56), and mean total hospital stay was 7.00 days (SE ± 0.21). In general, total charges and length of hospitalization were significantly elevated among patients in whom complications of BM occurred ([$272,885.20 [SE ± $107,564.70] and 16.66 days [SE ± 4.99] respectively, all p < 0.001). The most common disposition among survivors was routine-home discharge (50.11%). Among routine-home discharges, the overall 1-year follow-up readmission rate was 14.19%; however, 2.26% were readmitted within 30 days immediately following discharge. Generally, 30-day readmission rates in patients with CSF- versus non-CSF leaks (5.92 vs. 2.19%, p = 0.02), open versus closed SBF (0.45 vs. 2.39%, p = 0.01) and BM versus non-BM (18.69 vs. 2.21%, p < 0.001). Results of multivariate analysis revealed that patients in whom BM occurred were more likely for readmissions within 30 days post-index discharge (OR = 10.93; 95% CI = 2.23–53.43; p < 0.001).

Conclusion Skull-base fractures associated with head-trauma result in significant morbidity/mortality. Patients with open-SBF experience worse short-term outcomes. Complications of bacterial meningitis although rare, are more common in patients with cerebrospinal fluid leaks, and result in significantly increased hospital costs and readmissions among patients.