A Nationwide Analysis of Geriatric Odontoid Fracture Incidence, Complications, Mortality, and Cost

Ram Alluri, MD; Gabriel Bouz, MD; Samantha Solaru, BS; Hyunwoo Kang, MD; Jeffrey Wang, MD; Raymond J. Hah, MD


Spine. 2021;46(2):131-137. 

In This Article


The primary objective of this study was to compare the national incidence of operative versus nonoperative management of geriatric odontoid fractures between 2003 and 2017. Over the 15-year time period, operative management significantly increased 3.7% per year, nearly doubling from 46% in 2003 to 86% in 2017. The use of halo application correspondingly decreased, while the use of cervical collar immobilization remained relatively constant over the study period. The reason for this paradigm shift is likely multifactorial and largely driven by some studies demonstrating lower mortality rates and potentially better functional outcomes with operative management, and other studies demonstrating poor tolerance of halo application in the geriatric population.

In a prospective study, Vaccaro et al[16] demonstrated improved functional outcomes and a trend towards decreased mortality in patients older than 65 years with an odontoid fracture managed operatively versus nonoperatively. Chapman et al[15] retrospectively demonstrated a significant decrease in short and long-term mortality favoring operative treatment in patients older than 65 years. Several other studies have also suggested that nonoperative management may be associated with higher mortality rates in this patient population,[11,13,22,25–27] and two of these studies are among the most highly cited odontoid fracture management publications.[11,26,28] In addition to decreased mortality, complications secondary to halo application in the geriatric population also plausibly contributed to the decrease in nonoperative management seen in our study. Several studies have demonstrated pulmonary complications ranging from pneumonia to respiratory distress requiring tracheostomy,[27,29,30] and gastrointestinal complications ranging from dysphagia to severe malnutrition requiring gastrostomy tube placement.[19,20,30,31]

The subanalysis of patients treated nonoperatively demonstrated a significant decrease in halo application and increase in cervical collar utilization. In geriatric patients with significant comorbidities, the benefit of operative intervention may not outweigh the increased perioperative risk. In these patients, nonoperative management may be ideal, but halo application may result in an unacceptable complication rate, thus the demonstrated shift towards cervical collar application. A recent systematic review by Waqar et al[32] comparing halo application to a hard collar concluded that both had similar failure rates but the high morbidity associated with halo application may not be justified, lending further evidence to support the trend noted.

The present study is not the first study to utilize a national database to analyze odontoid fractures,[4,5,33–35] but only two prior studies have specifically assessed national trends in operative versus nonoperative management.[5,35] Both Daniels et al[5] and Kukreja et al[35] utilized the NIS database from 2000 to 2010 and 2002 to 2011, respectively. Both studies demonstrated a significant decrease in nonoperative management over the respective time periods assessed, however, only Daniels et al[5] demonstrated an increase in operative intervention which was similarly seen in the current study.[35] However, the incidence of operative and nonoperative management seen in these studies is not directly comparable to the current study for several reasons. First, we included cervical collar immobilization which was included in neither of the aforementioned studies. Second, we excluded patients who did not have an assigned ICD procedure code (operative or nonoperative) associated a diagnosed odontoid fracture. In the studies by Daniels et al[5] and Kukreja et al,[35] over 50% of patients had unknown treatment, but these patients were included in the denominator when calculating the incidence of treatment. Third, we only included patients 65 to 90, while Daniels et al[5] and Kukreja et al[35] included all patients greater than 18 years. Lastly, the current study included data up to 2017.

In addition to assessing temporal trends in geriatric odontoid fracture management, a secondary objective was to compare comorbidities, mortality rates, hospital characteristics, and cost between operative and nonoperative management. In general, patients with at least one major comorbidity underwent nonoperative management, however, patients with rheumatoid arthritis and pulmonary comorbidities were more likely to receive operative management, similar to a database study by Dhall et al.[34] Inpatient mortality was significantly lower (3.6%) in patients receiving operative versus nonoperative management (5.9%). Lower inpatient mortality associated with operative management was likely secondary to selection bias as sicker patients with increased baseline likelihood of inpatient mortality were likely to be treated nonoperatively, as seen in prior studies.[5,7,34] Inpatient complications were generally higher in all operatively managed patients, particularly pulmonary, gastrointestinal, and renal complications, consistent with several prior studies.[7,12,34] Lastly, with regard to hospital characteristics, patients at teaching hospitals were more likely to receive operative management compared with non-teaching hospitals. Similar associations between hospital teaching status and increased operative management have been seen in studies outside of the spine literature.[36]

Cost of care per episode between 2003 and 2017 was $131,855 for operative treatment and $65,374 for nonoperative treatment (P < 0.001), similar to prior studies.[35] The increased cost of care with operative treatment can likely largely be attributed to operating room utilization, hardware implantation costs, and greater LOS. Estimating the total cost of care for geriatric odontoid fractures with our methodology would not be accurate as we excluded 78% of patients without an associated treatment procedure code, and this would result in a significant underestimation of the true national cost of geriatric odontoid fracture management.

The overall results of this study must be interpreted within limitations of the methodology. First, we used a national database which renders our study susceptible to bias and errors associated with use of administrative databases for clinical analysis that have been previously described.[24] Second, the NIS database is dependent on analyzing ICD codes to isolate procedures and diagnosis which lends our study susceptible to coding error. Third, the NIS database is an inpatient-only database and therefore patients who presented in the outpatient setting were not captured with our estimates. Fourth, we could not assess fracture pattern, which is a critical factor when deciding to elect for operative or nonoperative management; however, the distribution of odontoid fracture patterns likely remained relatively constant over the 15-year period studied, and therefore, our results demonstrating increased rates of operative management for geriatric odontoid fractures are unlikely to be due to change in fracture characteristics at a population level.

This study demonstrates a clear national paradigm shift in the management of geriatric odontoid fractures, wherein operative management nearly doubled from 46% in 2003 to 86% in 2017, halo application correspondingly decreased, and cervical collar utilization remained relatively constant. The described trends in management can conceivably be attributed to a growing body of literature describing decreased mortality rates with operative treatment and intolerance of halo application in a geriatric population. However, there remains no evidence-based consensus regarding management of odontoid fractures in this patient population. Additional prospective studies are needed to investigate whether this trend towards increased operative management results in decreased mortality, improved outcomes, and decreased complications for the geriatric population.