Early mortality in elderly patients with isolated C2 odontoid fracture treated by halo-vest immobilization, anterior spinal xation, or posterior spinal xation: A generalized propensity score-based analysis using a nationwide database

Purpose To compare in-hospital mortality of three procedures in the treatment of elderly patients with isolated C2 odontoid fracture: halo-vest immobilization, anterior spinal xation (ASF), and posterior spinal xation (PSF). Methods We extracted data for elderly patients who were admitted with C2 odontoid fracture and treated with at least one of the three procedures (halo-vest immobilization, ASF, or PSF) during hospitalization. We conducted a generalized propensity score-based matching weight analysis to compare in-hospital mortality among the three procedures. We further investigated independent risk factors for in-hospital death. Results The study involved 891 patients (halo-vest, n = 463; ASF, n = 74; and PSF, n = 354) with a mean age of 78 years. In-hospital death occurred in 45 (5.1%) patients. Treatment type was not signicantly associated with in-hospital mortality. Male sex (odds ratio, 2.98; 95% condence interval, 1.32–6.73; p = 0.009) and a Charlson comorbidity index of ≥ 3 (odds ratio, 9.18; 95% condence interval, 3.25–25.92; p < 0.001) were independent risk factors for in-hospital mortality. events with odontoid fracture are considered less suitable for surgical treatment.


Introduction
The incidence of C2 odontoid fracture in elderly patients has increased during the past two decades because of expansion of the geriatric population worldwide [1]. In elderly patients, C2 odontoid fracture mostly results from low-energy impacts such as falls [2]. With this increase in the number of elderly patients sustaining C2 odontoid fracture, the number of conservative treatments has increased by two to three times in the last decade [3]. Because most elderly patients have comorbidities and high baseline mortality, optimal management of odontoid fracture has long been a major concern.
Conservative treatment of odontoid fracture requires external immobilization using a rigid cervical collar or halo-vest, whereas surgical treatment involves anterior spinal xation (ASF) or posterior spinal xation (PSF) [4]. Although halo-vest is widely used for conservative treatment in patients whose fracture should be stabilized, several studies showed that halo-vest immobilization had worse survival outcomes than surgery [5]. The mortality rate of elderly people treated with halo-vest ranged from 16-42%; thus, halo-vest was considered less suitable for elderly patients than young patients [6]. In contrast, several other studies showed no association between the treatment type and clinical outcome [2,7]. This controversy makes it di cult for clinicians to select the optimal treatment for odontoid fracture.
We previously reported that most patients who sustained isolated C2 odontoid fracture were elderly and treated conservatively without halo-vest [8]. However, many clinicians may have di culty deciding which treatment is optimal when the fracture requires stabilization, especially in elderly patients who are likely to have several comorbidities. Thus, we conducted a generalized propensity score-based analysis to compare clinical outcomes among halo-vest immobilization, ASF, and PSF in elderly patients with isolated C2 odontoid fracture.

Data source
Inpatient data were extracted from the Japanese Diagnosis Procedure Combination database, a national database containing administrative claims and discharge data [9]. All academic hospitals are obliged to participate in the database, and more than 1,000 community hospitals voluntarily contribute to the database. Overall, the database provides data for approximately 50% of all acute-care inpatients in Japan. The database contains the following information: encrypted unique identi ers; age and sex; body weight and height; admission and discharge dates; diagnoses coded according to the International Patient selection From July 2010 to March 2017, we screened all patients who were admitted with C2 fracture (ICD-10 code: S12.1) and further identi ed odontoid fracture using Japanese disease codes. The inclusion criteria were age of ≥ 65 years and admission for treatment of odontoid fracture by at least one of three procedures (halo-vest immobilization, ASF, or PSF) during hospitalization. We excluded patients with multiple fractures (any fractures other than odontoid fractures), with severe consciousness disturbance at admission, who underwent combined surgery (both ASF and PSF), or who died within 2 days of admission. The patients who were treated with halo-vest before or after ASF or PSF were included in the surgery group.
The primary endpoint was overall in-hospital mortality. The secondary endpoints were at least one complication after admission and the post-treatment length of stay (PLOS). We identi ed complications after admission from the diagnoses recorded after admission using the following ICD-10 codes and de ned at least one complication as at least one of the following complications during hospitalization: sepsis (A40-A41), pulmonary embolism (I26), respiratory complications [pneumonia (J12-J18, J69), respiratory failure (J96), respiratory disorders (J95)], acute coronary syndrome (I21-I24), heart failure (I50), stroke (I60-I64), urinary tract infection (N30, N34, N36-N37, N39), and renal failure (N17-N19). PLOS was de ned as the length of stay from the day treated with halo-vest, ASF, or PSF to discharge (or death).

Statistical analysis
We used a propensity score-based method to account for differences in observed factors that might affect either the treatment assignment or outcome [16].
The propensity score was de ned as the probability of a patient undergoing halo-vest immobilization, ASF, or PSF based on the patient's baseline covariates. Covariate selection was prespeci ed by using both potential confounding factors and variables that can serve as proxies for unknown or unmeasured confounding variables. The propensity score was estimated using a multinomial logistic model with the procedure received as the dependent variable and the following baseline factors as independent variables [17]: age; sex; BMI category; smoking status; ambulance use; emergency admission; admission to intensive care unit before treatment; oxygenation therapy before treatment; use of urinary catheter; pre-existence of diabetes mellitus, hypertension, or chronic lung disease; history of cerebrovascular disease, cardiac disease, hepatic disease, dementia, or osteoporosis; at least one comorbidity; Japan Coma Scale score category; Barthel index; and CCI category on admission.
To balance the patients' baseline characteristics among the three procedures, a matching weight approach was applied [18]. Matching weights is recommended for comparing outcomes across multiple treatment groups when the covariates' overlaps are relatively limited, outcomes are rare, or exposure distributions are unequal [19]. Each patient was weighted by the inverse probability with the lower propensity score of the three procedures as the numerator [19]. The patients would receive each of the treatments among halo-vest immobilization, ASF, or PSF, allowing average treatment effects to be estimated.
Baseline covariate balance was checked after weighting, using a p-value of > 0.05 calculated by analysis of variance or the chi-squared test among the three treatments.
We compared the following outcomes among the three groups (halo-vest immobilization, ASF, and PSF) using analysis of variance and the chi-square test in the matching weighted cohort: overall in-hospital death, complications after admission, PLOS, total hospitalization costs, and Barthel index at discharge. We further conducted logistic regression analyses to estimate the odds ratios (ORs) and 95% con dence intervals (CIs) for overall in-hospital death and at least one complication after admission. We also conducted a linear regression analysis to estimate the regression coe cient and 95% CI for the PLOS. Moreover, we conducted a multivariable logistic regression analysis with adjustment for age, sex, BMI category, smoking status, and CCI category in the non-weighted and weighted cohorts to identify risk factors for in-hospital death. The following sensitivity analyses were undertaken to assess the robustness of the results. We combined the ASF and PSF groups as the surgery group and compared halo-vest immobilization with the surgery group using propensity score-matching analysis and matching weight analysis to balance the baseline variables.
Statistical analyses were performed using Stata/MP version 15 software (StataCorp, College Station, TX, USA). A two-tailed signi cance level of p < 0.05 and 95% CIs were used in the analyses.

Results
We nally included 891 patients (halo-vest, n = 463; ASF, n = 74; and PSF, n = 354) with isolated C2 odontoid fracture (Fig. 1). More than half of the patients underwent halo-vest immobilization. Table 1 shows the patients' baseline characteristics before and after matching weight. Overall, 366 patients (41%) were male, and their mean age was 78 ± 7.5 years. Despite signi cant differences in emergency admission, ambulance use, urinary catheter use on admission, and Data are presented as n (%), mean ± standard deviation, or median (interquartile range).
ASF, anterior spinal xation; PSF, posterior spinal xation; BMI, body mass index; ICU, intensive care unit; JCS, Japan Coma Scale; CCI, Charlson comorbidity index Table 2 shows the clinical outcomes among the treatments before and after weighting. Overall and 30-day in-hospital death occurred in 45 (5.1%) and 10 (1.1%) patients, respectively. The proportion of patients with at least one complication was 15%, and the most common complications after admission were respiratory complications (7.4%). The halo-vest group had a signi cantly longer PLOS than the ASF and PSF groups and signi cantly lower total costs than the PSF group. Univariable analysis in the weighted cohort showed that (i) in-hospital death was higher in the halo-vest group (6.6%) than the ASF (4.1%) and PSF (4.7%) groups with no signi cant difference, (ii) at least one complication was not signi cantly different among the treatments, and (iii) the PLOS was signi cantly longer in the halo-vest group than in the ASF or PSF group. Regarding complications after admission in the weighted cohort, the proportion of respiratory complications, cardiac events, and stroke were lower in the halo-vest group than in the ASF or PSF group.     Tables 1-6).

Discussion
We used a nationwide database and conducted a propensity score-based matching weight analysis to compare clinical outcomes of halo-vest immobilization, ASF, and PSF for elderly patients with isolated C2 odontoid fracture. In-hospital mortality and the development of at least one complication were not signi cantly different among the three procedures, whereas the PLOS was longer in the halo-vest group than in the surgery groups. Male sex and a higher CCI were independent risk factors for in-hospital mortality.
Halo-vest has been considered to be associated with higher mortality than surgical treatment in patients with C2 odontoid fracture, especially elderly patients [6,20]. Furthermore, in the latest meta-analysis, conservative treatment showed a trend toward higher mortality than surgical treatment [21]. The present study also showed relatively higher mortality in the halo-vest group than in the ASF or PSF group. However, halo-vest immobilization was not an independent risk factor for in-hospital death. One reason for higher mortality with conservative treatment may be selection bias due to limited settings of the target population.
Most previous studies may have included critically ill patients with C2 fracture who could not be treated surgically. Furthermore, the sample sizes were small, even in the meta-analysis [1,2,6]. According to our results, the difference in in-hospital mortality between halo-vest immobilization and surgery may be slight.
Halo-vest immobilization can be an option for C2 odontoid fracture if the patient cannot be treated surgically even when the fracture should be initially stabilized with surgery.
Respiratory complications are a cause of increased mortality of elderly patients who undergo halo-vest immobilization, and surgical treatment can reportedly decrease the incidence of pneumonia, cardiac arrest, and respiratory failure [5]. However, several studies showed no signi cant difference in complications between conservative and surgical treatment [5,7]. In the present study, complications including pneumonia, heart failure, and stroke were less common in the halo-vest group than in the ASF and PSF groups. Respiratory and cardiac complications can also occur as a result of surgery or general anesthesia, especially in elderly patients, who tend to have higher comorbidities and lower cardiac function. Because surgical treatment may have more complications than halo-vest immobilization in elderly patients, careful attention is needed to avoid adverse events after surgical treatment of C2 odontoid fracture.
Optimal treatment for odontoid fracture has been discussed over the years. Previous studies have revealed that surgical treatment is more effective than conservative treatment for inducing bony fusion [23]. However, brous fusion is a more acceptable outcome than morbidity or mortality associated with surgery [23]. Thus, osseous union is not a prerequisite to obtaining satisfactory clinical outcomes in elderly patients. Additionally, the association between bony fusion and mortality remains inconsistent if neurological complications are absent [23]. In the present study, male sex and a higher CCI were strongly associated with in-hospital death in patients with isolated C2 odontoid fracture. Among elderly patients, pre-existing comorbidities themselves can be associated with mortality [24]. A comprehensive decision is necessary regardless of treatment type for C2 odontoid fracture, especially in terms of age, sex, and comorbidities.
This study has several limitations. First, we could not obtain data on the type of fracture, severity of instability, and degree of dislocation from the database.
Second, despite using propensity score-based analysis, unmeasured confounding may not have been completely removed. The above-mentioned unavailable data may have been an unmeasured potential confounder affecting the indication for each treatment type. However, because more severe conditions make clinicians more likely to choose surgery, the surgery group likely had patients with more severe fractures. We conducted a sensitivity analyses, and the results were unchanged. Third, the database provides no data on outcomes after discharge. However, we assume that we covered most of the early adverse events because of the relatively long length of index hospitalization in Japan (median LOS for odontoid fracture is 31 days) [25]. Despite these limitations, we believe that our ndings will have a signi cant impact on future treatment.
In conclusion, our study showed that the treatment type (halo-vest immobilization, ASF, or PSF) was not signi cantly associated with in-hospital mortality.
Because elderly people are susceptible to higher comorbidity and baseline mortality rates, careful management may be required when these patients are male or have a higher CCI, regardless of treatment type for isolated C2 odontoid fracture.

Declarations
This article has not been published or submitted for publication elsewhere.
No bene ts in any form have been received or will be received from any commercial party related directly or indirectly to the subject of this article.
Funding: This work was supported by grants from the Ministry of Health, Labour and Welfare, Japan (19AA2007 and 20AA2005) and the Ministry of Education, Culture, Sports, Science and Technology, Japan (20H03907).
Con icts of interest/Competing interests: All authors have no competing interests to declare.
Ethics approval: The study design was approved by the Institutional Review Board of The University of Tokyo.
Consent to participate: The requirement for informed consent was waived because of the anonymous nature of the data.
Consent for publication: The requirement for informed consent was waived because of the anonymous nature of the data.
Availability of data and material: The datasets analyzed during the current study are not publicly available because of contracts with the hospitals providing data to the database.
Code availability: All statistical analyses were performed with STATA/MP version 15 software (StataCorp, College Station, TX, USA). The codes are available to interested researchers upon request to the corresponding author.
Authors' contributions: All authors contributed to the study conception and design. AH and YI designed and executed the experiments and wrote the manuscript. NM and YI were major contributors to the writing of the manuscript. NM and KM contributed to introducing the concepts of clinical epidemiology and helped to conduct the statistical analysis and write the manuscript. TM, ET, SI, SI, and KI contributed to introducing the concepts of orthopedic surgery and helped to write the manuscript. HM and KF contributed to construction of the database. HY and HC were the study supervisors and edited the manuscript. All authors reviewed and approved the nal manuscript. Figure 1 Flow chart of patients. We screened all patients who were admitted with C2 fracture (ICD-10 code: S12.1) and further identi ed odontoid fracture with the Japanese disease code. After excluding patients who were treated conservatively without halo-vest and who were aged <65 years, we further excluded patients who had severe consciousness disturbance, had multiple injuries, and died within 2 days after admission to exclude critically ill patients. Finally, 891 patients with isolated C2 odontoid fracture were identi ed (halo-vest, n = 463; ASF, n = 74; PSF, n = 354). ICD-10, International Classi cation of Diseases, 10th revision; ASF, anterior spinal xation; PSF, posterior spinal xation