If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Clinical Neuroscience, Karolinska Institutet, Solnavägen 1, 113 65, Stockholm, SwedenStockholm Spine Center, Löwenströmska Hospital, Karlavägen 108, 115 26, Upplands-Väsby, Sweden
Department of Surgical Sciences, Akademiska sjukhuset ing 70, 1 tr, 751 85 Uppsala, Uppsala, SwedenStockholm Spine Center, Löwenströmska Hospital, Karlavägen 108, 115 26, Upplands-Väsby, Sweden
Department of Surgical Sciences, Akademiska sjukhuset ing 70, 1 tr, 751 85 Uppsala, Uppsala, SwedenClinical Science, Intervention and Technology (CLINTEC), Karolinska Institutet, K54, Karolinska Universitetssjukhuset, 141 86 Stockholm, SwedenDepartment of Orthopedics and Hand surgery, Uppsala University Hospital, von Kraemers allé 4, 752 37, Uppsala, Sweden
A longer duration of patient follow-up arguably provides more reliable data on the long-term effects of a treatment. However, the collection of long-term follow-up data is resource demanding and often complicated by missing data and patients being lost to follow-up. In surgical fixation for cervical spine fractures, data are lacking on the evolution of patient reported outcome measures (PROMs) beyond 1-year of follow-up. We hypothesized that the PROMs would remain stable beyond the 1-year postoperative follow-up mark, regardless of the surgical approach.
PURPOSE
To assess the trends in the evolution of patient-reported outcome measures (PROMs) at 1, 2-, and 5-years following surgery in patients with traumatic cervical spine injuries.
STUDY DESIGN
Nation-wide observational study on prospectively collected data.
PATIENT SAMPLE
Individuals treated for subaxial cervical spine fractures with anterior, posterior, or combined anteroposterior approaches, between 2006 and 2016 were identified in the Swedish Spine Registry (Swespine).
OUTCOME MEASURES
PROMs consisting of EQ-5D-3Lindex and the Neck Disability Index (NDI) were considered.
METHODS
PROMs data were available for 292 patients at 1 and 2 years postoperatively. Five-years PROMs data were available for 142 of these patients. A simultaneous within-group (longitudinal) and between group (approach-dependent) analysis was performed using mixed ANOVA. The predictive ability of 1-year PROMs was subsequently assessed using linear regression.
RESULTS
Mixed ANOVA revealed that PROMs remained stable from 1- to 2-years as well as from 2- to 5-years postoperatively and were not significantly affected by the surgical approach (p<0.05). A strong correlation was found between 1-year and both 2- and 5-years PROMs (R>0.7; p<0.001). Linear regression confirmed the accuracy of 1-year PROMs in predicting both 2- and 5-years PROMs (p<0.001).
CONCLUSION
PROMs remained stable beyond 1-year of follow-up in patients treated with anterior, posterior, or combined anteroposterior surgeries for subaxial cervical spine fractures. The 1-year PROMs were strong predictors of PROMs measured at 2, and 5 years. The 1-year PROMs were sufficient to assess the outcomes of subaxial cervical fixation irrespective of the surgical approach.
Long-term patient follow-ups have been preferred in clinical research settings, as they provide more definite data on the effectiveness, durability, and reliability of a treatment [
]. However, the collection of consistent long-term follow-up data is not a simple task; in addition to the required resources and costs, it is often complicated by missing data and patients being lost to follow-up. In surgery for subaxial cervical spine fractures, a time point is yet to be defined at which the patient reported outcome measures (PROMs) stabilize.
In degenerative lumbar spine procedures, the outcomes tend to plateau after 1 year of follow-up [
]. However, similar studies after traumatic injuries in the subaxial cervical spine have not been performed. As suggested by the findings in lumbar procedures, we hypothesized that the PROMs would remain stable beyond 1-year of postoperative follow-up, regardless of the surgical approach.
Using the Swedish Spine Registry (Swespine) which provides PROMs at several time points, the aim of this study was to investigate whether 1-year PROMs were sufficient to estimate the outcome at 2 and 5 years after anterior, posterior, or combined anteroposterior (360°) fixation of subaxial cervical spine fractures.
Materials and methods
Study design
This observational study on prospectively collected data from 2006 to 2016 collected by the Swedish Spine Registry (Swespine) is in accordance with the STROBE guidelines (Supplementary File 1).
Sources and type of data
In 2016, Swespine had a coverage of 98% (46 of 47 spinal clinics), the completeness was approximately 75% (30%–90%), and the 1-year follow-up was performed in approximately 75%, covering all types of spinal surgical procedures [
]. Surgical approach, fracture type, any preoperative neurological impairment, operated levels, perioperative complications (thromboembolisms, urinary tract infections, bleeding/hematomas, wound infections, iatrogenic dural tears, implant related events, and vascular injuries), and reoperations for various reasons are registered in Swespine by the treating physician. Additionally, patients are asked to fill out questionnaires 1, 2, 5 and 10 years after surgery. The follow-up questionnaires cover patient satisfaction, ambulatory function, use of pain medication, and 90-day postoperative complications. The questionnaires also include the Neck Disability Index (NDI) [
]. The NDI ranges from 0 (no disability) to 100 (maximal disability) and evaluates neck-related disability experienced by subjects in their everyday life. The EQ-5D-3L measures HRQoL and the respondents are asked to rate 5 different health-related dimensions (mobility, self-care, usual activities, pain/discomfort, and anxiety/depression) based on three severity levels (no, moderate, or severe problems) [
]. The answers can be translated to an index ranging from -0.56 (worst imaginable health) to 1.00 (best imaginable health) based on the UK-TTO tariff.
Study population and eligibility criteria
Patients surgically treated for a subaxial cervical spine fracture between 2006 and 2016 were identified using the Swedish Spine Registry (Swespine). All patients with a minimum postoperative follow-up of 1 year, with available PROMs were eligible for screening regardless of the surgical approach: anterior, posterior, or combined anteroposterior (360°) fixation. (Fig. 1). Patients with available both 1- and 2-years PROMs were included in the analysis. Almost half (49%) of the patients also had PROMs at 5 years. In an attempt to evaluate attrition bias in our cohort, statistical analysis comparing patients with and without 5-years PROMs was performed. There were no statistically significant differences in terms of demographics, surgical approach, injury mechanisms, overall surgical complications, and 1-, or 2-year PROMs (p≥ 0.05).
Fig. 1Flowchart showing available and included patients.
The descriptive data is presented as numbers with percentages for categorical variables and medians (IQR) for continuous variables. For between-group statistical comparison of categorical variables the Chi-squared test was used while a one-way ANOVA was used for continuous variables. A mixed ANOVA was used to evaluate the effect resulting from the interaction of both follow-up time and surgical approach on the PROMs. Linear regression was used to assess the predictive ability of 1-year PROMs in predicting outcome at 2 and 5 years. All statistical analyses were performed in the SPSS statistics software, version 26, and p-values of < 0.05 were considered statistically significant.
Ethical considerations
Swespine uses the opt-out method, meaning that patient consent is not needed to be part of the quality registry, but answering the questionnaires is voluntary. This study was approved by the Regional Ethical Review Board of Stockholm (Dnr 2016/897–31/1).
Results
Patient characteristics
There were 292 patients that met the inclusion criteria. The entire cohort was divided into three groups depending on the surgical approach (Table 1): anterior (n = 123), posterior (n = 138), and combined anteroposterior (360°) fixation (n = 31). Age (p < 0.001), but not sex (p = 0.782), differed significantly between groups. Patients in the posterior fixation group were the oldest, followed by those in the anteroposterior, and then the anterior group. The types of vertebral fractures did not differ between groups (p = 0.311). However, patients with a greater number of injured vertebrae were more likely to be managed with a posterior fixation (p = 0.016). Despite that, there were no significant differences between approaches in terms of neurologic function on admission, as measured in Frankel grades (p = 0.343). The rate of complications did not significantly differ between the groups (p = 0.154). Complications included implant related ones (n = 17) - encompassing implant dislocation, loosening, extraction, or the need of refusion - wound infections (n = 3), vascular injuries (n = 2), iatrogenic dural tears (n = 1), and others which were not specified (n = 7). The length of hospital stay was the longest for patients who were treated with an anteroposterior fixation, followed by those treated with a posterior fixation, and shortest for those treated with anterior fixation (p = 0.014).
The median EQ-5D-3LIndex score remained stable at 0.727 through 1-, 2-, and 5-years of follow-up. The median NDI scores at 1-, 2-, and 5-years follow-up were 20.0, 20.0, and 18.0, respectively (Table 2).
Table 2Median EQ-5D-3LIndex and NDI scores (IQR) in the overall cohort as well as depending on the fixation approach
Overall cohort
Anterior fixation
Posterior fixation
Anteroposterior (360) fixation
At 1-year FU
n1=292 / n2=284
n1=123 / n2=119
n1=138 / n2=136
n1=31 / n2=29
EQ-5D-3LIndex
0.727 (0.28)
0.760 (0.18)
0.727 (0.25)
0.725 (0.54)
NDI
20.0 (30.0)
16.0 (30.0)
21.0 (31.5)
28.0 (35.0)
At 2-years FU
n1=292 / n2=284
n1=123 / n2=119
n1=138 / n2=136
n1=31 / n2=29
EQ-5D-3LIndex
0.727 (0.21)
0.760 (0.23)
0.726 (0.28)
0.725 (0.60)
NDI
20.0 (31.5)
18.0 (30.0)
22.0 (32.0)
20.0 (35.0)
At 5-years FU
n1=142 / n2=137
n1=75 / n2=72
n1=55 / n2=54
n1=12 / n2=11
EQ-5D-3LIndex
0.727 (0.27)
0.760 (0.34)
0.725 (0.28)
0.743 (0.83)
NDI
18.0 (32.0)
15.0 (28.0)
25.0 (32.0)
16.0 (26.0)
n1 = sample of patients with available EQ-5D-3Lindex, n2 = sample of patients with available NDI.
No significant changes in the postoperative PROMs were observed beyond the first year of follow-up on the longitudinal within-group analysis (Table 3). The PROMs remained stable through years 1, 2, and 5 postoperatively (Supplementary file 2; Figures A and B). As shown by the between-group analysis, surgical approach did not significantly affect PROMs (Table 3). However, between 2- and 5-years postoperatively, a marginally significant difference in EQ-5D-3Lindex values was seen among the groups. During that time, patients who were treated with an anterior fixation performed better than the rest (p = 0.043). However, the difference between groups failed to reach the minimal clinically important difference (MCID) of 0.24 [
Assessment of the minimum clinically important difference in pain, disability, and quality of life after anterior cervical discectomy and fusion: clinical article.
Table 3Mixed-ANOVA showing the interaction between a two-level time factor in the within-group analysis and surgical approach (anterior / posterior / anteroposterior), as the between-group variable on changes in PROMs
Within-group
Between-group
Within- and between- group interaction
PROM
Time points
Sample size
pη2
p-value
pη2
p-value
pη2
p -value
EQ-5D-3LIndex
1y vs. 2y
123 / 138 / 31
< 0.001
0.947
0.014
0.125
0.007
0.349
2y vs. 5y
75 / 55 / 12
0.003
0.539
0.044
0.043
0.004
0.781
NDI
1y vs. 2y
119 / 136 / 29
0.001
0.528
0.010
0.244
0.011
0.217
2y vs. 5y
73 / 55 / 12
0.007
0.317
0.038
0.072
0.016
0.333
Bold represents significant p-values.
pη2 = partial Eta-squared; the proportion of variance explained by a given variable of the total variance remaining after accounting for variance explained by other variables in the model.
The between-group comparisons consider the changes in PROMs based on the fixation approach. The within- and between group interaction considers the changes in PROMs relating to the joined interaction between time and fixation approach.
According to the mixed ANOVA results, the interaction between both within- (time) and between-group (surgical approach) factors did not significantly affect the PROMs, meaning that overall, the PROM values evolved in a similar fashion between groups from 1- to 5-years postoperatively. After conduction of a similar analysis, we also found that the occurrence of postoperative complications did not have a significant impact on the evolution of EQ5D (p = 0.34; p = 0.14) or NDI scores (p = 0.61; p = 0.84; from 1 to 2, or 2 to 5 years postoperatively, respectively).
The correlation between 1-year and 2- or 5-year postoperative PROMs were all strong (R > 0.7 (Fig. 2). These results were true regardless of the surgical approach.
Fig. 2Scatterplot matrix illustrating the correlation between 1-year and both 2- and 5-years PROMs.
On linear regression analysis, 1-year postoperative EQ-5D-3Lindex score was an independent predictor of both 2-year and 5-year EQ-5D-3Lindex (both p < 0.001). The resulting prediction model revealed unstandardized beta-coefficients (B) of 0.796 (SE: 0.04) and 0.855 (SE: 0.06), respectively. The results remained significant regardless of the group being analyzed.
Likewise, the 1-year postoperative NDI score was also an independent predictor of both 2-year and 5-year NDI (both p< 0.001). The resulting prediction model revealed unstandardized beta-coefficients (B) of 0.803 (SE: 0.04) and 0.768 (SE: 0.06), respectively. The results remained significant regardless of the group being analyzed. The correlation between 1-year and 2- or 5-year postoperative PROMs were all strong (R > 0.7) regardless of the surgical approach (Fig. 2).
Discussion
The aim of this study was to assess the evolution of PROMs at 1-, 2- and 5-years follow-up after surgery for traumatic subaxial cervical spine fractures. The results of the analysis revealed that PROMs remained stable beyond one year of follow-up, regardless of the surgical approach.
The study population scored substantially worse on EQ-5D-3Lindex and NDI than published data on the general population [
]. This indicates that a substantial portion of patients do not return to baseline levels even after appropriate management. In fact, previous studies performed on the same registry showed that 44% of patients treated for traumatic subaxial spine injuries used pain medication several years after surgery [
Outcome and health related quality of life after combined anteroposterior surgery vs anterior surgery alone in subaxial cervical spine fractures. Analysis of a national multicenter dataset.
Outcome and health related quality of life after combined anteroposterior surgery vs anterior surgery alone in subaxial cervical spine fractures. Analysis of a national multicenter dataset.
Stabilizing potential of anterior, posterior, and circumferential fixation for multilevel cervical arthrodesis: an in vitro human cadaveric study of the operative and adjacent segment kinematics.
] outcomes following surgery for traumatic subaxial cervical spine injuries. However, there are only a limited number of studies addressing neck disability and HRQoL outcomes in patients treated for subaxial cervical spine fractures. Available studies often have low response rates and fail to present a coherent picture due to methodological differences [
]. In one study assessing the outcomes of surgical treatment following traumatic cervical spine instability, the average NDI at last follow-up reached around 20 [
] the average NDI at last follow-up (around 5 years) was 23.8. In that study, patients presenting with spinal cord injury had significantly worse NDI. Koller et al, [
] excluded patients with spinal cord injury and found significantly better NDI scores (12.4) at 5.5 years follow-up.
In this study on subaxial spine fractures, no changes in the PROMs were seen beyond the first postoperative year. The findings are in line with published data for nontraumatic disease in the thoracic and lumbar spine, where PROMs are shown to stabilize within the first year [
]. In anterior cervical discectomy and fusion for nontraumatic disease, self-reported outcomes were stable for up to 10 years after initial improvements [
] studying outcomes after surgery for traumatic cervical instability, found improvements in the reported NDI between the preoperative and the 6- and 12-months scores, but no further improvements beyond 12 months. Self-reported pain outcomes after surgery in the cervical spine also follow the same pattern [
Hermansen A. Clinical and patient-reported outcomes after anterior cervical decompression and fusion surgery : a focus on functioning and daily life. 1443. [e-pub ahead of print May 22, 2015]. https://doi.org/10.3384/DISS.DIVA-117347. accessed February 2, 2023.
]. Our data validate these findings and show that the effect remains stable also when comparing 2- and 5-years after surgery. Interestingly, the occurrence of postoperative complications did not affect the evolution of PROMs in our material. Patients with or without complications maintained their one-year PROM scores at 2 and 5-years. Moreover, the between group comparison did not reveal any significant differences in scores between those with and without complications. This strengthens the hypothesis that postoperative changes affecting the HRQoL measures seem to occur within the first-year post-surgery.
Strengths and limitations
While previous studies have established the validity of short-term follow-up outcomes as proxies for long-term outcomes for various spinal procedures, the corresponding data was unavailable for the subaxial cervical spine. This study is, to the best of our knowledge, the first study to present evidence for the utility of one-year PROMs in predicting longer-term outcomes after surgery for subaxial traumatic injuries. The large sample size with repeated postoperative follow-ups, the prospectively collected data, the nationwide character of the study, as well as the use of validated outcome questionnaires constitute the strengths of this study.
The study suffers from limitations due to the moderate loss to follow-up at 1 and 2 years and the more accentuated loss at 5 years postoperatively, which may induce a risk of attrition bias. However, previous studies on Swedish, Norwegian, and Danish spine registries have shown that the nonrespondents did not significantly influence the study outcomes [
Non-respondents do not bias outcome assessment after cervical spine surgery: a multicenter observational study from the Norwegian registry for spine surgery (NORspine).
]. Additionally, statistical analysis of our data revealed no differences between the patients that were lost to follow-up and those who were not, in terms of demographics, surgical approach, injury mechanisms, overall surgical complications, and 1-, or 2-year PROMs (p ≥ 0.05). Moreover, stratification based on neurological function on admission was not performed due to the small sample size of the 5-year cohort. However, this parameter did not significantly differ between surgical approaches, and had minimal to no impact on the findings (p= 0.343). Furthermore, there are limitations inherent to patient-reported data originating from voluntary surveys; there may be a selection bias between those who respond to questionnaires and those who opt out. The Swespine registry has a nationwide coverage including 46 of 47 spinal clinics in the country. The registry is directed towards all types of spinal surgical procedures, and in the context of fractures, amasses all kinds of spine injuries regardless of the degree of severity. In other words, there are no exclusion criteria precluding the registration of any patient group. Prospective surgical data is entered by the operating surgeon, while patients are asked to fill out questionnaires at the different follow-up times, regardless of discharge disposition and functional status. Patients may however opt out from inclusion in the registry at any time. The large coverage of the registry and its wide inclusiveness minimize the risk of selection bias and ensure the generalizability of the resulting findings. Finally, the small sample size associated with the anteroposterior group warrants a cautious interpretation of the associated results.
Conclusion
PROMs remained stable beyond 1-year of follow-up in patients treated with anterior, posterior, or combined anteroposterior surgeries for acute subaxial cervical spine fractures. The 1-year PROMs were strong predictors of PROMs measured at 2, and 5 years. Our findings suggest that the 1-year PROMs are sufficient to assess the outcomes of cervical fixation following subaxial cervical spine fractures irrespective of the surgical approach.
Availability of data
The datasets generated during the current study are available from the corresponding author on reasonable request.
Code availability
No specific software codes used.
Ethical Considerations
Swespine uses the opt-out method, meaning that patient consent is not needed to be part of the quality registry, but answering the questionnaires is voluntary. The Regional Ethical Review Board in Stockholm has authorized the study and the use of the data collection (Dnr 2016/897–31/1).
Acknowledgments
Paul Gerdhem was supported by Region Stockholm in a clinical research appointment, by CIMED, Karolinska Institutet, Uppsala University, and the Swedish Research Council (Dnr 2020-00943). Adrian Elmi-Terander was supported by Region Stockholm in a clinical research appointment. The funding sources had no role in the study design, analysis, or interpretation of the data, in the manuscript writing, or in the decision to submit the paper for publication. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. The authors have no conflicts of interest. The authors acknowledge all the patients and surgeons contributing data to the Swedish Spine Registry, and Carina Blom, registry secretary for assistance with database extraction.
Assessment of the minimum clinically important difference in pain, disability, and quality of life after anterior cervical discectomy and fusion: clinical article.
Outcome and health related quality of life after combined anteroposterior surgery vs anterior surgery alone in subaxial cervical spine fractures. Analysis of a national multicenter dataset.
Stabilizing potential of anterior, posterior, and circumferential fixation for multilevel cervical arthrodesis: an in vitro human cadaveric study of the operative and adjacent segment kinematics.
Hermansen A. Clinical and patient-reported outcomes after anterior cervical decompression and fusion surgery : a focus on functioning and daily life. 1443. [e-pub ahead of print May 22, 2015]. https://doi.org/10.3384/DISS.DIVA-117347. accessed February 2, 2023.
Non-respondents do not bias outcome assessment after cervical spine surgery: a multicenter observational study from the Norwegian registry for spine surgery (NORspine).
Author disclosures: VE: Nothing to disclose. AS: Nothing to disclose. EE: Nothing to disclose. AE: Grants: Stockholm County (D). PG: Nothing to disclose.
00175-4&id=gr1.jpg){kind=link}
00175-4&id=gr2.jpg){kind=link}