ABSTRACT
Background:
Postlaminoplasty kyphotic deformity (PKD) is a complication affecting roughly 20% of patients undergoing cervical laminoplasty. Identification of preoperative risk factors for PKD could allow surgeons to adapt treatment plans to reduce PKD.
Objective:
The aim of this study was to investigate the ability of the Charlson Comorbidity Index (CCI), 5-item Modified Frailty Index (5i-mFi), and Administrative Risk Analysis Index (RAI-A) to predict for the development of PKD in patients with cervical spondylotic myelopathy (CSM) undergoing laminoplasty.
Materials and Methods:
We retrospectively reviewed CSM patients who underwent laminoplasty at an academic tertiary care center between January 1, 2016, and January 30, 2022, and had a complete set of anterolateral cervical X-rays at 1-year follow-up. Angular kyphosis was defined as the loss of cervical lordosis by more than − 10° after surgery when measuring the difference between pre- and post-operative C2-7 Cobb angles. Regression and receiver operating characteristic (ROC) curve analysis were used to assess the ability of the frailty assessments to predict for PKD.
Results:
Seventy-six CMS patients were eligible, 11.8% of which developed PKD. The cohort consisted of 54 males and 22 females with a mean age of 59.5 years and body mass index of 29.2 kg/m2. No CCI, 59-mFi, or RAI-A frailty subgroup was associated with kyphotic development and ROC curve analysis showed that neither CCI (P = 0.81), 5i-mFi (P = 0.59), nor RAI-A (P = 0.63) predicted for PKD. None of these assessments were a superior prognosticator of PKD.
Conclusion:
CCI, 5i-mFi, and RAI-A frailty assessments were not associated with the development of PKD in CSM patients.
Keywords: Cervical laminoplasty, cervical myelopathy, frailty, postoperative kyphosis
INTRODUCTION
Cervical spondylotic myelopathy (CSM) is one of the most common degenerative diseases of the cervical spine, causing compression of the spinal cord and potentially leading to substantial cord or nerve damage if left untreated.[1] Cervical laminoplasty for the treatment of CSM is a surgical treatment option that is able to preserve a cervical range of motion while maintaining a lower risk of postoperative adjacent segment kyphosis,[1,2,3,4,5] and provide good long-term clinical outcomes.[5,6,7,8,9] Nevertheless, approximately 20% of patients undergoing laminoplasty develop postoperative loss of cervical lordosis or kyphotic deformity.[10] As cervical lordosis is associated with posterior migration of the cervical spinal cord, kyphotic deformity could lead to the draping of the spinal cord over the vertebral bodies, greatly affecting neurological function and causing adjacent segment disease, pain, and disability.[11,12] The identification of preoperative risk factors for the development of postlaminoplasty kyphotic deformity (PKD) would be useful for surgeons to guide preoperative surgical decision-making and modify treatment plans to prevent this complication.
Understanding a patient’s frailty index through several frailty scores has become increasingly utilized across surgical disciplines to predict postoperative complications and adverse events.[13,14,15] Since its inception as a clinical tool, different assessments have been developed to characterize frailty, taking into consideration gender, age, weight, resilience, medical comorbidities, and physical function, among other characteristics.[16,17] In patients who underwent spine surgery, multiple frailty indices have shown that increased frailty is associated with increased risk of mortality and morbidity, surgical complications, extended length of hospital stay, readmission rates, and other adverse events.[17,18] Frailty assessments vary considerably in length and complexity, impacting their clinical feasibility. As a result, and specifically in the context of spine surgery, short yet effective frailty assessments may have the most clinical utility due to their clinical feasibility.[15]
Despite the suggestive predictive ability of these indices across a wide range of spine surgeries, their ability to predict postoperative complications specifically within the context of correction of cervical spinal deformity has yet to be explored. In this study, we aimed to study the performance of widely used frailty indexes the Charlson Comorbidity Index (CCI),[19] the 5-item Modified Frailty Index (5i-mFi),[20] and the Administrative Risk Analysis Index (RAI-A)[21] in predicting postlaminoplasty PKD in patients diagnosed with CSM.
MATERIALS AND METHODS
Data source
After the institutional review board (IRB) approved the waiver of consent for this single institutional, retrospective study (IRB 202009133), we reviewed the electronic health records (EHRs) of all adult patients who underwent unilateral expansive open-door cervical laminoplasty (CPT codes 6305 and 63051) at the C4-C6 vertebral levels for the treatment of CSM at a single academic tertiary care center between January 1, 2016, and January 30, 2022, and who had a complete set of anterolateral cervical X-rays 1 year after surgery. We excluded patients who had <18 years of age as well as those who underwent laminoplasty for diagnoses other than CSM.
Demographic variables collected included age, gender, body mass index (BMI), and occupation. Radiographic measures included the preoperative and postoperative C2-7 Cobb angle. Outcome measures were defined as the development of kyphotic deformity. In this study, angular kyphosis was specifically defined as a loss of cervical lordosis of over −10° after surgery when measuring the difference between the preoperative and postoperative C2-7 Cobb angle[22,23,24] [Figure 1]. This study was performed following the Strengthening the Reporting of Observational Studies in Epidemiology guidelines.[25]
Figure 1.
Radiographic assessment of the cervical spine in patients with cervical spondylotic myelopathy. (Left) Preoperative X-rays in a 60-year old male (Right) Postoperative X-rays in a 60-year old male demonstrating kyphotic changes based on the C2-7 Cobb angle
Frailty index measurements
The CCI is based on 19 items. The CCI-weighted scoring system is seen in E-Table 1, and CCI severity is stratified as none = 0, mild = 1–2, moderate = 3–4, and severe ≥5.[26]
The 5i-mFi is based on 5 variables where each variable is given 1 point for a total score between 0–5 points. The scoring system is seen in E-Table 2, with frailty status stratified as nonfrail = 0, prefrail = 1, frail = 2, and severely frail ≥3.[27]
The RAI-A is based on the following 11 variables. The RAI-A weighted scoring system is seen in E-Table 3a-d. Frailty status based on RAI-A scoring is stratified as nonfrail ≤10, prefrail = 11–20, frail = 21–30, and severely frail ≥31.[27]
While the VASQIP/ACS-NSQIP datasets, from which the RAI-A and the 5i-mFi were originally created, have designated variables for overall functional status as well as functional status across various activities of daily living, these variables were not specifically recorded in our study population’s EHRs in a manner that directly aligned with their option labels on the frailty assessments. As a result, we decided on each patient’s functional status overall and across the various activities of daily living by thoroughly investigating progress notes and clinical summaries for each patient. We consolidated the information presented in these documents to best designate their functional status in accordance with the options presented in the RAI-A and 5i-mFi frailty assessments.
Data analysis
All data were first inputted into an electronic spreadsheet (Microsoft Excel, Microsoft Office, Redmond, Washington). Initial descriptive statistics were performed and the Wilcoxon signed-rank test was used to compare continuous variables (i.e., age, BMI, and overall frailty assessment scores) between the kyphotic and the nonkyphotic patient groups. Univariate and multivariate binary logistic regressions were performed for CCI, 5i-mFi, and RAI-A for the development of PKD, with outcomes summarized with the odds ratio (OR), 95% confidence interval (CI), and statistical significance. We also performed receiver operating characteristic (ROC) curve analysis with the area under the curve (AUC)/C-statistic calculations to examine the predictive abilities of the CCI, 5i-mFi, and RAI-A for the development of PKD. A DeLong test was used to compare the resulting C-statistic between CCI, 5i-mFi, and RAI-A. Statistical analysis was performed in IBM SPSS statistics version 28.0 (IBM Co., Armonk, NY, USA) and MedCalc for Windows, version 20.218 (MedCalc Software, Ostend, Belgium) with statistical significance defined as P < 0.05. All tables were produced and modified in Microsoft Word (Microsoft Word, Microsoft Office, Redmond, Washington).
RESULTS
Study population demographic and clinical information
We identified 76 patients that met our inclusion and exclusion criteria, of which 9 developed PKD (11.8%). The cohort included 54 males and 22 females with a mean age of 59.5 ± 12.6 years and BMI of 29.2 ± 4.5. The patient population was 73.7% White, 23.7% black, and 2.6% Asian. Within the overall study population, the mean CCI score was 2.9 ± 2.1, the 5i-mFi frailty score was 1.6 ± 1.0, and the RAI-A frailty score was 10.8 ± 8.1. Within the nonkyphotic patients (n = 67), the distribution of among the different frailty tiers was as follows: CCI – none (8, 11.9%), mild (22, 32.9%), moderate (23, 34.3%), severe (14, 20.9%); 5i-mFi – not frail (18, 26.9%), prefrail (26, 38.8%), frail (17, 25.4%), severely frail (6, 9.0%); RAI-A – not frail (47, 70.1%), prefrail (11, 16.4%), frail (6, 8.9%), severely frail (3, 4.5%). Within the kyphotic patients, the distribution of among the different frailty tiers was as follows: CCI – none (1, 11.1%), mild (3, 33.3%), moderate (3, 33.3%), severe (2, 22.2%); 5i-mFi – not frail (3, 33.3%), prefrail (3, 33.3%), frail (3, 33.3%), severely frail (0, 0.0%); RAI-A – not frail (5, 55.6%), prefrail (4, 44.4%), frail (0, 0.0%), severely frail (0, 0.0%).
There was no significant difference between the kyphotic and nonkyphotic patients in terms of age (P = 0.37), BMI (P = 0.75), overall CCI score (P = 0.79), overall 5i-mFi score (P = 0.69), or overall RAI-A score (P = 0.84). The kyphotic patients encompassed significantly more males than the nonkyphotic group (P = 0.01). All patient demographics and clinical information are summarized in Table 1.
Table 1.
Patient demographic and clinical information
| Variable | Value | P | ||
|---|---|---|---|---|
|
| ||||
| Overall (n=76) | No kyphotic deformity (n=67) | Kyphotic deformity (n=9) | ||
| Female sex | 22 (28.9) | 20 (29.9) | 2 (22.2) | 0.01* |
| Race | ||||
| White race | 56 (73.7) | 49 (73.1) | 7 (77.8) | 1.00 |
| Black or African American | 18 (23.7) | 16 (23.9) | 2 (22.2) | 1.00 |
| Asian | 2 (2.6) | 2 (3.0) | 0 | 1.00 |
| Mean age (years) | 59.47±12.59 | 58.97±12.57 | 63.22±12.83 | 0.37 |
| Mean BMI (kg/m2) | 29.18±4.51 | 29.12±4.54 | 29.64±4.52 | 0.75 |
| Occupation | ||||
| Currently employed | 25 (32.9) | 23 (34.3) | 2 (22.2) | 0.73 |
| Retired/on disability | 22 (28.9) | 19 (28.4) | 3 (33.3) | 0.39 |
| Not on file/not recorded | 29 (38.2) | 25 (37.3) | 4 (44.4) | 1.00 |
| CCI | 2.93±2.10 | 2.95±2.09 | 2.78±2.28 | 0.79 |
| None (CCI=0) | 9 (11.8) | 8 (11.9) | 1 (11.1) | 0.78 |
| Mild (CCI=1–2) | 25 (32.9) | 22 (32.9) | 3 (33.3) | 1.00 |
| Moderate (CCI=3–4) | 26 (34.2) | 23 (34.3) | 3 (33.3) | 1.00 |
| Severe (CCI≥5) | 16 (21.1) | 14 (20.9) | 2 (22.2) | 1.00 |
| 5i-mFi frailty total | 1.58±0.95 | 1.18±0.97 | 0.67±0.86 | 0.69 |
| Not frail (m5i-Fi=0) | 21 (27.6) | 18 (26.9) | 3 (33.3) | 0.99 |
| Prefrail (m5i-Fi=1) | 29 (38.2) | 26 (38.8) | 3 (33.3) | 1.00 |
| Frail (m5i-Fi=2) | 20 (26.3) | 17 (25.4) | 3 (33.3) | 0.92 |
| Severely frail (m5i-Fi ≥3) | 6 (7.9) | 6 (9.0) | 0 | 0.78 |
| RAI-A frailty total | 10.76±8.05 | 10.93±8.32 | 9.56±5.85 | 0.84 |
| Not frail (RAI-A ≤10) | 52 (68.4) | 47 (70.1) | 5 (55.6) | 0.62 |
| Prefrail (RAI-A=11–20) | 15 (19.7) | 11 (16.4) | 4 (44.4) | 0.12 |
| Frail (RAI-A=21–30) | 6 (7.9) | 6 (8.9) | 0 | 0.78 |
| Severely frail (RAI-A ≥31) | 3 (3.9) | 3 (4.5) | 0 | 1.00 |
| Preoperative comorbidities | ||||
| Functional dependance† | 11 (14.4) | 11 (16.4) | 0 | 0.42 |
| Diabetes mellitus | 20 (26.3) | 18 (26.9) | 2 (22.2) | 1.00 |
| COPD | 9 (11.8) | 9 (13.4) | 0 | 0.53 |
| Hypertension | 43 (56.6) | 37 (55.2) | 6 (66.6) | 0.77 |
| CHF | 5 (6.6) | 4 (6.0) | 1 (11.1) | 1.00 |
| Former smoker | 30 (39.5) | 26 (38.8) | 4 (44.4) | 1.00 |
| Current smoker | 12 (15.8) | 12 (17.9) | 0 | 0.37 |
| Cancer | 6 (7.9) | 6 (9.0.5) | 0 | 0.78 |
| Weight loss | 11 (14.5) | 9 (13.4) | 2 (22.2) | 0.84 |
| Poor appetite | 9 (11.8) | 7 (10.4) | 2 (22.2) | 0.63 |
| Renal failure | 3 (3.9) | 2 (3.0) | 1 (11.1) | 0.79 |
| SOB at rest | 10 (13.2) | 9 (13.4) | 1 (11.1) | 1.00 |
| Cognitive decline | 9 (11.8) | 7 (10.4) | 2 (22.2) | 0.63 |
| Mobility functional status | ||||
| Independent | 57 (75.0) | 49 (73.1) | 8 (88.8) | 0.54 |
| Supervised | 15 (19.7) | 14 (20.9) | 1 (11.1) | 0.81 |
| Limited assistance | 3 (3.9) | 3 (4.5) | 0 | 1.00 |
| Extensive assistance | 0 | 0 | 0 | N/A |
| Total dependence | 1 (1.3) | 1 (1.5) | 0 | 1.00 |
| Eating functional status | ||||
| Independent | 71 (93.4) | 62 (92.5) | 9 (100.0) | 0.90 |
| Supervised | 1 (1.3) | 1 (1.5) | 0 | 1.00 |
| Limited assistance | 3 (3.9) | 3 (4.5) | 0 | 1.00 |
| Extensive assistance | 1 (1.3) | 1 (1.5) | 0 | 1.00 |
| Total dependence | 0 | 0 | 0 | N/A |
| Toilet use functional status | ||||
| Independent | 68 (89.5) | 59 (88.1) | 9 (100) | 0.60 |
| Supervised | 4 (5.3) | 4 (6.0) | 0 | 1.00 |
| Limited assistance | 2 (2.6) | 2 (3.0) | 0 | 1.00 |
| Extensive assistance | 1 (1.3) | 1 (1.5) | 0 | 1.00 |
| Total dependence | 1 (1.3) | 1 (1.5) | 0 | 1.00 |
| Personal hygiene functional status | ||||
| Independent | 69 (90.8) | 60 (89.6) | 9 (100) | 0.69 |
| Supervised | 3 (3.9) | 3 (4.5) | 0 | 1.00 |
| Limited assistance | 2 (2.6) | 2 (3.0) | 0 | 1.00 |
| Extensive assistance | 2 (2.6) | 2 (3.0) | 0 | 1.00 |
| Total dependence | 0 | 0 | 0 | N/A |
*Statistically significant P<0.05, †Includes partial or total functional dependence. Values represent the number of patients (%) or mean±SD. CCI - Charlson Comorbidity Index; 5i-mFi - 5 item modified frailty index; RAI-A - Administrative risk analysis index; COPD - Chronic obstructive pulmonary disease; CHF - Congestive heart failure; SOB - Shortness of breath; N/A - Not available; SD - Standard deviation
Univariate analysis of Charlson Comorbidity Index, 5-item modified frailty index, and administrative risk analysis index on postlaminoplasty kyphotic deformity
We conducted a univariate binomial logistic regression using the overall CCI, 5i-mFi, and RAI-A scores as well as using each of their respective stratified scoring tiers to predict for the development of PKD. Neither CCI (OR 0.96, P = 0.81), 5i-mFi (OR 0.81, P = 0.60), nor RAI-A (OR 0.98, P = 0.63) were associated with the development of kyphosis. When the CCI was stratified into tiers, the severe group had the highest OR of 1.08 (95% CI, 0.15–5.10), but this was not significant (P = 0.93). Likewise, when the 5i-mFi score was stratified into frailty tiers, the frail group had the highest OR of 1.47 (95% CI, 0.29–6.25), but this was not significant (P = 0.61). Lastly, when the RAI-A score was stratified into frailty tiers, the prefrail group had the highest OR of 4.07 (95% CI, 0.89–17.94), but this also was not significantly associated with the development of kyphosis (P = 0.06). Output from the univariate analysis is summarized in Table 2.
Table 2.
Univariate binary logistic regression analysis of Charlson Comorbidity Index, 5-item modified frailty index, and administrative risk analysis index as predictors for postlaminoplasty kyphotic deformity
| Variable | Postlaminoplasty kyphotic deformity | P |
|---|---|---|
| CCI total | 0.96 (0.66–1.33) | 0.81 |
| None (CCI=0) | 0.00 (0.00–0.00) | 1.00 |
| Mild (CCI=1–2) | 0.96 (0.19–3.98) | 0.95 |
| Moderate (CCI=3–4) | 1.02 (0.20–4.27) | 0.98 |
| Severe (CCI≥5) | 1.08 (0.15–5.10) | 0.93 |
| 5i-mFi frailty total | 0.81 (0.35–1.69) | 0.60 |
| Not frail (m5i-Fi=0) | 1.36 (0.26–5.80) | 0.68 |
| Prefrail (m5i-Fi=1) | 0.79 (0.16–3.27) | 0.75 |
| Frail (m5i-Fi=2) | 1.47 (0.29–6.25) | 0.61 |
| Severely frail (m5i-Fi≥11) | 0.00 (0.00–0.00) | 1.00 |
| RAI-A frailty total | 0.98 (0.86–1.06) | 0.63 |
| Not frail (RAI-A≤10) | 0.53 (0.12–2.34) | 0.38 |
| Prefrail (RAI-A=11–20) | 4.07 (0.89–17.94) | 0.06 |
| Frail (RAI-A=21–30) | 0.00 (0.00–0.00) | 1.00 |
| Severely frail (RAI-A≥31) | 0.00 (0.00–0.00) | 1.00 |
No statistically significant values P<0.05. Values are presented as OR (95% CI). CI - Confidence interval; OR - Odds ratio; CCI - Charlson Comorbidity Index; 5i-mFi - 5 item modified frailty index; RAI-A - Administrative risk analysis index
Multivariate analysis of Charlson Comorbidity Index, 5-item modified frailty index, and administrative risk analysis index on postlaminoplasty kyphotic deformity
Our multivariate binomial logistic regression analysis adjusting for age, gender, and BMI showed similar results to our univariate analysis, with none of the overall CCI (P = 0.27), 5i-mFi (P = 0.22), or RAI-A (P = 0.42) scores being significantly associated with the development of PKD. Addition of the covariates increased the OR to 1.15 within the CCI mild group (95% CI, 0.21–5.57; P = 0.86) while slightly decreasing the OR to 1.10 within the 5i-mFi frail group (95% CI, 0.17–5.74; P = 0.92). Both frailty tiers remained insignificantly associated with kyphotic deformity. In the multivariate analysis, the 5i-mFi not frail group had the highest OR of 1.94 (95% CI, 0.34–10.36), but this also was not significantly associated with kyphotic deformity (P = 0.43). Adjusting for confounding factors decreased the OR to 3.62 within the RAI-A prefrail group (95% CI, 0.74–16.97), and this group was not a significant predictor for the development of kyphosis (P = 0.09). Output from the multivariate analysis is summarized in Table 3.
Table 3.
Multivariate binary logistic regression analysis of Charlson Comorbidity Index, 5-item modified frailty index, and risk analysis index administrative as predictors for postlaminoplasty kyphotic deformity
| Variable | Postlaminoplasty kyphotic deformity | P |
|---|---|---|
| CCI total | 0.72 (0.37–1.20) | 0.27 |
| None (CCI=0) | 0.00 (0.00–0.00) | 0.99 |
| Mild (CCI=1–2) | 1.15 (0.21–5.57) | 0.86 |
| Moderate (CCI=3–4) | 0.89 (0.16–4.14) | 0.89 |
| Severe (CCI ≥ 5) | 0.71 (0.08–4.24) | 0.73 |
| 5i-mFi frailty total | 0.57 (0.21–1.33) | 0.22 |
| Not frail (m5i-Fi=0) | 1.94 (0.34–10.36) | 0.43 |
| Prefrail (m5i-Fi=1) | 0.98 (0.19–4.39) | 0.98 |
| Frail (m5i-Fi=2) | 1.10 (0.17–5.74) | 0.92 |
| Severely frail (m5i-Fi ≥11) | 0.00 (0.00–0.00) | 0.99 |
| RAI-A frailty total | 0.96 (0.82–1.05) | 0.42 |
| Not frail (RAI-A ≤10) | 0.60 (0.13–2.87) | 0.50 |
| Prefrail (RAI-A=11–20) | 3.63 (0.74–16.97) | 0.09 |
| Frail (RAI-A=21–30) | 0.00 (0.00–0.00) | 0.99 |
| Severely frail (RAI-A ≥31) | 0.00 (0.00–0.00) | 0.99 |
Values are presented as OR (95% CI). The multivariate model was controlled for covariates: Age, gender, BMI. CCI - Charlson Comorbidity Index; 5i-mFi - 5 item modified frailty index; RAI-A - Administrative risk analysis index; CI - Confidence interval; OR - Odds ratio; BMI - Body mass index
Receiver operating characteristic and C-statistics analysis for Charlson Comorbidity Index, 5-item modified frailty index, and administrative risk analysis index
We conducted ROC curve analysis including AUC/C-statistic calculations to determine the relative predictive value of CCI, 5i-mFi, and RAI-A for the development of PKD [Figure 2]. While 5i-mFi had the highest discriminative performance (C-statistic = 0.54, 95% CI, 0.34–0.74) followed by CCI (C-statistic = 0.53, 95% CI, 0.20–0.75) and then RAI-A (C-statistic = 0.52, 95% CI, 0.31–0.71), none of these scoring systems were a significant predictor for kyphosis with P = 0.81, 0.56, and 0.63, respectively [Table 4]. Similarly, a DeLong test revealed that neither of the three frailty scoring systems were significantly better than another at predicting for the development of PKD.
Figure 2.
Receiver operating characteristic/area under the curve analysis for the relative predictive abilities of the (Red) Charlson Comorbidity Index, (Blue) 5-item Modified Frailty Index, and (Green) Administrative Risk Analysis Index on postlaminoplasty kyphotic deformity
Table 4.
Receiver operating characteristic/curve -statistics analysis for the relative predictive abilities of the Charlson Comorbidity Index, 5-item modified frailty index, and risk analysis index - administrative for postlaminoplasty kyphotic deformity
| Variable | C-statistic | P |
|---|---|---|
| CCI | 0.53 (0.30–0.75) | 0.81 |
| 5i-mFi | 0.54 (0.34–0.74) | 0.56 |
| RAI-A | 0.52 (0.31–0.71) | 0.63 |
Values are presented as C-statistic (95% CI). CCI - Charlson Comorbidity Index; 5i-mFi - 5-item modified frailty index; RAI-A - Administrative risk analysis index; CI - Confidence interval
DISCUSSION
This study was the first to our knowledge to investigate the ability of short, clinically feasible frailty assessments to predict for the development of kyphotic deformity following cervical laminoplasty in patients diagnosed with CSM. Our findings showed that the CCI, the 5i-mFi frailty index, and the RAI-A frailty index were not able to predict the development of PKD. In addition, none of the frailty score indices had a significantly stronger association with kyphosis than another. When patients were stratified into their relative comorbidity and frailty tiers within each frailty assessment modality, no frailty group was shown to be significantly associated with kyphosis. Similarly, no frailty group was significantly associated with kyphosis development when the binomial logistic regression was adjusted for confounding factors. As a result, our study highlights the need for investigation into preoperative risk factors for the development of PKD separate from the factors captured in standard frailty assessment measures.
The association between preoperative frailty measures and postoperative complications, adverse events, and increased risk of mortality and morbidity has historically been well established in the context of spine surgery.[17,18] This association has been consistently observed across various indications for spine surgery, even when utilizing shorter and more practical frailty assessments such as the CCI, 5i-mFi, and RAI-A. For example, CCI scores have since been shown to be associated with higher mortality in patients with cervical spine fractures,[28] clinical and functional outcomes of posterior lumbar interbody fusion surgery,[29] adverse events and 30-day complication rates in spine metastasis surgery,[30,31] and risk of readmission or reoperation in far lateral lumbar discectomy.[32] Similarly, the utility of the 5i-mFi has been explored across a wide range of spine surgeries, and it has been shown to be associated with increased complication and readmission rates, mortality, and morbidity in cervical, thoracic, and lumbar posterior fusion.[33,34] The 5i-mFi has also been shown to predict for severe adverse events in adult spine deformity surgery[35] and associate with 30-day adverse outcomes of anterior cervical discectomy and fusion.[36] In addition, the RAI-A has shown prognostic utility for operative outcomes such as 30-day mortality and morbidity in patients undergoing surgery for traumatic spine injury as well as with postoperative morbidity after spine deformity surgery in adults.[27,37] It was also associated with surgical complications and adverse events following brain tumor resection.[38]
While the CCI, 5i-mFi, and RAI-A have previously been shown to be associated with multiple postoperative complications and adverse events, this association is sensitive to the surgical population as well as the postoperative outcome measures under investigation. In patients undergoing posterior lumbar fusion, for example, the modified frailty index and CCI had similar or worse discriminative ability for perioperative adverse outcomes than simply using patient age.[39] Similarly, the CCI was shown to be an inaccurate differentiator for patient symptom improvement following minimally invasive lumbar spinal fusion.[40] This high sensitivity for the utility of frailty assessments can result in mixed findings regarding its predictive value even for similar spine surgeries carried out within different spinal regions. While the 5i-mFi was associated with 30-day adverse outcomes of anterior cervical discectomy and fusion,[36] it was not able to predict for hospital length of stay, 30-day adverse outcomes, or 30-day unplanned readmission in patients undergoing lumbar spinal decompression and fusion for spondylolisthesis.[41] These mixed findings regarding the utility of frailty assessments highlight the need to validate the predictive properties of different frailty assessments in the context of each specific surgical population and outcome measure of interest as opposed to applying frailty assessment broadly across surgical fields.
Of the frailty assessments analyzed in our study, the RAI was the most comprehensive with 14 different items covering five different domains.[15] When patients were stratified by frailty tier, this was the only assessment that isolated a group of patients, namely the prefrail group, which was trending towards having a significant associated with the development of kyphosis. Compared to the others, this risk assessment had a much greater emphasis on defining a patients’ preoperative independence level across multiple activities of daily living. While these data points are generally harder to retrospectively capture through a patient’s EHR, they may independently have more value in their ability to predict for the development of PKD following laminoplasty. As the items captured by a more granular frailty index specifically looking at activities of daily living may have more potential to be closely associated with kyphosis development, future investigation into the use of the multidisciplinary, 40-item ASD frailty index within this surgical population is needed.[15] In addition, future studies should investigate clinical factors not explicitly captured within frailty assessments, particularly those involving the biomechanical parameters of the spine. These parameters, such as radiographic X-ray measurements and paraspinal muscle fat infiltration, used either in isolation or in combination with frailty assessments may have better predictive potential for the development of PKD.
Limitations
One of the primary limitations of this study is the relatively low number of patients who underwent cervical laminoplasty and had all of the appropriate clinical and radiological follow-ups. As a result, we observed PKD in only nine patients (11.8%). While this PKD rate is in line with past studies,[10] it left us with a small number of patients from which we were able to conduct our analysis. Studies investigating the use of CCI, 5i-mFI, and the RAI-A frailty assessments in a greater number of CSM patients are necessary to determine the full utility of their ability to predict for PKD. Similarly, the laminoplasty surgeries as well as the EHR chart review were all performed at a single institution. Therefore, our results may not be generalized across different institutions or surgeons. Nevertheless, this is the first study to investigate the use of short, effective, and clinical feasible frailty assessments at predicting for the development of kyphotic deformity in CSM patients who undergo cervical laminoplasty and provide a foundation for future research into preoperative measures that can effectively predict PKD.
Another limitation is the granularity of the data available in the EHRs. For example, educated judgments on a patient’s functional status across activities of daily living had to be made using progress notes and clinical summaries to best stratify patients within the frailty assessments. If our assumptions did not align correctly with the VASQIP/ACS-NSQIP dataset variable designations, this would have an impact on our populations’ resulting 5i-mFI and RAI-A frailty scores. Future research looking specifically at the use of prospective frailty assessments, such as the RAI-C,[21] has the potential to overcome this limitation by allowing for the documentation of clinical variables of interest in real-time. Similarly, we observed a lack of validated consensus among various studies regarding score cutoffs for stratifying frailty and morbidity tiers using the CCI, 5i-mFI, and RAI-A frailty assessments.[27,42,43,44] Although we employed what we deemed to be the most extensive and widely used stratification tiers in previous frailty research,[27] our findings could be affected by alternative score cutoffs and differently defined frailty tiers.
CONCLUSION
While they may be utilized successfully for the prediction of adverse outcomes and postoperative complications across various other surgical indications, our study demonstrates that the CCI, 5i-mFi, and RAI-A frailty assessment tools are not significantly associated with the development of kyphotic deformity in CSM patients who undergo cervical laminoplasty. Our results suggest that the preoperative variables most predictive for the development of PKD are not effectively captured within these three frailty assessments. As a result, our future research aims to identify additional preoperative risk factors associated with the development of kyphosis and examine how these measures can be best incorporated into preoperative surgical decision-making to optimize patient outcomes.
Conflicts of interest
Dr. Camilo Molina is a consultant for Augmedics, Stryker, DePuy Synthes, Kuros Biosciences.
Supplementary Tables
E-Table 1.
Charlson comorbidity index scoring
| CCI variable | Score if present |
|---|---|
| Age (years) | |
| <50 | 0 |
| 50–59 | +1 |
| 60–69 | +2 |
| 70–79 | +3 |
| >80 | +4 |
| History of myocardial infarction | +1 |
| CHF | +1 |
| Peripheral vascular disease | +1 |
| History of CVA or TIA | +1 |
| Dementia | +1 |
| COPD | +1 |
| Connective tissue disease | +1 |
| Peptic ulcer disease | +1 |
| Liver disease | |
| Mild | +1 |
| Moderate to severe | +3 |
| Diabetes mellitus | |
| None | +0 |
| Uncomplicated | +1 |
| End-organ damage | +2 |
| Hemiplegia | +2 |
| Moderate to severe chronic kidney disease | +2 |
| History of solid tumor | |
| Localized | +2 |
| Metastatic | +6 |
| History of leukemia | +2 |
| History of lymphoma | +2 |
| AIDS | +6 |
| Total score (least frail-most frail) | 0–37 |
CCI - Charlson Comorbidity Index; COPD - Chronic obstructive pulmonary disease; CHF - Congestive heart failure; TIA - Transient ischemic attack; CVA - Cerebrovascular accident
E-Table 2.
Modified 5-item frailty index scoring
| Clinical variable | Score if present |
|---|---|
| Nonindependent functional status* | +1 |
| Diabetes mellitus with oral agents or insulin | +1 |
| COPD | +1 |
| Hypertension requiring medication | +1 |
| CHF | +1 |
| Total score (least frail-most frail) | 0–5 |
*Including both partial and complete dependence. COPD - Chronic obstructive pulmonary disease; CHF - Congestive heart failure
E-Table 3a.
Administrative risk analysis index scoring
| RAI-A variable | Score if present |
|---|---|
| Male Gender | +5 |
| Age and cancer sub score* | +2–20 |
| Unintentional weightless (>10 lbs) in the last 3 months | +5 |
| Renal failure | +6 |
| Chronis or CHF | +4 |
| Poor appetite | +4 |
| SOB | +8 |
| Nonindependent residential setting (skilled nursing facility, assisted living, or nursing home) | +8 |
| ADL and cognitive decline sub score† | +2–21 |
| RAI-A score total (least frail-most frail) | 4–81 |
*View supplemental Table 3b, †View supplemental Table 3c and supplemental Table 3d. RAI-A - Administrative risk analysis index; SOB - Shortness of breath; CHF - Congestive heart failure
E-Table 3b.
Administrative risk analysis index age and cancer subscoring
| Age | Score without cancer | Score with cancer* |
|---|---|---|
| <69 | 2 | 20 |
| 70–74 | 3 | 19 |
| 75–79 | 4 | 18 |
| 80–84 | 5 | 17 |
| 85–89 | 6 | 16 |
| 90–94 | 7 | 15 |
| 95–99 | 8 | 14 |
| 100+ | 9 | 13 |
| Age and cancer sub score total | 2–20 | |
*Excluding skin cancer, except for melanoma
E-Table 3c.
Administrative risk analysis index activities of daily living subscoring
| Mobility/locomotion | Eating | Toilet use | Personal hygiene |
|---|---|---|---|
| Independent +0 | Independent +0 | Independent +0 | Independent +0 |
| Supervised +1 | Supervised +1 | Supervised +1 | Supervised +1 |
| Limited assistance +2 | Limited assistance +2 | Limited assistance +2 | Limited assistance +2 |
| Significant assistance +3 | Significant assistance +3 | Significant assistance +3 | Significant assistance +3 |
| Totally dependent +4 | Totally dependent +4 | Totally dependent +4 | Totally dependent +4 |
| ADL score total | 0–16 | ||
ADL - Activities of daily living
E-Table 3d.
Administrative risk analysis index activities of daily living and cognitive decline sub scoring
| ADL score without cognitive decline in the past 3 months | ADL score with cognitive decline in the past 3 months |
|---|---|
| 0 | ADL score − 2 |
| 1, 2 | ADL score − 1 |
| 3, 4 | ADL score 0 |
| 5, 6, 7 | ADL score +1 |
| 8, 9 | ADL score +2 |
| 10, 11 | ADL score +3 |
| 12, 13 | ADL score +4 |
| 14, 15, 16 | ADL score +5 |
| ADL and cognitive decline subscore total | −2–21 |
ADL - Activities of daily living
Funding Statement
Nil.
REFERENCES
- 1.Alam I, Sharma R, Borkar SA, Goda R, Katiyar V, Kale SS. Factors predicting loss of cervical lordosis following cervical laminoplasty: A critical review. J Craniovertebr Junction Spine. 2020;11:163–8. doi: 10.4103/jcvjs.JCVJS_70_20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Sharma R, Borkar SA, Goda R, Kale SS. Which factors predict the loss of cervical lordosis following cervical laminoplasty?A review of various indices and their clinical implications. Surg Neurol Int. 2019;10:147. doi: 10.25259/SNI_339_2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Ratliff JK, Cooper PR. Cervical laminoplasty: A critical review. J Neurosurg. 2003;98:230–8. doi: 10.3171/spi.2003.98.3.0230. [DOI] [PubMed] [Google Scholar]
- 4.Matsunaga S, Sakou T, Nakanisi K. Analysis of the cervical spine alignment following laminoplasty and laminectomy. Spinal Cord. 1999;37:20–4. doi: 10.1038/sj.sc.3100749. [DOI] [PubMed] [Google Scholar]
- 5.Zhao H, Ren R, Ma W, Xu S, Peng L, Zhong Z, et al. Comparison of laminoplasty versus laminectomy for cervical spondylotic myelopathy: A systematic review and meta-analysis. Front Surg. 2021;8:790593. doi: 10.3389/fsurg.2021.790593. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Kawaguchi Y, Kanamori M, Ishihara H, Ohmori K, Nakamura H, Kimura T. Minimum 10-Year Followup After En Bloc Cervical Laminoplasty. Clinical Orthopaedics and Related Research®. 2003;411:129. doi: 10.1097/01.blo.0000069889.31220.62. [DOI] [PubMed] [Google Scholar]
- 7.van Geest S, de Vormer AM, Arts MP, Peul WC, Vleggeert-Lankamp CL. Long-term follow-up of clinical and radiological outcome after cervical laminectomy. Eur Spine J. 2015;24(Suppl 2):229–35. doi: 10.1007/s00586-013-3089-6. [DOI] [PubMed] [Google Scholar]
- 8.Seichi A, Takeshita K, Ohishi I, Kawaguchi H, Akune T, Anamizu Y, et al. Long-term results of double-door laminoplasty for cervical stenotic myelopathy. Spine (Phila Pa 1976) 2001;26:479–87. doi: 10.1097/00007632-200103010-00010. [DOI] [PubMed] [Google Scholar]
- 9.Nurboja B, Kachramanoglou C, Choi D. Cervical laminectomy versus laminoplasty: Is there a difference in outcome and postoperative pain? Neurosurgery. 2012;70:965–70. doi: 10.1227/NEU.0b013e31823cf16b. [DOI] [PubMed] [Google Scholar]
- 10.Pettersson SD, Skrzypkowska P, Ali S, Szmuda T, Krakowiak M, Počivavšek T, et al. Predictors for cervical kyphotic deformity following laminoplasty: A systematic review and meta-analysis. J Neurosurg Spine. 2023;38:4–13. doi: 10.3171/2022.4.SPINE22182. [DOI] [PubMed] [Google Scholar]
- 11.Scheer JK, Tang JA, Smith JS, Acosta FL, Jr, Protopsaltis TS, Blondel B, et al. Cervical spine alignment, sagittal deformity, and clinical implications: A review. J Neurosurg Spine. 2013;19:141–59. doi: 10.3171/2013.4.SPINE12838. [DOI] [PubMed] [Google Scholar]
- 12.Tang JA, Scheer JK, Smith JS, Deviren V, Bess S, Hart RA, et al. The impact of standing regional cervical sagittal alignment on outcomes in posterior cervical fusion surgery. Neurosurgery. 2012;71:662–9. doi: 10.1227/NEU.0b013e31826100c9. [DOI] [PubMed] [Google Scholar]
- 13.Lin HS, Watts JN, Peel NM, Hubbard RE. Frailty and post-operative outcomes in older surgical patients: A systematic review. BMC Geriatr. 2016;16:157. doi: 10.1186/s12877-016-0329-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Panayi AC, Orkaby AR, Sakthivel D, Endo Y, Varon D, Roh D, et al. Impact of frailty on outcomes in surgical patients: A systematic review and meta-analysis. Am J Surg. 2019;218:393–400. doi: 10.1016/j.amjsurg.2018.11.020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Kitamura K, van Hooff M, Jacobs W, Watanabe K, de Kleuver M. Which frailty scales for patients with adult spinal deformity are feasible and adequate?A systematic review. Spine J. 2022;22:1191–204. doi: 10.1016/j.spinee.2022.01.017. [DOI] [PubMed] [Google Scholar]
- 16.Dent E, Kowal P, Hoogendijk EO. Frailty measurement in research and clinical practice: A review. Eur J Intern Med. 2016;31:3–10. doi: 10.1016/j.ejim.2016.03.007. [DOI] [PubMed] [Google Scholar]
- 17.Veronesi F, Borsari V, Martini L, Visani A, Gasbarrini A, Brodano GB, et al. The impact of frailty on spine surgery: Systematic review on 10 years clinical studies. Aging Dis. 2021;12:625–45. doi: 10.14336/AD.2020.0904. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Chan V, Wilson JR, Ravinsky R, Badhiwala JH, Jiang F, Anderson M, et al. Frailty adversely affects outcomes of patients undergoing spine surgery: A systematic review. Spine J. 2021;21:988–1000. doi: 10.1016/j.spinee.2021.01.028. [DOI] [PubMed] [Google Scholar]
- 19.Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. J Chronic Dis. 1987;40:373–83. doi: 10.1016/0021-9681(87)90171-8. [DOI] [PubMed] [Google Scholar]
- 20.Velanovich V, Antoine H, Swartz A, Peters D, Rubinfeld I. Accumulating deficits model of frailty and postoperative mortality and morbidity: Its application to a national database. J Surg Res. 2013;183:104–10. doi: 10.1016/j.jss.2013.01.021. [DOI] [PubMed] [Google Scholar]
- 21.Hall DE, Arya S, Schmid KK, Blaser C, Carlson MA, Bailey TL, et al. Development and initial validation of the risk analysis index for measuring frailty in surgical populations. JAMA Surg. 2017;152:175–82. doi: 10.1001/jamasurg.2016.4202. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Kim B, Yoon DH, Ha Y, Yi S, Shin DA, Lee CK, et al. Relationship between T1 slope and loss of lordosis after laminoplasty in patients with cervical ossification of the posterior longitudinal ligament. Spine J. 2016;16:219–25. doi: 10.1016/j.spinee.2015.10.042. [DOI] [PubMed] [Google Scholar]
- 23.Lee SH, Son DW, Lee JS, Sung SK, Lee SW, Song GS. Does extension dysfunction affect postoperative loss of cervical lordosis in patients who undergo laminoplasty? Spine (Phila Pa 1976) 2019;44:E456–64. doi: 10.1097/BRS.0000000000002887. [DOI] [PubMed] [Google Scholar]
- 24.Kim TH, Lee SY, Kim YC, Park MS, Kim SW. T1 slope as a predictor of kyphotic alignment change after laminoplasty in patients with cervical myelopathy. Spine (Phila Pa 1976) 2013;38:E992–7. doi: 10.1097/BRS.0b013e3182972e1b. [DOI] [PubMed] [Google Scholar]
- 25.von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP, et al. Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: Guidelines for reporting observational studies. BMJ. 2007;335:806–8. doi: 10.1136/bmj.39335.541782.AD. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Huang YQ, Gou R, Diao YS, Yin QH, Fan WX, Liang YP, et al. Charlson comorbidity index helps predict the risk of mortality for patients with type 2 diabetic nephropathy. J Zhejiang Univ Sci B. 2014;15:58–66. doi: 10.1631/jzus.B1300109. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Conlon M, Thommen R, Kazim SF, Dicpinigaitis AJ, Schmidt MH, McKee RG, et al. Risk analysis index and its recalibrated version predict postoperative outcomes better than 5-factor modified frailty index in traumatic spinal injury. Neurospine. 2022;19:1039–48. doi: 10.14245/ns.2244326.163. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Harris MB, Reichmann WM, Bono CM, Bouchard K, Corbett KL, Warholic N, et al. Mortality in elderly patients after cervical spine fractures. J Bone Joint Surg Am. 2010;92:567–74. doi: 10.2106/JBJS.I.00003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Shinonara K, Ugawa R, Arataki S, Nakahara S, Takeuchi K. Charlson comorbidity index is predictive of postoperative clinical outcome after single-level posterior lumbar interbody fusion surgery. J Orthop Surg Res. 2021;16:235. doi: 10.1186/s13018-021-02377-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Arrigo RT, Kalanithi P, Cheng I, Alamin T, Carragee EJ, Mindea SA, et al. Charlson score is a robust predictor of 30-day complications following spinal metastasis surgery. Spine (Phila Pa 1976) 2011;36:E1274–80. doi: 10.1097/BRS.0b013e318206cda3. [DOI] [PubMed] [Google Scholar]
- 31.Elsamadicy AA, Havlik JL, Reeves B, Sherman J, Koo AB, Pennington Z, et al. Assessment of frailty indices and Charlson comorbidity index for predicting adverse outcomes in patients undergoing surgery for spine metastases: A national database analysis. World Neurosurg. 2022;164:e1058–70. doi: 10.1016/j.wneu.2022.05.101. [DOI] [PubMed] [Google Scholar]
- 32.Borja AJ, Connolly J, Kvint S, Detchou DK, Glauser G, Strouz K, et al. Charlson comorbidity index score predicts adverse post-operative outcomes after far lateral lumbar discectomy. Clin Neurol Neurosurg. 2021;206:106697. doi: 10.1016/j.clineuro.2021.106697. [DOI] [PubMed] [Google Scholar]
- 33.Weaver DJ, Malik AT, Jain N, Yu E, Kim J, Khan SN. The modified 5-item frailty index: A concise and useful tool for assessing the impact of frailty on postoperative morbidity following elective posterior lumbar fusions. World Neurosurg. 2019;124:e626–32. doi: 10.1016/j.wneu.2018.12.168. [DOI] [PubMed] [Google Scholar]
- 34.Camino-Willhuber G, Choi J, Holc F, Oyadomari S, Guiroy A, Bow H, et al. Utility of the Modified 5-Items Frailty Index to Predict Complications and Mortality After Elective Cervical, Thoracic and Lumbar Posterior Spine Fusion Surgery: Multicentric Analysis From ACS-NSQIP Database. Global Spine Journal. 2022;14:839–45. doi: 10.1177/21925682221124101. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Yagi M, Michikawa T, Hosogane N, Fujita N, Okada E, Suzuki S, et al. The 5-item modified frailty index is predictive of severe adverse events in patients undergoing surgery for adult spinal deformity. Spine (Phila Pa 1976) 2019;44:E1083–91. doi: 10.1097/BRS.0000000000003063. [DOI] [PubMed] [Google Scholar]
- 36.Zreik J, Alvi MA, Yolcu YU, Sebastian AS, Freedman BA, Bydon M. Utility of the 5-item modified frailty index for predicting adverse outcomes following elective anterior cervical discectomy and fusion. World Neurosurg. 2021;146:e670–7. doi: 10.1016/j.wneu.2020.10.154. [DOI] [PubMed] [Google Scholar]
- 37.Covell MM, Rumalla K, Kassicieh AJ, Segura AC, Kazim SF, Schmidt MH, et al. Frailty measured by risk analysis index and adverse discharge outcomes after adult spine deformity surgery: Analysis of 3104 patients from a prospective surgical registry (2011-2020) Spine J. 2023;23:739–45. doi: 10.1016/j.spinee.2022.12.014. [DOI] [PubMed] [Google Scholar]
- 38.Thommen R, Kazim SF, Rumalla K, Kassicieh AJ, Kalakoti P, Schmidt MH, et al. Preoperative frailty measured by risk analysis index predicts complications and poor discharge outcomes after brain tumor resection in a large multi-center analysis. J Neurooncol. 2022;160:285–97. doi: 10.1007/s11060-022-04135-z. [DOI] [PubMed] [Google Scholar]
- 39.Ondeck NT, Bohl DD, Bovonratwet P, McLynn RP, Cui JJ, Shultz BN, et al. Discriminative ability of commonly used indices to predict adverse outcomes after poster lumbar fusion: A comparison of demographics, ASA, the modified Charlson comorbidity index, and the modified frailty index. Spine J. 2018;18:44–52. doi: 10.1016/j.spinee.2017.05.028. [DOI] [PubMed] [Google Scholar]
- 40.Jenkins NW, Parrish JM, Nolte MT, Jadczak CN, Geoghegan CE, Mohan S, et al. Charlson comorbidity index: An inaccurate predictor of minimally invasive lumbar spinal fusion outcomes. Int J Spine Surg. 2021;15:770–9. doi: 10.14444/8099. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Elsamadicy AA, Freedman IG, Koo AB, David WB, Reeves BC, Havlik J, et al. Modified-frailty index does not independently predict complications, hospital length of stay or 30-day readmission rates following posterior lumbar decompression and fusion for spondylolisthesis. Spine J. 2021;21:1812–21. doi: 10.1016/j.spinee.2021.05.011. [DOI] [PubMed] [Google Scholar]
- 42.Estes EM, Rumalla K, Kazim SF, Kassicieh AJ, Segura AC, Kogan M, et al. Frailty measured by the risk analysis index predicts nonhome discharge and mortality after resection in refractory epilepsy: Analysis of 1236 patients from a prospective surgical registry, 2012 to 2020. Neurosurgery. 2023;93:267–73. doi: 10.1227/neu.0000000000002439. [DOI] [PubMed] [Google Scholar]
- 43.Agarwal N, Goldschmidt E, Taylor T, Roy S, Dunn SC, Bilderback A, et al. Impact of frailty on outcomes following spine surgery: A prospective cohort analysis of 668 patients. Neurosurgery. 2021;88:552–7. doi: 10.1093/neuros/nyaa468. [DOI] [PubMed] [Google Scholar]
- 44.Rumalla K, Thommen R, Kazim SF, Segura AC, Kassicieh AJ, Schmidt MH, et al. Risk analysis index and 30-day mortality after brain tumor resection: A multicenter frailty analysis of 31,776 patients from, 2012 to 2020. J Neurol Surg B Skull Base. 2024;85:168–71. doi: 10.1055/a-2015-1162. [DOI] [PMC free article] [PubMed] [Google Scholar]