Prediction of 2-year clinical outcome trajectories in patients undergoing anterior cervical discectomy and fusion for spondylotic radiculopathy

Jeffrey J. Hébert PhD1,2, Tyler Adams BScKin1, Erin Cunningham BScKin1, Dana El-Mughayyar BSc3, Neil Manson MD, FRCSC3,4,5, Edward Abraham MD, FRCSC3,4,5, Niels Wedderkopp MD, PhD6,7, Erin Bigney MA3,8, Eden Richardson BA3,8, Amanda Vandewint BSc3,8, Chris Small MD, FRCSC3,4,5, George Kolyvas MD, FRCSC3,5, Andre le Roux MD, FRCSC3,5,9, Aaron Robichaud MD, FRCSC3,5,9, Michael H. Weber MD, FRCSC, PhD10,21, Charles Fisher MHSc, MD, FRCSC11, Nicolas Dea MD, MSc, FRCSC11, Stephan du Plessis MD, MMed12, Raphaele Charest-Morin MD, FRCSC11, Sean D. Christie MD, FRCSC13, Christopher S. Bailey MD, MSc, FRCSC14, Y. Raja Rampersaud MD, FRCSC15, Michael G. Johnson MD, FRCSC16, Jerome Paquet MD, FRCSC17, Andrew Nataraj MSc, MD, FRCSC18, Bernard LaRue MD, FRCSC19, Hamilton Hall MD, FRCSC20, and Najmedden Attabib MD, FRCSC3,5,9
View More View Less
  • 1 Faculty of Kinesiology, University of New Brunswick, Fredericton, New Brunswick, Canada;
  • | 2 School of Psychology and Exercise Science, Murdoch University, Murdoch, Western Australia, Australia;
  • | 3 Canada East Spine Centre, Saint John, New Brunswick, Canada;
  • | 4 Division of Orthopaedic Surgery, Zone 2, Horizon Health Network, Saint John, New Brunswick, Canada;
  • | 5 Faculty of Medicine, Dalhousie University, Halifax, Nova Scotia, Canada;
  • | 6 Department of Regional Health Research, University of Southern Denmark, Odense, Denmark;
  • | 7 The Orthopedic Department, Hospital of Southwestern Jutland, Esbjerg, Denmark;
  • | 8 Research Services, Zone 2, Horizon Health Network, Saint John, New Brunswick, Canada;
  • | 9 Division of Neurosurgery, Zone 2, Horizon Health Network, Saint John, New Brunswick, Canada;
  • | 10 Department of Neurology and Neurosurgery, McGill University, Montréal, Québec, Canada;
  • | 11 Combined Neurosurgical and Orthopaedic Spine Program, Department of Orthopaedic Surgery, University of British Columbia, Vancouver, British Columbia, Canada;
  • | 12 Department of Orthopaedics, University of Calgary, Alberta, Canada;
  • | 13 Division of Neurosurgery, Halifax Infirmary, Halifax, Nova Scotia, Canada;
  • | 14 Department of Orthopaedic Surgery, London Health Science Centre, Western University, London, Ontario, Canada;
  • | 15 Division of Orthopaedic Surgery, Department of Surgery, University Health Network, University of Toronto, Ontario, Canada;
  • | 16 Department of Orthopaedics, University of Manitoba, Winnipeg, Manitoba, Canada;
  • | 17 Centre de Recherche CHU de Québec, CHU de Québec-Université Laval, Québec City, Québec, Canada;
  • | 18 Division of Neurosurgery, Department of Surgery, University of Alberta Hospital, Edmonton, Alberta, Canada;
  • | 19 Département de Chirurgie, Faculté de Médecine et des Sciences de la Santé, Université de Sherbrooke, Québec, Canada;
  • | 20 Department of Surgery, University of Toronto, Ontario, Canada; and
  • | 21 Department of Surgery, Montréal General Hospital, McGill University, Montréal, Québec, Canada
Open access

OBJECTIVE

Anterior cervical discectomy and fusion (ACDF) is often described as the gold standard surgical technique for cervical spondylotic radiculopathy. Although outcomes are considered favorable, there is little prognostic evidence to guide patient selection for ACDF. This study aimed to 1) describe the 24-month postoperative trajectories of arm pain, neck pain, and pain-related disability; and 2) identify perioperative prognostic factors that predict trajectories representing poor clinical outcomes.

METHODS

In this retrospective cohort study, patients with cervical spondylotic radiculopathy who underwent ACDF at 1 of 12 orthopedic or neurological surgery centers were recruited. Potential outcome predictors included demographic, health, clinical, and surgery-related prognostic factors. Surgical outcomes were classified by trajectories of arm pain intensity, neck pain intensity (numeric pain rating scales), and pain-related disability (Neck Disability Index) from before surgery to 24 months postsurgery. Trajectories of postoperative pain and disability were estimated with latent class growth analysis, and prognostic factors associated with poor outcome trajectory were identified with robust Poisson models.

RESULTS

The authors included data from 352 patients (mean age 50.9 [SD 9.5] years; 43.8% female). The models estimated that 15.5%–23.5% of patients followed a trajectory consistent with a poor clinical outcome. Lower physical and mental health–related quality of life, moderate to severe risk of depression, and longer surgical wait time and procedure time predicted poor postoperative trajectories for all outcomes. Receiving compensation and smoking additionally predicted a poor neck pain outcome. Regular exercise, physiotherapy, and spinal injections before surgery were associated with a lower risk of poor disability outcome. Patients who used daily opioids, those with worse general health, or those who reported predominant neck pain or a history of depression were at greater risk of poor disability outcome.

CONCLUSIONS

Patients who undergo ACDF for cervical spondylotic radiculopathy experience heterogeneous postoperative trajectories of pain and disability, with 15.5%–23.5% of patients experiencing poor outcomes. Demographic, health, clinical, and surgery-related prognostic factors can predict ACDF outcomes. This information may further assist surgeons with patient selection and with setting realistic expectations. Future studies are needed to replicate and validate these findings prior to confident clinical implementation.

ABBREVIATIONS

ACDF = anterior cervical discectomy and fusion; ASA = American Society of Anesthesiologists; CSORN = Canadian Spine Outcomes and Research Network; HRQOL = health-related quality of life; NDI = Neck Disability Index; NRS = numeric pain rating scale; PHQ-8 = Patient Health Questionnaire–8; SF-12v2 = 12-Item Short Form Health Survey version 2.

OBJECTIVE

Anterior cervical discectomy and fusion (ACDF) is often described as the gold standard surgical technique for cervical spondylotic radiculopathy. Although outcomes are considered favorable, there is little prognostic evidence to guide patient selection for ACDF. This study aimed to 1) describe the 24-month postoperative trajectories of arm pain, neck pain, and pain-related disability; and 2) identify perioperative prognostic factors that predict trajectories representing poor clinical outcomes.

METHODS

In this retrospective cohort study, patients with cervical spondylotic radiculopathy who underwent ACDF at 1 of 12 orthopedic or neurological surgery centers were recruited. Potential outcome predictors included demographic, health, clinical, and surgery-related prognostic factors. Surgical outcomes were classified by trajectories of arm pain intensity, neck pain intensity (numeric pain rating scales), and pain-related disability (Neck Disability Index) from before surgery to 24 months postsurgery. Trajectories of postoperative pain and disability were estimated with latent class growth analysis, and prognostic factors associated with poor outcome trajectory were identified with robust Poisson models.

RESULTS

The authors included data from 352 patients (mean age 50.9 [SD 9.5] years; 43.8% female). The models estimated that 15.5%–23.5% of patients followed a trajectory consistent with a poor clinical outcome. Lower physical and mental health–related quality of life, moderate to severe risk of depression, and longer surgical wait time and procedure time predicted poor postoperative trajectories for all outcomes. Receiving compensation and smoking additionally predicted a poor neck pain outcome. Regular exercise, physiotherapy, and spinal injections before surgery were associated with a lower risk of poor disability outcome. Patients who used daily opioids, those with worse general health, or those who reported predominant neck pain or a history of depression were at greater risk of poor disability outcome.

CONCLUSIONS

Patients who undergo ACDF for cervical spondylotic radiculopathy experience heterogeneous postoperative trajectories of pain and disability, with 15.5%–23.5% of patients experiencing poor outcomes. Demographic, health, clinical, and surgery-related prognostic factors can predict ACDF outcomes. This information may further assist surgeons with patient selection and with setting realistic expectations. Future studies are needed to replicate and validate these findings prior to confident clinical implementation.

In Brief

The authors described the 24-month postoperative trajectories of arm pain, neck pain, and pain-related disability in patients undergoing anterior cervical discectomy and fusion, and they identified the predictors of poor outcome. Outcome trajectories were variable, with 15.5%–23.5% of patients experiencing a poor result. Demographic, health, clinical, and surgery-related prognostic factors predicted outcomes. This information informs future research and may assist surgeons with patient selection and in setting realistic expectations with patients.

Neck pain is a leading source of disability worldwide.1 Cervical spondylotic radiculopathy is a common neck disorder caused by spinal degeneration and compression of the cervical nerve roots.2 Patients with cervical spondylotic radiculopathy who do not respond to conservative therapies may require surgical nerve root decompression. Anterior cervical discectomy and fusion (ACDF) is often described as the gold standard surgical technique for cervical radiculopathy.3 Although the results of ACDF are considered favorable, the variability in outcomes and occurrence of adverse events underscores the importance of appropriate patient selection.4,5 Prognostic factors provide important context to the process of clinical decision-making.6 However, there is little prognostic evidence to guide patient selection for ACDF.

The few prognostic studies to date have reported inconsistent and conflicting findings. For example, greater baseline pain and disability have been reported to predict a better7,8 and a worse9,10 outcome with ACDF. Similarly, older age was a predictor of either better10 or worse11,12 outcome. Although education,10 workers’ compensation, weak narcotic use, and sensory function7 have been reported to predict outcomes, other studies have reported no associations with these factors.8,1113 The inconsistent findings to date may be due to limitations such as small sample sizes810,13 and an emphasis on multivariable models that do not demonstrate the potential value of individual prognostic factors.7,8 To date, prognostic studies have measured the average outcome of the total sample at a single point in time, an approach that is likely to conflict with the perspectives of patients and clinicians who consider the evolving course of symptoms.14

Given the current lack of prognostic evidence, identification of prognostic factors for ACDF outcome is essential to inform clinical practice and future risk prediction and clinical studies. Therefore, this exploratory prognostic study aimed to describe the clinical outcome trajectories and investigate the perioperative factors associated with clinical outcomes following ACDF for cervical spondylotic radiculopathy. Specifically, the first study objective was to describe the postoperative trajectories of arm pain, neck pain, and pain-related disability. The second study objective was to identify perioperative prognostic factors that predict trajectories representing poor clinical outcomes.

Methods

Study Design and Participants

This was a retrospective cohort study including patients enrolled in the Canadian Spine Outcomes and Research Network (CSORN). The CSORN is a multicenter initiative of orthopedic and neurological spine surgeons that comprises a registry of clinical outcomes for patients undergoing spine surgery. Patients were enrolled prior to surgery and outcomes were measured at the preoperative baseline and at 3, 12, and 24 months after surgery.

We included data from all patients who underwent ACDF for cervical radiculopathy resulting from degenerative spinal pathology. We excluded patients diagnosed with myelopathy, including those diagnosed with concomitant radiculopathy and myelopathy, as well as patients with nondegenerative pathology (e.g., fracture, infection, tumor) or with a previous cervical spine fusion. The CSORN project was approved by research ethics boards local to each data collection site. The research ethics boards at the Horizon Health Network and the University of New Brunswick approved the current study protocol. All patients provided written informed consent to participate before study enrollment.

Potential Predictors of Outcome

Patients completed a preoperative assessment, including standardized forms and questionnaires to collect demographic, health-related, and clinical information. Medical staff collected additional clinical (e.g., adverse events) and surgery-related data. All data were contemporaneously entered into the CSORN surgical registry.

Demographic and Health-Related Prognostic Factors

Patients reported their age, sex, height, and weight at baseline. Body mass index was calculated as weight (kg)/height (m)2. We defined compensation status as preoperative participation in legal consultation, workers’ compensation, or other insurance claim related to the patient’s spinal complaint. Patients were considered to participate in regular exercise when they reported ≥ 20 minutes of nonstop activity such as swimming, jogging, rapid walking, or resistance training, ≥ 2 times per week. Current smoking was self-reported.

Health-related quality of life (HRQOL) was quantified with the Medical Outcomes Study’s 12-Item Short Form Health Survey version 2 (SF-12v2). Survey answers are divided into physical and mental component subscales, which are purported to be reliable and valid for people living with noncancer pain.15

Clinical Prognostic Factors

Patients reported their history of preoperative treatments for their presenting complaint, including medications, therapeutic spinal injections (with or without imaging guidance), and treatment by a physiotherapist or chiropractor. Frequency of medication use was classified as intermittent or daily, except for antidepressants because so few patients reported intermittent use. Clinical staff from each center collected additional clinical details from patients, including their history of depression, the duration of the presenting complaint, and sensory or motor deficits related to their spine problem.

Overall preoperative health was categorized using the American Society of Anesthesiologists (ASA) physical status classification.16 The ASA scores range from I (normal healthy patient) to VI (patient declared brain dead, organs are being removed for donor purposes) and are used to predict preoperative health risk.17,18 Patients were categorized as having normal to mild systemic disease (ASA score I–II) or severe systemic disease or worse health status (ASA score ≥ III).

We screened for depression risk with the Patient Health Questionnaire–8 (PHQ-8). The PHQ-8 is a valid measure of depression19 and appropriate for use with patients undergoing spine surgery.20,21 Scores ≥ 10 indicate moderate to severe risk of depression, a cutpoint with 88% sensitivity and 88% specificity for major depression.22

We calculated the ratio of arm pain intensity to neck pain intensity (arm pain/neck pain); a similar leg pain/arm pain ratio is associated with outcomes for lumbar discectomy.23 We categorized patients with a ratio < 1 as experiencing predominant neck pain.

Surgery-Related Prognostic Factors

Surgical wait time was measured as the period from the first surgical evaluation to the date of surgery. The attending spine surgeon and medical staff recorded surgical details including the number of operative spinal levels, total time to complete the surgery, and the occurrence of intraoperative or postoperative adverse events.

Clinical Outcomes

Pain and disability outcomes were collected at the preoperative baseline and at 3, 12, and 24 months after surgery. Arm and neck pain intensity over the preceding 24 hours were measured with separate 0- to 10-point (0, no pain; 10, worst pain imaginable) numeric pain rating scales (NRSs).24 The NRS has excellent test–retest reliability and responsiveness.25,26

We quantified disability related to arm and neck pain with the neck disability questionnaire.27 Patients rated the difficulty of performing 10 activities of daily living. Total scores ranged from 0 to 100 points, with higher values indicating greater disability. The Neck Disability Index (NDI) has adequate test–retest reliability and responsiveness.28,29

Data Analysis

Surgical Outcome Trajectories

We modeled trajectories of arm pain, neck pain, and disability from the preoperative baseline to 24 postoperative months, with separate latent class growth models applying a censored normal distribution. This is a specialized application of finite mixture modeling used to identify meaningful trajectory subgroups when outcomes are heterogeneous.14,30

We included data from patients with outcomes measured at ≥ 2 time points. The models handled missing outcome data with maximum likelihood estimation, resulting in asymptotically unbiased parameter estimates when data are missing at random.14 Because patients can primarily experience either arm or neck symptoms, we excluded data from patients with minimal (i.e., < 3/10) arm pain (for the arm pain model) or neck pain (for the neck pain model) scores at baseline. Similarly, when modeling neck pain–related disability, we excluded data from patients with preoperative NDI scores < 21 (i.e., minimal disability).

We first constructed a single-class model and then increased the number of latent classes and the complexity of the polynomial distributions until optimal models were identified. We used several criteria to define optimal model fit, because decision-making should not hinge on a single metric.14 Initial decisions were made using the Bayesian information criterion and judgment to identify clinically meaningful subgroups. We also considered 4 a priori diagnostic criteria: 1) a minimum average posterior probability of group membership ≥ 0.7; 2) minimum odds of correct classification > 5; 3) precision of confidence intervals around estimates of group membership probabilities; and 4) close correspondence between the estimated group membership probability and the proportion of participants assigned to each group based on the posterior probability.14,30 The subgroups identified by the model were assigned descriptive labels based on clinical judgment.

We investigated the construct validity of the trajectory results descriptively and calculated the proportion of patients meeting clinical benchmarks for minimum important improvement and success. There is no consensus on minimum important change in pain and disability or clinical success in patients undergoing ACDF. Therefore, we applied cutpoints of 30% improvement for minimum important change for pain and disability outcomes31,32 and 50% improvement for clinical success for the disability outcome,33 based on prior evidence for patients with lumbar spine disorders. Differences between trajectory subgroups in the proportion of patients meeting each clinical benchmark were examined with Fisher’s exact test.

Prognostic Factors

We explored for associations between the potential prognostic factors and a binary dependent variable with robust Poisson models. This approach avoids the inflation of odds ratios from logistic regression,34 and yields unbiased estimates of risk with frequently occurring outcomes.35 The binary dependent variable was poor surgical outcome trajectory, which was constructed by collapsing the excellent and good trajectory subgroups for neck and arm pain models or the excellent and fair trajectory groups for the disability model.

Because the aims of this exploratory prognostic study concern description (outcome trajectories) and prediction (prognostic factor identification), and not causal inference, we did not adjust the models for potential confounding variables.36 We confirmed linearity between continuous prognostic factors and the probability of poor outcome graphically and with Box-Tidwell models.37 We investigated the potential for clustering attributable to individual surgeons by adding a surgeon identifier as a random effect. Given that there were no appreciable differences between mixed- and fixed-effects model results, we reported the simpler (fixed-effects) strategy. Model results were reported with risk ratios and 95% confidence intervals. All analyses were performed with Stata 15.1 software (StataCorp).

Results

We considered data from 388 patients and included 352 patients (43.8% female) who enrolled in the CSORN registry at 1 of 12 surgical spine centers between 2015 and 2018. In total, 342 patients contributed to the arm pain trajectory model, 334 to the neck pain model, and 329 to the disability model (Fig. 1). Preoperative demographic, health, and clinical information as well as surgical details are reported in Table 1.

FIG. 1.
FIG. 1.

Study flow diagram. *Patients were included in ≥ 1 outcome model.

TABLE 1.

Descriptive preoperative characteristics and surgical details in 352 patients who underwent ACDF for spondylotic radiculopathy

VariableSample SizeValue
Age, yrs35250.9 ± 9.5
Female sex352154 (43.8%)
Education346
 High school degree or less123 (35.6%)
 Technical/associate’s degree81 (23.4%)
 Undergraduate/graduate degree142 (41.0%)
Compensation32399 (30.7%)
Regular exercise339161 (47.5%)
Body mass index categories343
 Normal weight76 (22.2%)
 Overweight132 (38.5%)
 Obese135 (39.4%)
Current smoking34377 (22.5%)
Physical HRQOL33734.1 ± 8.5
Mental HRQOL33741.8 ± 11.5
Comorbid depression34028 (8.2%)
Moderate to severe depression risk351183 (52.1%)
Predominant neck pain35290 (25.6%)
Complaint duration351
 ≤1 yr106 (30.2%)
 >1–2 yrs80 (22.8%)
 >2 yrs165 (47.0%)
Neurological deficit351
 Sensory only73 (20.8%)
 Motor w/ or w/o sensory135 (38.5%)
ASA score >II34958 (16.6%)
Chiropractic35296 (27.3%)
Physiotherapy351173 (49.3%)
Spinal injection34199 (29.0%)
Medication, opioids343
 Intermittent74 (21.6%)
 Daily96 (28.0%)
Medication, anticonvulsants340
 Intermittent36 (10.6%)
 Daily134 (39.4%)
Medication, antidepressant33681 (24.1%)
Surgery wait time, days34881.2 ± 102.3
No. of spinal levels treated352
 1 level223 (63.4%)
 2 levels98 (27.8%)
 ≥3 levels31 (8.8%)
Procedure time, mins321134.8 ± 72.5
Adverse event35040 (11.4%)

Values are expressed as number (%) or mean ± SD.

Clinical Outcome Trajectories

All models identified 3 distinct outcome trajectory subgroups (Fig. 2). Depending on the outcome measure, between 22.8% and 23.1% of patients reported outcome data at 2 time points, 29.2%–30.2% at 3 time points, and 46.7%–48.0% at all 4 time points. The arm pain model showed that 23.5% of patients followed a poor outcome trajectory, whereas 52.0% and 24.5% of patients were classified as experiencing a good or excellent outcome, respectively. Similarly, the neck pain model classified 23.2% of patients as having a poor outcome, whereas most patients were classified as members of the good (63.1%) or excellent (13.7%) trajectory subgroups. Patients in the neck pain–related disability trajectory subgroups were labeled as members of an excellent (45.3%), fair (39.2%), or poor (15.5%) trajectory subgroup.

FIG. 2.
FIG. 2.

Clinical outcome trajectories for arm pain intensity (A), neck pain intensity (B), and neck pain–related disability (C) with prevalence estimates. Point estimates are average outcome scores. Shaded areas represent 95% CIs. Figure is available in color online only.

The mean outcome scores for all trajectory subgroups are reported in Table 2. The proportion of patients meeting each clinical benchmark are reported in Table 3; all between-subgroup outcome differences were statistically significant (p < 0.001). The latent class growth models met all diagnostic criteria except for the odds of correct classification for the good outcome subgroup from the neck pain model (4.30), which was less than the recommended level of 5.00 (Table 4).

TABLE 2.

Descriptive clinical outcomes stratified by trajectory group in patients who underwent ACDF for spondylotic radiculopathy

Preop3 Mos12 Mos24 Mos
Arm pain trajectory groups (0–10 arm NRS score)
 1, excellent7.1 ± 1.90.2 ± 0.60.3 ± 0.60.3 ± 0.6
 2, good–gradual6.8 ± 1.92.3 ± 1.52.2 ± 1.72.5 ± 1.9
 3, poor8.0 ± 1.45.9 ± 1.96.0 ± 1.86.5 ± 1.9
Neck pain trajectory groups (0–10 neck NRS score)
 1, excellent7.0 ± 1.90.9 ± 1.10.0 ± 0.20.0 ± 0.0
 2, good6.7 ± 1.82.7 ± 1.62.4 ± 1.52.6 ± 1.9
 3, poor7.9 ± 1.55.9 ± 1.75.7 ± 1.96.0 ± 2.1
Disability trajectory groups (0–100 NDI score)
 1, excellent40.6 ± 12.718.4 ± 12.710.2 ± 7.88.7 ± 7.3
 2, fair49.4 ± 11.035.1 ± 11.028.7 ± 10.734.2 ± 11.2
 3, poor62.9 ± 10.156.7 ± 11.155.0 ± 10.754.5 ± 8.7

Values are expressed as the mean ± SD.

TABLE 3.

Proportion of patients meeting 12- and 24-month clinical outcome benchmarks, stratified by trajectory group

Arm Pain MIC*Neck Pain MICNDI MICNDI Success§
12-mo outcomes
 Arm pain trajectory groups
  Excellent100.0%95.2%95.2%85.5%
  Good86.0%82.4%69.3%52.6%
  Poor41.4%56.9%38.6%21.1%
 Neck pain trajectory groups
  Excellent94.3%100.0%94.1%94.1%
  Good84.1%85.8%75.2%57.1%
  Poor55.6%47.3%32.1%14.3%
 Disability trajectory groups
  Excellent92.3%93.1%94.0%82.9%
  Fair76.6%71.3%63.4%39.8%
  Poor50.0%54.1%10.8%0.0%
24-mo outcomes
 Arm pain trajectory groups
  Excellent100.0%90.2%90.4%84.6%
  Good83.2%75.5%66.3%51.6%
  Poor33.3%54.8%26.8%17.1%
 Neck pain trajectory groups
  Excellent91.7%100.0%95.8%95.8%
  Good85.5%83.6%74.4%62.4%
  Poor41.5%39.0%22.0%4.9%
 Disability trajectory groups
  Excellent91.2%93.3%95.7%90.2%
  Fair66.7%56.5%42.7%22.1%
  Poor50.0%54.6%9.1%4.6%

MIC = minimum important change.

All differences in the proportion of patients meeting each of the clinical benchmarks were statistically significant (p < 0.001).

Set at ≥ 30% reduction in NRS score for arm pain.

Set at ≥ 30% reduction in NRS score for neck pain.

Set at ≥ 30% reduction in NDI.

Set at ≥ 50% reduction in NDI.

TABLE 4.

Latent class growth model diagnostics in patients who underwent ACDF for spondylotic radiculopathy

Avg Posterior Probability*Odds of Correct Classification% Est Membership (95% CI)Assigned Membership
Arm pain model (n = 342)
 1, excellent0.8112.9624.5 (14.9–34.2)24.9%
 2, good0.865.6852.0 (42.3–61.6)52.3%
 3, poor0.9030.5723.5 (17.7–29.3)22.8%
Neck pain model (n = 334)
 1, excellent0.8330.3613.7 (8.4–19.0)12.8%
 2, good0.884.3063.1 (54.8–71.4)65.6%
 3, poor0.8519.0623.2 (15.6–30.8)21.6%
Disability model (n = 329)
 1, excellent0.888.8445.3 (36.6–54.0)45.9%
 2, fair0.827.0839.2 (30.8–47.6)38.6%
 3, poor0.8735.7015.5 (10.2–20.8)15.5%

Avg = average; est = estimated.

Minimum threshold ≥ 0.70.

Minimum threshold > 5.0.

Perioperative Prognostic Factors Associated With Surgical Outcomes

We identified 14 demographic/health, clinical, and surgical prognostic factors that predicted postoperative outcomes (Fig. 3).

FIG. 3.
FIG. 3.

Potential prognostic factors associated with poor outcomes for arm pain, neck pain, and disability. *Risk per SD. †Risk per 180 days. ‡Risk per 30 minutes. Boldface type indicates statistical significance. MCS = SF-12 mental component summary; PCS = SF-12 physical component summary. Figure is available in color online only.

Demographic and Health-Related Prognostic Factors

Preoperative HRQOL predicted all 3 clinical outcomes. Higher SF-12v2 physical component summary scores were associated with reduced risks of poor outcome for arm pain (per 1 SD change: RR 0.80 [95% CI 0.65–0.99]); neck pain (per 1 SD change: RR 0.74 [95% CI 0.58–0.93]); and disability (per 1 SD change: RR 0.40 [95% CI 0.30–0.53]). Similarly, higher SF-12v2 mental component summary scores were associated with reduced risk of poor outcome for arm pain (per 1 SD change: RR 0.73 [95% CI 0.59–0.89]); neck pain (per 1 SD change: RR 0.67 [95% CI 0.53–0.84]); and disability (per 1 SD change: RR 0.47 [95% CI 0.36–0.62]).

Patients involved in legal consultation, workers’ compensation, or other insurance claims before surgery were more likely to experience a poor neck pain outcome (26.8% vs 17.1%: RR 1.56 [95% CI 1.00–2.44]), as were patients who smoked (32.0% vs 18.3%: RR 1.75 [95% CI 1.15–2.66]). Patients who engaged in regular exercise before surgery were less likely to experience a poor outcome for disability (9.1% vs 20.8%: RR 0.44 [95% CI 0.24–0.79]).

Clinical Predictors

Depression risk predicted all clinical outcomes. Moderate to severe depression risk was associated with greater risk of poor outcome for arm pain (28.7% vs 16.0%: RR 1.80 [95% CI 1.18–2.74]); neck pain (29.4% vs 12.8%: RR 2.29 [95% CI 1.43–3.66]); and disability (26.4% vs 2.1%: RR 12.84 [95% CI 4.07–40.44]). Self-reported depression was associated with an increased risk of poor disability outcome (40.7% vs 13.5%: RR 3.03 [95% CI 1.76–5.21]), but not arm or neck pain outcomes, although the latter result was borderline (35.7% vs 20.8%: RR 1.72 [95% CI 1.00–2.97]).

Patients with greater neck pain than arm pain intensity were more likely to experience a poor disability outcome (24.1% vs 12.4%: RR 1.95 [95% CI 1.18–3.22]), as were patients with ASA scores > II (28.9% vs 12.8%: RR 2.26 [95% CI 1.33–3.83]) and patients who used opioids daily (25.0% vs 12.2%: RR 2.05 [95% CI 1.18–3.56]). Conversely, patients receiving preoperative physiotherapy (10.5% vs 20.5%: RR 0.51 [95% CI 0.30–0.88]) and spinal injections (8.5% vs 17.9%: RR 0.48 [95% CI 0.23–0.98]) were less likely to experience a poor disability outcome.

Surgery-Related Predictors

Surgical wait time and procedure time were associated with all outcomes. Patients with longer wait times for surgery were more likely to experience poor outcomes for arm pain (per 180 days: RR 1.31 [95% CI 1.05–1.63]); neck pain (per 180 days: RR 1.54 [95% CI 1.22–1.94]); and disability (per 180 days: RR 1.64 [95% CI 1.20–2.24]). Similarly, patients who experienced longer procedure times were more likely to experience poor outcomes for arm pain (per 30 minutes: 1.07 [95% CI 1.03–1.10]); neck pain (per 30 minutes: RR 1.08 [95% CI 1.04–1.11]); and disability (per 30 minutes: RR 1.08 [95% CI 1.05–1.12]). All other demographic, health, clinical, and surgical prognostic factors were not associated with clinical outcomes.

Discussion

Patients who undergo ACDF for degenerative spondylotic radiculopathy experience heterogeneous courses of postoperative pain and disability. Although most patients in this study followed a favorable 24-month outcome trajectory, 15.5%–23.5% of patients experienced little to no improvement in arm pain, neck pain, or pain-related disability. Differences in the proportion of patients who met clinical benchmarks for minimally important change and clinical success supported the construct validity of the trajectory subgroups. We identified 14 demographic, health-related, clinical, and surgery-related prognostic factors that predicted trajectory group membership. Lower physical and mental HRQOL, moderate to severe risk of depression, and longer surgical wait time and procedure time predicted poor postoperative trajectories for all outcomes. Receiving compensation and smoking additionally predicted a poor neck pain outcome. Regular exercise, physiotherapy, and spinal injections before surgery were associated with a lower risk of poor disability outcome. Patients who used daily opioids, those with worse general health, and those who reported predominant neck pain or a history of depression were at greater risk of poor disability outcome.

Clinical Implications

The characteristics and prevalence of these pain and disability trajectories demonstrate the outcomes that can be expected following ACDF, and this information may assist patients with setting appropriate expectations for recovery. Prognostic factors require replication and validation prior to confident clinical implementation; however, given the current lack of prognostic evidence, surgeons may wish to consider these prognostic factors when assessing potential candidates for ACDF.

Although data for many of these factors are routinely collected, the measurement of some factors may require a change in practice. For example, we found moderate to severe depression risk measured with the PHQ-8 to be a more consistent and stronger outcome predictor than self-reported depression. Although we identified 52.1% of patients at moderate to high risk of depression, only 8.2% of patients reported an existing depression diagnosis. This may indicate a lack of diagnosis or personal awareness about depression or a reluctance to disclose a diagnosis of comorbid depression. Therefore, clinicians may gain more clinical utility with the routine measurement of depression risk than they would by relying on patients to report a previous diagnosis of depression.

Research Implications

These study results can inform future prognostic and intervention studies. These factors can serve as candidate predictors in future studies that develop and test multivariable risk prediction models. Additionally, the prognostic factors identified in the current study can help identify prognostic balance in future observational studies and clinical trials.6 Beyond their prognostic value, future studies may also investigate these factors as potential causal determinants of outcome. Such studies could use causal inference methods with observational data or randomized trials. For example, we found that moderate to severe risk of depression predicted a poor outcome for pain and disability. An appropriately designed study applying causal inference methods (e.g., directed acyclic graphs, target trial framework) could distinguish whether depression risk is a predictor or a determinant (i.e., cause) of outcome. Furthermore, a randomized or target trial could test the effectiveness of intervening on depression risk and the impact on surgical outcome.

Strengths and Limitations

The primary strengths of this study include the relatively large sample of patients from 12 surgical spine centers and the implementation of standardized outcome assessment over 24 months of follow-up. Whereas traditional outcome assessment considers the group-averaged score at a single point in time, our models applied a person-centered approach to identify the courses of pain and disability from before surgery to 24 months after surgery, a perspective that is likely to accord with clinicians’ and patients’ conceptualization of outcome.14

Although 8 of 9 trajectory subgroups met all model diagnostic criteria, patients assigned to the good neck pain outcome trajectory had odds of correct classification that were lower than the recommended threshold (4.30 vs 5.00). This indicates greater uncertainty around the assignment of patients to this subgroup. Although our trajectory model estimated missing outcome data, we did not estimate or impute missing values for prognostic factors due to early differences in the implementation of CSORN procedures. However, the prevalence of missing baseline data was low (< 5%) for the large majority of prognostic factors, and it is unlikely that the mechanism and amount of missing data significantly impacted the model outcomes. Finally, we relied on self-reported information for some prognostic factors (e.g., physical activity behavior, medication use), which is a potential source of bias.

Conclusions

In this exploratory prognostic study, we defined the different 24-month pain and disability trajectories experienced by patients with cervical spondylotic radiculopathy who underwent ACDF. We identified 14 demographic, health-related, clinical, or surgery-related prognostic factors that predicted a poor outcome trajectory. The presence of some factors was associated with an increased risk of poor outcome, whereas other factors were associated with a decreased risk of poor outcome. The most consistent outcome predictors were as follows: moderate to severe depression risk, HRQOL, surgical wait time, and procedure time. Understanding the different courses of pain and disability following ACDF can assist patients with appropriate expectation setting, and the prognostic factors identified in this study may further assist surgeons with patient selection. Given its potential prognostic importance, surgeons should consider measuring depression risk in patients with cervical radiculopathy. Future studies are needed to replicate and validate these findings prior to confident clinical implementation.

Disclosures

Dr. Hébert receives support of non–study-related clinical or research effort that he oversees from New Brunswick Health Research Foundation and the Canadian Chiropractic Research Foundation. Ms. Cunningham received clinical or research support for the study described (includes equipment or material) from Canadian Institutes of Health Research and the New Brunswick Health Research Foundation. Dr. Manson is a consultant for Medtronic Canada, and he also receives that company’s support of a non–study-related clinical or research effort that he oversees. Dr. Fisher is a consultant for Medtronic and NuVasive; receives royalties from Medtronic; and receives fellowship support (paid to his institution) from Medtronic and AO Spine. Dr. Dea is a consultant for Medtronic, Stryker, and Baxter. He has direct stock ownership in Medtronic. Dr. Rampersaud receives royalties from Medtronic. Dr. Johnson received clinical or research support for the study described (includes equipment or material) from Stryker. Dr. LaRue receives support of non–study-related clinical or research effort that he oversees from DePuy-Synthes Spine.

Author Contributions

Conception and design: Hébert, Manson, Wedderkopp. Acquisition of data: Manson, Abraham, Kolyvas, le Roux, Robichaud, Weber, Fisher, Dea, du Plessis, Charest-Morin, Christie, Bailey, Rampersaud, Johnson, Paquet, Nataraj, LaRue, Hall, Attabib. Analysis and interpretation of data: Hébert, Adams, Cunningham, Wedderkopp, Small. Drafting the article: Hébert, Adams, Cunningham. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Hébert. Statistical analysis: Hébert. Administrative/technical/material support: Mughayyar, Bigney, Richardson, Vandewint.

References

  • 1

    GBD 2019 Diseases and Injuries Collaborators. Diseases and Injuries Collaborators. Global burden of 369 diseases and injuries in 204 countries and territories, 1990-2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet. 2020;396(10258):12041222.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 2

    Turner DA. Degenerative cervical spine disease. Accessed August 3, 2022. https://bestpractice.bmj.com/topics/en-us/577

  • 3

    Lopez CD, Boddapati V, Lombardi JM, et al. Recent trends in Medicare utilization and reimbursement for anterior cervical discectomy and fusion. Spine J. 2020;20(11):17371743.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 4

    Gornet MF, Lanman TH, Burkus JK, et al. Two-level cervical disc arthroplasty versus anterior cervical discectomy and fusion: 10-year outcomes of a prospective, randomized investigational device exemption clinical trial. J Neurosurg Spine. 2019;31(4):508518.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 5

    Andresen AK, Paulsen RT, Busch F, Isenberg-Jørgensen A, Carreon LY, Andersen MO. Patient-reported outcomes and patient-reported satisfaction after surgical treatment for cervical radiculopathy. Global Spine J. 2018;8(7):703708.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 6

    Riley RD, Hayden JA, Steyerberg EW, et al. Prognosis Research Strategy (PROGRESS) 2: prognostic factor research. PLoS Med. 2013;10(2):e1001380.

  • 7

    Anderson PA, Subach BR, Riew KD. Predictors of outcome after anterior cervical discectomy and fusion: a multivariate analysis. Spine (Phila Pa 1976). 2009;34(2):161166.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 8

    Hermansen A, Hedlund R, Vavruch L, Peolsson A. Positive predictive factors and subgroup analysis of clinically relevant improvement after anterior cervical decompression and fusion for cervical disc disease: a 10- to 13-year follow-up of a prospective randomized study: clinical article. J Neurosurg Spine. 2013;19(4):403411.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 9

    Peolsson A, Peolsson M. Predictive factors for long-term outcome of anterior cervical decompression and fusion: a multivariate data analysis. Eur Spine J. 2008;17(3):406414.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 10

    Peolsson A, Hedlund R, Vavruch L, Oberg B. Predictive factors for the outcome of anterior cervical decompression and fusion. Eur Spine J. 2003;12(3):274280.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 11

    Bertalanffy H, Eggert HR. Clinical long-term results of anterior discectomy without fusion for treatment of cervical radiculopathy and myelopathy. A follow-up of 164 cases. Acta Neurochir (Wien). 1988;90(3-4):127135.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 12

    Eriksen EF, Buhl M, Fode K, et al. Treatment of cervical disc disease using Cloward’s technique. The prognostic value of clinical preoperative data in 1,106 patients. Acta Neurochir (Wien). 1984;70(3-4):181197.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 13

    Peolsson A, Vavruch L, Oberg B. Predictive factors for arm pain, neck pain, neck specific disability and health after anterior cervical decompression and fusion. Acta Neurochir (Wien). 2006;148(2):167173.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 14

    Nagin DS, Odgers CL. Group-based trajectory modeling in clinical research. Annu Rev Clin Psychol. 2010;6:109138.

  • 15

    Hayes CJ, Bhandari NR, Kathe N, Payakachat N. Reliability and validity of the medical outcomes study Short Form-12 version 2 (SF-12v2) in adults with non-cancer pain. Healthcare (Basel). 2017;5(2):E22.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 16

    Saklad M. Grading of patients for surgical procedures. Anesthesiology. 1941;2(3):281284.

  • 17

    Dalton JE, Kurz A, Turan A, Mascha EJ, Sessler DI, Saager L. Development and validation of a risk quantification index for 30-day postoperative mortality and morbidity in noncardiac surgical patients. Anesthesiology. 2011;114(6):13361344.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 18

    Bjorgul K, Novicoff WM, Saleh KJ. American Society of Anesthesiologist Physical Status score may be used as a comorbidity index in hip fracture surgery. J Arthroplasty. 2010;25(6 suppl):134137.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 19

    Kroenke K, Strine TW, Spitzer RL, Williams JB, Berry JT, Mokdad AH. The PHQ-8 as a measure of current depression in the general population. J Affect Disord. 2009;114(1-3):163173.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 20

    Purvis TE, Neuman BJ, Riley LH, Skolasky RL. Comparison of PROMIS Anxiety and Depression, PHQ-8, and GAD-7 to screen for anxiety and depression among patients presenting for spine surgery. J Neurosurg Spine. 2019;30(4):524531.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 21

    Babington JR, Edwards A, Wright AK, Dykstra T, Friedman AS, Sethi RK. Patient-reported outcome measures: utility for predicting spinal surgery in an integrated spine practice. PM R. 2018;10(7):724729.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 22

    Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med. 2001;16(9):606613.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 23

    Hebert JJ, Fritz JM, Koppenhaver SL, Thackeray A, Kjaer P. Predictors of clinical outcome following lumbar disc surgery: the value of historical, physical examination, and muscle function variables. Eur Spine J. 2016;25(1):310317.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 24

    McCaffery M, Beebe A. Pain: Clinical Manual for Nursing Practice. Mosby;1989.

  • 25

    Childs JD, Piva SR, Fritz JM. Responsiveness of the numeric pain rating scale in patients with low back pain. Spine (Phila Pa 1976). 2005;30(11):13311334.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 26

    Jensen MP, Turner JA, Romano JM. What is the maximum number of levels needed in pain intensity measurement?. Pain. 1994;58(3):387392.

  • 27

    Vernon H. The Neck Disability Index: state-of-the-art, 1991-2008. J Manipulative Physiol Ther. 2008;31(7):491502.

  • 28

    MacDermid JC, Walton DM, Avery S, et al. Measurement properties of the neck disability index: a systematic review. J Orthop Sports Phys Ther. 2009;39(5):400417.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 29

    Young BA, Walker MJ, Strunce JB, Boyles RE, Whitman JM, Childs JD. Responsiveness of the Neck Disability Index in patients with mechanical neck disorders. Spine J. 2009;9(10):802808.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 30

    Nagin DS. Group-Based Modeling of Development. Harvard University Press;2005.

  • 31

    Ostelo RW, Deyo RA, Stratford P, et al. Interpreting change scores for pain and functional status in low back pain: towards international consensus regarding minimal important change. Spine (Phila Pa 1976). 2008;33(1):9094.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 32

    Khan I, Pennings JS, Devin CJ, et al. Clinically meaningful improvement following cervical spine surgery: 30% reduction versus absolute point-change MCID values. Spine (Phila Pa 1976). 2021;46(11):717725.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 33

    Fritz JM, Hebert J, Koppenhaver S, Parent E. Beyond minimally important change: defining a successful outcome of physical therapy for patients with low back pain. Spine (Phila Pa 1976). 2009;34(25):28032809.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 34

    Greenland S. Interpretation and choice of effect measures in epidemiologic analyses. Am J Epidemiol. 1987;125(5):761768.

  • 35

    Chen W, Qian L, Shi J, Franklin M. Comparing performance between log-binomial and robust Poisson regression models for estimating risk ratios under model misspecification. BMC Med Res Methodol. 2018;18(1):63.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 36

    Hernán MA, Hsu J, Healy B. A second chance to get causal inference right: a classification of data science tasks. CHANCE. 2019;32(1):4249.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 37

    Box GEP, Tidwell PW. Transformation of the independent variables. Technometrics. 1962;4(4):531550.

  • View in gallery

    Study flow diagram. *Patients were included in ≥ 1 outcome model.

  • View in gallery

    Clinical outcome trajectories for arm pain intensity (A), neck pain intensity (B), and neck pain–related disability (C) with prevalence estimates. Point estimates are average outcome scores. Shaded areas represent 95% CIs. Figure is available in color online only.

  • View in gallery

    Potential prognostic factors associated with poor outcomes for arm pain, neck pain, and disability. *Risk per SD. †Risk per 180 days. ‡Risk per 30 minutes. Boldface type indicates statistical significance. MCS = SF-12 mental component summary; PCS = SF-12 physical component summary. Figure is available in color online only.

  • 1

    GBD 2019 Diseases and Injuries Collaborators. Diseases and Injuries Collaborators. Global burden of 369 diseases and injuries in 204 countries and territories, 1990-2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet. 2020;396(10258):12041222.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 2

    Turner DA. Degenerative cervical spine disease. Accessed August 3, 2022. https://bestpractice.bmj.com/topics/en-us/577

  • 3

    Lopez CD, Boddapati V, Lombardi JM, et al. Recent trends in Medicare utilization and reimbursement for anterior cervical discectomy and fusion. Spine J. 2020;20(11):17371743.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 4

    Gornet MF, Lanman TH, Burkus JK, et al. Two-level cervical disc arthroplasty versus anterior cervical discectomy and fusion: 10-year outcomes of a prospective, randomized investigational device exemption clinical trial. J Neurosurg Spine. 2019;31(4):508518.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 5

    Andresen AK, Paulsen RT, Busch F, Isenberg-Jørgensen A, Carreon LY, Andersen MO. Patient-reported outcomes and patient-reported satisfaction after surgical treatment for cervical radiculopathy. Global Spine J. 2018;8(7):703708.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 6

    Riley RD, Hayden JA, Steyerberg EW, et al. Prognosis Research Strategy (PROGRESS) 2: prognostic factor research. PLoS Med. 2013;10(2):e1001380.

  • 7

    Anderson PA, Subach BR, Riew KD. Predictors of outcome after anterior cervical discectomy and fusion: a multivariate analysis. Spine (Phila Pa 1976). 2009;34(2):161166.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 8

    Hermansen A, Hedlund R, Vavruch L, Peolsson A. Positive predictive factors and subgroup analysis of clinically relevant improvement after anterior cervical decompression and fusion for cervical disc disease: a 10- to 13-year follow-up of a prospective randomized study: clinical article. J Neurosurg Spine. 2013;19(4):403411.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 9

    Peolsson A, Peolsson M. Predictive factors for long-term outcome of anterior cervical decompression and fusion: a multivariate data analysis. Eur Spine J. 2008;17(3):406414.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 10

    Peolsson A, Hedlund R, Vavruch L, Oberg B. Predictive factors for the outcome of anterior cervical decompression and fusion. Eur Spine J. 2003;12(3):274280.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 11

    Bertalanffy H, Eggert HR. Clinical long-term results of anterior discectomy without fusion for treatment of cervical radiculopathy and myelopathy. A follow-up of 164 cases. Acta Neurochir (Wien). 1988;90(3-4):127135.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 12

    Eriksen EF, Buhl M, Fode K, et al. Treatment of cervical disc disease using Cloward’s technique. The prognostic value of clinical preoperative data in 1,106 patients. Acta Neurochir (Wien). 1984;70(3-4):181197.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 13

    Peolsson A, Vavruch L, Oberg B. Predictive factors for arm pain, neck pain, neck specific disability and health after anterior cervical decompression and fusion. Acta Neurochir (Wien). 2006;148(2):167173.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 14

    Nagin DS, Odgers CL. Group-based trajectory modeling in clinical research. Annu Rev Clin Psychol. 2010;6:109138.

  • 15

    Hayes CJ, Bhandari NR, Kathe N, Payakachat N. Reliability and validity of the medical outcomes study Short Form-12 version 2 (SF-12v2) in adults with non-cancer pain. Healthcare (Basel). 2017;5(2):E22.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 16

    Saklad M. Grading of patients for surgical procedures. Anesthesiology. 1941;2(3):281284.

  • 17

    Dalton JE, Kurz A, Turan A, Mascha EJ, Sessler DI, Saager L. Development and validation of a risk quantification index for 30-day postoperative mortality and morbidity in noncardiac surgical patients. Anesthesiology. 2011;114(6):13361344.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 18

    Bjorgul K, Novicoff WM, Saleh KJ. American Society of Anesthesiologist Physical Status score may be used as a comorbidity index in hip fracture surgery. J Arthroplasty. 2010;25(6 suppl):134137.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 19

    Kroenke K, Strine TW, Spitzer RL, Williams JB, Berry JT, Mokdad AH. The PHQ-8 as a measure of current depression in the general population. J Affect Disord. 2009;114(1-3):163173.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 20

    Purvis TE, Neuman BJ, Riley LH, Skolasky RL. Comparison of PROMIS Anxiety and Depression, PHQ-8, and GAD-7 to screen for anxiety and depression among patients presenting for spine surgery. J Neurosurg Spine. 2019;30(4):524531.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 21

    Babington JR, Edwards A, Wright AK, Dykstra T, Friedman AS, Sethi RK. Patient-reported outcome measures: utility for predicting spinal surgery in an integrated spine practice. PM R. 2018;10(7):724729.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 22

    Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med. 2001;16(9):606613.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 23

    Hebert JJ, Fritz JM, Koppenhaver SL, Thackeray A, Kjaer P. Predictors of clinical outcome following lumbar disc surgery: the value of historical, physical examination, and muscle function variables. Eur Spine J. 2016;25(1):310317.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 24

    McCaffery M, Beebe A. Pain: Clinical Manual for Nursing Practice. Mosby;1989.

  • 25

    Childs JD, Piva SR, Fritz JM. Responsiveness of the numeric pain rating scale in patients with low back pain. Spine (Phila Pa 1976). 2005;30(11):13311334.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 26

    Jensen MP, Turner JA, Romano JM. What is the maximum number of levels needed in pain intensity measurement?. Pain. 1994;58(3):387392.

  • 27

    Vernon H. The Neck Disability Index: state-of-the-art, 1991-2008. J Manipulative Physiol Ther. 2008;31(7):491502.

  • 28

    MacDermid JC, Walton DM, Avery S, et al. Measurement properties of the neck disability index: a systematic review. J Orthop Sports Phys Ther. 2009;39(5):400417.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 29

    Young BA, Walker MJ, Strunce JB, Boyles RE, Whitman JM, Childs JD. Responsiveness of the Neck Disability Index in patients with mechanical neck disorders. Spine J. 2009;9(10):802808.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 30

    Nagin DS. Group-Based Modeling of Development. Harvard University Press;2005.

  • 31

    Ostelo RW, Deyo RA, Stratford P, et al. Interpreting change scores for pain and functional status in low back pain: towards international consensus regarding minimal important change. Spine (Phila Pa 1976). 2008;33(1):9094.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 32

    Khan I, Pennings JS, Devin CJ, et al. Clinically meaningful improvement following cervical spine surgery: 30% reduction versus absolute point-change MCID values. Spine (Phila Pa 1976). 2021;46(11):717725.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 33

    Fritz JM, Hebert J, Koppenhaver S, Parent E. Beyond minimally important change: defining a successful outcome of physical therapy for patients with low back pain. Spine (Phila Pa 1976). 2009;34(25):28032809.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 34

    Greenland S. Interpretation and choice of effect measures in epidemiologic analyses. Am J Epidemiol. 1987;125(5):761768.

  • 35

    Chen W, Qian L, Shi J, Franklin M. Comparing performance between log-binomial and robust Poisson regression models for estimating risk ratios under model misspecification. BMC Med Res Methodol. 2018;18(1):63.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 36

    Hernán MA, Hsu J, Healy B. A second chance to get causal inference right: a classification of data science tasks. CHANCE. 2019;32(1):4249.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • 37

    Box GEP, Tidwell PW. Transformation of the independent variables. Technometrics. 1962;4(4):531550.

Metrics

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 930 930 930
PDF Downloads 680 680 680
EPUB Downloads 0 0 0