SPINE Volume 40, Number 10, pp 748-756 ©2015, Wolters Kluwer Health, Inc. All rights reserved.
OCCUPATIONAL HEALTH/ERGONOMICS
Clinical Depression Is a Strong Predictor of Poor Lumbar Fusion Outcomes Among Workers’ Compensation Subjects Joshua T. Anderson, BS,* Arnold R. Haas, BS, BA,† Rick Percy, PhD,† Stephen T. Woods, MD,† Uri M. Ahn, MD‡ and Nicholas U. Ahn, MD§
Study Design. Retrospective cohort study. Objective. Determine how psychosocial factors, particularly depression, impact lumbar fusion outcomes in a workers’ compensation (WC) setting. Summary of Background Data. WC patients are less likely to return to work (RTW) after fusion. Few studies evaluate risk factors within this clinically distinct population. Methods. A total of 2799 Ohio WC subjects were identified who underwent lumbar fusion between 1993 and 2013 using Current Procedural Terminology (CPT) procedural and International Classification of Diseases, Ninth Revision (ICD-9) diagnosis codes. A total of 123 subjects were diagnosed with depression before fusion. Subjects with a smoking history, prior lumbar surgery, permanent disability, and failed back syndrome were excluded. The primary outcome was whether subjects returned to work within 2 years of fusion and sustained this RTW for more than 6 months of the following year. To determine the impact depression had on RTW status, we performed a multivariate logistic regression analysis. We also compared time absent from work and other secondary outcomes using χ2 and t tests. Results. Subjects with preoperative depression had significantly higher rates of legal representation, degenerative lumbar disease, and higher medical costs, and used opioid analgesics for considerably longer before and after fusion (P < 0.001). From the *University Hospitals Case Medical Center Department of Orthopaedics, Case Western Reserve University School of Medicine, Cleveland, OH; †Ohio Bureau of Workers’ Compensation, Columbus, OH; ‡New Hampshire NeuroSpine Institute, Bedford, NH; and §University Hospitals Case Medical Center Department of Orthopaedics, Cleveland, OH. Acknowledgment date: July 14, 2014. First revision date: November 11, 2014. Second revision date: January 1, 2015. Acceptance date: January 10, 2015. The manuscript submitted does not contain information about medical device(s)/drug(s). No funds were received in support of this work. Relevant financial activities outside the submitted work: patents, consultancy, grants/grants pending, expert testimony, payment for lectures, royalties, travel accommodations, stock/stock options. Address correspondence and reprint requests to Joshua T. Anderson, BS, University Hospitals Case Medical Center Department of Orthopaedics, Case Western Reserve University School of Medicine, 11100 Euclid Ave, Cleveland, OH 44106; E-mail: [email protected] DOI: 10.1097/BRS.0000000000000863
748
Depression group (10.6% [13/123]) and controls (33.0% [884/2676]) met our RTW criteria (P < 0.001). Preoperative depression was a negative predictor of RTW status (P < 0.001; odds ratio [OR]: 0.38). Additional predictors included working during same week as fusion (OR: 2.15), age more than 50 years (OR: 0.58), chronic preoperative opioid analgesia (OR: 0.58), and legal representation (OR: 0.64). After surgery, depression subjects were absent from work 184 more days compared with controls (P < 0.001). Conclusion. Overall, RTW rates after fusion were low, which was especially true for those with pre-existing depression. Depression was a strong negative predictor of postoperative RTW status. Psychological screening and treatment may be beneficial in these subjects. The poor outcomes in this study may highlight a more limited role for fusion among WC subjects with chronic low back pain where RTW is the treatment goal. Key words: workers’ compensation, lumbar fusion, depression, biopsychosocial, return to work, failed back syndrome, opioid analgesics, psychosocial, risk factors, predictors, legal representation, degenerative. Level of Evidence: 3 Spine 2015;40:748–756
M
ultiple studies have demonstrated that US workers’ compensation (WC) subjects tend to have worse clinical outcomes after lumbar fusion than the general population.1–7 A 2010 case-control study reported that WC subjects, when compared with the general population, had significantly less improvement in mean Oswestry Disability Index and 36-Item Short Form Health Survey scores after fusion.3 Studies of subjects from several WC jurisdictions report return to work (RTW) rates ranging from 26% to 36% and reoperation rates from 22% to 27% after fusion.1,5,8–10 Despite these poor outcomes, fusion rates within the WC population are steadily rising.11 The estimated costs of treating low back pain in the United States exceed $100 billion a year, two-thirds of which are thought to arise from lost wages and reduced productivity.12,13 Mounting evidence highlights the importance of using a biopsychosocial model of care when treating patients with low back pain, including those with chronic discogenic low back pain receiving WC.5,6,8,14–19
www.spinejournal.com
May 2015
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE140922_LR 748
14/04/15 8:55 PM
OCCUPATIONAL HEALTH/ERGONOMICS TABLE 1. Coding Used for Study Fusion Technique
CPT Code
PLF Single level
22612
Multilevel
22612 + 22614
PLIF Single level
22630 or 22633
Multilevel
22630 or 22633 + 22632 or 22634
ALIF Single level
22558
Multilevel
22558 + 22585
360° fusion (ALIF + PLF/PLIF) Single level
SL ALIF + SL PLF/PLIF
Multilevel
ML ALIF + ML PLF/PLIF
Lumbar Comorbidity Lumbar disc degeneration
ICD-9 Code 722.52
Spondylolisthesis
738.4 or 756.12
Scoliosis
737.30–737.39
Spinal stenosis
724.02–724.03
Disc herniation
722.10
Spondylosis
721.3
Spondylolysis
756.11
Lumbar sprain
847.2
Radiculopathy
724.4 or 729.2
Depression Depressive disorder Major depressive disorder Dysthymic disorder
ICD-9 Code 311 296.20-296.26 or 296.30-296.36 300.4
This table includes the International Classification of Diseases, Ninth Revision (ICD-9) codes with which we characterized each subject’s lumbar comorbidities and identified which subjects were diagnosed with depression before fusion. It also contains the Current Procedural Terminology (CPT) coding we used to characterize each subject's fusion surgery. CPT indicates Current Procedural Terminology; PLF, posterior lumbar fusion; SL, single level; ML, multilevel; PLIF, posterior lumbar interbody fusion; ALIF, anterior lumbar interbody fusion; ICD, International Classification of Diseases.
One potential strategy to improve fusion outcomes within this clinically distinct subset of patients is to identify presurgical risk factors. Optimizing subject selection criteria for lumbar fusion would not only help provide the best and most appropriate care for these patients but it could also help curb medical and disability payments. Recent studies have Spine
Depression’s Impact on Lumbar Fusion Outcomes • Anderson et al
demonstrated the importance of psychosocial factors to outcomes after lumbar surgery.9,20–23 However, relatively few studies have been published evaluating specific predictors of lumbar fusion outcomes in US WC subjects.4,5,8,12 Furthermore, depression is a leading cause of disability as measured by years lived with disability as well as the fourth leading contributor to the global burden of disease.24 Epidemiologic studies support a lifetime prevalence of major clinical depression ranging from 16.2% to 17.9% within the United States.25–28 Studies have also shown that individuals experiencing depression are less likely to call on social supports or follow a plan of rehabilitation.29 Therefore, the objective of this study was to determine how psychosocial factors, depression in particular, impact lumbar fusion outcomes, primarily the ability to RTW within a reasonable time line postoperatively, in a WC setting. We performed this study at the University Hospitals Case Medical Center in Cleveland, Ohio. All data were collected in late July of 2013 from the Ohio Bureau of Workers’ Compensation’s (BWC’s) database.
MATERIALS AND METHODS Data Source Our retrospective cohort study includes 2799 WC subjects from the Ohio BWC who underwent lumbar fusion between 1993 and 2013. We used the administrative database at the Ohio BWC to acquire all data. This is the same database that Nguyen et al1 used to compare fusion versus nonoperative outcomes in 2011.
Subject Selection Initially, we identified 14,640 WC subjects diagnosed with lumbar conditions using International Classification of Diseases, Ninth Revision (ICD-9) codes. From this, we excluded 11,841 subjects who did not undergo lumbar fusion, were permanently disabled before fusion, underwent prior lumbar surgery before fusion, had a positive smoking history, and/or had a history of failed back syndrome because all of these factors have previously been reported to negatively impact fusion outcomes.6,17,30–33 Using Current Procedural Terminology (CPT) codes, we included subjects who underwent an anterior, posterior, or 360° lumbar fusion with at least 3 years of follow-up. Thus, we identified our final study population of 2799 subjects. Using ICD-9 codes, we further characterized each subject’s lumbar comorbidities. Next, we used ICD-9 codes to identify 123 subjects who were clinically diagnosed with depression before undergoing fusion, forming the depression group. We treated the remaining 2676 subjects as controls. See Table 1 for the ICD-9 and CPT codes used to characterize each subject’s fusion surgery, lumbar comorbidities, and identify depression. See the Supplemental Digital Content Appendix, available at: http://links.lww.com/ BRS/A967, for more specific information on how we derived our study population. Figure 1 illustrates our subject selection process. www.spinejournal.com
749
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE140922_LR 749
14/04/15 8:55 PM
OCCUPATIONAL HEALTH/ERGONOMICS
Depression’s Impact on Lumbar Fusion Outcomes • Anderson et al
Figure 1. Subject selection flow diagram. We used a combination of International Classification of Diseases, Ninth Revision (ICD-9) diagnosis and Current Procedural Terminology (CPT) procedural codes to arrive at our final study population.
Outcome Measures We measured outcomes after each subject’s index fusion surgery after WC-qualifying workplace injury. The primary outcome measure for this study was whether or not each subject made a sustainable RTW within a reasonable time line after lumbar fusion. Specifically, if a subject returned to work within 2 years of fusion and sustained this RTW for 6 months or more of the following year, we considered that subject as returned to work. Similar to the WC study conducted by Nguyen et al,1 we used 2 years as a general cutoff for RTW. However, we also included the additional requirement that each worker must sustain his or her RTW. For example, a patient may fulfill the criteria of making an RTW within 2 years but after a few days of work may leave again because he is still too limited. We did not wish to conclude that such a subject was satisfactorily returned to work. We measured a number of secondary outcomes at 3 years after fusion. These include number of days each subject was absent from work, net postoperative medical costs paid for by the BWC, prescription opioid analgesic utilization, new psychological comorbidity, number of psychotherapy sessions, and if these sessions were a continuation from sessions occurring before fusion, additional major lumbar surgical procedures, and rates of postsurgical infection, failed back surgery syndrome, pseudoarthrosis, all-cause mortality, and new permanent disability. We defined major lumbar surgical procedures as additional fusions and/or decompression procedures. See the Supplemental Digital Content Appendix, available at: http://links.lww.com/BRS/A967, for more detailed descriptions of how we measured each secondary outcome.
Additional Data Collection Prescription opioid data were limited to that administered by the oral route. We calculated the total number of days each 750
subject was supplied with opioid analgesics. Each opioid prescription was converted to morphine equivalent units, and the total morphine equivalent units supplied were measured. Also, we determined the average morphine equivalent units per day for each subject. We measured these variables both before index fusion (BIF) and at 3 years after index fusion (AIF). We further collected information on each subject’s age at index fusion, sex, approximated income, and presence of obesity. Specific incomes for each subject were unavailable. So, each subject’s zip code of residence was correlated to a mean per capita income value on the basis of the 2010 United States Census. In addition to depression, we identified subjects diagnosed with other psychological comorbidity, both BIF and within 3 years AIF. We quantified the number of psychotherapy sessions each subject used BIF and within 3 years AIF. We determined which subjects were still working within the same week as their index fusion, because many WC subjects are still able to work leading up to their surgery, albeit often on lighter duty and more sedentary jobs. Also, we determined preoperative rates of discography and legal representation.
Analysis To determine the impact preoperative depression had on RTW status after fusion, we used a multivariate logistic regression analysis. The dependent variable was whether or not our RTW criteria were met. We adjusted for a number of covariates within the regression model. We adjusted for the following binary variables: each lumbar and psychological comorbidity, preoperative discography, ability to continue working within the same week as fusion, obesity, sex, age more than 50 years at the time of fusion, legal representation, single versus multilevel fusion, concurrent decompression with fusion, additional lumbar surgery within 3 years AIF, and prescription opioid analgesia for more than 1 year
www.spinejournal.com
May 2015
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE140922_LR 750
14/04/15 8:55 PM
OCCUPATIONAL HEALTH/ERGONOMICS before fusion. We adjusted for the following continuous variables: approximated income, net morphine equivalent units supplied BIF, daily morphine equivalent units supplied BIF, and the number of days between each subject’s date of injury and date of index fusion. We adjusted for the following categorical variables: type of fusion surgery, instrumentation, and graft at index fusion. To compare secondary outcomes and baseline characteristics between the depression and control groups, we used χ2 tests for binary and categorical variables
Depression’s Impact on Lumbar Fusion Outcomes • Anderson et al
and t tests for continuous variables. P value of less than 0.05 was considered statistically significant. For all analyses, we used Statgraphics Centurion XVI software (Statpoint Technologies, Inc.; Warrenton).
RESULTS Table 2 includes important prefusion population characteristics. Compared with controls, subjects in the depression group differed significantly in the following ways: subjects
TABLE 2. Population Characteristics Before Fusion Characteristic
P
Depression Group
Control Group
123
2676
Age at index fusion (mean ± SD)
45.1 ± 9.5
42.9 ± 9.5
50 yr at fusion
36 (29.3%)
576 (21.5%)
OCCUPATIONAL HEALTH/ERGONOMICS
Clinical Depression Is a Strong Predictor of Poor Lumbar Fusion Outcomes Among Workers’ Compensation Subjects Joshua T. Anderson, BS,* Arnold R. Haas, BS, BA,† Rick Percy, PhD,† Stephen T. Woods, MD,† Uri M. Ahn, MD‡ and Nicholas U. Ahn, MD§
Study Design. Retrospective cohort study. Objective. Determine how psychosocial factors, particularly depression, impact lumbar fusion outcomes in a workers’ compensation (WC) setting. Summary of Background Data. WC patients are less likely to return to work (RTW) after fusion. Few studies evaluate risk factors within this clinically distinct population. Methods. A total of 2799 Ohio WC subjects were identified who underwent lumbar fusion between 1993 and 2013 using Current Procedural Terminology (CPT) procedural and International Classification of Diseases, Ninth Revision (ICD-9) diagnosis codes. A total of 123 subjects were diagnosed with depression before fusion. Subjects with a smoking history, prior lumbar surgery, permanent disability, and failed back syndrome were excluded. The primary outcome was whether subjects returned to work within 2 years of fusion and sustained this RTW for more than 6 months of the following year. To determine the impact depression had on RTW status, we performed a multivariate logistic regression analysis. We also compared time absent from work and other secondary outcomes using χ2 and t tests. Results. Subjects with preoperative depression had significantly higher rates of legal representation, degenerative lumbar disease, and higher medical costs, and used opioid analgesics for considerably longer before and after fusion (P < 0.001). From the *University Hospitals Case Medical Center Department of Orthopaedics, Case Western Reserve University School of Medicine, Cleveland, OH; †Ohio Bureau of Workers’ Compensation, Columbus, OH; ‡New Hampshire NeuroSpine Institute, Bedford, NH; and §University Hospitals Case Medical Center Department of Orthopaedics, Cleveland, OH. Acknowledgment date: July 14, 2014. First revision date: November 11, 2014. Second revision date: January 1, 2015. Acceptance date: January 10, 2015. The manuscript submitted does not contain information about medical device(s)/drug(s). No funds were received in support of this work. Relevant financial activities outside the submitted work: patents, consultancy, grants/grants pending, expert testimony, payment for lectures, royalties, travel accommodations, stock/stock options. Address correspondence and reprint requests to Joshua T. Anderson, BS, University Hospitals Case Medical Center Department of Orthopaedics, Case Western Reserve University School of Medicine, 11100 Euclid Ave, Cleveland, OH 44106; E-mail: [email protected] DOI: 10.1097/BRS.0000000000000863
748
Depression group (10.6% [13/123]) and controls (33.0% [884/2676]) met our RTW criteria (P < 0.001). Preoperative depression was a negative predictor of RTW status (P < 0.001; odds ratio [OR]: 0.38). Additional predictors included working during same week as fusion (OR: 2.15), age more than 50 years (OR: 0.58), chronic preoperative opioid analgesia (OR: 0.58), and legal representation (OR: 0.64). After surgery, depression subjects were absent from work 184 more days compared with controls (P < 0.001). Conclusion. Overall, RTW rates after fusion were low, which was especially true for those with pre-existing depression. Depression was a strong negative predictor of postoperative RTW status. Psychological screening and treatment may be beneficial in these subjects. The poor outcomes in this study may highlight a more limited role for fusion among WC subjects with chronic low back pain where RTW is the treatment goal. Key words: workers’ compensation, lumbar fusion, depression, biopsychosocial, return to work, failed back syndrome, opioid analgesics, psychosocial, risk factors, predictors, legal representation, degenerative. Level of Evidence: 3 Spine 2015;40:748–756
M
ultiple studies have demonstrated that US workers’ compensation (WC) subjects tend to have worse clinical outcomes after lumbar fusion than the general population.1–7 A 2010 case-control study reported that WC subjects, when compared with the general population, had significantly less improvement in mean Oswestry Disability Index and 36-Item Short Form Health Survey scores after fusion.3 Studies of subjects from several WC jurisdictions report return to work (RTW) rates ranging from 26% to 36% and reoperation rates from 22% to 27% after fusion.1,5,8–10 Despite these poor outcomes, fusion rates within the WC population are steadily rising.11 The estimated costs of treating low back pain in the United States exceed $100 billion a year, two-thirds of which are thought to arise from lost wages and reduced productivity.12,13 Mounting evidence highlights the importance of using a biopsychosocial model of care when treating patients with low back pain, including those with chronic discogenic low back pain receiving WC.5,6,8,14–19
www.spinejournal.com
May 2015
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE140922_LR 748
14/04/15 8:55 PM
OCCUPATIONAL HEALTH/ERGONOMICS TABLE 1. Coding Used for Study Fusion Technique
CPT Code
PLF Single level
22612
Multilevel
22612 + 22614
PLIF Single level
22630 or 22633
Multilevel
22630 or 22633 + 22632 or 22634
ALIF Single level
22558
Multilevel
22558 + 22585
360° fusion (ALIF + PLF/PLIF) Single level
SL ALIF + SL PLF/PLIF
Multilevel
ML ALIF + ML PLF/PLIF
Lumbar Comorbidity Lumbar disc degeneration
ICD-9 Code 722.52
Spondylolisthesis
738.4 or 756.12
Scoliosis
737.30–737.39
Spinal stenosis
724.02–724.03
Disc herniation
722.10
Spondylosis
721.3
Spondylolysis
756.11
Lumbar sprain
847.2
Radiculopathy
724.4 or 729.2
Depression Depressive disorder Major depressive disorder Dysthymic disorder
ICD-9 Code 311 296.20-296.26 or 296.30-296.36 300.4
This table includes the International Classification of Diseases, Ninth Revision (ICD-9) codes with which we characterized each subject’s lumbar comorbidities and identified which subjects were diagnosed with depression before fusion. It also contains the Current Procedural Terminology (CPT) coding we used to characterize each subject's fusion surgery. CPT indicates Current Procedural Terminology; PLF, posterior lumbar fusion; SL, single level; ML, multilevel; PLIF, posterior lumbar interbody fusion; ALIF, anterior lumbar interbody fusion; ICD, International Classification of Diseases.
One potential strategy to improve fusion outcomes within this clinically distinct subset of patients is to identify presurgical risk factors. Optimizing subject selection criteria for lumbar fusion would not only help provide the best and most appropriate care for these patients but it could also help curb medical and disability payments. Recent studies have Spine
Depression’s Impact on Lumbar Fusion Outcomes • Anderson et al
demonstrated the importance of psychosocial factors to outcomes after lumbar surgery.9,20–23 However, relatively few studies have been published evaluating specific predictors of lumbar fusion outcomes in US WC subjects.4,5,8,12 Furthermore, depression is a leading cause of disability as measured by years lived with disability as well as the fourth leading contributor to the global burden of disease.24 Epidemiologic studies support a lifetime prevalence of major clinical depression ranging from 16.2% to 17.9% within the United States.25–28 Studies have also shown that individuals experiencing depression are less likely to call on social supports or follow a plan of rehabilitation.29 Therefore, the objective of this study was to determine how psychosocial factors, depression in particular, impact lumbar fusion outcomes, primarily the ability to RTW within a reasonable time line postoperatively, in a WC setting. We performed this study at the University Hospitals Case Medical Center in Cleveland, Ohio. All data were collected in late July of 2013 from the Ohio Bureau of Workers’ Compensation’s (BWC’s) database.
MATERIALS AND METHODS Data Source Our retrospective cohort study includes 2799 WC subjects from the Ohio BWC who underwent lumbar fusion between 1993 and 2013. We used the administrative database at the Ohio BWC to acquire all data. This is the same database that Nguyen et al1 used to compare fusion versus nonoperative outcomes in 2011.
Subject Selection Initially, we identified 14,640 WC subjects diagnosed with lumbar conditions using International Classification of Diseases, Ninth Revision (ICD-9) codes. From this, we excluded 11,841 subjects who did not undergo lumbar fusion, were permanently disabled before fusion, underwent prior lumbar surgery before fusion, had a positive smoking history, and/or had a history of failed back syndrome because all of these factors have previously been reported to negatively impact fusion outcomes.6,17,30–33 Using Current Procedural Terminology (CPT) codes, we included subjects who underwent an anterior, posterior, or 360° lumbar fusion with at least 3 years of follow-up. Thus, we identified our final study population of 2799 subjects. Using ICD-9 codes, we further characterized each subject’s lumbar comorbidities. Next, we used ICD-9 codes to identify 123 subjects who were clinically diagnosed with depression before undergoing fusion, forming the depression group. We treated the remaining 2676 subjects as controls. See Table 1 for the ICD-9 and CPT codes used to characterize each subject’s fusion surgery, lumbar comorbidities, and identify depression. See the Supplemental Digital Content Appendix, available at: http://links.lww.com/ BRS/A967, for more specific information on how we derived our study population. Figure 1 illustrates our subject selection process. www.spinejournal.com
749
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE140922_LR 749
14/04/15 8:55 PM
OCCUPATIONAL HEALTH/ERGONOMICS
Depression’s Impact on Lumbar Fusion Outcomes • Anderson et al
Figure 1. Subject selection flow diagram. We used a combination of International Classification of Diseases, Ninth Revision (ICD-9) diagnosis and Current Procedural Terminology (CPT) procedural codes to arrive at our final study population.
Outcome Measures We measured outcomes after each subject’s index fusion surgery after WC-qualifying workplace injury. The primary outcome measure for this study was whether or not each subject made a sustainable RTW within a reasonable time line after lumbar fusion. Specifically, if a subject returned to work within 2 years of fusion and sustained this RTW for 6 months or more of the following year, we considered that subject as returned to work. Similar to the WC study conducted by Nguyen et al,1 we used 2 years as a general cutoff for RTW. However, we also included the additional requirement that each worker must sustain his or her RTW. For example, a patient may fulfill the criteria of making an RTW within 2 years but after a few days of work may leave again because he is still too limited. We did not wish to conclude that such a subject was satisfactorily returned to work. We measured a number of secondary outcomes at 3 years after fusion. These include number of days each subject was absent from work, net postoperative medical costs paid for by the BWC, prescription opioid analgesic utilization, new psychological comorbidity, number of psychotherapy sessions, and if these sessions were a continuation from sessions occurring before fusion, additional major lumbar surgical procedures, and rates of postsurgical infection, failed back surgery syndrome, pseudoarthrosis, all-cause mortality, and new permanent disability. We defined major lumbar surgical procedures as additional fusions and/or decompression procedures. See the Supplemental Digital Content Appendix, available at: http://links.lww.com/BRS/A967, for more detailed descriptions of how we measured each secondary outcome.
Additional Data Collection Prescription opioid data were limited to that administered by the oral route. We calculated the total number of days each 750
subject was supplied with opioid analgesics. Each opioid prescription was converted to morphine equivalent units, and the total morphine equivalent units supplied were measured. Also, we determined the average morphine equivalent units per day for each subject. We measured these variables both before index fusion (BIF) and at 3 years after index fusion (AIF). We further collected information on each subject’s age at index fusion, sex, approximated income, and presence of obesity. Specific incomes for each subject were unavailable. So, each subject’s zip code of residence was correlated to a mean per capita income value on the basis of the 2010 United States Census. In addition to depression, we identified subjects diagnosed with other psychological comorbidity, both BIF and within 3 years AIF. We quantified the number of psychotherapy sessions each subject used BIF and within 3 years AIF. We determined which subjects were still working within the same week as their index fusion, because many WC subjects are still able to work leading up to their surgery, albeit often on lighter duty and more sedentary jobs. Also, we determined preoperative rates of discography and legal representation.
Analysis To determine the impact preoperative depression had on RTW status after fusion, we used a multivariate logistic regression analysis. The dependent variable was whether or not our RTW criteria were met. We adjusted for a number of covariates within the regression model. We adjusted for the following binary variables: each lumbar and psychological comorbidity, preoperative discography, ability to continue working within the same week as fusion, obesity, sex, age more than 50 years at the time of fusion, legal representation, single versus multilevel fusion, concurrent decompression with fusion, additional lumbar surgery within 3 years AIF, and prescription opioid analgesia for more than 1 year
www.spinejournal.com
May 2015
Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. SPINE140922_LR 750
14/04/15 8:55 PM
OCCUPATIONAL HEALTH/ERGONOMICS before fusion. We adjusted for the following continuous variables: approximated income, net morphine equivalent units supplied BIF, daily morphine equivalent units supplied BIF, and the number of days between each subject’s date of injury and date of index fusion. We adjusted for the following categorical variables: type of fusion surgery, instrumentation, and graft at index fusion. To compare secondary outcomes and baseline characteristics between the depression and control groups, we used χ2 tests for binary and categorical variables
Depression’s Impact on Lumbar Fusion Outcomes • Anderson et al
and t tests for continuous variables. P value of less than 0.05 was considered statistically significant. For all analyses, we used Statgraphics Centurion XVI software (Statpoint Technologies, Inc.; Warrenton).
RESULTS Table 2 includes important prefusion population characteristics. Compared with controls, subjects in the depression group differed significantly in the following ways: subjects
TABLE 2. Population Characteristics Before Fusion Characteristic
P
Depression Group
Control Group
123
2676
Age at index fusion (mean ± SD)
45.1 ± 9.5
42.9 ± 9.5
50 yr at fusion
36 (29.3%)
576 (21.5%)
