RESEARCH—HUMAN—CLINICAL STUDIES RESEARCH—HUMAN—CLINICAL STUDIES
The Relationship Between Preoperative General Mental Health and Postoperative Quality of Life in Minimally Invasive Lumbar Spine Surgery BACKGROUND: In assessing poor lumbar surgery outcomes, researchers continue to investigate psychosocial predictors of patient postoperative quality of life. This is the first study of its kind to investigate this relationship in an exclusively minimally invasive patient sample. OBJECTIVE: To determine the association between preoperative mental health and postoperative patient-centered outcomes in patients undergoing minimally invasive lumbar surgery. METHODS: In 83 adults undergoing single-level minimally invasive lumbar spine surgery, Pearson correlation and partial correlation analyses were conducted between all demographic and clinical baseline variables and Oswestry Disability Index (ODI), Visual Analogue Scale (VAS), and 36-item Short-Form Health Survey Version 2.0 (SF-36v2) scores at 6 to 12 months postoperatively. SF-36v2 mental component summary scores (MCS) were used to assess pre- and postoperative general mental health. Post hoc analysis consisted of Pearson correlations between baseline SF-36v2, ODI, and VAS scores, and an identical set of correlations at outcomes. RESULTS: Preoperative MCS showed no significant association with outcomes VAS, ODI, or physical component summary scores. Baseline disability correlated significantly and more strongly with baseline MCS (P , .001, r = 20.40) than baseline pain levels (VAS back not significant, VAS leg P = .015, r = 0.27). Outcomes disability correlated significantly and more strongly with outcome back and leg pain levels (P , .001, r = 0.60 and 0.66) than outcome MCS (P = .031, r = 20.24). CONCLUSION: In a patient sample with mental health scores comparable to the population mean, there is no relationship between preoperative general mental health and postoperative patient-centered outcomes. Surgeons should consider the dynamic relationships between patient disability, mental health, and pain levels in assessing quality of life at different time points.
Rachel Asher, BS* Ashley E. Mason, PhD‡ Joseph Weiner, BS§ Richard G. Fessler, MD, PhD* *Department of Neurological Surgery, Rush University Medical Center, Chicago, Illinois; ‡Osher Center for Integrative Medicine, Department of Medicine, University of California at San Francisco, San Francisco, California; §Northwestern University Feinberg School of Medicine, Chicago, Illinois Correspondence: Richard G. Fessler, MD, PhD, Department of Neurological Surgery, Rush University Medical Center, 1725 W. Harrison St, Suite 855, Chicago, IL 60612. E-mail: [email protected] Received, September 11, 2014. Accepted, January 14, 2015. Published Online, February 14, 2015. Copyright © 2015 by the Congress of Neurological Surgeons.
KEY WORDS: Lumbar surgery outcomes, Mental health, Minimally invasive Neurosurgery 76:672–679, 2015
DOI: 10.1227/NEU.0000000000000695
I
n the United States, low back pain is one of the most common presenting concerns in clinical visits. The 2004 National Health Interview
ABBREVIATIONS: MCS, 36 Item Short-Form Health Survey Version 2.0 mental component summary score; MH, mental health; MIS, minimally invasive spine surgery; ODI, Oswestry Disability Index Version 1.3; PCS, 36 Item Short-Form Health Survey Version 2.0 physical component summary score; QOL, quality of life; SF36v2, 36 Item Short-Form Health Survey Version 2.0; VAS, Visual Analogue Scale
672 | VOLUME 76 | NUMBER 6 | JUNE 2015
www.neurosurgery-online.com
Survey indicated more than 25% prevalence of low back pain among respondents.1 Between 1995 and 1999, spine surgery rates in the Medicare population increased by 40%.2 From 1998 to 2008, data from the Nationwide Inpatient Sample indicates a 137% increase in spinal fusion rates.3 In comparison with open spinal procedures, the advent of minimally invasive lumbar surgery is associated with more rapid mobilization and discharge,4 as well as decreased perioperative surgical site infection5 and blood loss.6 However, concomitant changes
www.neurosurgery-online.com
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
MENTAL HEALTH AND QUALITY OF LIFE MIS
in longer-term quality of life do not always follow such improvements in short-term outcomes. This suggests that factors other than surgical approach may contribute to variations in postoperative quality of life. Although psychological predictors of postoperative quality of life have frequently been investigated in open procedures, the literature in minimally invasive populations is very limited. The intent of our study was to assess the relationship between preoperative mental health and postoperative quality of life (QOL) in minimally invasive populations. Since the observation of an association between hypochondria, hysteria, depression, and spinal surgery outcomes,7 investigation of preoperative psychological status as an outcome predictor has increased. Current measurement tools assess a diverse array of psychological constructs, such as patient expectations, fearavoidance behavior, depression, and somatic perception. Although predictive value or association between depression and lumbar surgery outcomes has been validated in many studies,7-12 others have found no such association.7-9,13 This lack of consensus reflects a larger trend in the literature on psychological distress and lumbar surgical outcomes. Reviews address the difficulty in coming to consensus regarding the role of preoperative psychological status in outcomes, given the variety and quality of psychological measurement tools,9 and the confounding influence of psychological distress during diagnostic testing for surgically remediable pathology.7-9 Some research suggests that, during discography, psychological distress correlates with more frequent pain reporting.14 The assessment of the literature on mental health and postoperative QOL is further complicated by the clinical heterogeneity of lumbar surgical patients. In particular, before surgical intervention, patients undergoing fusion tend to have been in pain much longer and had more previous surgeries than patients undergoing decompression and discectomy. A fusion is also a comparatively larger operation with a more painful, longer recovery. While comparing the results of studies with exclusively fusion or nonfusion samples, it is necessary to consider how such differences may affect the relationship between preoperative mental health and postoperative QOL. Although psychometrics specific to a particular psychological diagnosis or construct are common in the literature, using a general-health measure presents the advantage of minimizing respondent and data collection burdens.15 The SF-36 QOL measurement tool has a mental component summary score (MCS) that assesses general mental health. While important specific conditions such as depressed mood are a component of the SF-36, the metric tests overall mental health rather than depression specifically. The SF-36 has frequently been used to assess the relationship between preoperative mental health and QOL outcomes.16-21 Five of 6 studies found a significant relationship between SF-36 mental health scores and outcome.18 Although the relationship between preoperative mental health and outcomes in open lumbar surgeries is frequently addressed, particular focus on minimally invasive approaches is rare. To our knowledge, this is the first study to investigate the relationship
NEUROSURGERY
between preoperative general mental health and postoperative QOL in a minimally invasive lumbar surgical population. In this study, we investigated the relationships between preoperative mental health and patient-centered outcomes in minimally invasive lumbar surgical procedures.
METHODS Patient Population We conducted a retrospective analysis of prospectively collected data from 83 adult patients undergoing minimally invasive lumbar surgery (ages 29-82) with baseline data between the dates of March 31, 2008 and August 31, 2012. Fifty-one fusions, 18 decompressions, and 14 discectomies, all single-level, were performed. The institutional review board at Northwestern University approved the present study. The inclusion criteria were patients over age 18 who underwent single-level minimally invasive lumbar spine surgery performed by the senior author. Patients meeting one or more of the following conditions were excluded: infection, multilevel operations, previous fusion procedures, pending or past spine litigation, and workers’ compensation recipients. Baseline demographic, clinical, and surgical characteristics of the patient sample are shown in Tables 1 and 2. Baseline and outcomes Oswestry Disability Index Version 1.3 (ODI), Visual Analogue Scale (VAS), MCS, and physical component summary (PCS) scores are listed in Table 2.
Measures Mental Health The 36-item Short-Form Health Survey Version 2.022 (SF-36v2) was used to assess mental health. The scoring algorithm to calculate MCS uses scores from each subdomain to generate a final score. The 4 subdomains most closely related to mental health are given the heaviest weighting in this calculation. Of the 8 health domains measured in the SF-36, the 5-item Mental Health Index, which made up the Mental Health (MH) domain, had been most extensively evaluated by the time of the initial SF36 publications.23 It contains 5 items that assess 4 major MH components—anxiety, depression, loss of behavioral or emotional control, and psychological well-being. In the validation studies published in the 1990s by Ware et al,22 the MH domain was shown to be the most valid predictor of MH. All SF-36v2 scores were obtained using norm-based scoring algorithms; scores range from 0 to 100 with a population mean of 50. Scores lower than 50 represent poorer health relative to the population mean.
Patient-Centered Outcomes Measures All patient-centered outcomes were obtained at baseline and 6 to 12 months postoperatively. The SF-36v2 MCS, PCS, and the ODI version 1.324 were used to assess general physical health. ODI scores range from 0 to 100, with higher scores indicating increasing disability. Back and leg pain were measured by the VAS.
Statistical Analysis Data were analyzed with SAS (SAS Institute Inc, 2011). Groups were collapsed when possible to minimize the influences of sample size. Continuous variables were examined for extreme departures from normality. Missing data points were managed by using multiple imputation by chained equations (also called fully conditional specification).25 PROC
VOLUME 76 | NUMBER 6 | JUNE 2015 | 673
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
ASHER ET AL
TABLE 1. Patient Sample Characteristics at Baselinea
Sex, female Marital status Unmarried Married/engaged Level of education Less than college 4-year college degree Graduate/professional degree Work status Currently working Short-term disability Not employed Retired Homemaker/other Smoker Alcohol use Never 1-7 drinks/wk .8 drinks/wk OTC/NSAID medication use Never Intermittent Daily Narcotic pain medication use Never Intermittent Daily Exercise frequency Never, disability prevents Never Once or less/wk Twice/wk Three times/wk Four or more times/wk Comorbidities History of cancer Previous lumbar surgery (all decompression or discectomy) Primary diagnosis Disc pathology Spondylolisthesis Stenosis Other Surgical category Decompression Discectomy Fusion a
n
%
42
51
15 68
19 82
21 18 29
31 26 43
36 2 2 29 7 4
47 3 3 38 9 5
19 38 11
28 56 16
10 34 36
12 42 45
46 15 17
59 19 22
25 9 4 7 7 14 65 12 24
38 14 6 11 11 21 78 14 29
33 27 21 1
40 33 26 1
18 14 51
22 17 61
NSAID, nonsteroidal anti-inflammatory drug; OTC, other the counter.
MI in SAS was used. All potential predictors and outcomes were used in the imputation models. Logistic regression was used to impute the dichotomous variables and regression or regression with predictive mean matching was used to impute the continuous variables. Ten imputed data sets were created.
674 | VOLUME 76 | NUMBER 6 | JUNE 2015
From the imputed data sets, 23 sociodemographic, clinical, and QOL baseline variables were selected for correlation analysis. Bivariate correlations were performed between each variable and patient-centered outcomes metric (Table 3). Partial correlations were also performed between outcome variables and potential predictors, after partialling out baseline score for corresponding outcome (Table 4). The significance level was set at P , .05 for entry into regression analysis. Because baseline MCS did not significantly correlate with patient-centered outcomes in either analysis (with the exception of outcomes MCS), multivariate regressions were not performed. Given the relatively high baseline and outcomes MCS score in comparison with the literature, post hoc analysis was performed. t tests between the lowest 25% and highest 25% of baseline MCS, and each outcomes variable, were conducted. Correlations between baseline ODI, VAS back and leg, PCS, and baseline MCS were assessed (Table 5). An identical set of correlations was performed between corresponding metrics at outcome and outcomes MCS scores (Table 6). Because this study included fusion, decompression, and discectomy procedures, additional post hoc analysis was conducted to investigate the possible moderating effects of surgical category on the relationship between preoperative MCS and postoperative QOL. Because regressions were underpowered owing to small sample size, surgical category was collapsed to fusion vs nonfusion. A series of t tests with the Levene test for equality of variances was performed between surgical category and baseline MCS, PCS, VAS back, VAS leg, and ODI. An identical set of t tests was performed with corresponding metrics at outcomes. Maineffect regressions on outcomes QOL (VAS back, VAS leg, PCS, and ODI) were conducted for surgical category, controlling for baseline QOL and MCS. The interaction between surgical category and baseline MCS was investigated in an identical set of regressions, differing only by inclusion of the interaction term.
RESULTS Mean baseline scores for VAS back and leg pain, ODI, SF-36v2, PCS, and MCS were 5.4, 5.9, 34.2, 34.5, and 50.2, respectively. Mean outcomes scores were 2.2, 2.2, 18.0, 44.5, and 55.1, respectively. Patients improved from moderately disabled at baseline (ODI range 20-40), to minimally disabled at outcomes (ODI range 0-20). Likewise, patients improved 10 points in PCS scores, equivalent to 1 standard deviation in the SF-36v2 metric. MCS scores improved 4.9 points, to approximately 0.5 standard deviations above the population mean. Preoperative MCS did not significantly correlate with VAS back, leg, ODI, or PCS outcomes scores. Thus, regressions between MCS, other significant predictors, and outcomes were not performed. Results of post hoc t tests between lower and upper quartiles of MCS and outcomes were also insignificant. Results for post hoc correlations between each QOL metric (ODI, PCS, VAS back, VAS leg, and MCS) are reported both in terms of r and P values. R values represent the correlation coefficient. Correlations between baseline MCS and baseline ODI were significant at P , .001, r = 20.40. Correlations between baseline ODI and baseline PCS were significant at P , .001, r = 20.60. However, correlations between baseline VAS back pain and baseline ODI were insignificant. Correlations between baseline VAS leg pain and baseline ODI were significant at P = .015, r = 0.27.
www.neurosurgery-online.com
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
MENTAL HEALTH AND QUALITY OF LIFE MIS
TABLE 2. Baseline and Postoperative Patient Sample Characteristicsa Preoperative Mean (Standard Deviation) Age, y BMI ODI score VAS back score VAS leg score SF-36v2 MCS score SF-36v2 PCS score
60.7 27.4 34.2 5.4 5.9 50.2
Postoperative Mean (Standard Deviation)
(13.7) (5.1) (13.2) (2.5) (2.7) (11.4)
— — 18.0 (16.2) 2.2 (2.2) 2.2 (2.4) 55.1 (6.9)
34.5 (7.6)
44.5 (10.2)
a
BMI, body mass index; ODI, Oswestry Disability Index Version 1.3; VAS, Visual Analogue Scale; SF36v2, 36 Item Short-Form Health Survey Version 2.0; MCS, SF36v2 mental component summary score; PCS, SF36v2 physical component summary score.
Correlations between each QOL measure at outcomes changed relative to baseline. Correlations between outcomes VAS back and leg and outcomes ODI increased to r = 0.60 (P , .001) and r = 0.66
(P , .001), respectively. With the use of the Cohen guidelines for effect size,26 this represents a change from nonsignificant or small correlations at baseline to high correlations at outcomes. The correlation between MCS and ODI decreased from r = 20.40 (P , .001) at baseline, to r = 20.24 (P = .031) at outcomes. This represents a change from a medium correlation preoperatively to a small correlation postoperatively. The correlation between outcomes PCS and outcomes ODI increased to 20.87 (P , .001). Correlations between MCS and VAS (back and leg) were not found to be significant at baseline or outcomes. Post hoc t tests on surgical category and QOL metrics were found to be insignificant, with the exception of outcomes MCS. Although patients not undergoing fusion were found to have significantly higher outcomes MH (P = .038 equal variances assumed, P = .030 equal variances not assumed) than patients undergoing fusion, this outcome fell to insignificance with Bonferroni correction for multiple comparisons. Results for regressions examining the main effect of surgical category on outcomes QOL were insignificant. Regressions on outcomes QOL examining the interaction between surgical category and baseline MCS were also insignificant.
TABLE 3. Bivariate Correlations Between Outcomes and Potential Predictorsa Sample Characteristics Baseline ODI Baseline PCS Baseline MCS Baseline VAS-back Baseline VAS-leg Sex Age Marital status Education level Work status Alcohol consumption Exercise OTC/NSAID use Narcotic use BMI Comorbidities History of cancer Previous lumbar surgery Primary indication—back pain Primary indication—leg pain Primary diagnosis Type of surgery Blood loss
Outcome ODI b
0.33 20.33b 20.15 0.05 20.02 0.03 0.05 20.09 20.07 20.06 20.06 20.17 0.19d 0.02 0.19 0.16 0.17 0.11 0.18d 0.03 0.04 20.03 20.03
Outcome PCS
Outcome MCS
Outcome VAS-Back
Outcome VAS-Leg
20.16 0.30b 0.06 20.06 20.01 0.05 20.09 0.10 0.08 0.07 0.06 0.10 20.28c 0.08 20.21 20.11 20.15 20.10 20.14 20.03 20.09 0.03 0.03
20.28 0.06 0.56e 20.27c 20.09 20.12 0.08 20.05 0.19 0.26c 0.23d 0.21d 0.08 20.04 0.02 0.01 20.06 0.08 20.11 20.02 0.11 20.23c 20.02
20.01 20.06 0.04 0.32b 20.01 20.06 20.01 20.09 20.26c 20.01 0.03 20.24d 0.18 0.15 0.20 0.17 0.12 0.00 0.35b 20.02 0.07 20.09 20.04
0.16 20.20d 20.01 0.09 0.17 0.12 0.18 20.15 20.16 20.04 20.17 20.31c 0.17 20.13 0.19 0.22d 0.05 20.02 20.02 0.11 20.01 20.17 20.06
c
a
BMI, body mass index; ODI, Oswestry Disability Index Version 1.3; VAS, Visual Analogue Scale; SF36v2, 36 Item Short-Form Health Survey Version 2.0; MCS, SF36v2 mental component summary score; PCS, SF36v2 physical component summary score; NSAID, nonsteroidal anti-inflammatory drug; OTC, over the counter. P , .01. c P , .05. d P , .10. e P , .001. b
NEUROSURGERY
VOLUME 76 | NUMBER 6 | JUNE 2015 | 675
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
ASHER ET AL
TABLE 4. Bivariate Correlations Between Outcomes and Potential Predictors, After Partialling Out Baseline Score for Corresponding Outcomea Sample Characteristics Baseline ODI Baseline PCS Baseline MCS Baseline VAS-back Baseline VAS-leg Gender Age Marital status Education level Work status Alcohol consumption Exercise OTC/NSAID use Narcotics use BMI Comorbidities History of cancer Previous lumbar surgery Primary indication—back pain Primary indication—leg pain Primary diagnosis Type of Surgery Blood loss
Outcome ODI
Outcome PCS
Outcome MCS
Outcome VAS-Back
Outcome VAS-Leg
— 20.18 20.02 0.02 20.12 20.09 0.06 20.07 20.07 0.02 20.05 20.10 0.13 20.07 0.21c 0.15 0.23d 0.14 0.18 20.04 0.01 20.07 20.02
0.04 — 0.10 0.00 0.07 0.14 20.07 0.09 0.07 0.04 0.03 20.05 20.29d 0.13 20.18 20.11 20.17 20.11 20.14 0.03 20.06 0.04 0.04
20.07 0.15 — 20.32b 20.05 20.14 20.04 20.12 0.14 0.16 0.15 0.30d 0.21c 0.08 20.08 0.00 20.07 0.20 20.13 0.01 0.15 20.18 20.06
20.06 20.01 0.06 — 20.12 20.13 20.02 20.10 20.20c 0.01 0.07 20.18 0.21c 0.12 0.26d 0.17 0.10 0.04 0.23d 20.05 0.10 20.12 0.00
0.12 20.16 0.01 0.04 — 0.08 0.15 20.15 20.14 20.01 20.15 20.25d 0.16 20.12 0.17 0.21c 0.08 20.02 20.01 0.04 0.01 20.16 20.03
a
BMI, body mass index; ODI, Oswestry Disability Index Version 1.3; VAS, Visual Analogue Scale; SF36v2, 36 Item Short-Form Health Survey Version 2.0; MCS, SF36v2 mental component summary score; PCS, SF36v2 physical component summary score; NSAID, nonsteroidal anti-inflammatory drug; OTC, over the counter. b P , .01. c P , .10. d P , .05.
DISCUSSION In this retrospective analysis of prospectively collected data, preoperative general MH was not significantly associated with postoperative physical patient-centered outcomes. Preoperative MH showed no statistically significant relationship with ODI, VAS, or SF-36v2 PCS scores at 6- to 12-month outcomes. Results suggest that, contrary to much of the literature, preoperative
TABLE 5. Bivariate Correlations Between Patient-Centered Metrics at Baselinea
BaseODI BasePCS BaseMCS
BaseODI BasePCS BaseMCS
BaseVAS Back
BaseVAS Leg
20.5971b 20.3992b — 20.1351 — —
0.0983 20.1676 20.0101
0.2690c 20.2773c 20.0999
— — —
a
ODI, Oswestry Disability Index Version 1.3; VAS, Visual Analogue Scale; SF36v2, 36 Item Short-Form Health Survey Version 2.0; MCS, SF36v2 mental component summary score; PCS, SF36v2 physical component summary score. b P , .001. c P , .05.
676 | VOLUME 76 | NUMBER 6 | JUNE 2015
psychological status may not play as significant a role in postoperative physical QOL outcome. Relative to other studies that used the SF-36 MCS as a predictor, this study population has better overall MH. Thus, it could be stated that, in a population of patients with relatively normal preoperative MCS scores, MCS is not significantly related to postoperative outcome. However, certain elements of both study design and patient sample necessitate moderation when interpreting these conclusions in the context of the literature. Variations in statistical methodology, particularly methodological lapses in some studies, are key limiting factors in the interpretation of conclusions. A review by Celestin et al9 on psychosocial predictors of lumbar surgery and spinal cord stimulation outcomes cited that many studies did not control for preoperative levels of pain and disability. As an example, Moraci et al27 found preoperative depression to be a significant predictor of postoperative pain levels in lumbar microdiskectomy, but did not control for preoperative pain levels in statistical analyses. Thus, the reliability of these results is questionable. To our knowledge, this is the only study to examine the relationship between preoperative emotional health and postoperative QOL in an exclusively minimally invasive population. Furthermore, there are many different questionnaires and constructs used to assess MH as an outcomes predictor. Although
www.neurosurgery-online.com
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
MENTAL HEALTH AND QUALITY OF LIFE MIS
TABLE 6. Bivariate correlations Between Patient-Centered Metrics at Outcomesa
Outcomes ODI Outcomes PCS Outcomes MCS
Outcomes ODI
OutcomesPCS
Outcomes MCS
Outcomes VAS Back
Outcomes VAS Leg
— — —
20.8701b — —
20.2427c 0.1157 —
0.6046b 20.5843b 20.1944
0.6639b 20.6208b 20.1899
a
ODI, Oswestry Disability Index Version 1.3; VAS, Visual Analogue Scale; SF36v2, 36 Item Short-Form Health Survey Version 2.0; MCS, SF36v2 mental component summary score; PCS, SF36v2 physical component summary score. b P , .001. c P , .05.
general MH may not show strong association with outcome, the relationship between more specific psychological conditions and outcome may differ. Studies which assessed somatization, depression, and anxiety were investigated often in literature reviews, with results frequently being significant.9,28,29 In terms of study design, the limitations of not using an open control include the possibility that our results may generalize to open studies. We are not aware of any minimally invasive spine surgery (MIS) studies on the relationship between preoperative MH and postoperative QOL that have used an open control. However, our group is currently performing a prospective investigation of patient expectations in open vs MIS which includes the possibility for future MCS comparative analysis. Of the 4 studies in open lumbar surgical populations which used the SF-36, mean preoperative MCS scores were 41.54,21 37.05,17 35.37,20 and 43.77.19 Edwards et al16 did not use norm-based scoring, and compared their mean baseline MH score of 63.2 to the mean MH score of 80 in a general population of patients with degenerative lumbar spines. The distribution of MH scores at baseline was described as similar to patients with mild depressive disorders.16 Each of these studies found higher preoperative MCS scores to be significantly associated with better QOL outcomes. Of the 83 patients in this study, the mean SF-36 MCS score at baseline was 50.2, which is not statistically distinct from the population norm of 50. The divergence in conclusions in comparison with this study may be a function of the distribution of emotional distress in the studied populations. It is possible that, in order to begin finding a significant association between baseline MH and outcome, our patient population would require lower mean MH scores. The purpose of the post hoc t tests between the upper and lower MCS quartiles and QOL outcomes was to test this possibility. However, t tests were not significant. Although this provides stronger evidence for a lack of association between preoperative MH and QOL outcomes, it is not necessarily conclusive. The 25th percentile of baseline MCS was 40.6, still greater than the mean MCS for 2 of the studies referenced above. t tests between lower and higher MCS percentiles could not be performed because of small sample size. With the exception of the discectomy sample in Edwards et al,16 the comparable open studies described above are inclusive only of
NEUROSURGERY
patients undergoing fusion. Because this study included fusion, decompression, and discectomy procedures, post hoc tests were performed to facilitate comparison among these groups. Small sample size necessitated collapsing the 3 categories to fusion vs nonfusion. Post hoc regressions on QOL outcomes did not indicate a significant interaction between baseline MH and surgical category. Given these limitations, we were unable to draw conclusions on different effects of MH across minimally invasive fusions, decompressions, and discectomies. To our knowledge, these moderating relationships have not yet been investigated in the literature. Notably, post hoc t tests revealed no significant differences in baseline or outcomes MCS, PCS, ODI, or VAS for fusion vs nonfusion groups. Although fusion and nonfusion groups are often quite distinct clinically, QOL metrics are responsive to both sociodemographic and clinical differences. It is certainly possible that the lack of significant differences in QOL observed across fusion vs nonfusion groups is partially reflective of patient demographics and inclusion criteria. First, the selected patient population reflects that of the surgeon’s practice, which in this sample is distinctive in terms of education level, smoking status, employment status, baseline MH, disability, and pain levels. The patient population for this study is highly educated, almost exclusively white, and mostly retired or currently working. Only 4 of 83 patients are smokers. Very few patients are unemployed, and none have past/pending litigation or workers’ compensation status. In particular, work status, sick leave, workers’ compensation, and spine litigation have been shown to associate with poor outcomes.30-32 For studies that do include these populations, the role of MH in determining QOL outcomes may be more significant. The exclusion of patients with multilevel fusions, history of prior fusion, and spine litigation or workers’ compensation claims could have created a fusion sample with a QOL profile more similar to patients undergoing decompression and discectomy. These criteria, in combination with the favorable employment and smoking status of patients, could have further minimized the differences between fusion and nonfusion groups. Furthermore, mean baseline disability, leg pain, and back pain are moderate overall, at 34.2, 5.4, and 5.9 points, respectively. The mean preoperative disability is lower,17,18,20,21 VAS back and leg
VOLUME 76 | NUMBER 6 | JUNE 2015 | 677
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
ASHER ET AL
pain levels lower,17,20,21 and QOL higher17-19,21 in comparison with similar open studies using the SF-36 MCS as a predictor. Because these studies did find a direct relationship between MH and postoperative QOL outcomes, it is possible that the results of this study could differ markedly in a more severely disabled patient population with poorer MH. Specifically, a significant interaction between baseline disability and MH could arise, such that patients with more severe disabilities could exhibit a stronger relationship between baseline MH and postoperative QOL outcomes. In this study, baseline disability correlated significantly and more strongly with baseline MH than baseline pain levels. Surprisingly, correlations between VAS back pain and disability at baseline were insignificant. Correlations between baseline leg pain and baseline disability were significant but with small effect size. However, the correlations between MH, disability, and pain levels changed dramatically at outcome. Correlations between pain levels and disability changed from nonsignificant or small at baseline to high correlations at outcomes. The correlation between MH and disability decreased from medium preoperatively to small postoperatively. It is necessary to note that our results do not allow for conclusions regarding causality. However, despite the lack of association between baseline MH and patient-centered outcomes, MH still played an important role in the interpretation of patient disability both preoperatively and postoperatively. Baseline correlations indicate that MH, rather than pain, is more closely associated with baseline disability. The relationship between pain and disability in low back pain populations is somewhat controversial in the literature, with studies finding associations ranging from modest to strong.33 The variation in this relationship could be due to associations between disability and other psychosocial or demographic variables. Therefore, the preoperative assessment of MH by surgeons could facilitate the task of interpreting and understanding disability as a measure of QOL. Given the prevalence of MH literature in lumbar surgery, some surgeons administer MH questionnaires preoperatively to screen for patients at risk for poor outcomes. Theoretically, this separation of patients into more easily classifiable groups facilitates the interpretation of outcomes. In doing so, however, surgeons run the risk of assuming that the psychological features of these groups are relatively static. The results of this study suggest that surgeons should be cautious in interpreting patient disability and pain in the context of MH. A patient whose MH is believed to play a significant role in his or her disability at baseline may not exhibit those same tendencies at outcome.
CONCLUSION The primary objective of our study was to assess the relationship between preoperative MH and postoperative QOL in patients undergoing minimally invasive lumbar surgery. In a patient sample with MH scores comparable to the population mean, there is no association between preoperative
678 | VOLUME 76 | NUMBER 6 | JUNE 2015
general MH and postoperative patient-centered outcomes. Our results demonstrate that, although MH is more strongly correlated to disability than pain at baseline, the reverse trend is observed at outcome. This suggests that the surgical procedure itself may act as a psychological as well as physical intervention,34 modifying the relationship between MH and disability. To more effectively interpret disability as a clinical construct, spine surgeons must take into account the dynamic nature of the relationships between pain, disability, and MH. Future prospective research on other psychological constructs, such as patient expectations, and postoperative QOL outcomes in MIS lumbar surgery is warranted. Disclosure The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article.
REFERENCES 1. Health, United States, 2006: With Chartbook on Trends in the Health of Americans. Hyattsville, MD: National Center for Health Statistics: 2006. 2. Deyo RA, Mirza SK. The case for restraint in spinal surgery: does quality management have a role to play? Eur Spine J. 2009;18(suppl 3):331-337. 3. Rajaee SS, Bae HW, Kanim LE, Delamarter RB. Spinal fusion in the United States: analysis of trends from 1998 to 2008. Spine (Phila Pa 1976). 2012;37(1): 67-76. 4. Wang MY, Lerner J, Lesko J, McGirt MJ. Acute hospital Costs after minimally invasive versus open lumbar interbody fusion: data from a US National Database with 6106 patients. J Spinal Disord Tech. 2012;25(6):324-328. 5. McGirt MJ, Parker SL, Lerner J, Engelhart L, Knight T, Wang MY. Comparative analysis of perioperative surgical site infection after minimally invasive versus open posterior/transforaminal lumbar interbody fusion: analysis of hospital billing and discharge data from 5170 patients. J Neurosurg Spine. 2011;14(6):771-778. 6. Karikari IO, Isaacs RE. Minimally invasive transforaminal lumbar interbody fusion: a review of techniques and outcomes. Spine (Phila Pa 1976). 2010;35 (26 suppl):S294-S301. 7. Mannion AF, Elfering A, Staerkle R, et al. Predictors of surgical outcome and their assessment. Eur Spine J. 2006;15(suppl 1):S93-S108. 8. Mannion AF, Elfering A, Staerkle R, et al. Predictors of multidimensional outcome after spinal surgery. Eur Spine J. 2007;16(6):777-786. 9. Celestin J, Edwards RR, Jamison RN. Pretreatment psychosocial variables as predictors of outcomes following lumbar surgery and spinal cord stimulation: a systematic review and literature synthesis. Pain Med. 2009;10(4):639-653. 10. Chaichana KL, Mukherjee D, Adogwa O, Cheng JS, McGirt MJ. Correlation of preoperative depression and somatic perception scales with postoperative disability and quality of life after lumbar discectomy. J Neurosurg Spine 2011;14(2):261-267. 11. Sinikallio S, Aalto T, Lehto SM, et al. Depressive symptoms predict postoperative disability among patients with lumbar spinal stenosis: a two-year prospective study comparing two age groups. Disabil Rehabil. 2010;32(6):462-468. 12. Adogwa O, Parker SL, Shau DN, et al. Preoperative Zung Depression Scale predicts outcome after revision lumbar surgery for adjacent segment disease, recurrent stenosis, and pseudarthrosis. Spine J. 2012;12(3):179-185. 13. Hobby JL, Lutchman LN, Powell JM, Sharp DJ. The distress and risk assessment method (DRAM). J Bone Joint Surg Br. 2001;83(1):19-21. 14. Block AR, Vanharanta H, Ohnmeiss DD, Guyer RD. Discographic pain report. Influence of psychological factors. Spine (Phila Pa 1976). 1996;21(3):334-338. 15. Walsh TL, Hanscom B, Lurie JD, Weinstein JN. Is a condition-specific instrument for patients with low back pain/leg symptoms really necessary? The responsiveness of the Oswestry Disability Index, MODEMS, and the SF-36. Spine (Phila Pa 1976). 2003;28(6):607-615. 16. Edwards RR, Klick B, Buenaver L, et al. Symptoms of distress as prospective predictors of pain-related sciatica treatment outcomes. Pain. 2007;130(1-2): 47-55.
www.neurosurgery-online.com
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
MENTAL HEALTH AND QUALITY OF LIFE MIS
17. Carreon LY, Glassman SD, Djurasovic M, et al. Are preoperative health-related quality of life scores predictive of clinical outcomes after lumbar fusion? Spine (Phila Pa 1976). 2009;34(7):725-730. 18. Abbott AD, Tyni-Lenne R, Hedlund R. Leg pain and psychological variables predict outcome 2-3 years after lumbar fusion surgery. Eur Spine J. 2011;20(10): 1626-1634. 19. Derby R, Lettice JJ, Kula TA, Lee SH, Seo KS, Kim BJ. Single-level lumbar fusion in chronic discogenic low-back pain: psychological and emotional status as a predictor of outcome measured using the 36-item short form. J Neurosurg Spine. 2005;3(4):255-261. 20. Cobo Soriano J, Sendino Revuelta M, Fabregate Fuente M, Cimarra Diaz I, Martinez Urena P, Deglane Meneses R. Predictors of outcome after decompressive lumbar surgery and instrumented posterolateral fusion. Eur Spine J. 2010;19(11): 1841-1848. 21. Trief PM, Ploutz-Snyder R, Fredrickson BE. Emotional health predicts pain and function after fusion: a prospective multicenter study. Spine (Phila Pa 1976). 2006;31(7):823-830. 22. Ware JE Jr. SF-36 health survey update. Spine (Phila Pa 1976). 2000;25(24): 3130-3139. 23. Tarlov AR, Ware JE Jr, Greenfield S, Nelson EC, Perrin E, Zubkoff M. The Medical Outcomes Study. An application of methods for monitoring the results of medical care. JAMA. 1989;262(7):925-930. 24. Fritz JM, Irrgang JJ. A comparison of a modified Oswestry Low Back Pain Disability Questionnaire and the Quebec Back Pain Disability Scale. Phys Ther. 2001;81(2):776-788.
NEUROSURGERY
25. White IR, Royston P, Wood AM. Multiple imputation using chained equations: issues and guidance for practice. Stat Med. 2011;30(4):377-399. 26. Cohen JA. A power primer. Psychol Bull. 1992;112(1):155-159. 27. Arpino L, Iavarone A, Parlato C, Moraci A. Prognostic role of depression after lumbar disc surgery. Neurol Sci. 2004;25(3):145-147. 28. den Boer JJ, Oostendorp RA, Beems T, Munneke M, Oerlemans M, Evers AW. A systematic review of bio-psychosocial risk factors for an unfavourable outcome after lumbar disc surgery. Eur Spine J. 2006;15(5):527-536. 29. Aalto TJ, Malmivaara A, Kovacs F, et al. Preoperative predictors for postoperative clinical outcome in lumbar spinal stenosis: systematic review. Spine (Phila Pa 1976). 2006;31(18):E648-E663. 30. Gum JL, Glassman SD, Carreon LY. Is type of compensation a predictor of outcome after lumbar fusion? Spine (Phila Pa 1976). 2013;38(5):443-448. 31. DeBerard MS, LaCaille RA, Spielmans G, Colledge A, Parlin MA. Outcomes and presurgery correlates of lumbar discectomy in Utah Workers’ Compensation patients. Spine J. 2009;9(3):193-203. 32. Silverplats K, Lind B, Zoëga B, et al. Clinical factors of importance for outcome after lumbar disc herniation surgery: long-term follow-up. Eur Spine J. 2010;19(9): 1459-1467. 33. Peters ML, Vlaeyen JW, Weber WE. The joint contribution of physical pathology, pain-related fear and catastrophizing to chronic back pain disability. Pain. 2005; 113(1-2):45-50. 34. Lebow R, Parker SL, Adogwa O, et al. Microdiscectomy improves pain-associated depression, somatic anxiety, and mental well-being in patients with herniated lumbar disc. Neurosurgery. 2012;70(2):306-311; discussion 311.
VOLUME 76 | NUMBER 6 | JUNE 2015 | 679
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
The Relationship Between Preoperative General Mental Health and Postoperative Quality of Life in Minimally Invasive Lumbar Spine Surgery BACKGROUND: In assessing poor lumbar surgery outcomes, researchers continue to investigate psychosocial predictors of patient postoperative quality of life. This is the first study of its kind to investigate this relationship in an exclusively minimally invasive patient sample. OBJECTIVE: To determine the association between preoperative mental health and postoperative patient-centered outcomes in patients undergoing minimally invasive lumbar surgery. METHODS: In 83 adults undergoing single-level minimally invasive lumbar spine surgery, Pearson correlation and partial correlation analyses were conducted between all demographic and clinical baseline variables and Oswestry Disability Index (ODI), Visual Analogue Scale (VAS), and 36-item Short-Form Health Survey Version 2.0 (SF-36v2) scores at 6 to 12 months postoperatively. SF-36v2 mental component summary scores (MCS) were used to assess pre- and postoperative general mental health. Post hoc analysis consisted of Pearson correlations between baseline SF-36v2, ODI, and VAS scores, and an identical set of correlations at outcomes. RESULTS: Preoperative MCS showed no significant association with outcomes VAS, ODI, or physical component summary scores. Baseline disability correlated significantly and more strongly with baseline MCS (P , .001, r = 20.40) than baseline pain levels (VAS back not significant, VAS leg P = .015, r = 0.27). Outcomes disability correlated significantly and more strongly with outcome back and leg pain levels (P , .001, r = 0.60 and 0.66) than outcome MCS (P = .031, r = 20.24). CONCLUSION: In a patient sample with mental health scores comparable to the population mean, there is no relationship between preoperative general mental health and postoperative patient-centered outcomes. Surgeons should consider the dynamic relationships between patient disability, mental health, and pain levels in assessing quality of life at different time points.
Rachel Asher, BS* Ashley E. Mason, PhD‡ Joseph Weiner, BS§ Richard G. Fessler, MD, PhD* *Department of Neurological Surgery, Rush University Medical Center, Chicago, Illinois; ‡Osher Center for Integrative Medicine, Department of Medicine, University of California at San Francisco, San Francisco, California; §Northwestern University Feinberg School of Medicine, Chicago, Illinois Correspondence: Richard G. Fessler, MD, PhD, Department of Neurological Surgery, Rush University Medical Center, 1725 W. Harrison St, Suite 855, Chicago, IL 60612. E-mail: [email protected] Received, September 11, 2014. Accepted, January 14, 2015. Published Online, February 14, 2015. Copyright © 2015 by the Congress of Neurological Surgeons.
KEY WORDS: Lumbar surgery outcomes, Mental health, Minimally invasive Neurosurgery 76:672–679, 2015
DOI: 10.1227/NEU.0000000000000695
I
n the United States, low back pain is one of the most common presenting concerns in clinical visits. The 2004 National Health Interview
ABBREVIATIONS: MCS, 36 Item Short-Form Health Survey Version 2.0 mental component summary score; MH, mental health; MIS, minimally invasive spine surgery; ODI, Oswestry Disability Index Version 1.3; PCS, 36 Item Short-Form Health Survey Version 2.0 physical component summary score; QOL, quality of life; SF36v2, 36 Item Short-Form Health Survey Version 2.0; VAS, Visual Analogue Scale
672 | VOLUME 76 | NUMBER 6 | JUNE 2015
www.neurosurgery-online.com
Survey indicated more than 25% prevalence of low back pain among respondents.1 Between 1995 and 1999, spine surgery rates in the Medicare population increased by 40%.2 From 1998 to 2008, data from the Nationwide Inpatient Sample indicates a 137% increase in spinal fusion rates.3 In comparison with open spinal procedures, the advent of minimally invasive lumbar surgery is associated with more rapid mobilization and discharge,4 as well as decreased perioperative surgical site infection5 and blood loss.6 However, concomitant changes
www.neurosurgery-online.com
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
MENTAL HEALTH AND QUALITY OF LIFE MIS
in longer-term quality of life do not always follow such improvements in short-term outcomes. This suggests that factors other than surgical approach may contribute to variations in postoperative quality of life. Although psychological predictors of postoperative quality of life have frequently been investigated in open procedures, the literature in minimally invasive populations is very limited. The intent of our study was to assess the relationship between preoperative mental health and postoperative quality of life (QOL) in minimally invasive populations. Since the observation of an association between hypochondria, hysteria, depression, and spinal surgery outcomes,7 investigation of preoperative psychological status as an outcome predictor has increased. Current measurement tools assess a diverse array of psychological constructs, such as patient expectations, fearavoidance behavior, depression, and somatic perception. Although predictive value or association between depression and lumbar surgery outcomes has been validated in many studies,7-12 others have found no such association.7-9,13 This lack of consensus reflects a larger trend in the literature on psychological distress and lumbar surgical outcomes. Reviews address the difficulty in coming to consensus regarding the role of preoperative psychological status in outcomes, given the variety and quality of psychological measurement tools,9 and the confounding influence of psychological distress during diagnostic testing for surgically remediable pathology.7-9 Some research suggests that, during discography, psychological distress correlates with more frequent pain reporting.14 The assessment of the literature on mental health and postoperative QOL is further complicated by the clinical heterogeneity of lumbar surgical patients. In particular, before surgical intervention, patients undergoing fusion tend to have been in pain much longer and had more previous surgeries than patients undergoing decompression and discectomy. A fusion is also a comparatively larger operation with a more painful, longer recovery. While comparing the results of studies with exclusively fusion or nonfusion samples, it is necessary to consider how such differences may affect the relationship between preoperative mental health and postoperative QOL. Although psychometrics specific to a particular psychological diagnosis or construct are common in the literature, using a general-health measure presents the advantage of minimizing respondent and data collection burdens.15 The SF-36 QOL measurement tool has a mental component summary score (MCS) that assesses general mental health. While important specific conditions such as depressed mood are a component of the SF-36, the metric tests overall mental health rather than depression specifically. The SF-36 has frequently been used to assess the relationship between preoperative mental health and QOL outcomes.16-21 Five of 6 studies found a significant relationship between SF-36 mental health scores and outcome.18 Although the relationship between preoperative mental health and outcomes in open lumbar surgeries is frequently addressed, particular focus on minimally invasive approaches is rare. To our knowledge, this is the first study to investigate the relationship
NEUROSURGERY
between preoperative general mental health and postoperative QOL in a minimally invasive lumbar surgical population. In this study, we investigated the relationships between preoperative mental health and patient-centered outcomes in minimally invasive lumbar surgical procedures.
METHODS Patient Population We conducted a retrospective analysis of prospectively collected data from 83 adult patients undergoing minimally invasive lumbar surgery (ages 29-82) with baseline data between the dates of March 31, 2008 and August 31, 2012. Fifty-one fusions, 18 decompressions, and 14 discectomies, all single-level, were performed. The institutional review board at Northwestern University approved the present study. The inclusion criteria were patients over age 18 who underwent single-level minimally invasive lumbar spine surgery performed by the senior author. Patients meeting one or more of the following conditions were excluded: infection, multilevel operations, previous fusion procedures, pending or past spine litigation, and workers’ compensation recipients. Baseline demographic, clinical, and surgical characteristics of the patient sample are shown in Tables 1 and 2. Baseline and outcomes Oswestry Disability Index Version 1.3 (ODI), Visual Analogue Scale (VAS), MCS, and physical component summary (PCS) scores are listed in Table 2.
Measures Mental Health The 36-item Short-Form Health Survey Version 2.022 (SF-36v2) was used to assess mental health. The scoring algorithm to calculate MCS uses scores from each subdomain to generate a final score. The 4 subdomains most closely related to mental health are given the heaviest weighting in this calculation. Of the 8 health domains measured in the SF-36, the 5-item Mental Health Index, which made up the Mental Health (MH) domain, had been most extensively evaluated by the time of the initial SF36 publications.23 It contains 5 items that assess 4 major MH components—anxiety, depression, loss of behavioral or emotional control, and psychological well-being. In the validation studies published in the 1990s by Ware et al,22 the MH domain was shown to be the most valid predictor of MH. All SF-36v2 scores were obtained using norm-based scoring algorithms; scores range from 0 to 100 with a population mean of 50. Scores lower than 50 represent poorer health relative to the population mean.
Patient-Centered Outcomes Measures All patient-centered outcomes were obtained at baseline and 6 to 12 months postoperatively. The SF-36v2 MCS, PCS, and the ODI version 1.324 were used to assess general physical health. ODI scores range from 0 to 100, with higher scores indicating increasing disability. Back and leg pain were measured by the VAS.
Statistical Analysis Data were analyzed with SAS (SAS Institute Inc, 2011). Groups were collapsed when possible to minimize the influences of sample size. Continuous variables were examined for extreme departures from normality. Missing data points were managed by using multiple imputation by chained equations (also called fully conditional specification).25 PROC
VOLUME 76 | NUMBER 6 | JUNE 2015 | 673
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
ASHER ET AL
TABLE 1. Patient Sample Characteristics at Baselinea
Sex, female Marital status Unmarried Married/engaged Level of education Less than college 4-year college degree Graduate/professional degree Work status Currently working Short-term disability Not employed Retired Homemaker/other Smoker Alcohol use Never 1-7 drinks/wk .8 drinks/wk OTC/NSAID medication use Never Intermittent Daily Narcotic pain medication use Never Intermittent Daily Exercise frequency Never, disability prevents Never Once or less/wk Twice/wk Three times/wk Four or more times/wk Comorbidities History of cancer Previous lumbar surgery (all decompression or discectomy) Primary diagnosis Disc pathology Spondylolisthesis Stenosis Other Surgical category Decompression Discectomy Fusion a
n
%
42
51
15 68
19 82
21 18 29
31 26 43
36 2 2 29 7 4
47 3 3 38 9 5
19 38 11
28 56 16
10 34 36
12 42 45
46 15 17
59 19 22
25 9 4 7 7 14 65 12 24
38 14 6 11 11 21 78 14 29
33 27 21 1
40 33 26 1
18 14 51
22 17 61
NSAID, nonsteroidal anti-inflammatory drug; OTC, other the counter.
MI in SAS was used. All potential predictors and outcomes were used in the imputation models. Logistic regression was used to impute the dichotomous variables and regression or regression with predictive mean matching was used to impute the continuous variables. Ten imputed data sets were created.
674 | VOLUME 76 | NUMBER 6 | JUNE 2015
From the imputed data sets, 23 sociodemographic, clinical, and QOL baseline variables were selected for correlation analysis. Bivariate correlations were performed between each variable and patient-centered outcomes metric (Table 3). Partial correlations were also performed between outcome variables and potential predictors, after partialling out baseline score for corresponding outcome (Table 4). The significance level was set at P , .05 for entry into regression analysis. Because baseline MCS did not significantly correlate with patient-centered outcomes in either analysis (with the exception of outcomes MCS), multivariate regressions were not performed. Given the relatively high baseline and outcomes MCS score in comparison with the literature, post hoc analysis was performed. t tests between the lowest 25% and highest 25% of baseline MCS, and each outcomes variable, were conducted. Correlations between baseline ODI, VAS back and leg, PCS, and baseline MCS were assessed (Table 5). An identical set of correlations was performed between corresponding metrics at outcome and outcomes MCS scores (Table 6). Because this study included fusion, decompression, and discectomy procedures, additional post hoc analysis was conducted to investigate the possible moderating effects of surgical category on the relationship between preoperative MCS and postoperative QOL. Because regressions were underpowered owing to small sample size, surgical category was collapsed to fusion vs nonfusion. A series of t tests with the Levene test for equality of variances was performed between surgical category and baseline MCS, PCS, VAS back, VAS leg, and ODI. An identical set of t tests was performed with corresponding metrics at outcomes. Maineffect regressions on outcomes QOL (VAS back, VAS leg, PCS, and ODI) were conducted for surgical category, controlling for baseline QOL and MCS. The interaction between surgical category and baseline MCS was investigated in an identical set of regressions, differing only by inclusion of the interaction term.
RESULTS Mean baseline scores for VAS back and leg pain, ODI, SF-36v2, PCS, and MCS were 5.4, 5.9, 34.2, 34.5, and 50.2, respectively. Mean outcomes scores were 2.2, 2.2, 18.0, 44.5, and 55.1, respectively. Patients improved from moderately disabled at baseline (ODI range 20-40), to minimally disabled at outcomes (ODI range 0-20). Likewise, patients improved 10 points in PCS scores, equivalent to 1 standard deviation in the SF-36v2 metric. MCS scores improved 4.9 points, to approximately 0.5 standard deviations above the population mean. Preoperative MCS did not significantly correlate with VAS back, leg, ODI, or PCS outcomes scores. Thus, regressions between MCS, other significant predictors, and outcomes were not performed. Results of post hoc t tests between lower and upper quartiles of MCS and outcomes were also insignificant. Results for post hoc correlations between each QOL metric (ODI, PCS, VAS back, VAS leg, and MCS) are reported both in terms of r and P values. R values represent the correlation coefficient. Correlations between baseline MCS and baseline ODI were significant at P , .001, r = 20.40. Correlations between baseline ODI and baseline PCS were significant at P , .001, r = 20.60. However, correlations between baseline VAS back pain and baseline ODI were insignificant. Correlations between baseline VAS leg pain and baseline ODI were significant at P = .015, r = 0.27.
www.neurosurgery-online.com
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
MENTAL HEALTH AND QUALITY OF LIFE MIS
TABLE 2. Baseline and Postoperative Patient Sample Characteristicsa Preoperative Mean (Standard Deviation) Age, y BMI ODI score VAS back score VAS leg score SF-36v2 MCS score SF-36v2 PCS score
60.7 27.4 34.2 5.4 5.9 50.2
Postoperative Mean (Standard Deviation)
(13.7) (5.1) (13.2) (2.5) (2.7) (11.4)
— — 18.0 (16.2) 2.2 (2.2) 2.2 (2.4) 55.1 (6.9)
34.5 (7.6)
44.5 (10.2)
a
BMI, body mass index; ODI, Oswestry Disability Index Version 1.3; VAS, Visual Analogue Scale; SF36v2, 36 Item Short-Form Health Survey Version 2.0; MCS, SF36v2 mental component summary score; PCS, SF36v2 physical component summary score.
Correlations between each QOL measure at outcomes changed relative to baseline. Correlations between outcomes VAS back and leg and outcomes ODI increased to r = 0.60 (P , .001) and r = 0.66
(P , .001), respectively. With the use of the Cohen guidelines for effect size,26 this represents a change from nonsignificant or small correlations at baseline to high correlations at outcomes. The correlation between MCS and ODI decreased from r = 20.40 (P , .001) at baseline, to r = 20.24 (P = .031) at outcomes. This represents a change from a medium correlation preoperatively to a small correlation postoperatively. The correlation between outcomes PCS and outcomes ODI increased to 20.87 (P , .001). Correlations between MCS and VAS (back and leg) were not found to be significant at baseline or outcomes. Post hoc t tests on surgical category and QOL metrics were found to be insignificant, with the exception of outcomes MCS. Although patients not undergoing fusion were found to have significantly higher outcomes MH (P = .038 equal variances assumed, P = .030 equal variances not assumed) than patients undergoing fusion, this outcome fell to insignificance with Bonferroni correction for multiple comparisons. Results for regressions examining the main effect of surgical category on outcomes QOL were insignificant. Regressions on outcomes QOL examining the interaction between surgical category and baseline MCS were also insignificant.
TABLE 3. Bivariate Correlations Between Outcomes and Potential Predictorsa Sample Characteristics Baseline ODI Baseline PCS Baseline MCS Baseline VAS-back Baseline VAS-leg Sex Age Marital status Education level Work status Alcohol consumption Exercise OTC/NSAID use Narcotic use BMI Comorbidities History of cancer Previous lumbar surgery Primary indication—back pain Primary indication—leg pain Primary diagnosis Type of surgery Blood loss
Outcome ODI b
0.33 20.33b 20.15 0.05 20.02 0.03 0.05 20.09 20.07 20.06 20.06 20.17 0.19d 0.02 0.19 0.16 0.17 0.11 0.18d 0.03 0.04 20.03 20.03
Outcome PCS
Outcome MCS
Outcome VAS-Back
Outcome VAS-Leg
20.16 0.30b 0.06 20.06 20.01 0.05 20.09 0.10 0.08 0.07 0.06 0.10 20.28c 0.08 20.21 20.11 20.15 20.10 20.14 20.03 20.09 0.03 0.03
20.28 0.06 0.56e 20.27c 20.09 20.12 0.08 20.05 0.19 0.26c 0.23d 0.21d 0.08 20.04 0.02 0.01 20.06 0.08 20.11 20.02 0.11 20.23c 20.02
20.01 20.06 0.04 0.32b 20.01 20.06 20.01 20.09 20.26c 20.01 0.03 20.24d 0.18 0.15 0.20 0.17 0.12 0.00 0.35b 20.02 0.07 20.09 20.04
0.16 20.20d 20.01 0.09 0.17 0.12 0.18 20.15 20.16 20.04 20.17 20.31c 0.17 20.13 0.19 0.22d 0.05 20.02 20.02 0.11 20.01 20.17 20.06
c
a
BMI, body mass index; ODI, Oswestry Disability Index Version 1.3; VAS, Visual Analogue Scale; SF36v2, 36 Item Short-Form Health Survey Version 2.0; MCS, SF36v2 mental component summary score; PCS, SF36v2 physical component summary score; NSAID, nonsteroidal anti-inflammatory drug; OTC, over the counter. P , .01. c P , .05. d P , .10. e P , .001. b
NEUROSURGERY
VOLUME 76 | NUMBER 6 | JUNE 2015 | 675
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
ASHER ET AL
TABLE 4. Bivariate Correlations Between Outcomes and Potential Predictors, After Partialling Out Baseline Score for Corresponding Outcomea Sample Characteristics Baseline ODI Baseline PCS Baseline MCS Baseline VAS-back Baseline VAS-leg Gender Age Marital status Education level Work status Alcohol consumption Exercise OTC/NSAID use Narcotics use BMI Comorbidities History of cancer Previous lumbar surgery Primary indication—back pain Primary indication—leg pain Primary diagnosis Type of Surgery Blood loss
Outcome ODI
Outcome PCS
Outcome MCS
Outcome VAS-Back
Outcome VAS-Leg
— 20.18 20.02 0.02 20.12 20.09 0.06 20.07 20.07 0.02 20.05 20.10 0.13 20.07 0.21c 0.15 0.23d 0.14 0.18 20.04 0.01 20.07 20.02
0.04 — 0.10 0.00 0.07 0.14 20.07 0.09 0.07 0.04 0.03 20.05 20.29d 0.13 20.18 20.11 20.17 20.11 20.14 0.03 20.06 0.04 0.04
20.07 0.15 — 20.32b 20.05 20.14 20.04 20.12 0.14 0.16 0.15 0.30d 0.21c 0.08 20.08 0.00 20.07 0.20 20.13 0.01 0.15 20.18 20.06
20.06 20.01 0.06 — 20.12 20.13 20.02 20.10 20.20c 0.01 0.07 20.18 0.21c 0.12 0.26d 0.17 0.10 0.04 0.23d 20.05 0.10 20.12 0.00
0.12 20.16 0.01 0.04 — 0.08 0.15 20.15 20.14 20.01 20.15 20.25d 0.16 20.12 0.17 0.21c 0.08 20.02 20.01 0.04 0.01 20.16 20.03
a
BMI, body mass index; ODI, Oswestry Disability Index Version 1.3; VAS, Visual Analogue Scale; SF36v2, 36 Item Short-Form Health Survey Version 2.0; MCS, SF36v2 mental component summary score; PCS, SF36v2 physical component summary score; NSAID, nonsteroidal anti-inflammatory drug; OTC, over the counter. b P , .01. c P , .10. d P , .05.
DISCUSSION In this retrospective analysis of prospectively collected data, preoperative general MH was not significantly associated with postoperative physical patient-centered outcomes. Preoperative MH showed no statistically significant relationship with ODI, VAS, or SF-36v2 PCS scores at 6- to 12-month outcomes. Results suggest that, contrary to much of the literature, preoperative
TABLE 5. Bivariate Correlations Between Patient-Centered Metrics at Baselinea
BaseODI BasePCS BaseMCS
BaseODI BasePCS BaseMCS
BaseVAS Back
BaseVAS Leg
20.5971b 20.3992b — 20.1351 — —
0.0983 20.1676 20.0101
0.2690c 20.2773c 20.0999
— — —
a
ODI, Oswestry Disability Index Version 1.3; VAS, Visual Analogue Scale; SF36v2, 36 Item Short-Form Health Survey Version 2.0; MCS, SF36v2 mental component summary score; PCS, SF36v2 physical component summary score. b P , .001. c P , .05.
676 | VOLUME 76 | NUMBER 6 | JUNE 2015
psychological status may not play as significant a role in postoperative physical QOL outcome. Relative to other studies that used the SF-36 MCS as a predictor, this study population has better overall MH. Thus, it could be stated that, in a population of patients with relatively normal preoperative MCS scores, MCS is not significantly related to postoperative outcome. However, certain elements of both study design and patient sample necessitate moderation when interpreting these conclusions in the context of the literature. Variations in statistical methodology, particularly methodological lapses in some studies, are key limiting factors in the interpretation of conclusions. A review by Celestin et al9 on psychosocial predictors of lumbar surgery and spinal cord stimulation outcomes cited that many studies did not control for preoperative levels of pain and disability. As an example, Moraci et al27 found preoperative depression to be a significant predictor of postoperative pain levels in lumbar microdiskectomy, but did not control for preoperative pain levels in statistical analyses. Thus, the reliability of these results is questionable. To our knowledge, this is the only study to examine the relationship between preoperative emotional health and postoperative QOL in an exclusively minimally invasive population. Furthermore, there are many different questionnaires and constructs used to assess MH as an outcomes predictor. Although
www.neurosurgery-online.com
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
MENTAL HEALTH AND QUALITY OF LIFE MIS
TABLE 6. Bivariate correlations Between Patient-Centered Metrics at Outcomesa
Outcomes ODI Outcomes PCS Outcomes MCS
Outcomes ODI
OutcomesPCS
Outcomes MCS
Outcomes VAS Back
Outcomes VAS Leg
— — —
20.8701b — —
20.2427c 0.1157 —
0.6046b 20.5843b 20.1944
0.6639b 20.6208b 20.1899
a
ODI, Oswestry Disability Index Version 1.3; VAS, Visual Analogue Scale; SF36v2, 36 Item Short-Form Health Survey Version 2.0; MCS, SF36v2 mental component summary score; PCS, SF36v2 physical component summary score. b P , .001. c P , .05.
general MH may not show strong association with outcome, the relationship between more specific psychological conditions and outcome may differ. Studies which assessed somatization, depression, and anxiety were investigated often in literature reviews, with results frequently being significant.9,28,29 In terms of study design, the limitations of not using an open control include the possibility that our results may generalize to open studies. We are not aware of any minimally invasive spine surgery (MIS) studies on the relationship between preoperative MH and postoperative QOL that have used an open control. However, our group is currently performing a prospective investigation of patient expectations in open vs MIS which includes the possibility for future MCS comparative analysis. Of the 4 studies in open lumbar surgical populations which used the SF-36, mean preoperative MCS scores were 41.54,21 37.05,17 35.37,20 and 43.77.19 Edwards et al16 did not use norm-based scoring, and compared their mean baseline MH score of 63.2 to the mean MH score of 80 in a general population of patients with degenerative lumbar spines. The distribution of MH scores at baseline was described as similar to patients with mild depressive disorders.16 Each of these studies found higher preoperative MCS scores to be significantly associated with better QOL outcomes. Of the 83 patients in this study, the mean SF-36 MCS score at baseline was 50.2, which is not statistically distinct from the population norm of 50. The divergence in conclusions in comparison with this study may be a function of the distribution of emotional distress in the studied populations. It is possible that, in order to begin finding a significant association between baseline MH and outcome, our patient population would require lower mean MH scores. The purpose of the post hoc t tests between the upper and lower MCS quartiles and QOL outcomes was to test this possibility. However, t tests were not significant. Although this provides stronger evidence for a lack of association between preoperative MH and QOL outcomes, it is not necessarily conclusive. The 25th percentile of baseline MCS was 40.6, still greater than the mean MCS for 2 of the studies referenced above. t tests between lower and higher MCS percentiles could not be performed because of small sample size. With the exception of the discectomy sample in Edwards et al,16 the comparable open studies described above are inclusive only of
NEUROSURGERY
patients undergoing fusion. Because this study included fusion, decompression, and discectomy procedures, post hoc tests were performed to facilitate comparison among these groups. Small sample size necessitated collapsing the 3 categories to fusion vs nonfusion. Post hoc regressions on QOL outcomes did not indicate a significant interaction between baseline MH and surgical category. Given these limitations, we were unable to draw conclusions on different effects of MH across minimally invasive fusions, decompressions, and discectomies. To our knowledge, these moderating relationships have not yet been investigated in the literature. Notably, post hoc t tests revealed no significant differences in baseline or outcomes MCS, PCS, ODI, or VAS for fusion vs nonfusion groups. Although fusion and nonfusion groups are often quite distinct clinically, QOL metrics are responsive to both sociodemographic and clinical differences. It is certainly possible that the lack of significant differences in QOL observed across fusion vs nonfusion groups is partially reflective of patient demographics and inclusion criteria. First, the selected patient population reflects that of the surgeon’s practice, which in this sample is distinctive in terms of education level, smoking status, employment status, baseline MH, disability, and pain levels. The patient population for this study is highly educated, almost exclusively white, and mostly retired or currently working. Only 4 of 83 patients are smokers. Very few patients are unemployed, and none have past/pending litigation or workers’ compensation status. In particular, work status, sick leave, workers’ compensation, and spine litigation have been shown to associate with poor outcomes.30-32 For studies that do include these populations, the role of MH in determining QOL outcomes may be more significant. The exclusion of patients with multilevel fusions, history of prior fusion, and spine litigation or workers’ compensation claims could have created a fusion sample with a QOL profile more similar to patients undergoing decompression and discectomy. These criteria, in combination with the favorable employment and smoking status of patients, could have further minimized the differences between fusion and nonfusion groups. Furthermore, mean baseline disability, leg pain, and back pain are moderate overall, at 34.2, 5.4, and 5.9 points, respectively. The mean preoperative disability is lower,17,18,20,21 VAS back and leg
VOLUME 76 | NUMBER 6 | JUNE 2015 | 677
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
ASHER ET AL
pain levels lower,17,20,21 and QOL higher17-19,21 in comparison with similar open studies using the SF-36 MCS as a predictor. Because these studies did find a direct relationship between MH and postoperative QOL outcomes, it is possible that the results of this study could differ markedly in a more severely disabled patient population with poorer MH. Specifically, a significant interaction between baseline disability and MH could arise, such that patients with more severe disabilities could exhibit a stronger relationship between baseline MH and postoperative QOL outcomes. In this study, baseline disability correlated significantly and more strongly with baseline MH than baseline pain levels. Surprisingly, correlations between VAS back pain and disability at baseline were insignificant. Correlations between baseline leg pain and baseline disability were significant but with small effect size. However, the correlations between MH, disability, and pain levels changed dramatically at outcome. Correlations between pain levels and disability changed from nonsignificant or small at baseline to high correlations at outcomes. The correlation between MH and disability decreased from medium preoperatively to small postoperatively. It is necessary to note that our results do not allow for conclusions regarding causality. However, despite the lack of association between baseline MH and patient-centered outcomes, MH still played an important role in the interpretation of patient disability both preoperatively and postoperatively. Baseline correlations indicate that MH, rather than pain, is more closely associated with baseline disability. The relationship between pain and disability in low back pain populations is somewhat controversial in the literature, with studies finding associations ranging from modest to strong.33 The variation in this relationship could be due to associations between disability and other psychosocial or demographic variables. Therefore, the preoperative assessment of MH by surgeons could facilitate the task of interpreting and understanding disability as a measure of QOL. Given the prevalence of MH literature in lumbar surgery, some surgeons administer MH questionnaires preoperatively to screen for patients at risk for poor outcomes. Theoretically, this separation of patients into more easily classifiable groups facilitates the interpretation of outcomes. In doing so, however, surgeons run the risk of assuming that the psychological features of these groups are relatively static. The results of this study suggest that surgeons should be cautious in interpreting patient disability and pain in the context of MH. A patient whose MH is believed to play a significant role in his or her disability at baseline may not exhibit those same tendencies at outcome.
CONCLUSION The primary objective of our study was to assess the relationship between preoperative MH and postoperative QOL in patients undergoing minimally invasive lumbar surgery. In a patient sample with MH scores comparable to the population mean, there is no association between preoperative
678 | VOLUME 76 | NUMBER 6 | JUNE 2015
general MH and postoperative patient-centered outcomes. Our results demonstrate that, although MH is more strongly correlated to disability than pain at baseline, the reverse trend is observed at outcome. This suggests that the surgical procedure itself may act as a psychological as well as physical intervention,34 modifying the relationship between MH and disability. To more effectively interpret disability as a clinical construct, spine surgeons must take into account the dynamic nature of the relationships between pain, disability, and MH. Future prospective research on other psychological constructs, such as patient expectations, and postoperative QOL outcomes in MIS lumbar surgery is warranted. Disclosure The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article.
REFERENCES 1. Health, United States, 2006: With Chartbook on Trends in the Health of Americans. Hyattsville, MD: National Center for Health Statistics: 2006. 2. Deyo RA, Mirza SK. The case for restraint in spinal surgery: does quality management have a role to play? Eur Spine J. 2009;18(suppl 3):331-337. 3. Rajaee SS, Bae HW, Kanim LE, Delamarter RB. Spinal fusion in the United States: analysis of trends from 1998 to 2008. Spine (Phila Pa 1976). 2012;37(1): 67-76. 4. Wang MY, Lerner J, Lesko J, McGirt MJ. Acute hospital Costs after minimally invasive versus open lumbar interbody fusion: data from a US National Database with 6106 patients. J Spinal Disord Tech. 2012;25(6):324-328. 5. McGirt MJ, Parker SL, Lerner J, Engelhart L, Knight T, Wang MY. Comparative analysis of perioperative surgical site infection after minimally invasive versus open posterior/transforaminal lumbar interbody fusion: analysis of hospital billing and discharge data from 5170 patients. J Neurosurg Spine. 2011;14(6):771-778. 6. Karikari IO, Isaacs RE. Minimally invasive transforaminal lumbar interbody fusion: a review of techniques and outcomes. Spine (Phila Pa 1976). 2010;35 (26 suppl):S294-S301. 7. Mannion AF, Elfering A, Staerkle R, et al. Predictors of surgical outcome and their assessment. Eur Spine J. 2006;15(suppl 1):S93-S108. 8. Mannion AF, Elfering A, Staerkle R, et al. Predictors of multidimensional outcome after spinal surgery. Eur Spine J. 2007;16(6):777-786. 9. Celestin J, Edwards RR, Jamison RN. Pretreatment psychosocial variables as predictors of outcomes following lumbar surgery and spinal cord stimulation: a systematic review and literature synthesis. Pain Med. 2009;10(4):639-653. 10. Chaichana KL, Mukherjee D, Adogwa O, Cheng JS, McGirt MJ. Correlation of preoperative depression and somatic perception scales with postoperative disability and quality of life after lumbar discectomy. J Neurosurg Spine 2011;14(2):261-267. 11. Sinikallio S, Aalto T, Lehto SM, et al. Depressive symptoms predict postoperative disability among patients with lumbar spinal stenosis: a two-year prospective study comparing two age groups. Disabil Rehabil. 2010;32(6):462-468. 12. Adogwa O, Parker SL, Shau DN, et al. Preoperative Zung Depression Scale predicts outcome after revision lumbar surgery for adjacent segment disease, recurrent stenosis, and pseudarthrosis. Spine J. 2012;12(3):179-185. 13. Hobby JL, Lutchman LN, Powell JM, Sharp DJ. The distress and risk assessment method (DRAM). J Bone Joint Surg Br. 2001;83(1):19-21. 14. Block AR, Vanharanta H, Ohnmeiss DD, Guyer RD. Discographic pain report. Influence of psychological factors. Spine (Phila Pa 1976). 1996;21(3):334-338. 15. Walsh TL, Hanscom B, Lurie JD, Weinstein JN. Is a condition-specific instrument for patients with low back pain/leg symptoms really necessary? The responsiveness of the Oswestry Disability Index, MODEMS, and the SF-36. Spine (Phila Pa 1976). 2003;28(6):607-615. 16. Edwards RR, Klick B, Buenaver L, et al. Symptoms of distress as prospective predictors of pain-related sciatica treatment outcomes. Pain. 2007;130(1-2): 47-55.
www.neurosurgery-online.com
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
MENTAL HEALTH AND QUALITY OF LIFE MIS
17. Carreon LY, Glassman SD, Djurasovic M, et al. Are preoperative health-related quality of life scores predictive of clinical outcomes after lumbar fusion? Spine (Phila Pa 1976). 2009;34(7):725-730. 18. Abbott AD, Tyni-Lenne R, Hedlund R. Leg pain and psychological variables predict outcome 2-3 years after lumbar fusion surgery. Eur Spine J. 2011;20(10): 1626-1634. 19. Derby R, Lettice JJ, Kula TA, Lee SH, Seo KS, Kim BJ. Single-level lumbar fusion in chronic discogenic low-back pain: psychological and emotional status as a predictor of outcome measured using the 36-item short form. J Neurosurg Spine. 2005;3(4):255-261. 20. Cobo Soriano J, Sendino Revuelta M, Fabregate Fuente M, Cimarra Diaz I, Martinez Urena P, Deglane Meneses R. Predictors of outcome after decompressive lumbar surgery and instrumented posterolateral fusion. Eur Spine J. 2010;19(11): 1841-1848. 21. Trief PM, Ploutz-Snyder R, Fredrickson BE. Emotional health predicts pain and function after fusion: a prospective multicenter study. Spine (Phila Pa 1976). 2006;31(7):823-830. 22. Ware JE Jr. SF-36 health survey update. Spine (Phila Pa 1976). 2000;25(24): 3130-3139. 23. Tarlov AR, Ware JE Jr, Greenfield S, Nelson EC, Perrin E, Zubkoff M. The Medical Outcomes Study. An application of methods for monitoring the results of medical care. JAMA. 1989;262(7):925-930. 24. Fritz JM, Irrgang JJ. A comparison of a modified Oswestry Low Back Pain Disability Questionnaire and the Quebec Back Pain Disability Scale. Phys Ther. 2001;81(2):776-788.
NEUROSURGERY
25. White IR, Royston P, Wood AM. Multiple imputation using chained equations: issues and guidance for practice. Stat Med. 2011;30(4):377-399. 26. Cohen JA. A power primer. Psychol Bull. 1992;112(1):155-159. 27. Arpino L, Iavarone A, Parlato C, Moraci A. Prognostic role of depression after lumbar disc surgery. Neurol Sci. 2004;25(3):145-147. 28. den Boer JJ, Oostendorp RA, Beems T, Munneke M, Oerlemans M, Evers AW. A systematic review of bio-psychosocial risk factors for an unfavourable outcome after lumbar disc surgery. Eur Spine J. 2006;15(5):527-536. 29. Aalto TJ, Malmivaara A, Kovacs F, et al. Preoperative predictors for postoperative clinical outcome in lumbar spinal stenosis: systematic review. Spine (Phila Pa 1976). 2006;31(18):E648-E663. 30. Gum JL, Glassman SD, Carreon LY. Is type of compensation a predictor of outcome after lumbar fusion? Spine (Phila Pa 1976). 2013;38(5):443-448. 31. DeBerard MS, LaCaille RA, Spielmans G, Colledge A, Parlin MA. Outcomes and presurgery correlates of lumbar discectomy in Utah Workers’ Compensation patients. Spine J. 2009;9(3):193-203. 32. Silverplats K, Lind B, Zoëga B, et al. Clinical factors of importance for outcome after lumbar disc herniation surgery: long-term follow-up. Eur Spine J. 2010;19(9): 1459-1467. 33. Peters ML, Vlaeyen JW, Weber WE. The joint contribution of physical pathology, pain-related fear and catastrophizing to chronic back pain disability. Pain. 2005; 113(1-2):45-50. 34. Lebow R, Parker SL, Adogwa O, et al. Microdiscectomy improves pain-associated depression, somatic anxiety, and mental well-being in patients with herniated lumbar disc. Neurosurgery. 2012;70(2):306-311; discussion 311.
VOLUME 76 | NUMBER 6 | JUNE 2015 | 679
Copyright © Congress of Neurological Surgeons. Unauthorized reproduction of this article is prohibited.
