by Darlene L. Shaw, Ph.D (Medical University of South Carolina, Charleston, SC), Shannon S. Voor, Ph.D (Frazier Rehab Center, Louisville, KY), George L. Cogar, Ph.D (Carolina Spine Institute, Charleston, SC), Donald R. Johnson, M.D. (Carolina Spine Institute, Charleston, SC), and Steven C. Polletti, M.D. (Carolina Spine Institute, Charleston, SC)
Abstract
Study Design: Hierarchical regression analyses were used to examine the relationship of post-surgery depression as measured by the Beck Depression Inventory (BDI) and retrospective reports of childhood trauma to spinal fusion surgery outcome. Surgery outcome was measured by four independent sources: two physician ratings, patient self-ratings, and ratings based on an independent chart review. Childhood trauma history, depression, and patients' ratings of surgery outcome were assessed using a post-surgery, self-report questionnaire mailed to patients' homes.
Objectives: The study sought to determine whether patients' post-surgery self-reports of childhood trauma and depression were significantly correlated with spinal fusion surgery outcome when the demographic variables of age, gender, and post-surgery employment status were statistically controlled.
Methods: Twelve to thirty months after their surgery, fifty-six lumbar spinal fusion surgery patients returned completed questionnaires reporting self-ratings of surgery outcome, depression, and recollections of childhood trauma. After the fifty-six questionnaires had been received, the success of each patient's surgery was independently assessed by the operating physician and a non-operating physician, using a 5-point behaviorally anchored rating scale (1 = definitely unsuccessful, 5 = definitely successful). A fourth measure of surgery success was provided by an independent review of objective criteria in the patient's medical chart.
Results: After controlling for the effects of age, gender and post-surgery employment status, regression analyses revealed that both depression and childhood trauma accounted for a significant proportion of the variance in surgery outcome as rated by the operating physician, a non-operating physician, and the patient. Patients' post-surgery self-reports of depression accounted for a significant proportion of the variance in ratings of surgery outcome based on an objective chart review.
Conclusions: These results indicate that patients' post-surgery self-reports of depression on the BDI and childhood trauma history are significantly correlated with unsuccessful outcome of lumbar spinal fusion surgery.
Background Data
There is a growing consensus among orthopaedic surgeons that the outcome of lumbar spinal fusion surgery is determined by the interaction of physical, demographic, and psychological factors. Demographic variables such as age and employment status1 and involvement in Workers Compensation and other litigation2 have been found to predict treatment success in back pain patients receiving non-surgical conservative care. Other investigators have found that psychological status is correlated with spinal surgery outcome.3 For example, several investigations have found a strong correlation between elevations on MMPI scales 1 (Hs), 2 (D), and 3 (Hy) and poor surgical outcome.4 Depression, another psychological variable, has also been found to predict medication intake and pain behaviors in lumbar disc disease patients.5 More recently, childhood trauma (physical abuse, sexual abuse, caregiver substance abuse, abandonment, or emotional neglect) was found to be positively correlated with unsuccessful fusion and nonfusion lumbar spinal surgery.6
Objective
The present study investigated the association between patients' post-surgery self-reports of depression on the Beck Depression Inventory (BDI) and childhood trauma history and the outcome of spinal fusion surgery during adulthood. It was hypothesized that patients' self-reports of depression and childhood trauma would account for a significant amount of the variance in ratings of surgery success, beyond the variance accounted for by demographic variables and post-surgery employment status.
Subjects
N = 56 patients S/P lumbar spinal fusion surgery
Men = 28, Women = 28
x Age = 44 years (sd = 12.03)
Marital Status / Pre-Surgical Employment
Married / 79% / Unemployed / =33%
Other / 21% / Heavy Labor / =39%
Sedentary labor / =26%
Light labor / =2%
Educational Level / Pre-Surgical Diagnosis
11th Grade of less / =9% / Spondylolistheses / =33%
High school graduate / =69% / Degenerative Disk Disease / =39%
college degree / =21%
Graduate degree / =2%
Surgical Procedure
Posterior lumbar interbody fusion / =62% / One-level fusion / =71%
Lateral transverse process
fusion with segmental spinal
instrumentation / =38% / Two-level fusion / =29%
Methods
Patient Survey Procedures
· In October 1994, a post-surgery questionnaire was mailed to all 120 patients who underwent lumbar spinal fusion surgery between December 1991 and June 1993 at Carolina Spine Institute in Charleston, SC.
· Twelve questionnaires were not deliverable.
· Fifty-six questionnaires were completed and returned, yielding a 52% response rate.
· Patients completed questionnaires twelve to thirty months (mean = 17 months) after their surgery.
Post-surgery Patient Questionnaire Data:
· Self-report of post-surgery employment status (Employed vs Unemployed)
· Check-list (Yes or No) indicating retrospective recall of each of five types of childhood trauma (childhood sexual abuse, childhood physical abuse, caregiver alcohol/drug abuse, childhood abandonment, and childhood neglect)
· Beck Depression Inventory
· Self-rating of success of surgery using a 5-point scale ranging from 1 (definitely unsuccessful) to 5 (definitely successful).
Assessment of Surgical Outcome
· Patients' self-ratings of surgery success (See Patient Questionnaire)
· The success of each patient's surgery was rated independently by the operating physician and a non-operating physician using a 5-point scale:
1. (definitely unsuccessful) = patient not improved subjectively and regularly required narcotic pain medication.
2. (probably unsuccessful) = patient felt subjectively better compared to his properative state but had not returned to gainful employment and regularly required narcotic pain medication.
3. (cannot determine) = patient returned to light-duty work and occasionally used narcotic pain medication.
4. (probably successful) = patient had full-time-light-duty work and used anti-inflammatory medication daily.
5. (definitely successful) = patient had returned to gainful employment and was taking no pain medication.
· An independent rating of surgery success was based on an objective chart review performed by a physical therapy student who was blind to the purpose of the study. The surgery was rated as "unsuccessful" if the patient had one or more of the following:
o Repeat surgery
o Use of opioid analgesics six months or more post-surgery
o CT or MRI of the lumbar spine more than six months post-surgery
o Epidural corticosteroid injection more than six months post-surgery
o Failure to return to work or household duties by one year post-surgery
o None of the raters of surgery outcome (with the exception of the patients) had access to patients' self-reports or surgery, success, childhood trauma histories, or BDI scores.
o Interrater reliability across all four raters of surgery outcome was adequate (Kappa = .64)
Statistical Analyses:
· Predictor variables included: 1) Age; 2) Gender; 3) Post-surgery employment status; 4) Post-surgery score on the Beck Depression Inventory; and 5) Number of types of childhood trauma.
· Outcome variables included: 10 Operating physicians' ratings of surgery outcome; 2) Non-operating physicians' ratings of surgery outcome; 3) Patients' self-ratings of surgery outcome; and 4) Ratings of surgery outcome based upon a chart review of objective criteria of success.
· A separate hierarchical regression analysis was calculated for each of the 4 outcome variables.
Results
· T-test analyses revealed no significant differences between responders (28 men and 28 women), and non-responders (27 men and 25 women) on the patient questionnaire for employment status (t(1,107) = .43, p > 7.51), involvement in Workers Compensation (t(1,107) = 77 p > .38), or litigation status (t(1,107) = .22, p > .64).
· Averaging across raters 68% of surgeries were rated as successful and 12% were rated as unsuccessful (See Table 1).
· Forty percent of the patients in the study reported a history of childhood trauma (See Table 1).
· Statistically significant positive correlations were found between the operating physician, non-operating physician, patient, and objective outcome ratings of surgery success (See Table 2).
· Depression and childhood trauma predicted a significant proportion of the variance in the patients' self-ratings of surgery success, operating physicians' ratings of surgery success, and non-operating physicians' ratings of surgery success (See Table 3).
· Depression predicted a significant proportion of the variance in the ratings of surgery success based upon a chart review of objective criteria (See Table 3).
· A statistically significant positive correlation was found between depression and each of the four ratings of surgery success (See Table 4).
· Childhood trauma was significantly positively correlated with patients', operating physicians', and non-operating physicians', and non-operating physicians' ratings of surgery success (See Table 4).
· A statistically significant positive correlation (r=.29) was found between patients' post-surgery self-reports of childhood trauma and depression.
Table 1
Percentages of Surgeries Rated Successful/Unsuccessful
and mean BDI and Childhood Trauma Ratings for These Groups
Successful
Surgery / Cannot
Determine / Unsuccessful
Surgery
Average rating for
total group of patients / 68% / 10% / 12%
Spondylolistheses
patients / 59% / 26% / 15%
Degenerative disc
disease patients / 72% / 7% / 21%
Mean BDI* / 6.2** / 18**
Mean number of types
of childhood trauma*** / .27**** / 1.1****
* / Mean BDI total group = 12.7
** / t(1,46) = 6.05,p< .02
*** / 40% of total group reported childhood trauma
**** / t(1,46) = 8.08,p< .007
Table 2
Cross-Correlations Between Outcome Variables
Predictor / Operating
Patients'
Ratings / Non-Operating
Physicians'
Ratings / Physicians'
Ratings / Objective
Measures
Patient
Operating
Physician / .43*
Non-Operating
Physician / .39* / .88*
Objective / .35* / .80* / .74*
*p < .05
References
1. Fredreickson BE, Trief PM, Van Beveren P, Yuan HA, Baum G. Rehabilitation of the patient with chronic back pain: A search for outcome predictors. Spine 1988; 13:351-353. back
2. Trief P, Stein N. Pending litigation and rehabilitation outcome of chronic back pain. Arch Phys Med Rehabil 1985; 66:95-99. back
3. Spengler DM, Oullette EA, Battie M, Zeh J. Elective discectomy for herniation of a lumbar disc: Additional experience with an objective method. J Bone Joint Surg 1990; 72A:230-237. back
4. Herron LD, Turner J, Weiner P. A comparison of the Millon Clinical Multiaxial Inventory and the Minnesota Multiphasic Personality Inventory as predictors of successful treatment by lumbar laminectormy. Clin Orthop 1986; 203:232-238. back
5. Keefe FJ, Wilkins RH, Cook WA, Crisson JE, Muhlbaier, LH. Depression, pain, and pain behavior. J Consult Clin Psychol 1986; 54:665-669. back
6. Schoffeman J, Anderson D, Hines R, Smith G, White A. Childhood psychological trauma correlates with unsuccessful lumbar spine surgery. Spine 1992; 17:S138-S144. back
Table 3
Regression Analyses Results
Variables / Total R2 / df / F ratio
for R2 / F ratio for
R2 change
Patient Rating / .80 / 4,55 / 48.43
Age / 1.05
Gender / 0.00
Employment / 0.51
BDI Score / 6.12*
Trauma / 146.29*
Operating Physician Rating / .26 / 4.55 / 4.44*
Age / 0.02
Gender / 0.34
Employment / 0.07
BDI Score / 5.63*
Trauma / 6.04*
Non-Operating Physician
Rating / .25 / 4,55 / 4.17
Age / 0.78
Gender / 1.03
Employment / 0.02
BDI Score / 4.37*
Trauma / 6.89*
Objective Measures / .17 / 4,55 / 2.55*
Age / 0.00
Gender / 2.61
Employment / 1.14
BDI Score / 4.40*
Trauma / 1.51
*p<.05.; BDI = Beck Depression Inventory
Table 4
Correlations Between Predictor Variables and
Ratings of Surgery Outcome
Predictor / Patients'
Rating / Operating
Physicians'
Rating / Non-Operating
Physicians'
Rating / Objective
Measures
Age / .14 / .09 / -.05 / .05
Gender / -.01 / .03 / .10 / .17
Employment / .09 / .08 / .04 / -.06
Depression / -.43* / -.40* / -.33* / -.31*
Trauma / -.87* / -.41* / -.41* / -.25
*p<.05.
Discussion
· Lumbar spinal fusion surgery outcome is significantly related to patients' post-surgery self-reports of childhood trauma and depression as measured by the BDI.
· Patients' post-surgery reports of depression (BDI) and childhood trauma accounted for a significant amount of variance in ratings of surgical outcome even after controlling for the effects of age, gender, and post-surgery employment status.
· Depression and retrospective self-reports of childhood trauma were moderately correlated in the lumbar spinal fusion surgery patients in this study.
· On average, patients with successful lumbar spinal fusion surgery reported less than one childhood trauma and had post-surgery BDI scores in the normal range.
· Because the present study was correlational in nature, no causal link may be assumed between post-surgery depression and childhood trauma and the outcome of lumbar spinal fusion surgery.
· Prospective studies are needed to asses the ability of patients' pre-surgical self-reports of depression and childhood trauma history to predict the outcome of lumbar spinal fusion surgery.
· Studies are also needed to determine whether pre-surgical intervention with patients suffering from psychological problems would increase the success of lumbar spinal fusion surgery.
· The results hold promise that a brief self-report screening measure of depression (the BDI) and a checklist of childhood trauma could prove helpful in predicting lumbar spinal surgery outcome. A screening measure of this type could help reduce the number of unsuccessful surgeries, leading to better containment of health care costs.
· The results suggest that more collaboration between surgeons and psychologists could enhance patient care and lead to cost savings.
5