Cognitive Behavioral Therapy (CBT) After Lumbar Spinal Fusion in Patients With High Pain Catastrophizing

The incidence of moderate to severe chronic postsurgical pain (CPSP) at 12 months after surgery is circa 12% in Europe, depending in part upon the surgical procedure. Worldwide, more than 230 million people undergo major surgery every year; the global annual cost of new cases of CPSP is in the hundreds of billions of dollars. The response to this costly tragedy has been unacceptably slow.

Perioperative protocols have historically not incorporated routine screening for patients at-risk for post-surgery persistent pain or compromised function. However, early detection and management of patients at high risk for CPSP may modify the postsurgical pain trajectory and reduce the incidence of CPSP.

In particular, persistent pain and disability after spinal surgery poses a large problem. Around 20% of the patients have persistent or recurrent pain in the back or limbs after surgical interventions for spinal disease. Several risk factors for such unfavorable outcomes have been identified, among which psychological factors. The most robust psychological predictors of high levels of acute postsurgical pain and of persistent pain in the back or limbs after spinal surgery are high levels of preoperative anxiety and pain catastrophizing.

Psychological variables are modifiable and could be target for intervention. Cognitive Behavioral Therapy (CBT) is the leading psychological treatment for chronic pain. CBT aims to reduce maladaptive cognitions and behaviors and replace these with more adaptive ones. CBT has been shown to reduce anxiety and catastrophizing and to increase self-efficacy and feelings of control beliefs and through this lowers pain-related disability and pain interference (including sleep problems and depression).

This study is a prospective, randomized, controlled multicenter trial with 1:1 allocation, comparing lumbar spinal fusion surgery outcome between 2 groups with high pain catastrophizing, 1 that receives perioperative CBT plus usual care (experimental group) and 1 that receives usual care (control group). To minimize expectation bias in patients, an educational intervention is added in the control group.

Patients will be enrolled in the Department of Neurosurgery at the Neurocenter of Southern Switzerland, Lugano Regional Hospital and the department of Neurosurgery of Zuyderland Medical Center, Heerlen, The Netherlands.

Patients with as score of ≥ 24 on the PCS will be eligible to be randomized to perioperative CBT plus usual care or education plus usual care.

Data collection sessions will occur for each patient randomized, at baseline, at 8 weeks, and at 6 and 12 months following surgery. A third, non-randomized group is also included, consisting of low catastrophizing patients. These patients will not undergo any additional intervention besides usual care, but will undergo the same data-collection as the high catastrophizing randomized groups. This offers the possibility to examine whether CBT can reduce the negative effects of catastrophizing on outcome to the level of improvement one would expect in non-catastrophizing patients. Staff blinded to group assignment will contact patients by phone at circa one week before data collection time points to remind them to fill in the questionnaires.

Primary hypothesis of the investigators is that patients randomized to the CBT group show a greater improvement on the Core Outcome Measures Index (COMI) at 12 months post-surgery compared to patients randomized to the control group.

Secondary hypotheses are that patients randomized to the CBT group show a larger reduction in disability (ODI), leg/back pain (NRS), depression and pain catastrophizing (PCS) at 12 months post-surgery compared to patients randomized to the control group. Moreover, patients in the CBT group report higher global impression of chance, less reliance on analgesics and better employment status compared to patients in the control group. Finally, the investigators hypothesize that the high catastrophizing group receiving the control intervention will have worse outcomes on all measures compared to the low catastrophizing group receiving standard care, whereas the high catastrophizing group receiving the CBT intervention will have comparable outcomes as the low catastrophizing group.

All included patients will be assessed before surgery for the outcomes of interest. Patients assigned to the CBT intervention condition will attend six individual 60-min CBT sessions (2 before and 4 after surgery). Patients randomized to the education condition will receive six session of biomedical education and an exercise program. Non-randomized low catastrophizing patients do not receive an additional intervention. All patients will receive care that they would have routinely received had they not been entered in the study (i.e. standard postoperative physical therapy). The three groups of patients will undergo the same data collection procedures.

The primary outcome is the difference in the COMI between the study groups, 12 months after surgery. A difference of two points is considered clinically meaningful [44]. With an assumed SD of 2.5, the effect size of this difference, quantified as Cohen's d is 0.80. Inclusion of 42 patients per intervention group will yield a power of 0.95 at a type-I error rate of 0.05. In addition, 42 patients with low pain catastrophizing scores will undergo the same assessments, but not receive any intervention besides care as usual. Taking into account a drop-out rate of 15%, 150 patients will be included in total.

Baseline patient characteristics will be stratified by group and presented as mean and standard deviation (SD), median and first and third quartile, and count and percentage, as appropriate. Differences in baseline characteristics between randomized groups will not be statistically tested, whereas differences with the separate control cohort will be tested using the independent-samples t-test for continuous variables, and Pearson's chi-square test for categorical variables. In case of expected cell counts of less than 5, Fisher's Exact test will be used instead.

All patients randomized (i.e. those with a PCS score > 24) will be analyzed in an intention-to-treat analysis. To test for a difference between the CBT and control group in the COMI at 12 months, linear regression will be used, with correction for baseline COMI and center. In addition, the difference in the trajectory of COMI over time will be assessed by means of linear mixed-effects regression, taking all follow-up measurements into account. Adjustment for center, type of surgery, age and gender will be made and, if necessary, for other prognostic variables significantly associated with the outcome (i.e., acute postoperative pain intensity, preoperative depression, preoperative surgical fear, length of current sick leave, pain duration). Stepwise backward elimination using the Wald test will be used to select prognostic variables significantly associated with the outcome. Any control variables that are incomplete will be imputed if the proportion of incomplete patients exceeds 0.05. Investigators will use multiple imputation with fully conditional specification with the number of imputations set to the percentage of incomplete patients. Predictive mean matching will be used to draw values to be imputed. Secondary outcome variables will be analyzed similar to the primary outcome. For these, logistic regression analysis adjusted for center will be used to test for differences in proportions at 12 months postoperatively.

Exploratory, investigators will compare the outcomes of the patients in the two intervention groups with the outcomes of patients with low pain catastrophizing scores (i.e. those with a PCS score < 24) undergoing care as usual. Linear mixed-effects regression for the three groups will be performed with primary and secondary outcomes at the three postoperative assessment periods and controlling for preoperative values. This analysis will indicate whether high catastrophizing patients follow a similar trajectory of pain and functioning after spinal fusion surgery compared non-catastrophizing patients.