The Questionnaire the Spine Clinic Skips: Catastrophising and Six-Month Outcomes in Cervical Disc Herniation
- Among 118 patients treated for single-level cervical disc herniation in Sichuan, 27.1 percent (32 of 118) scored 30 or above on the Pain Catastrophizing Scale before the procedure, with a group mean of 36.7 ± 5.2 against 15.8 ± 6.5 in the rest of the cohort.
- At six months, 79 of 86 low catastrophisers (91.9 percent) reported satisfaction on the NASS Patient Satisfaction Index, against 15 of 32 high catastrophisers (46.9 percent); chi-square 25.4, p < 0.001.
- Residual burden split the same way: neck pain VAS 1.4 ± 1.5 versus 3.5 ± 2.1, arm pain VAS 0.9 ± 1.3 versus 2.8 ± 2.3, Neck Disability Index 7.5 ± 5.8 versus 16.8 ± 9.2, all p < 0.001. Both groups improved substantially from baseline.
- Adjusted for age, sex, BMI, smoking, symptom duration, treated level and baseline pain and disability, a PCS score of 30 or above carried an odds ratio for dissatisfaction of 4.5; the confidence interval published for that estimate is internally inconsistent and is not reproduced here. Symptom duration beyond six months added an independent OR of 2.05 (95% CI 1.12–3.75).
A pain department in Nanchong, Sichuan, did something most interventional spine services do not: it handed every consecutive patient a 13-item questionnaire before the procedure and then followed them for six months. What separated the cohort was the questionnaire, not the demographic or procedural data collected alongside it. For anyone taking referrals from pain medicine or orthopaedics, this is a concrete argument about where in the pathway a psychologist belongs.
What the Sichuan Cohort Measured
Between January and December 2024 the Affiliated Hospital of North Sichuan Medical College consecutively enrolled patients aged 18 to 65 with single-level cervical disc herniation (C3-4 through C6-7) who had failed at least three months of standardised conservative treatment and were scheduled for CT-guided percutaneous ozone chemoneurolysis. Of 150 screened, 125 consented; 7 were lost to follow-up (5.6 percent), leaving 118. All completed the validated Chinese version of the Pain Catastrophizing Scale preoperatively, and were split at the conventional cut-off of 30, roughly the 75th percentile in chronic pain samples.
The two groups were indistinguishable on everything the surgeon usually looks at: age (45.2 ± 9.8 years overall), sex (55.1 percent male), BMI, smoking, symptom duration (10.5 ± 8.2 months) and treated level. They differed on everything the patient reports. At baseline the high catastrophisers already had worse neck pain (VAS 7.4 ± 1.5 versus 5.2 ± 1.6), worse arm pain (8.5 ± 1.1 versus 7.1 ± 1.1) and worse function (NDI 35.1 ± 7.6 versus 25.9 ± 7.1).
Both groups then improved, and that is the part worth holding onto: pain and disability fell in both, though with no comparison group the cohort cannot attribute that change to the procedure itself. What diverged was the endpoint. Satisfaction ran at 91.9 percent in the low group and 46.9 percent in the high group, and residual symptoms at six months were roughly twice as heavy in the high group on every measure. The adjusted model kept catastrophising in the picture after baseline severity was controlled for, and added symptom duration over six months as a second, weaker risk factor.
What to Do With a PCS Score of 36
Notice what this design does and does not license. It is an observational cohort, so catastrophising is a marker rather than a demonstrated cause, and the high group started in worse shape, so part of their worse endpoint is simply where they began. The adjusted odds ratio adds less than it appears to. The unadjusted contrast, though, is large enough to act on: five minutes of paperwork identified a quarter of the caseload in which this procedure would satisfy fewer than half of patients.
Two practical consequences follow. The first is timing. Referral for psychological work after a disappointing procedure is the usual sequence and it is the wrong one; the PCS was informative before anything had been done. Administering it at the point of the treatment decision costs nothing and opens a window in which expectation is still negotiable.
The second is target selection. The authors read their result through two different mechanisms, and the two call for different work. One is the fear-avoidance loop: magnification and helplessness feed kinesiophobia, avoidance, deconditioning and new pain, which responds to graded exposure and activity scheduling. The other they call the expectation-satisfaction gap: a catastrophising patient forms an all-or-nothing forecast, and every real procedure leaves some residue that then reads as failure. That one responds to explicit, numerical preoperative counselling about what a good outcome will actually feel like at six months. This cohort supplies the numbers: a neck VAS around 1.4 for a low catastrophiser, around 3.5 for a high one, with function improved in both. Telling a patient that in advance is not managing expectations downward. It is replacing a binary forecast with a calibrated one.
A thirteen-item questionnaire, filled in before the procedure, separated a group in which nine of ten patients ended up satisfied from one in which fewer than half did.
Single centre, 118 patients, and only about 24 dissatisfaction events against the eleven covariates carried in the adjusted model, roughly two events per variable, so the adjusted odds ratio is fragile and the authors acknowledge the overfitting risk; the Methods describe adjustment for age, sex and baseline arm pain alone while the model table lists eleven covariates, a second internal inconsistency. Because the high-catastrophising group also began with worse pain and disability, baseline severity and catastrophising cannot be cleanly separated here. The reported main effect (OR 4.5, 95% CI 1.08–9.78, p < 0.001) is not internally consistent, since an interval running from 1.08 is neither centred on 4.5 nor compatible with p < 0.001, and the discrepancy could not be resolved from the published text.