Objective: This study was designed to investigate whether the PainDETECT Questionnaire (PDQ) classification was predictive of outcomes at 3 and 12 months follow-up in low back pain (LBP) patients with associated leg pain. Identification of clinically important subgroups and targeted treatment is believed to be important in LBP care. The PD-Q is designed to classify whether a person has neuropathic pain, based on their self-reported pain characteristics. However, it is unknown whether this classification is a prognostic factor or predicts treatment response.

Methods: One hundred forty-five participants were recruited in the secondary care. Inclusion criteria were 3 to 12 months of LBP and related leg pain. Baseline PDQ scores classified participants into 3 groups ("likely" to have neuropathic pain, "uncertain," and "unlikely") but did not affect treatment decisions. The outcome measures were LBP, leg pain, activity limitation, and self-reported general health. Scores were compared between those with "likely" neuropathic pain (neuropathic group) and "unlikely" (non-neuropathic group), using Mann-Whitney, Friedman, and χ tests.

Results: At baseline, the neuropathic group had worse scores on all outcome measures, and analgesic use, sick leave, and sense of coherence (P = 0.000 to 0.044). At 3 months and 12 months, both groups improved (P = 0.001 to 0.032). However, the groups remained different at each time point on all outcome measures (P = 0.000 to 0.033) except LBP (P = 0.054 to 0.214).

Conclusions: The PDQ classification was a prognostic factor but was not predictive of response to treatment that was not targeted to neuropathic pain. Further studies should investigate whether PD-Q groups are predictive of treatment response when targeted to neuropathic pain.