Pulsed radiofrequency in clinical practice - A retrospective analysis of 238 patients with chronic non-cancer pain treated at an academic tertiary pain centre

Background and aims: Pulsed radiofrequency is a non-neurodestructive invasive pain treatment which, in contrast to conventional continuous radiofrequency treatment, does not entail nerve tissue destruction. The aim of this study was to retrospectively analyse the short-term benefits of a broad use of pulsed radiofrequency in clinical practice.

Methods: The medical records of all patients treated with pulsed radiofrequency, or who received a diagnostic test block with a local anaesthetic in view of such a treatment, were retrospectively analysed. The patients had been referred to a tertiary pain centre in Sweden. The treatment effect one month after pulsed radiofrequency was retrospectively graded as follows, based on the wordings of the medical records: major improvement; minor improvement; no change; or worsened.

Results: A total of 238 patients received 587 interventions from 2009 to 2014. Chronic low back pain (CLBP) was by far the most common treatment indication (57% of patients), followed by CLBP with sciatica (9%). The age at first pulsed radiofrequency was 55 (15-94) years (mean, range), and 65% were female. Thirty-six patients (15%) underwent only a diagnostic test block using a local anaesthetic, i.e., the test block did not lead to treatment with pulsed radiofrequency. A total of 445 pulsed radiofrequency interventions were performed on 202 patients. Dichotomizing data into responders (i.e., minor or major improvement) and non-responders (i.e., worsened or no change), we found that, out of 63 responders to a median branch diagnostic test block (either at the cervical or lumbar level), 33 were responders to the first following median branch pulsed radiofrequency. Hence the positive predictive value of a median branch test block was 52%. In 127 patients, the lumbar level was targeted for median branch pulsed radiofrequency because of clinically suspected lumbar facetogenic pain. Looking at the first treatment, 30% experienced major improvement after 1 month, 16% minor improvement, 36% no change, 5% a worsened situation, and the effect was not assessable in 13% of patients. Lone dorsal root ganglion L2-treatment for suspected discogenic lumbar pain was done on 39 patients and, after one month, the effect was not assessable in 17% of patients, 14% had major improvement, 14% minor improvement, and 55% had no change. In 40 patients, a dorsal root ganglion or a peripheral nerve was targeted because of a non-axial chronic pain condition. There was a plethora of indications, but the most common was by far related to some form of neuropathic pain (52% of interventions, mainly because of neuralgia), followed by chronic nociceptive shoulder pain (8% of interventions).

Conclusions: This study shows that, after one month, the effect size of a broad and indiscriminate clinical use of pulsed radiofrequency is rather small.

Implications: The clinical effectiveness of pulsed radiofrequency has to be investigated further in carefully selected and more homogenous patient groups, in order to define effective treatment niches for this nondestructive invasive treatment method.