Prognostic nerve blocks: meaningful or meaningless?
Updated: Jun 9, 2022
Prognostic blocks address the question: will a subsequent treatment be effective or not?
Prognostic blocks are a screening for a subsequent treatment, and are different from diagnostic and test blocks. Learn more here about the differences.
When performing a prognostic nerve block, a local anesthetic is injected in the proximity of a nerve. If the patient experiences pain relief, a subsequent treatment that targets that nerve or the tissues supplied by the nerve may be considered to achieve long-lasting improvement. In some cases, this treatment could be an ablation of the nerve using heat (radiofrequency ablation), cold (cryoablation), chemical agents (such as alcohol or phenol), or even surgical resection of the nerve. If the pain is not reduced after the prognostic nerve block, those treatments would not be indicated.
Research on the validity of prognostic nerve blocks is sparse. In order to fully validate nerve blocks as being truly prognostic, studies that are able to fill the table below would be required.
The table illustrates the four situations that may arise:
The prognostic block provides pain relief, and the subsequent treatment is effective (blue cell on the upper row)
The prognostic block provides pain relief, but the subsequent treatment is ineffective (red cell on the upper row)
The prognostic block does not provide pain relief, and the subsequent treatment is ineffective (blue cell on the lower row)
The prognostic block does not provide pain relief, but the subsequent treatment is effective (red cell on the lower row)
The higher the numbers in the blue cells, and the lower the numbers in the red cells, the better the validity of a prognostic block. In research, this is quantified by the sensitivity, specificity, positive and negative predictive value, and accuracy (the formulae are presented in the uncolored cells).
Research has been limited to the positive predictive value
The positive predictive value (upper row of the table) considers only patients who experience pain relief with the prognostic block; it tells us the percentage of these patients that will benefit from the subsequent treatment (e.g., a nerve ablation). The best studied prognostic blocks are those of the nerves that supply the cervical and lumbar facet joints. Patients who experience pain relief subsequently undergo radiofrequency ablation of those nerves. Reviews on the prognostic value of these blocks for cervical and lumbar procedures are available.
What about negative predictive value, sensitivity, and specificity?
The positive predictive value alone does not consider those patients who do not have pain relief with the prognostic block (2nd row of the table). Therefore, studying only patients with pain relief does not allow the calculation of the negative predictive value, sensitivity, specificity, and accuracy.
These data are very hard to be produced, and, to my knowledge, still unavailable. One hurdle is the ability to perform a therapeutic procedure in patients who do not benefit from prognostic block. For instance, it is ethically questionable to offer a nerve ablation in patients who do not experience pain relief with an anesthesia of that nerve.
Do negative predictive value, sensitivity, and specificity matter?
It depends. They are of crucial importance for diagnostic tests. However, I argue that they are of limited importance for prognostic nerve blocks.
Negative predictive value. For most treatments, there is no meaningful support for a potential efficacy of a treatment in patients who do not have pain relief with a prognostic block. This makes the negative predictive value of questionable interest.
Sensitivity and specificity. Unlike predictive values, sensitivity and specificity do not depend on the success rate of the treatment. This is important if we wanted to generalize the results of a prognostic block for a treatment with high success rate to prognostic blocks for treatments with low success rate, or the opposite. If this is not done, sensitivity and specificity are not crucial.
What we learn from facet joint pain and knee pain
If we consider radiofrequency ablation of the lumbar facet joints, data indicate that the procedure is successful in 33% of patients who do not undergo prognostic blocks (see here). This is mostly due to the fact that pain stems from the facet joints in a minority of patients with low back pain, and therefore performing the ablation in all of them benefits only a small proportion. The study showed that for patients who had pain relief after two prognostic blocks on two different days, the nerve ablation was successful in 64%. Therefore, the prognostic blocks, while having an unimpressive positive predictive value of 64%, significantly increases the odds of success of a subsequent ablation.
For knee pain due to osteoarthritis, genicular nerve blocks are performed to screen patients for a subsequent nerve ablation. However, performing or not performing the prognostic block does not change the outcome of the subsequent nerve ablation (see here for a comparative study). The most likely explanation is that there is only one main source of pain in knee osteoarthritis, and in fact most patients who receive the nerve block experience pain relief. If pain relief occurs in almost all patients, there is no point in performing the prognostic block.
Meaningful or meaningless?
From these examples it is clear that the usefulness of prognostic blocks is not universal, and depends on different factors, such as the prevalence of the disease and the success rate of the subsequent treatment. The prognostic block has clinical usefulness if it increases significantly the odds of success of the treatment, as compared to offering the treatment without performing the prognostic block. For procedures with high success rate, or when the prevalence of the pain condition is high, prognostic blocks do not add much, if at all, to the chance of success of the treatment.
The positive predictive value is a very useful parameter, provided that the success rate of the subsequent treatment is known. Studies that provide information on negative predictive value, sensitivity and specificity are desirable, but in most cases not strictly necessary and frequently unfeasible.