to the editor: Although the article “Herpes Zoster and Postherpetic Neuralgia: Prevention and Managment”¹ in the September 15, 2005, issue of American Family Physician was generally well done and informative, it did not describe sources or selection criteria for articles to be used in generating evidence-based recommendations. We are concerned about the strength of evidence of the single citation² provided to support the use of the 5% lidocaine patch (Lidoderm) in the treatment of postherpetic neuralgia. The citation reported benefit in a neuropathic pain scale score in 96 patients participating in a randomized placebo-controlled trial.² In contrast to the authors' reviews¹ of other conventional pharmacotherapies for postherpetic neuralgia (tricyclic antidepressants, gabapentin [Neurontin], opiates, and capsaicin [Zostrix]), there is no discussion of methodology, numbers needed to treat, or side effects for the lidocaine patch.

We conducted a systematic review of treatments for postherpetic neuralgia³ and found insufficient published data to support the use of the lidocaine patch; however, at that time, it was the only drug that the U.S. Food and Drug Administration (FDA) had indicated for treatment of postherpetic neuralgia. We searched the publicly available information on the new drug application for the lidocaine patch.⁴ The original trial (still unpublished) had 150 patients and found a large placebo response in pain scores throughout the three- to four-week trial duration that was not significantly different from that demonstrated with the lidocaine patch.

Three findings from this trial were used to argue for FDA approval, but the relevance of these findings is unclear. A significant difference was reported in pain relief at the final visit (2.6 versus 2.1 on a 0 to 5 scale). Allodynia was improved at the beginning of the trial in patients receiving lidocaine, but this was based on investigators' sensory skin testing and was not evaluated in relation to patient disability. Finally, on trial exit, the increase in pain scores was greater among those receiving lidocaine.

The FDA denied approval and required “one additional efficacy study” demonstrating benefit before approval. Efficacy was then demonstrated in a placebo-controlled trial in 32 participants who were selected based on prolonged successful use of lidocaine patches.⁵ This finding met the technical requirement for FDA approval but has not been substantiated in unselected patients.

We encourage the article's authors and the readership of American Family Physician to adopt a more critical view of the lidocaine patch for the treatment of postherpetic neuralgia.

in reply: I would like to thank Drs. Alper and Lewis for their comments on our article.¹ The criteria for a level B Strength of Recommendation Taxonomy (SORT) recommendation are: inconsistent or limited-quality patient-oriented evidence. The outcome examined in the Galer study² is pain. We did not think that the evidence from this study² was strong enough to assign an evidence level of A (consistent, good-quality patient-oriented evidence), but because it was a randomized controlled trial, it certainly warranted a higher evidence level than C (consensus, disease-oriented evidence, usual practice, expert opinion, or case series). We agree with Drs. Alper and Lewis that the evidence for lidocaine patches in the treatment of postherpetic neuralgia is “inconsistent and of limited quality,” which, according to the SORT criteria,³ would give it a recommendation level of B.