Skip navigation

Does treating acute herpes zoster in primary care stop post herpetic neuralgia?

There is no consensus on how acute herpes zoster (shingles) should be managed in general practice. A particular problem is that patients with the disease go on to have persistent pain afterwards. It is thought that about 15% of patients have pain one month after healing of an acute herpetic rash, with about one quarter of these still have pain at one year.

This risk of postherpetic neuralgia increases with age. It can be as high as 50% in patients aged over 60 years and 75% in those aged over 75 years.

Treatment of the acute rash can include antiviral agents and corticosteroids. These have a cost (a week's treatment with acyclovir can cost over £100) and some adverse effects. Postherpetic neuralgia imposes a significant burden of suffering, and effective remedies that prevented postherpetic pain would be welcomed.

The question of whether treatments now used are effective is largely answered by a thorough systematic review and meta-analysis [1].

The study from Oxford and Flinders identified studies of the treatment of herpes zoster using standard searching techniques. Twenty-one trials were identified in this way; they examined several treatments, including acyclovir, other antivirals and corticosteroids.

The main outcome sought was prevalence of pain at one, three and six months after onset of acute herpetic rash.


Acyclovir was studied in eight trials with over 900 patients followed up who had received acyclovir for between five and ten days. At six months, 76 of 466 (16%) of patients receiving placebo had pain, compared with 61 of 455 (13%) of patients who received acyclovir.

No statistically significant reduction in pain was detected at one month (odds ratio 0.85, 95%CI 0.61 - 1.19) or six months (0.70, 0.47 - 1.06), but at three months there was a significant benefit (0.65, 0.46 - 0.93). Similar results were obtained for the studies which used only higher doses of acyclovir.

Other antivirals

These studies generally failed to demonstrate any differences at six months, or were of low numbers or quality, making conclusions suspect.


These were also of generally poor quality, and failed to show any significant differences.

How big is the effect

Remembering that there was no statistical difference between acyclovir and placebo, what would be the size of any effect if it were present but the studies were insufficiently large to demonstrate it? The best interpretation is that if acyclovir truly reduced the prevalence of postherpetic neuralgia by 30%, then 15 patients would have to receive a course of treatment to prevent one developing postherpetic neuralgia. To demonstrate an effect of 30% would need a study with over 1,500 patients.

The argument is not complex. Studies have been done which have had low power to detect even a relatively significant effect. Overall, the combined results in over 900 patients are not convincing. Getting a definitive answer would be long and costly, and the result even then would be likely to be clinically insignificant even if it were statistically significant.

Why bother?

With limited resources, the implication of all this is that different treatments should be explored.


  1. T Lancaster, C Silagy, S Gray. Primary care management of acute herpes zoster: systematic review of evidence from randomized controlled trials. British Journal of General Practice 1995 45: 39-45.

previous or next story in this issue