A few weeks' treatment with an opioid can help relieve neuropathic pain

During 15 years of research on the impact of opioids on neuropathic pain, only 607 patients have been randomised in comparative trials of reasonable quality, the longest of which lasted only eight weeks. That's not much to go on but, according to this meta-analysis, it's enough to show fairly conclusively that opioids such as morphine and oxycodone can relieve chronic neuropathic pain if taken regularly for several weeks. Short term treatments such as with fentanyl or intravenous morphine looked less effective, but it's impossible to say for certain because the trials were so small (median 13 patients) and hard to combine in any meaningful way.

Most of the 403 patients in the longer trials had post-herpetic neuralgia, diabetic neuropathy, or phantom pain. Taking opioids for a few weeks reduced their pain by a mean of 14 points on a scale of 1-100 compared with placebo—a 20-30% improvement, which the authors think is clinically meaningful and similar to the pain relief these patients get from other effective treatments such as gabapentin. The commonest side effects were nausea (number needed to harm 3.6), constipation (4.6), drowsiness (5.3), vomiting (6.2), and dizziness (6.7).

None of the trials in this review mentioned or measured the risk of addiction or misuse, possibly because in most of them the treatment period was too short for addiction to be a problem. Bigger, better, and longer trials are urgently needed to find out if opioids work long term for these patients, if the risks are acceptable, and if the moderate pain relief reported in this review translates to a better quality of life for patients.

Death from idiopathic pulmonary fibrosis is often preceded by rapid and unpredictable decline

Pulmonary fibrosis is one of the commonest idiopathic interstitial pneumonias, but we known little about it beyond the fact that prognosis is generally poor. Studying patients in the placebo arms of …