In robust trials side effects are noted meticulously but in acupuncture trials under-reporting of harm is extensive. For example, in a review of back pain, only 14 of 35 trials reported on harms (5% of patients receiving acupuncture and none of those receiving placebo).
在强有力的试验中，副作用经常被人精心指出，但在针灸试验中，漏报的伤害却非常广泛。例如，在对背痛的综述中，35个试验中只有14个被报道有害（5%的患者接受针灸治疗，没有接受安慰剂治疗）。

The cost of acupuncture to the NHS isn't peanuts, sessions range from ?25 to ?70. The total cost to the NHS may be ?25million a year, though reliable figures are hard to come by. I've never seen any convincing evidence of the effectiveness of acupuncture for treating pain.
对英国国民健康体系而言，针灸费用并非一笔小数目，疗程从25英镑至70英镑不等。每年英国国民健康体系花费的总费用可达2500万英镑，尽管可靠的数据仍旧难以获得。我从未见过任何令人信服的证据，证明针灸能有效治疗疼痛。

Edzard Ernst, emeritus professor in Complementary Medicine at Exeter University, and the UK's preeminent expert, goes further, saying: "Doctors should not recommend acupuncture for pain because there is insufficient evidence it is clinically worthwhile."
埃克塞特大学补充医学的名誉教授、英国杰出专家埃扎德·恩斯特进一步表示："医生不该建议用针灸缓解疼痛，因为尚未有足够的证据证明其是临床有效的。"

Cochrane Reviews are the gold standard for proving medical effectiveness. There have been 12 of them on acupuncture for pain and not one of them has shown an effect beyond placebo in a raft of conditions.
《考科兰回顾》（Cochrane Reviews）是证明医疗效果的黄金标准。已有12篇关于针灸治疗疼痛的报道，但却未有一篇证明在一系列病症中，针灸的疗效比安慰剂好。

These include lower back pain, rheumatoid arthritis, cancer pain, dysmenorrhoea, lateral elbow pain, endometriosis, peripheral joint osteoarthritis, prevention of migraine and tension type headache, shoulder pain, fibromyalgia, and pain in labour. The reviews on back pain, migraine, and tension headache considered acupuncture as a possible treatment but were based mostly on trials with non-blinded patients.
这些包括腰痛、类风湿性关节炎、癌痛、痛经、肘外侧疼痛、子宫内膜异位症、外周关节骨关节炎、预防偏头痛和紧张型头痛、肩痛、纤维肌痛和产痛。关于背痛、偏头痛和紧张型头痛的回顾认为针灸是一种可行的治疗方法，但主要是基于对拥有正常视力的患者的试验。

This weakens the results because patients knew they were receiving acupuncture which itself has a placebo effect. Effects beyond placebo were "small". Many trials are weak, that is, given to error. Patient blinding, crucial for meaningful results, is often inadequate in acupuncture trials resulting in small to moderate false positives.
这削弱了结果，因为患者知道他们在接受针灸，而针灸本身就具有安慰剂的效果。超出安慰剂的效果"极微"。很多试验都很弱，也就是说，有一定的误差。盲人患者对于有意义的结果至关重要，但针灸试验的盲人患者却很少，因此导致小至中等的虚假积极结果。

As Professor Ernst explains in the British Medical Journal, acupuncture is often regarded as harmless. Needling, however, may cause side effects such as pain, haemorrhages, infection, pneumothorax (rupture of the lung) and even death.
恩斯特教授在《英国医学杂志》上解释，针灸通常被认为是无害的。然而，针刺可能导致副作用，比如疼痛、出血、感染、气胸（肺部破裂），甚至会导致死亡。

In Denmark, for instance, four cases of pneumothorax, one fatal, were revealed in 2017. Such complications might be rare, but gathering exact numbers is stymied by unreliable information.
例如，2017年丹麦揭露了4例气胸和1例死亡事件。这些并发症可能并不多见，但对确切数字的收集却会受到不可靠信息的阻碍。