胡大一 北京大學人民醫院教授，中國控煙協會會長、著名心血管專家、醫學教育家，預防醫學家
對穩定冠心病患者，支架不可能預防心肌梗死，也不能降低總死亡率。
如能戒菸限酒，認真用好他汀，阿司匹林等藥物，參加心臟康復，可長治久安。
冠狀動脈CT發現70%-80%左右狹窄，無症狀，應做運動心電圖負荷試驗（運動平板或踏車）。如運動試驗未見心肌缺血，或有缺血，但不重，不要被過度支架。
目前，臨床上一個極大的問題是信息不對稱。醫院心臟科病歷上把很多無胸痛症狀的患者也診斷成“不穩定性心絞痛”；本是穩定的心絞痛，也故意寫成“不穩定心絞痛”。把一些焦慮抑鬱的胸痛表現也診為“不穩定性心絞痛”，成為一種令人氣憤的“集體撒謊”，誤導誘導許多不需做支架，或做支架弊大於利的患者，花大錢買風險。
【紐約時報 2018年2月12日刊文】
心臟支架對於大多數穩定冠心病患者無效，但仍然被廣泛使用
當我的孩子們小時候，如果他們抱怨疼痛，我有時會在他們身上抹些潤膚霜，告訴他們“奶油”會有幫助。通常情況下，安慰劑效應出奇有效。潤膚霜很便宜，幾乎沒有副作用，它實現了預期的效果，這是個完美的解決方案。
其他一些療法也有安慰劑效應，讓人感覺更好。但大多情況下這是有危險的，我們必須權衡利弊。
許多美國人由於心肌缺血、缺氧而引起胸部疼痛，這就是所謂的心絞痛。幾十年來，最常見的治療方法之一是在心臟供血冠狀動脈置入網狀支架。支架使血管開通，增加血液流向心臟，從理論上解決了這個問題。
插入這些支架的心臟病專家發現他們的患者感覺好些了。他們似乎更健康。許多人認為這些支架可以防止心肌梗死甚至死亡。經皮冠狀動脈介入治療植入支架非常普遍。
2007年發表在《新英格蘭醫學雜誌》的一項隨機對照試驗，主要研究終點是心肌梗死和死亡。研究人員收集了近2300名有嚴重冠狀動脈疾病並診斷有心肌缺血的患者，隨機將他們分成置入支架合併藥物治療組或單純藥物治療組。隨訪患者多年。結果呢？支架在預防這些患者不良後果方面沒有任何區別。
這讓人難以置信。因此隨後進行了更多的研究。
2012，研究者彙集並在JAMA Internal Medicine雜誌發佈了一項薈萃分析，三項針對心肌梗死後穩定患者進行的研究，另五項評價尚未發生過心肌梗死，而有穩定心絞痛或心肌缺血的患者。薈萃分析顯示，支架對預防穩定冠心病患者心肌梗死或死亡沒有效果。
然而，許多心臟病專家認為，支架改善了患者的疼痛，改善了他們的生活質量。即使沒有達到預期，即降低患者為中心的預後結果，但對那些植入支架的患者能改善生活。
問題是，很難知道支架是否導致疼痛緩解，或者是否是安慰劑效應。畢竟，通過植入手術這一流程，安慰劑效應非常明顯。因此需要一個假的控制流程試驗，這個過程可使患者不清楚他們是否置入了支架。
許多醫生反對這種研究。他們認為，大量心臟病學專家的經驗已證明了支架的效果，因此讓隨機分組的一些患者不接受支架是不道德的。另一些人認為，將患者暴露在假手術中是錯誤的，因為這會使他們受到潛在的傷害而毫無益處。更多持懷疑態度的觀察者可能會注意到，一些醫生和醫院在執行這一程序時也獲得了經濟回報。
無論如何，這樣的試驗已經完成，結果在今年公佈。
研究人員在英國的五個地方徵集患有嚴重冠心病的患者，並將其隨機分為兩組。所有患者按要求服用一段時間藥物。然後，第一組患者接受了支架。第二組患者麻醉鎮靜至少15分鐘，但沒有置入支架。
六週後，所有患者在運動平板進行測試。運動會使患者感到疼痛，而運動負荷監測是檢查心絞痛的常用方法。在測試時，患者和心臟病專家都不知道（雙盲）是否放置了支架。根據測試結果，他們甚至在測試後也無法得出結論：干預組和安慰劑組之間的結果沒有差別。支架甚至沒有減輕疼痛。
一些值得注意的事項：所有患者在接受手術前都接受藥物治療，所以很多患者在接受支架之前都有了顯著的改善。現實世界中，一些患者也許不會堅持規範頻繁的醫治，因此對這些患者來說，可能會受益於支架（我們尚不知道）。本研究隨訪時間僅為六週，因此長期效果尚不清楚。這些結果只適用於穩定型心絞痛患者。可能對有多處病變的嚴重患者，或藥物治療無效的患者會從支架獲益。
但很多人，也許大多數患者，可能並不需要他們。這對患者和醫生來說是很困難的。因為在他們的經驗中，置入支架的患者得到了改善，他們似乎從手術中受益。但這種獲益似乎是因為安慰劑效應，而不是來自血液流動改善的任何物理變化。研究中的患者感覺和我的孩子在擦“潤膚霜”時的感覺一樣好。
不同的是，潤膚霜不會真的引起傷害，但是支架置入會引起。在這項研究中，即使2%的患者有重大出血事件，但要記住，每年都要植入數十萬個支架。同時支架也很貴，可能增加至少10000美元的治療費用。
支架在醫療中仍然佔有一席之地，但比我們過去認為的要少得多。然而，許多醫生和患者仍然會要求使用支架，指出支架會使一些人得到改善，即使這種改善來自安慰劑效應。
在這方面，支架手術並不是孤例。可能許多醫療手術的效果都未必比假手術安慰效果好。雖然我們永遠不會批准那些獲益不超過安慰劑效果的藥物，但我們在醫療裝置方面沒有相同的標準。2014年，Rita Redberg在新英格蘭醫學雜誌指出，只有1%被批准的醫療設備被要求提交臨床數據的批文，而這些數據幾乎都是小規模試驗與有限時間的隨訪。以證明沒有安慰劑效應的隨機對照試驗非常罕見。
似乎有一個強烈的觀點，即我們應該更加意識到我們願意冒什麼風險，以及我們願意為安慰劑效應付出什麼。如果我們不想放棄利益，我們是否應該設計更便宜、更安全的假手術程序來達到同樣的效果？這是道德的嗎？難道比向那些收費五位數並把他們置於嚴重不良事件的風險更不道德嗎？
看起來穩定的單支血管病變患者應該被告知，支架並不比假手術更好，也不比藥物治療好。有些人可能仍然選擇支架，他們至少應該知道他們要付多少錢。
【原文】
Heart Stents Are Useless for Most Stable Patients. They’re Still Widely Used.
When my children were little, if they complained about aches and pains, I’d sometimes rub some moisturizer on them and tell them the “cream” would help. It often did. The placebo effect is surprisingly effective.
Moisturizer is cheap, it has almost no side effects, and it got the job done. It was a perfect solution.
Other treatments also have a placebo effect, and make people feel better. Many of these are dangerous, though, and we have to weigh the downsides against that benefit.
Lots of Americans have chest pain because of a lack of blood and oxygen reaching the heart. This is known as angina. For decades, one of the most common ways to treat this was to insert a mesh tube known as a stent into arteries supplying the heart. The stents held the vessels open and increased blood flow to the heart, theoretically fixing the problem.
Cardiologists who inserted these stents found that their patients reported feeling better. They seemed to be healthier. Many believed that these stents prevented heart attacks and maybe even death. Percutaneous coronary intervention, the procedure by which a stent can be placed, became very common.
Then in 2007, a randomized controlled trial was published in The New England Journal of Medicine. The main outcomes of interest were heart attacks and death. Researchers gathered almost 2,300 patients with significant coronary artery disease and proof of reduced blood flow to the heart. They assigned them randomly to a stent with medical therapy or to medical therapy alone.
They followed the patients for years. The result? The stents didn’t make a difference beyond medical treatment in preventing these bad outcomes.
This was hard to believe. So more such studies were conducted.
In 2012, the studies were collected in a meta-analysis in JAMA Internal Medicine. Three studies looked at patients who were stable after a heart attack. Five more examined patients who had stable angina or ischemia but had not yet had a heart attack. The meta-analysis showed that stents delivered no benefit over medical therapy for preventing heart attacks or death for patients with stable coronary artery disease.
Still, many cardiologists argued, stents improved patients’ pain. It improved their quality of life. Even if we didn’t reduce the outcomes that physicians cared about, these so-called patient-centered outcomes mattered, and patients who had stents reported improvements in these domains in studies.
The problem was that it was difficult to know whether the stents were leading to pain relief, or whether it was the placebo effect. The placebo effect is very strong with respect to procedures, after all. What was needed was a trial with a sham control, a procedure that left patients unclear whether they’d had a stent placed.
Many physicians opposed such a study. They argued that the vast experience of cardiologists showed that stents worked, and therefore randomizing some patients not to receive them was unethical. Others argued that exposing patients to a sham procedure was also wrong because it left them subject to potential harm with no benefit. More skeptical observers might note that some doctors and hospitals were also financially rewarded for performing this procedure.
Regardless, such a trial was done, and the results were published this year.
Researchers gathered patients with severe coronary disease at five sites in Britain, and randomized them to one of two groups. All were given medication according to a protocol for a period of time. Then, the first group of patients received a stent. In the second, patients were kept sedated for at least 15 minutes, but no stent was placed.
Six weeks later, all the patients were tested on a treadmill. Exercise tends to bring out pain in such patients, and monitoring them while they’re under stress is a common way to check for angina. At the time of testing, neither the patient nor the cardiologist knew whether a stent had been placed. And, based on the results, they couldn’t figure it out even after testing: There was no difference in the outcomes of interest between the intervention and placebo groups. Stents didn’t appear even to relieve pain.
Some caveats: All the patients were treated rigorously with medication before getting their procedures, so many had improved significantly before getting (or not getting) a stent. Some patients in the real world won’t stick to the intensive medical therapies, so there may be a benefit from stents for those patients (we don’t know). The follow-up was only at six weeks, so longer-term outcomes aren’t known. These results also apply only to those with stable angina. There may be more of a place for stents in patients who are sicker, who have disease in more than one blood vessel, or who fail to respond to medical therapy.
But many, if not most patients, probably don’t need them. This is hard for patients and physicians to wrap their heads around because, in their experience, patients who got stents got better. They seemed to receive a benefit from the procedure. But that benefit appears to be because of the placebo effect, not any physical change from improved blood flow. The patients in the study felt better from a procedure in the same way that my children did when I rubbed moisturizer on them.
The difference is that while the moisturizer can’t really harm, stent placement can. Even in this study, 2 percent of patients had a major bleeding event. Remember that hundreds of thousands of stents are placed every year. Stents are also expensive. They can add at least $10,000 to the cost of therapy.
Stents still have a place in care, but much less of one than we used to think. Yet many physicians as well as patients will still demand them, pointing out that they lead to improvements in some people, even if that improvement is from a placebo effect.
Stents are probably not alone in this respect. It’s possible that many procedures aren’t better than shams. Although we would never approve a drug without knowing its benefits above a placebo, we don’t hold devices to the same standard. As Rita Redberg noted in The New England Journal of Medicine in 2014, only 1 percent of approved medical devices are approved by a process that requires the submission of clinical data, and that data is almost always from one small trial with limited follow-up. Randomized controlled trials are very rare. The placebo effect is not.
There seems to be a strong argument that we should be more conscious of what we are willing to risk, and what we are willing to pay, for a placebo effect. If we don’t want to give up the benefit, should we design cheaper, safer fake procedures to achieve the same results? Is that ethical? Is it more unethical than charging people five figures and putting them at risk for serious adverse events?
It surely seems reasonable that stable patients with single-vessel disease should be informed that stents work no better than fake procedures, and no better than medical therapy. Some may still choose a stent. They should at least know what they’re paying for.
原文地址：
https://www.nytimes.com/2018/02/12/upshot/heart-stents-are-useless-for-most-stable-patients-theyre-still-widely-used.html?partner=rss&emc=rss
瞭解更多名醫的健康科普，可關注“大醫論道（dayilundao2017）”公眾號。
閱讀更多 大醫論道 的文章
