本年度迄今最震撼危重病界的文章在6月初的《Annals of internal medicine》发表—— Association between Critical Care Physician Management and Patient Mortality in the Intensive Care Unit
Mitchell M. Levy, John Rapoport, Stanley Lemeshow, Donald B. Chalfin, Gary Phillips, AND Marion Danis
Annals 2008 148: 801-809.
这是一篇“颠覆性”的文章，撼动学界是肯定的，关键是如何看待以及分析。废话少说，先从吾友“风乱月影”的笔记中把事件介绍扒过来：
我们面临的尴尬——重症监护医生可能会增加患者死亡率

重症监护医生可能会增加患者死亡率
Critically Ill Patients Seen More Likely to Die With Treatment by Intensivists
NEW
YORK (Reuters Health) Jun 02 - Surprisingly, and contrary to previous
findings, new research indicates that critically ill patients who are
managed by a critical care physician are more likely to die than those
managed by other doctors.
路透社（纽约）：6月2日
新近一项研究得出令人惊讶的结论，经过重症监护医生治疗的患者比经过其他医生治疗的患者有更高的死亡率。这与先前的研究结果完全相反。
While
some of this difference reflects the fact that patients treated by
critical care physicians are generally sicker, even after adjusting for
this, mortality was still higher, according to the report in the Annals
of Internal Medicine for June 3.
据6月3日的《the Annals of Internal Medicine 》报道，虽然这个结果反应出，重症监护医师经治的病人可能患有更加危重的疾病，但是经过疾病种类调整后，死亡率仍然偏高。
As an accompanying editorial emphasizes, however, the mechanisms by which intensivist care might increase mortality are unclear.
在本文编者按中，杂志编辑强调，重症监护医师导致患者死亡率增高的原因目前还不明确
"Our
study, which to our knowledge is based on the largest cohort ever
analyzed to examine the relationship of critical care management to
survival of critically ill patients, found some unexpected results,"
Dr. Mitchell M. Levy, from Rhode Island Hospital in Providence, and
colleagues note.
来自Rhode Island Hospital in Providence的Dr.
Mitchell M.
Levy和他的同事说道，“我们进行的这项研究，基于重症监护诊疗与危重患者生存率的关系的调查，它是该领域目前为止最大规模的cohort汇总分析，却
发现了意外的结果。”
Further studies, the team adds, are needed to
determine if the findings might have been the result of some
unrecognized remaining confounder of illness severity.
该研究小组补充说，需要开展进一步的研究，辨析是否这些尚未被确认的研究结果被（不同临床部门之间）疾病严重程度的差异所混淆。
Analyzing
data from 101,832 patients in 123 ICUs in the US, the researchers found
that in addition to being sicker than patients treated by non-critical
care physicians, those treated by critical care doctors received more
procedures.
研究者分析了来自123ICU的101832名患者的资料，发现除了重症监护医生所诊疗的疾病的严重程度要大于非重症监护医生所诊疗的疾病之外，重症监护医生的患者还接受了更多的临床干预。
As
noted, hospital mortality was higher for patients managed by critical
care physicians. Adjustment for illness severity (Simplified Acute
Physiology Score) and the probability of selective referral to a
critical care physician reduced, but did not eliminate, the mortality
difference.
如上文所述，重症监护医生所诊疗的患者的院内死亡率较高。虽然将疾病严重程度（Simplified Acute Physiology Score）进行调整，减少重症监护医生选择治疗措施的可能性，但这种死亡率的差异并未减少。
Specifically,
the team reports, "Among patients who received CCM (critical care
management) in ICUs that managed 5% to 95% of patients, the
standardized mortality ratio was 1.09 for patients who received CCM for
the entire stay compared with 0.91 for patients who did not receive
CCM."
The researchers put forth some potential explanations for
their findings, including the possibility that intensivists are more
likely than other physicians to abandon standardized protocols and
follow their own clinical judgment and the possibility that because
they are more comfortable performing procedures, they do more of them,
thereby increasing the risk of complications.
In a related
editorial, Dr. Gordon D. Rubenfeld, from the University of Toronto, and
Dr. Derek C. Angus, from the University of Pittsburgh, comment that
"although Levy and colleagues speculate about mechanisms by which
intensivists might increase mortality, they do not provide evidence to
support a proposed mechanism."
They add that "until someone
replicates Levy and colleagues' results in another cohort and provides
evidence for a mechanism by which intensivist-staffed ICUs increase
mortality, their study will remain one observation against many."
Ann Intern Med 2008;148:801-809,877-880.

我要提醒读者的是不妨先看看随本文刊发的述评： Are Intensivists Safe?   Gordon D. Rubenfeld, MD, MSc, and Derek C. Angus, MD, MPH  3 June 2008 | Volume 148 Issue 11 | Pages 877-879

首先述评作者大家应该有所了解：
第
一作者Gordon D. Rubenfeld来自加拿大，他的职业受训生涯相当显赫，分别与Johns Hopkins
University/Jefferson Medical College\Duke University以及 University of
Washington相关。其主业就是流行病学，而且身兼多个重要学术组织的委员。最后的是著名的Augus教授，来自匹兹堡，尽管比较年轻，但已经是危重病界流
行病学，卫生经济学以及健康服务研究的NO.1
级人物，而且属于少壮派。因此他参与的本篇评论意义自然重大），尤其是这样一篇关乎危重病医生前途的“颠覆性”文献。

还有从本文的分组可以看出，
分组患者的基本性质还是有偏倚的，大家注意本研究重监护治疗的患者以黑人居多（有关种族差异，实际上是经济背景差异会对研究结论产生重大影响——早期目标
治疗饱受质疑的原因之一就有这一条）、外院转入居多、呼吸道病因居多而心血管病因偏少、创伤与感染患者居多...上述偏倚很多已具有显著差异——看本文尤
其困难得的是，作者并没有给出以上差异的统计分析，而上述差异对后面结论的影响是无庸置疑的。

我尤其有一个假想：既然作者能发现监护治疗与死亡的风险，那么他们的数据库应该也可能会有监护医生的性别选项，能否进一步证明是男性还是女性监护医生更像“杀手”呢？另一个假想是,很快又会有一个相关问题的荟萃分析出台的——然后，你相信吗,荟萃分析的结论肯定是，如同述评所说的——还需要另外一个大型研究进一步探究、证实......

2.原IMPACT研究中所调研的指标并不能很好的考量本研究结论中出现的监护医生增加病死率的原因，就连作者们自己作的几个假设（诸如
“危重医生更易于放弃标准化方案去实施个体化治疗或者危重医生更愿意实施某些有创性操作，而这些操作会增加并发症发生的风险..."）也无法自数据中找到支持他们胡说八道的线索—— 这也再次说明这次研究是有严重缺陷的，或者说这根本就不是一个前瞻性的研究——由于其实质是简单的数据挪用，因此不能，也不可能很好的通过所谓“前瞻性”的设计去考察那些潜在的影响因素，或者充分稀释潜在的影响因素，尤其是某些未被纳入的但很可能直接影响到考察因素以及结论的指标，例如基础疾病，APACHE评分，血糖等... 说到底，本次研究的实质是作者们冀图充分利用某个既往的大型观察前瞻性研究的数据，试图发现某些新鲜的结论。但结论却令他们自己很尴尬，这是一个从未被既往其他研究所发现的反面结论。