When numbers get serious.
When numbers get serious.
复制标题
当数字变得严重时。
DOI：
10.1007/bf02599661
复制
发表时间：
1994
影响因子：
5.7
通讯作者：
NeaseJr,RF
中科院分区：
文献类型：
--
作者：
Littenberg,B;NeaseJr,RF
数字对于医学话语来说相对较新。过去，医生不会说出数字：发烧是用文字来描述的，贫血是用病人皮肤的苍白程度来衡量的，低血压是用“脉搏微弱”来描述的，而不是用一堆数字来描述。随着温度计和血压计等可靠仪器的出现，定量描述变得司空见惯。如今，我们的患者迫切要求“了解他们的营养状况”，包括胆固醇、血压，有时还有数十种其他看不见的个人特征，而我们往往不愿意提供这些数字。提供者和患者都对数字感到满意。至少有一些数字。我们尚未看到医生和患者广泛使用概率来定量表达与临床决策相关的事件的可能性。为什么对将可能性量化为概率犹豫不决？首先，数字的使用凸显了医疗决策中固有的不确定性。它强调了一个简单的事实：对于特定患者，我们不知道治疗是否有效或是否会发生某些不良事件。其次，通常无法获得使临床医生能够轻松地为个体患者提供这种定量估计所需的数据。由于这些和其他原因，医生和患者继续用言语而不是概率来交流事件的可能性。在本期杂志中，Mazur 和 Merz z 报道了一项实验，该实验为医生和患者在日常临床护理中如何解释词语提供了一些见解。他们询问了 200 多名普通内科门诊患者，以量化他们认为医生所说的“罕见”在手术并发症中的含义。他们发现了很大的变异性，这部分是由重要的协变量解释的。 “罕见”一词的数字翻译随患者的年龄、向患者描述的具体临床状况（死亡或肺炎）以及用于引出翻译的量表类型而变化。
NUMBERS are relatively new to medical discourse. In the old days, doctors did not utter numbers: fevers were described with words, anemias were measured by the pallor of the patient's skin, and low blood pressure was described as" a thready pulse" rather than by a flock of digits. With the advent of reliable instrumentation such as the thermometer and the sphygmomanometer, quantitative descriptions became commonplace. Today, our patients eagerly demand to" know their nutlihers" for cholesterol, blood pressure, and sometimes dozens of other invisible personal characteristics, and we are more often than not willing to supply those numbers. Both providers and patients are comfortable with numbers. At least with some numbers. We have yet to see the widespread use by physicians and patients of probabilities as a quantitative expression of the likelihood of events relevant to clinical decision making. Why th'e hesitancy to quantify likelihood as a probability? First, the use of numbers brings the uncertainty inherent in medical decision making to the fore. It highlights the plain truth that we don't know for a particular patient whether the treatment will work or whether some adverse event will occur. Second, the data necessary to make clinicians comfortable with providing such a quantitative estimate for the individual patient are usually not available. For these and other reasons, physicians and patients continue to communicate about the likelihood of events with words rather than probabilities. In this issue of the Journal, Mazur and Merz z report an experiment that provides some insights into how doctors and patients interpret words in daily clinical care. They asked more than 200 general internal medicine outpatients to quantify what they thought the doctor meant by" rare" in the context of complications for surgery. They found substantial variability, which is partially explained by important covariates. The numeric translation of the word" rare" varied with the patient's age, the specific clinical condition described to the patient (death or pneumonia), and the type of scale used to elicit the translation.