现代流行病-Chapter3-Prevalence, Incidence, and Mean Duration

Posted 2021-04-29 现代流行病学3rd

Prevalence,Incidence, and Mean Duration 患病率、发病率和平均病程

Often, the study of prevalence in place of incidence is rationalized on the basis ofthe simple relation between the two measures that obtains under certain very special conditions. We will examine these conditions carefully, with the objective of explaining why they rarely if ever provide a secure basis for studying prevalence as a proxy for incidence.

通常，替代发病率的患病率研究是基于两个指标间简单的关系基础而变得合理的，这个简单的关系是由一定的特殊条件下获得的。我们将仔细检验这些条件，目的在于解释（如果有过的话）为什么它们很少为患病率替代发病率的研究提供安全的基础。

 

Recall that a stationary population has an equal number of people entering and exiting during any unit of time. Suppose that both the population at risk and the prevalence pool are stationary and that everyone is either at risk or has the disease. Then the number of people entering the prevalence pool in any time period will be balanced by the number exiting from it:

回忆一下，稳态人群在任何时间都有相同数量进入或退出的人。假设风险人群和患病池都是稳态的，那么，每个人不是处于风险之中，就是患病。这就是说，在任何时间内，进入患病池的人数都将等于退出患病池的人数。

People can enter the prevalence pool from the nondiseased population and by immigration from another population while diseased. Suppose there is no immigration into or emigration from the prevalence pool, so that no one enters or leaves the pool except by disease onset, death, or recovery. If the size of the population is N and the size of the prevalence pool is P, then the size of the populationat risk that feeds the prevalence pool will be N - P. Also, during any time interval of length Δ t, the number of people who enter the prevalence pool will be

人们能从非患病人群进入患病池，还可以从另一个患病人群移入这个患病池。假设患病池没有移入或移出，那么除了发病、死亡或者康复，没人进入或者退出患病池。如果人群规模是N，患病池规模是P，那么满足患病池的风险人群规模将是N-P。也就是说，在任何时间段Δt，进入患病池的人数将是：

where I is the incidence rate, and the outflow from the prevalence pool will be

I是发病率，而患病池流出的人数将是：

where I' represents the incidence rate of exiting from the prevalence pool, that is, the number who exit divided by the person-time experience of those in the prevalence pool. Therefore, in the absence of migration, the reciprocal of I' will equal the mean duration of the disease, D, which is the mean time until death or recovery. It follows that

I’表示从患病池退出的发生率，即患病池中的退出人数除以他们的人时。因此，在没有迁移的情况下，I’的倒数等于平均患病时间D，这是直至死亡或者康复的平均时间。由此可得：

which yields 推出：

P/(N -P) is the ratio of diseased to nondiseased people in the population or, equivalently, the ratio of the prevalence proportion to the nondiseased proportion. (We could call those who are nondiseased healthy except that we mean they do not have a specific illness, which does not imply an absence of all illness.)

P/(N -P)是患病人数与非患病人数的比，或者相当于，患病比例与非患病比例的比。（我们可以把那些没有患病的人称为健康的人，只是我们的意思是他们没有特定的疾病，这并不意味着没有任何疾病）。

 

The ratio P/(N -P) is called the prevalence odds; it is the ratio of the proportion of a population that has a disease to the proportion that does not have the disease. As shown above, the prevalence odds equals the incidence rate times the mean duration of illness.If the prevalence is small, say <0.1, then

P/(N -P)称为患病比（prevalence odds），它是患病人群的构成比与非患病人群的构成比的比。如前所述，患病比等于发病率乘以平均患病时间。如果患病率很小，例如<0.1，那么

because the prevalence proportion will approximate the prevalence odds for small values of prevalence. Under the assumption of stationarity and no migration in or out of the prevalence pool (Freeman and Hutchison, 1980),

因为在患病率低的时候，患病比例将约等于患病比。在这个稳态假设下，而且患病池没有移入或移出(Freeman and Hutchison, 1980)，

which can be obtained from the above expression for the prevalence odds, P/(N -P).

这可由上式患病比P/(N -P)的表达式推出。

 

Like the incidence proportion, the prevalence proportion is dimensionless, with a range of 0 to 1. The above equations are in accord with these requirements, because in each of them the incidence rate, with a dimensionality of the reciprocal of time, is multiplied by the mean duration of illness, which has the dimensionality of time, giving a dimensionless product. Furthermore, the product I ·has the range of 0 to infinity, which corresponds to the range of prevalence odds, whereas the expression

像发病比一样，患病比是无量纲的，取值范围为0到1。上述等式与这些要求一致，因为在每个里边，发病率（有时间倒数的量纲）乘以平均患病时间（有时间量纲）获得一个无量纲的乘积。此外，乘积I·D的取值范围为0到无穷，它对应的是患病比的范围，而表达式是

is always in the range 0 to 1, corresponding to the range of aproportion.

取值范围总是0到1，与构成比范围一致。

 

Unfortunately, the above formulas have limited practical utility because of the nomigration assumption and because they do not apply to age-specific prevalence (Miettinen, 1976a). If we consider the prevalence pool of, say, diabetics who are 60 to 64 years of age, we can see that this pool experiences considerable immigration from younger diabetics aging into the pool, and considerable emigration from members aging out of the pool. More generally, because of the very strong relation of age to most diseases, we almost always need to consider age-specific subpopulations when studying patterns of occurrence. Under such conditions, proper analysis requires more elaborate formulas that give prevalence as a function of age-specific incidence, duration, and other population parameters (Preston, 1987; Manton and Stallard, 1988; Keiding, 1991; Alho, 1992).

不幸的是，因为没有迁移的假设，而且因为他们不能用于年龄别患病率,上述公式的实用价值有限(Miettinen, 1976a)。考虑60-64岁的糖尿病患病池，我们可以看到这个患病池中相当多的年轻糖尿病患者因年纪增大而进入患病池，同时相当多的患者因年纪增大而移出患病池。更一般化的说，由于由于年龄与大多数疾病的关系非常密切，我们几乎总是需要在研究发病模式时考虑年龄别亚组人群。在这个情况下，适当的分析需要更详细的公式，使患病率作为年龄别发病率、病程和其他人群参数的函数(Preston, 1987; Manton and Stallard, 1988; Keiding, 1991; Alho, 1992)。