N**********y 发帖数: 61 | 1 一个城市有5家医院,想COMPARE QUALITY, 有一个 INDICATOR "QUALITY NUMBER"
。 现在因为考虑每个医院ADMIT 病人不一样,想ADJUST "QUALITY NUMBER" WITH "
SICKNESS"。请问怎么做?
谢谢。太DUMMY 了不好意思 |
P****D 发帖数: 11146 | 2 看不懂。你到底是要adjust for某种疾病(sickness难道不是这个意思?)还是病人人
数?
什么叫“病人不一样”?哪些方面不一样?你们打算考虑哪些方面? |
N**********y 发帖数: 61 | 3 adjust for how sick the patients are
sickness is a variable, every patient has a score from 1 to 4.
and we have quality variable for each patient visit (numeric).
each patient has one record (one quality score, one sickness score) and the
patients are from different hospitals.
【在 P****D 的大作中提到】 : 看不懂。你到底是要adjust for某种疾病(sickness难道不是这个意思?)还是病人人 : 数? : 什么叫“病人不一样”?哪些方面不一样?你们打算考虑哪些方面?
|
y*******u 发帖数: 930 | 4 assign each score with one variable
then do regression.
the
【在 N**********y 的大作中提到】 : adjust for how sick the patients are : sickness is a variable, every patient has a score from 1 to 4. : and we have quality variable for each patient visit (numeric). : each patient has one record (one quality score, one sickness score) and the : patients are from different hospitals.
|
P****D 发帖数: 11146 | 5 我猜你那个quality score,是病人self report的吧。你这样的,弄个简单的ANCOVA就
可以。Y是quality score,X是医院的identifier和sickness score。
至于结果会成什么样,做了很久self-reported data(如果你的score是这样得来)的
人预先对你表示一下祝福…… |
N**********y 发帖数: 61 | 6 一个城市有5家医院,想COMPARE QUALITY, 有一个 INDICATOR "QUALITY NUMBER"
。 现在因为考虑每个医院ADMIT 病人不一样,想ADJUST "QUALITY NUMBER" WITH "
SICKNESS"。请问怎么做?
谢谢。太DUMMY 了不好意思 |
P****D 发帖数: 11146 | 7 看不懂。你到底是要adjust for某种疾病(sickness难道不是这个意思?)还是病人人
数?
什么叫“病人不一样”?哪些方面不一样?你们打算考虑哪些方面? |
N**********y 发帖数: 61 | 8 adjust for how sick the patients are
sickness is a variable, every patient has a score from 1 to 4.
and we have quality variable for each patient visit (numeric).
each patient has one record (one quality score, one sickness score) and the
patients are from different hospitals.
【在 P****D 的大作中提到】 : 看不懂。你到底是要adjust for某种疾病(sickness难道不是这个意思?)还是病人人 : 数? : 什么叫“病人不一样”?哪些方面不一样?你们打算考虑哪些方面?
|
y*******u 发帖数: 930 | 9 assign each score with one variable
then do regression.
the
【在 N**********y 的大作中提到】 : adjust for how sick the patients are : sickness is a variable, every patient has a score from 1 to 4. : and we have quality variable for each patient visit (numeric). : each patient has one record (one quality score, one sickness score) and the : patients are from different hospitals.
|
P****D 发帖数: 11146 | 10 我猜你那个quality score,是病人self report的吧。你这样的,弄个简单的ANCOVA就
可以。Y是quality score,X是医院的identifier和sickness score。
至于结果会成什么样,做了很久self-reported data(如果你的score是这样得来)的
人预先对你表示一下祝福…… |
w******4 发帖数: 488 | 11 how about quantile regression? I suppose there are outliers for quality
score. |
P****D 发帖数: 11146 | 12 如果他的quality score是我们做过的那个,那就没有outlier。是一个survey,对于每
一个问题病人从“很不好”、“很好”之类的选项里面选答案,不同的选项有不同的分
数0-4,最后把所有问题的分数加在一起就是这个病人给医院的quality打的总分。
【在 w******4 的大作中提到】 : how about quantile regression? I suppose there are outliers for quality : score.
|