Hi,
I am working on problem 2 of Chapter 8 in Data Analysis and Graphics Using R and
don't know how to approach the second half of the question:
In the data set (an artificial one of 3121 patients, that is similar to a subset
of the data analyzed in Stiell et al., 2001) head.injury, obtain a logistic
regression model relating clinically.important.brain.injury to other variables.
Patients whose risk is sufficiently high will be sent for CT (computed
tomography). Using a risk threshold of 0.025 (2.5%), turn the result into a
decision rule for use of CT.
This is what I have so far:
> names(head.injury)
[1] "age.65" "amnesia.before"
[3] "basal.skull.fracture" "GCS.decrease"
[5] "GCS.13" "GCS.15.2hours"
[7] "high.risk"
"loss.of.consciousness"
[9] "open.skull.fracture" "vomiting"
[11] "clinically.important.brain.injury"> attach(head.injury)
> head.glm = glm(clinically.important.brain.injury ~ ., family=binomial,
data=head.injury)
> summary(head.glm)
Call:
glm(formula = clinically.important.brain.injury ~ ., family = binomial,
data = head.injury)
Deviance Residuals:
Min 1Q Median 3Q Max
-2.2774 -0.3511 -0.2095 -0.1489 3.0028
Coefficients:
Estimate Std. Error z value Pr(>|z|)
(Intercept) -4.4972 0.1629 -27.611 < 2e-16 ***
age.65 1.3734 0.1827 7.518 5.56e-14 ***
amnesia.before 0.6893 0.1725 3.996 6.45e-05 ***
basal.skull.fracture 1.9620 0.2064 9.504 < 2e-16 ***
GCS.decrease -0.2688 0.3680 -0.730 0.465152
GCS.13 1.0613 0.2820 3.764 0.000168 ***
GCS.15.2hours 1.9408 0.1663 11.669 < 2e-16 ***
high.risk 1.1115 0.1591 6.984 2.86e-12 ***
loss.of.consciousness 0.9554 0.1959 4.877 1.08e-06 ***
open.skull.fracture 0.6304 0.3151 2.001 0.045424 *
vomiting 1.2334 0.1961 6.290 3.17e-10 ***
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05
'.' 0.1 ' ' 1
(Dispersion parameter for binomial family taken to be 1)
Null deviance: 1741.6 on 3120 degrees of freedom
Residual deviance: 1201.3 on 3110 degrees of freedom
AIC: 1223.3
Number of Fisher Scoring iterations: 6
How do I assess which patients have a high risk level, and how does the risk
threshold play into that?
Thanks in advance,
Diana
---------------------------------
[[alternative HTML version deleted]]
