Question about Logistic Regression
1. A random sample of 200 patients admitted to an adult intensive care unit (ICU) was collected to examine factors associated with death during hospital stay for ICU patients. Data was also collected on patient's age (in years), race, whether the patient had an infection at the time of ICU admission, and whether the patient had CPR administered prior to the hospital admission. Of specific interest is whether or not infection at the time of admission is associated with increased probability of death during hospital stay. Logistic regression was employed to help answer the substantive question. Below find the estimated coefficients for infection status at time of admission from 4 different logistic regression models all relating the probability of death in the ICU to patient characteristics.
e. Which of the 3 adjusted odds ratios computed in part (d) are "statistically significant"?
f. How do the resulting adjusted odds ratio estimates and confidence intervals from part (d) compare across the 3 sets of logistic regression models with additional predictors?
g. How could you use these results to assess whether the relationship between death in ICU patients and infection at the time of admission is confounded by other patient characteristics?
h. What type of study design is this? Would it be possible to use the results from the 4th logistic regression model (with infection status, age, CPR, and race as predictors) estimate the probability (risk) of death for various groups of patients based on the reported patient characteristics?
This question has the following supporting file(s):
- question.doc
File Viewer (Click To Zoom)
$2.19 
This answer includes:
- Plain text
- Cited sources when necessary
- Attached file(s)
- solution.doc
Active since 2006
Responses 944 | eBooks 2
Extracted Content from Question Files:
- question.doc
1. A random sample of 200 patients admitted to an adult intensive care unit (ICU) was
collected to examine factors associated with death during hospital stay for ICU
patients. Data was also collected on patient’s age (in years), race, whether the patient
had an infection at the time of ICU admission, and whether the patient had CPR
administered prior to the hospital admission. Of specific interest is whether or not
infection at the time of admission is associated with increased probability of death
during hospital stay. Logistic regression was employed to help answer the substantive
question. Below find the estimated coefficients for infection status at time of
admission from 4 different logistic regression models all relating the probability of
death in the ICU to patient characteristics.
Predictors (Covariates) in Regression Coefficient Standard Error of
for Infection Status1
Model Regression Coefficient
Infection Status 0.92 0.36
Infection Status, Patient 0.80 0.37
Age (4 categories)
Infection Status, Patient 0.70 0.38
Age (4 categories), CPR
Infection Status, Patient 0.66 0.38
Age (4 categories), CPR,
Patient Race
1
Infection Status is a binary indicator taking on a value of 1 if
the patient had an infection at time of the admission, 0 if not
(for questions a-d, please see the posting titled “1-part 1”.)
e. Which of the 3 adjusted odds ratios computed in part (d) are “statistically
significant”?
f. How do the resulting adjusted odds ratio estimates and confidence intervals
from part (d) compare across the 3 sets of logistic regression models with
additional predictors?
g. How could you use these results to assess whether the relationship between
death in ICU patients and infection at the time of admission is confounded by
other patient characteristics?
h. What type of study design is this? Would it be possible to use the results from
the 4th logistic regression model (with infection status, age, CPR, and race as
predictors) estimate the probability (risk) of death for various groups of
patients based on the reported patient characteristics?
"Should Bold Italic each heading??"
"In your opinion, how long should the summary be?"
"Ok, here goes the re-write as per your suggestion: OLD:This is important to know and will require further research because the same occurrences in the Carver study does have a correlation between those who must have an HR degree in contrast who those with non-HR degree versus those with a professional certification in HR. The new re-write ::This is important to know and will require further research because the same occurrences in the Carver study does have a correlation between those who must have an HR degree to practice than for those with a non-HR degree who can still make a impact.:: You thoughts??"
"Hi Elisabeth, Thank you so much for your help! I was looking for help though with the manual calculations of the part below though so I would know how to do this problem, and cannot turn in my assignment with an online calculator used for the part below. Could you possibly answer the question below without the online calculator? We have to show our work. Thank you! A calculator to automatically solve these probabilities is here: http://www.vassarstats.net/textbook/ch5apx.html Fill in the table on that website with N = 4 and p = 0.5. Change the value of k to e 0, 1, 2, 3, and 4 to find the theoretical probabilities for each possible number of heads. They are: a. Zero Heads = 0.0625 = 6.25% b. One Head = 0.25 = 25% c. Two Heads = 0.375 = 37.5% d. Three Heads = 0.25 = 25% e. Four Heads = 0.0625 = 6.25%"
"Hi Nicole, I have chapter 2 (roughly about 20 pages) for editing, i.e., punctuation etc for your review. Are you still editing? Please advise."