Hypothesis: Early enteral nutrition (tube feeding) is advantageous in coronary intensive care units.

Specifically, I hope to find that

Early enteral nutrition within 72 hours of admission on mechanically vented CICU patients admitted with myocardial infarctions, cardiac arrest, endocardidtis, or congestive heart failure will have:
1. Decreased length of stay
2. Decreased days on vent
3. Increased albumin levels
4. Increased pre-albumin levels (if tested)
The goal of the study would be to promote a policy change for the unit based on results, to insert a corpak or small-bore naso-gastric feeding tube within 72 hours of intubation.

I was thinking I can look back on old medical records of patients from CICU (after IRB approval of course)who meet the above criteria and were feed after 72 hours and compare the variables to those who were fed before 72 hours (since this is rare, I will be encouraging physicians to begin tube feeding within 72 hours, so I will have patients to compare against.

 
This is a classic comparison kind of study.
Few of us in the class have experience with the ailments you describe, so your review of literature will be up to you to show that your project has merit.  Your literature review will have to show that early feeding, and specifically tube vs. iv feeding has a benefit on the variables you are looking at.  In the event there are few studies on those variables, you will want to extend your lit review.
 
I always suggest starting with the broad variables of interest, and moving from the general literature findings to the more specific ones.
 
So you’ll start with the two concepts – timing of feeding after admission, and then try to show how tube feeding has benefits beyond iv feeding.  So two things to consider there.
Your literature review would begin with an introduction to the topic under study then launch right into the general discussion of past findings about timing of feeding after admission.
 
Then the tube vs. iv part.
 
Then you’d need to go into each of the variables you mentioned as these are related to feeding.
1. Decreased length of stay
2. Decreased days on vent
3. Increased albumin levels
4. Increased pre-albumin levels (if tested)

A good section of literature review on each one, with the feeding by tube or iv at the end of each section.
 
A solid literature review will set up your hypotheses.  I’d suggest a hypothesis for each of the 4 variables. Although this appears tedious and wordy, your job is to be clear at each step of the research process, so that everyone who reads your research understands the logic of it. Your hypotheses should fall logically out of your sections of the literature review  like this (if I have this tube vs iv thing wrong, then just disregard the 1st hypothesis in each set):
 
H1a: Among mechanically vented CICU patients admitted with myocardial infarctions, cardiac arrest, endocardidtis, or congestive heart failure, those who have early nutrition (nutrition within 72 hours of admission) will have decreased length of stay (in ICU or Hospital time), when compared to those who do not receive early nutrition.
H1b: Among mechanically vented CICU patients admitted with myocardial infarctions, cardiac arrest, endocardidtis, or congestive heart failure, those who have early enteral nutrition (nutrition via tubefeeding within 72 hours of admission) will have decreased length of stay (in ICU or Hospital time), when compared to those who do not receive early entereal nutrition.
 
H2a: Among mechanically vented CICU patients admitted with myocardial infarctions, cardiac arrest, endocardidtis, or congestive heart failure, those who have early nutrition (nutrition within 72 hours of admission) will have Decreased days on vent (in ICU or Hospital time), when compared to those who do not receive early nutrition.
H2b: Among mechanically vented CICU patients admitted with myocardial infarctions, cardiac arrest, endocardidtis, or congestive heart failure, those who have early enteral nutrition (nutrition via tubefeeding within 72 hours of admission) will have Decreased days on vent (in ICU or Hospital time), when compared to those who do not receive early entereal nutrition.
 
H3a: Among mechanically vented CICU patients admitted with myocardial infarctions, cardiac arrest, endocardidtis, or congestive heart failure, those who have early nutrition (nutrition within 72 hours of admission) will have Increased albumin levels  (in ICU or Hospital time), when compared to those who do not receive early nutrition.
H3b: Among mechanically vented CICU patients admitted with myocardial infarctions, cardiac arrest, endocardidtis, or congestive heart failure, those who have early enteral nutrition (nutrition via tubefeeding within 72 hours of admission) will have Increased albumin levels (in ICU or Hospital time), when compared to those who do not receive early entereal nutrition.
 
H4a: Among mechanically vented CICU patients admitted with myocardial infarctions, cardiac arrest, endocardidtis, or congestive heart failure, those who have early nutrition (nutrition within 72 hours of admission) will have Increased pre-albumin levels (in ICU or Hospital time), when compared to those who do not receive early nutrition.
H4b: Among mechanically vented CICU patients admitted with myocardial infarctions, cardiac arrest, endocardidtis, or congestive heart failure, those who have early enteral nutrition (nutrition via tubefeeding within 72 hours of admission) will have Increased pre-albumin levels (in ICU or Hospital time), when compared to those who do not receive early entereal nutrition.
 
IRB will have plenty to say to you about privacy issues. I’ll leave this to them and you (and your thesis committee).
 
Sampling will come from existing records at the hospital for “mechanically vented CICU patients admitted with myocardial infarctions, cardiac arrest, endocardidtis, or congestive heart failure.” 
 
I assume you will have access to the hospital records, and can first select the records of all these people. 
 
Now records will have all kinds of information in them, but the key information you want, after defining your sample as all these folks over the past 5 years??? will include:
   
There could be other variables as well, such as Age, Gender, etc.
This data should be transferred to a spreadsheet with rows indicating cases and columns representing variables, like this
 
Patient type of   Timing of         Days    Days    Days    Albumin            Albumin
   ID                 Illness      Feeding         ICU     Hosp    Vent     at admit            at discharge
001                  1                      1          2          4          2          .034                 .001
 
Each variable would have to be defined:                   Here you’d subtract admit from discharge – should be a negative number Statistical tests - Right off you see you have two sets of t-tests here:
 
The 1st test will include all patients in the sample regardless of the type of feeding.
First Test          Mean of           Mean of           Mean of           Mean of
Groups             ICU Days         Hosp Days       Vent Days        Albumin Difference      
Within 72 hrs               
After 72 hrs     
This test will address the first of each set of hypotheses. And will establish whether or not early nutrition will have the hypothesized effects.
 
The 2nd t-tests will address differences between iv and tube
IV Patients ONLY
2nd Test           Mean of           Mean of           Mean of           Mean of
Groups             ICU Days         Hosp Days       Vent Days        Albumin Difference      
Within 72 hrs               
After 72 hrs     
 
TUBE Patients ONLY
2nd Test           Mean of           Mean of           Mean of           Mean of
Groups             ICU Days         Hosp Days       Vent Days        Albumin Difference      
Within 72 hrs               
After 72 hrs     
These tests, when compared will show whether or not there is a difference on dependent variables because of the type of feeding
 
Finally:
you could also test only the patients who received early nutrition by type of feeding.
3rd Test           Mean of           Mean of           Mean of           Mean of
Groups             ICU Days         Hosp Days       Vent Days        Albumin Difference      
IV
TUBE
Everyone here would have received early nutrition. What you are looking for here is improved performance because of type of feeding.
 
To be able to write the findings, you’d need to find studies where comparative t-tests were used.
Tables always work best for me.

I tend to compose tables of findings and work on writing the findings from those tables.
 
Here’s some examples of tabled findings:
 
Table 1 - Characteristics of the Sample
 
Variables                                Mean                          STD
ICU Days                                 9.145                           4.893
Hosp Day                                 2.709                           1.197
Vent Days                                2.291                           1.197
Albumin Difference                   2.927                           1.608
 
 
Table 14 - t-Test for Gender Differences on Attitudes Toward GMOs 
t-test for effects of early feediing   Mean Std. Dev t df Sig. (2-tailed)
ICU Days Within 72
  After 72
2.70
3.13
1.23
1.13
-2.370 166 .019
Hosp Days Within 72
  After 72
2.56
2.79
1.25
1.08
-1.294 166 .025
Vent Days Within 72
  After 72
2.66
3.13
1.29
1.19  
-2.444 166 .016
Albumin Difference Within 72
After 72
-.034
-.044
.05
.09
-2.56 166 .035
Significant at p<.05
 
You’d then construct t-tables for the remaining tests just like this one.
From here you can look to see the effects and note the ones that survived scrutiny. 
Then in your Discussion/ recommendations section you can discuss the relative benefits of early nutrition and tube over other forms of nutrition.
-ddw