I liked Gonneke's comment that suggested common sense will give us the answer for 
some things without needing a full scale study.

Every now and again I am struck by the question of "where did that procedure come 
from".

So, here's my questions on both common sense and on evidence grounds:

Common sense question:  If you have a more direct indicator that is readily at your 
disposal (e.g. test weighing a baby before and after a feeding) compared to an indirect 
untested indicator (e.g. test weighing a diaper), why would you ever choose the indirect 
indicator over the direct indicator?

Evidence based question:  Is there any evidence that shows that test weighing a diaper 
has adequate sensitivity and specificity to be useful as an indicator of intake? If number 
of diapers does not have good sensitivity and specificity, I don't see how tiny little 
increments in mass are going to be any better.  I'd love to see some studies on this to 
see what the ROC curves are like.

Then, digging further into this issue, I see two epidemiologically based problems with test 
weighing diapers:

1) Measurement error:  when you measure a lighter object that is closer to the limits of 
accuracy --- the percentage of measurement error compared to the weight of the object 
increases.  This means that if you weigh a baby which weighs more than a diaper --- you 
are likely to get a lower proportion of measurement error.  

2) Confounding factors:  when you are measuring output as a proxy indicator for input, 
there are many factors that could confound the situation.  For instance, IV fluids or kidney 
function.  

So, it seems to me that weighing a diaper might have been useful in the situation when I 
worked with a baby that was ultimately diagnosed as having a metabolic disorder who 
was drinking plenty (3-4 ounces 8-10x/day) but rapidly losing weight.  But I really can't 
quite figure out why you would choose the indirect over the direct measurement of intake 
if intake is what you want to measure.

Best, Susan

             ***********************************************

Archives: http://community.lsoft.com/archives/LACTNET.html
To reach list owners: [log in to unmask]
Mail all list management commands to: [log in to unmask]
COMMANDS:
1. To temporarily stop your subscription write in the body of an email: set lactnet nomail
2. To start it again: set lactnet mail
3. To unsubscribe: unsubscribe lactnet
4. To get a comprehensive list of rules and directions: get lactnet welcome