appropriate statistical method needed

[drfg2008]

For a Phase II drug trial an appropriate statistical method is sought.

The following Background:

In a research on the acidity of gastric juice during cardiopulmonary bypass surgery a drug has been used and tested for efficacy.

During the entire operation the acidity of gastric juice should remain in a non-hazardous area. Neither the pH value should be too high nor too low. Maximum and minimum limits are given.

The sample was relatively small. Patients were randomly assigned to two groups of 20 persons each group. One group did not receive the drug, the other group did. The 40 operations lasted different lengths. Every second a probe of the stomach pH was measured. There are therefore 40 time series of different lengths of pH values.

Question: It is to be examined whether the pH values could be stabilized due to the drug for the entire duration of the operation within the safe area. The effectiveness compared with the control group without drug is to be demonstrated.

Which procedure, which method would be in question here?

[Maguin, Eugene]

FUB,

Here is an initial idea. The allowable pH range is known. If the pH values
in a person's data series remain within the range, score=1; otherwise,
score=0. Test Tx proportion greater than control proportion by one sided
test of proportions. Is this adequate for a phase II drug trial. No idea but
I'd bet not.

I think you might be much better served by finding a list-serv that focuses
on medical statistics because this sort of question is quite odd, I think,
for social science people, which is who I think is mostly on this list. But,
perhaps, I'm wrong.

Gene Maguin


-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
drfg2008
Sent: Wednesday, February 23, 2011 4:04 PM
To: [hidden email]
Subject: appropriate statistical method needed

For a Phase II drug trial an appropriate statistical method is sought.

The following Background:

In a research on the acidity of gastric juice during cardiopulmonary bypass
surgery a drug has been used and tested for efficacy.

During the entire operation the acidity of gastric juice should remain in a
non-hazardous area. Neither the pH value should be too high nor too low.
Maximum and minimum limits are given.

The sample was relatively small. Patients were randomly assigned to two
groups of 20 persons each group. One group did not receive the drug, the
other group did. The 40 operations lasted different lengths. Every second a
probe of the stomach pH was measured. There are therefore 40 time series of
different lengths of pH values.

Question: It is to be examined whether the pH values could be stabilized due
to the drug for the entire duration of the operation within the safe area.
The effectiveness compared with the control group without drug is to be
demonstrated.

Which procedure, which method would be in question here?


-----
FUB

--
View this message in context:
http://spssx-discussion.1045642.n5.nabble.com/appropriate-statistical-method
-needed-tp3397707p3397707.html
Sent from the SPSSX Discussion mailing list archive at Nabble.com.

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[drfg2008]

you're probably right, thank you !

[Rich Ulrich]

First, look at the literature.  You will look bad if you ignore a great
solution that has been published.

Second, you need an expert to tell you what outcomes are worst,
among the sort of patterns that you observe.  Yes/No   for "always
in the range"  is about as simple and direct as you can get, but it
ignores a lot of variation.  And it is pretty useless if every patient
has some excursions outside the min-max range.

What would you say about "The worst 5 (5... 15... 30... 100... ?) 
seconds"  of data, for any patient? 

If you are expected to provide all the numerical reduction of the
data, you might find it useful to ask your expert(s)  to provide a
blind rating (no knowledge of which group)  to the outcomes series,
using three or more categories like  [hazardous/ so-so/ safe] . 
Then you use those ratings to discover numerical rules.

--
Rich Ulrich

Original post

> Date: Wed, 23 Feb 2011 13:03:44 -0800
> From: [hidden email]
> Subject: appropriate statistical method needed
> To: [hidden email]
>
> For a Phase II drug trial an appropriate statistical method is sought.
>
> The following Background:
>
> In a research on the acidity of gastric juice during cardiopulmonary bypass
> surgery a drug has been used and tested for efficacy.
>
> During the entire operation the acidity of gastric juice should remain in a
> non-hazardous area. Neither the pH value should be too high nor too low.
> Maximum and minimum limits are given.
>
> The sample was relatively small. Patients were randomly assigned to two
> groups of 20 persons each group. One group did not receive the drug, the
> other group did. The 40 operations lasted different lengths. Every second a
> probe of the stomach pH was measured. There are therefore 40 time series of
> different lengths of pH values.
>
> Question: It is to be examined whether the pH values could be stabilized due
> to the drug for the entire duration of the operation within the safe area.
> The effectiveness compared with the control group without drug is to be
> demonstrated.
>
> Which procedure, which method would be in question here?
>