d_labes
★★★

Berlin, Germany,
2008-12-15 16:19
(5582 d 15:04 ago)

Posting: # 2926
Views: 9,923
 

 AUC in Bear [🇷 for BE/BA]

Dear Rusers, dear Bear inventors, dear all,

after looking at the details of NCA analysis in Bear (see NCA_PK.txt) I wonder, if/why the time point zero is not included in the AUC(0-tlast) calculations.

Any reasons for that, if really not included?
Is this not an 'oral' application dataset?

Further question: What about missing values at t=0.25 h (f.i. subject 2, 4 ...) for the built-in dataset. Missing or <LLOQ?

Regards,

Detlew
yjlee168
★★★
avatar
Homepage
Kaohsiung, Taiwan,
2008-12-17 19:57
(5580 d 11:26 ago)

@ d_labes
Posting: # 2935
Views: 8,512
 

 AUC in Bear

dear SASophylic DLabes,

Yes, the built-in dataset (for demo purpose) was the 'oral' dataset. The time zero was absolutely included. The first piece AUC is not exactly a trapezoid, but a triangle. At the time zero with oral data, the drug concentration is supposed to be 'zero', unless the washout period is not long enough. If there is a missing value due to any reason, bear will treat it as 'zero' concentration. Very sorry about this delayed response.

❝ Any reasons for that, if really not included?

❝ Is this not an 'oral' application dataset?


❝ Further question: What about missing values at t=0.25 h (f.i. subject 2, 4

❝ ...) for the built-in dataset. Missing or <LLOQ?


All the best,
-- Yung-jin Lee
bear v2.9.1:- created by Hsin-ya Lee & Yung-jin Lee
Kaohsiung, Taiwan https://www.pkpd168.com/bear
Download link (updated) -> here
d_labes
★★★

Berlin, Germany,
2008-12-18 09:25
(5579 d 21:58 ago)

@ yjlee168
Posting: # 2937
Views: 8,477
 

 Degenerated trapezoid ;-)

Dear Yung-jin,

thanks for replying.

❝ [...] The time zero was absolutely included. The first piece AUC

❝ is not exactly a trapezoid, but a triangle. [...]


So it should be! But sorry I cannot recover this in your output on the Bear homepage.
Here a piece of your NCA_PK.txt
<< NCA Output:- Subject# 1  (Ref.)>>
--------------------------------------------------------------------------
   subj  time   conc   AUC(0-t)    AUMC(0-t)
1     1  0.25   36.1     0.0000     0.000000
2     1  0.50  125.0    20.1375     8.940625
3     1  0.75  567.0   106.6375    69.909375
4     1  1.00  932.0   294.0125   239.565625
[...]


Seems to me: No triangle from time zero to time 0.25 h :-P:
Or miss I some piece?

For that data I have calculated AUC(0-t)=14445.275.

Yours is AUC(0-t)=14440.76. Triangle (some sort of degenerate trapezoid ;-) ) from 0-0.25h=0.25*36.1/2=4.5125.
Sum both and you get mine.

BTW: The half-life calculation I was able to verify exactly using SAS code for the ARS method :ok:.

Regards,

Detlew
yjlee168
★★★
avatar
Homepage
Kaohsiung, Taiwan,
2008-12-18 09:44
(5579 d 21:39 ago)

@ d_labes
Posting: # 2939
Views: 8,429
 

 Degenerated trapezoid ;-)

dear DLabes,

Weird! I just run bear again and what I get is displayed as follows.
[...]                    Reference                     
---------------------------------------------------

<< NCA Output:- Subject# 1  (Ref.)>>
--------------------------------------------------------------------------
   subj  time   conc  AUC(0-t) AUMC(0-t)
1     1  0.00    0.0     0.000     0.000
2     1  0.25   36.1     4.513     1.128
3     1  0.50  125.0    24.650    10.069
4     1  0.75  567.0   111.150    71.037
5     1  1.00  932.0   298.525   240.694
6     1  1.50 1343.0   867.275   977.319
7     1  2.00 1739.0  1637.775  2350.444
8     1  3.00 1604.0  3309.275  6495.444
9     1  4.00 1460.0  4841.275 11821.444
10    1  8.00  797.0  9355.275 36253.444
11    1 12.00  383.0 11715.275 58197.444
12    1 24.00   72.0 14445.275 96141.444[...]


which is identical to what you got (14445.275]. Could you please update all your R packages (from R console, plz) again? This is because there seems a big change from v2.7.2 to v2.8.0. My desktop PC system is Win/Vista sp1,R v2.8.0 and bear v2.0.1. If you update all packages of R from its console, it will also automatically update all packages that will be called by bear. Sorry about this.

❝ [...]

❝ For that data I have calculated AUC(0-t)=14445.275.

❝ [...]


All the best,
-- Yung-jin Lee
bear v2.9.1:- created by Hsin-ya Lee & Yung-jin Lee
Kaohsiung, Taiwan https://www.pkpd168.com/bear
Download link (updated) -> here
d_labes
★★★

Berlin, Germany,
2008-12-18 10:02
(5579 d 21:21 ago)

@ yjlee168
Posting: # 2940
Views: 8,526
 

 Bear at Home

Dear Yung-jin,

I have no R or Bear installed. Because I have all my working days to struggle with the beast SAS.
I refer to your Bear homepage! I think it would be a good idea to update it.

Regards,

Detlew
yjlee168
★★★
avatar
Homepage
Kaohsiung, Taiwan,
2008-12-18 10:47
(5579 d 20:36 ago)

@ d_labes
Posting: # 2941
Views: 8,505
 

 Bear at Home

dear DLabes,

Thank you so much. You're right about this. I didn't keep the website updated. But, I have already updated bear website. It's the output from previous version (v1.x.x?).

All the best,
-- Yung-jin Lee
bear v2.9.1:- created by Hsin-ya Lee & Yung-jin Lee
Kaohsiung, Taiwan https://www.pkpd168.com/bear
Download link (updated) -> here
d_labes
★★★

Berlin, Germany,
2008-12-18 12:46
(5579 d 18:37 ago)

@ yjlee168
Posting: # 2944
Views: 8,429
 

 Bear - statistical summary

Dear Bear inventors,

I would suggest a change in your statistical summary output.
Here a piece:
         Statistical Summaries for Bioequivalence Study (N= 14 )
--------------------------------------------------------------------------

    parameters Test_Mean  Test_SD  Ref_Mean   Ref_SD Ratio
1         Cmax  1630.786  197.522  1539.643  220.015 1.079
2       AUC0-t 13975.284 1997.852 13158.790 1971.053 1.096
3     AUC0-inf 14659.822 2092.321 13776.930 2033.362 1.097
4     ln(Cmax)     7.390    0.122     7.329    0.147 1.009
5   ln(AUC0-t)     9.535    0.145     9.473    0.161 1.007
6 ln(AUC0-inf)     9.583    0.144     9.520    0.159 1.007


The ratio of meanTest/meanRef. in case of log-transformed values does not make sense.
It is misleading as it may be interpreted as point estimator of the bioavailability ratio by some 'naive' users. This is especially true because the following part, the 90% confidence intervals, does nor give any point estimator.

My suggestion therefore is:

1. Do not give the ratio in case of log-transformed values. Include the point estimator of bioavailability ratio as exp(LSMeanT-LSMeanR) in the part with the 90% CIs.

or

2. Give the ratio in case of log-transformed values as exp(meanT-meanR) as an approximation to the biovailability ratio. Include the point estimator of bioavailability ratio as exp(LSMeanT-LSMeanR) in the part with the 90% CIs.
Both estimates are identical only in case of balanced studies.
But giving both may confuse the 'naive' users or even :-D regulators.

Dear Yung-jin,
at all great job. :clap:

Regards,

Detlew
yjlee168
★★★
avatar
Homepage
Kaohsiung, Taiwan,
2008-12-18 13:13
(5579 d 18:10 ago)

@ d_labes
Posting: # 2945
Views: 8,456
 

 Bear - statistical summary

dear DLabes,

Indeed. you're absolutely right. We will put your suggestions in the list for next release. Actually, we should do this back to a while ago when we discussed which 90%CI should be implemented in Bear. Thanks again. :ok:

All the best,
-- Yung-jin Lee
bear v2.9.1:- created by Hsin-ya Lee & Yung-jin Lee
Kaohsiung, Taiwan https://www.pkpd168.com/bear
Download link (updated) -> here
UA Flag
Activity
 Admin contact
22,957 posts in 4,819 threads, 1,636 registered users;
105 visitors (0 registered, 105 guests [including 4 identified bots]).
Forum time: 07:23 CET (Europe/Vienna)

With four parameters I can fit an elephant,
and with five I can make him wiggle his trunk.    John von Neumann

The Bioequivalence and Bioavailability Forum is hosted by
BEBAC Ing. Helmut Schütz
HTML5