PharmPK Discussion - Bioanalysis: Integration of chromatographic peaks

• On 6 Apr 2011 at 11:49:11, "Jeanbaptiste, Bernard (Wavre)" (Bernard.Jeanbaptiste.aaa.sgs.com) sent the message
  

  
  Dear,
   It is a good practice that integration must be done automatically:  manual integration must be avoided.
  We are currently busy reviewing our sop.  Actually, we authorize manual  integration for a maximum of 4% of the chromatograms contained in a run.
  I would like to have some feedback from you, on how you proceed: Is 4%  manual integration is too high?  Should we allow manual integration for  incurred samples only?  Is the reintegration of calibrators and QC's  must be absolutely avoided?  Thanks for your advices.
   Bernard Jeanbaptiste
  
  

  Back to the Top

• On 7 Apr 2011 at 18:06:25, Helmut Schuetz (helmut.schuetz.aaa.bebac.at) sent the message
  

  
  The following message was posted to: PharmPK
  
  Dear Jeanbaptiste!
  
  > It is a good practice that integration must be done automatically:
  > manual integration must be avoided.
  >
  
  I wouldn't call that good practice, but rather laziness (excuse me).
  Find integrations parameters which allow automatic integration of most
  samples in a batch and reintegrate samples where the automatic method
  failed. Problems are common at low concentrations, where - caused by
  random fluctuations - either positive noise mimics a shoulder and
  triggers the end-of-peak signal too early or negative noise prevents
  end-of-peak detection.
  
  > We are currently busy reviewing our sop.  Actually, we authorize manual
  > integration for a maximum of 4% of the chromatograms contained in a run.
  >
  
  Where does this 4% come from?
  
  > I would like to have some feedback from you, on how you proceed: Is 4%
  > manual integration is too high?
  
  I think that any arbitrary percentage is unjustified.
  
  > Should we allow manual integration for incurred samples only?
  >
  
  No, that's bad. Smells like cherry-picking.
  
  > Is the reintegration of calibrators and QC's
  > must be absolutely avoided?
  >
  
  Why? Calibrators and QCs should be treated exactly as unknown samples.
  
  See these threads (including links and some references):
  http://forum.bebac.at/mix_entry.php?id=5698
  http://forum.bebac.at/mix_entry.php?id=6154
  
  All the best,
  Helmut
  
  -
  Ing. Helmut Schuetz
  BEBAC - Consultancy Services for
  Bioequivalence and Bioavailability Studies
  Neubaugasse 36/11
  1070 Vienna, Austria
  
  

  Back to the Top

• On 8 Apr 2011 at 11:48:51, Frank Bales (frankbales.-a-.msn.com) sent the message
  

  
  Dear Jeanbaptiste,
   The bracketing standards and QC's should be intergrated the same as the  samples they bracket. This can be done by adjusting the automatic  integration parameters for that bracket. This is still considered  automatic integration.
   Manual integration by for example, using a "Drop Line" to attempt to  separate partially merged peaks, will not result in accurate  quatitation.  Regards,  Frank Bales, Ph.D. mailto: frankbales.-at-.msn.com
  
  

  Back to the Top

• On 8 Apr 2011 at 13:03:23, "Edward O'Connor" (efoconnor.aaa.gmail.com) sent the message
  

  
  The following message was posted to: PharmPK
  
  It is only considerd automatic if it is applied to all samples
  
  

  Back to the Top

Copyright 1995-2011 David W. A. Bourne (david@boomer.org)