Bioequivalence and Bioavailability Forum

Dear Detlew,

❝ Kinetica? I beg you pardon ...

As long as you don’t want to perform the BE-stats it is OK (& if you can stand the strange user-inter­face). Competitive price.

❝ PK-Xpress seems (beside the spare information given) an "Eierlegende-Wollmilch-Sau".

Leo.

❝ Anybody out there with experiences with that piece of software?

No hands-on experience but last July I endured a two-hours WebEx presentation with the folk in Pune. Observations:

• Open port 80 (HTTP) during installation – pointing to IP 180.87.43.162; “enable the download of execut­able (.exe) file from web”. That’s a no-go for me.
  
• Eventually (!) there is a stand-alone installer. They were unsure about it…
• Requires Office 2007+ (Excel’s Solver Add-On). Didn’t manage to find out why. Back luck for people using OpenOffice, LibreOffice, or Office 2000 like myself. IMHO Solver is an undocu­mented piece of shit.
  
• Automatic algo for λ_z (based on max R²_adj) may include t_max (like in WinNonlin <v6). However, manual selection of time range possible.
  
• Only linear trapezoidal method implemented.
  
• Extrapolation by C_last/λ_z. Not possible to use Ĉ_last/λ_z.
  
• No partial AUCs.
  
• BE-stats. Not for a SASian like you, but might be interesting to others:
  • 2×2×2 cross-overs and two-group parallel. Full stop. No higher order crossovers, no replicate designs.
    
  • In crossovers subjects as a random effect (FDA). They were unaware of EMA’s all-fixed effects require­­ment.
    
  • Given the above, no reference-scaling; working on replicate designs. They were sur­prised to hear [sic] that the FDA and EMA require different models.

At the bottom of the webpage you will find

Validation Kit

Don’t know what that means. I’ve sent them our reference datasets…

On 31 July I received an e-mail:

As a part of data refining process, we are working on refining of BE program. As soon as this pro­cess is done we will work on the data provided by you and share the desired outputs.

Haven’t heart anything ever since.

❝ Expense?

They didn’t want to talk about it. Luv told me:

3,000 USD for each of DissoXpress, IvivcXpress, PK-Xpress. If you are going for whole package the cost would be even lesser. Yearly support charges: 20% of license cost per client.

Much cheaper than Phoenix and slightly below Kinetica.

Overall I would say it is an interesting product – which calls for further improvements. For NCA, not that bad. For BE-stats, no way.