Sample Size "Efficiency" of Replicate CO vs. 2-way CO for NTID [Power / Sample Size]
Hi Sereng,
I completely understand your question, but I think it is slightly the wrong question to ask.
If you look at it from the regulator's side, they have had some pretty nasty experiences with Levo through the years and it may have gone so pear-shaped that regulators as a precautionary measure had to use some degree of belt-and-suspenders thinking and where the focus was on de-risking anything related to the patient. Hence additional requirements, like some degree of similarity of T and R variabilities.
We started out with the good old work horse, the 222BE design with plain average BE. Great for many things.
But "hey, poor me", said the Sponsor, "now I need 326 subjects in my trial for Schützoycin tablets. I demand an easier approach!".
Then regulators meditated 10 years and held conference all the while and at some point it looked like average BE was being scrapped and fancy things like Pop BE and Individual BE were winning the crowd. But when the smoke had cleared average BE was still standing, this time in combination with replicate designs. PopBE and Individual BE were soon on their way out. Widened limits in different flavours saw daylight. Sample sizes for HVDPs got normal (and some people began bashing others on their heads with guidance collections, pharmacopeias and wooden clubs for confusing HVDPs and HVDs, but that's a story for another day).
Then, fueled by a grieving innovator industry, who realised that revenue was lost on the HVDPs, attention was suddenly drawn to NTIs. Claims of doom and gloom lurking around every corner for generics of those products. FDA dismissed the matter: "If we can widen the limits for HVDP's then we can just put the selector in reverse and step on the pedal for NTIs". The innovators lobbyed hard. "Dose dumping this, levothyroxine that, internal bleeding here, Hancock-Mortimer syndrome there, look what happened in Europe, and do you really want to murder the average epileptic patient when she is switched from innovator to generic just because the guidance overlooks an obvious problem?"
And so FDA for fear of lawsuits and further pressure from patient organisation went: "OK, erm... right... we'll slap on some additional requirements to make even more sure that T and R behave the same when we talk NTIs." Kind of disappointing to the generic industry (and possibly also select members of the legal profession).
And there you have it. I wonder what will come next.
❝ What are the sample size “benefits” or “efficiencies” of undertaking this study as a replicate CO design (e.g., Test 50 x 2; Ref 50 x 2) versus a 2-way CO (Test 100 x 2; Ref 100 x 2). Any help would be much appreciated.
I completely understand your question, but I think it is slightly the wrong question to ask.
If you look at it from the regulator's side, they have had some pretty nasty experiences with Levo through the years and it may have gone so pear-shaped that regulators as a precautionary measure had to use some degree of belt-and-suspenders thinking and where the focus was on de-risking anything related to the patient. Hence additional requirements, like some degree of similarity of T and R variabilities.
We started out with the good old work horse, the 222BE design with plain average BE. Great for many things.
But "hey, poor me", said the Sponsor, "now I need 326 subjects in my trial for Schützoycin tablets. I demand an easier approach!".
Then regulators meditated 10 years and held conference all the while and at some point it looked like average BE was being scrapped and fancy things like Pop BE and Individual BE were winning the crowd. But when the smoke had cleared average BE was still standing, this time in combination with replicate designs. PopBE and Individual BE were soon on their way out. Widened limits in different flavours saw daylight. Sample sizes for HVDPs got normal (and some people began bashing others on their heads with guidance collections, pharmacopeias and wooden clubs for confusing HVDPs and HVDs, but that's a story for another day).
Then, fueled by a grieving innovator industry, who realised that revenue was lost on the HVDPs, attention was suddenly drawn to NTIs. Claims of doom and gloom lurking around every corner for generics of those products. FDA dismissed the matter: "If we can widen the limits for HVDP's then we can just put the selector in reverse and step on the pedal for NTIs". The innovators lobbyed hard. "Dose dumping this, levothyroxine that, internal bleeding here, Hancock-Mortimer syndrome there, look what happened in Europe, and do you really want to murder the average epileptic patient when she is switched from innovator to generic just because the guidance overlooks an obvious problem?"
And so FDA for fear of lawsuits and further pressure from patient organisation went: "OK, erm... right... we'll slap on some additional requirements to make even more sure that T and R behave the same when we talk NTIs." Kind of disappointing to the generic industry (and possibly also select members of the legal profession).
And there you have it. I wonder what will come next.
—
Pass or fail!
ElMaestro
Pass or fail!
ElMaestro
Complete thread:
- Sample Size "Efficiency" of Replicate CO vs. 2-way CO for NTID Sereng 2022-01-30 23:11 [Power / Sample Size]
- Sample Size "Efficiency" of Replicate CO vs. 2-way CO for NTIDElMaestro 2022-01-30 23:50
- OT d_labes 2022-01-31 18:48
- Follow the guidance (if possible) Helmut 2022-01-30 23:52
- Follow the guidance (if possible) Sereng 2022-01-31 21:58
- Now I got it! Helmut 2022-01-31 23:53
- Follow the guidance (if possible) Sereng 2022-01-31 21:58
- Sample Size "Efficiency" of Replicate CO vs. 2-way CO for NTIDElMaestro 2022-01-30 23:50