dennis001
☆

Hong Kong,
2022-07-24 17:39
(21 d 16:00 ago)

(edited by dennis001 on 2022-07-24 17:52)
Posting: # 23163
Views: 447

Could we design the RCT as a non-inferiority trial at this situation? [Design Issues]

Dear all,

We plan to conduct a randomized controlled trial.
The goal of an RCT is to compare the difference of complication rates of CVD surgery between two groups (group A and group B).
The patients in group A will take medicine (traditionally treatment) and the patients in group B without taking medicine.

Notably, we want to design the RCT as a non-inferiority trial. That is, the effect of complication rates of patients without taking medicine in group B is not worse than the patients in group A with take medicine.

So, could we design the RCT as a non-inferiority trial?

Moreover, how to defining the non-inferiority margin at the situation?

Thanks a lot!
Helmut
★★★

Vienna, Austria,
2022-07-25 16:37
(20 d 17:01 ago)

@ dennis001
Posting: # 23176
Views: 383

Non-Superiority

Hi Dennis,

» The goal of an RCT is to compare the difference of complication rates of CVD surgery between two groups (group A and group B).
» The patients in group A will take medicine (traditionally treatment) and the patients in group B without taking medicine.
»
» Notably, we want to design the RCT as a non-inferiority trial. That is, the effect of complication rates of patients without taking medicine in group B is not worse than the patients in group A with take medicine.

Hhm.

» So, could we design the RCT as a non-inferiority trial?

The other way ’round.

Higher response is considered ‘better’ → Non-Inferiority
Lower response is considered ‘better’  → Non-Superiority

You hope that the complication rates in group A (with medicine) are lower than in group B (without medicine), right? Hence, that’s a Non-Superiority trial.

Note: x-axis in log-scale. $$\delta$$ made up out of thin air.

» Moreover, how to defining the non-inferiority margin at the situation?

You don’t need only to define the non-superiority margin $$\small{\delta}$$ but also assume the variability. No idea.
You are dealing with complication rates, which are $$\small{0\leq x \leq 1}$$. As an aside, you can’t use the function sampleN.noninf() of the -package PowerTOST, because it handles only lognormal $$(\small{x\in\mathbb{R^{+}}}$$) or normal $$(\small{-\infty\leq x+\infty}$$) distributed data. I’m sure, PASS has a fancy method for that.

Dif-tor heh smusma 🖖
Helmut Schütz

The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes
dennis001
☆

Hong Kong,
2022-07-26 02:32
(20 d 07:07 ago)

@ Helmut
Posting: # 23177
Views: 361

Non-Superiority

Dear Helmut
Thanks a lot!

»     Higher response is considered ‘better’ → Non-Inferiority
»     Lower response is considered ‘better’  → Non-Superiority

You really hit the nail on the head.
Beacause we will compare the difference of complication rates of CVD surgery between two groups (group A and group B), lower response is considered ‘better’.

» You hope that the complication rates in group A (with medicine) are lower than in group B (without medicine), right?

However, we try to show the complication rates in group B (without medicine) are lower (or not worse or non-inferiority) than in group A (with medicine), duo to the medicine is very expensive.

And thank you very much for providing the paper and explaing the role of PowerTOST and PASS when coduct the analysis in our study. We are ignoring the key point.

Is there any other probelm for the design?

Best regards
Dennis
Helmut
★★★

Vienna, Austria,
2022-07-26 15:04
(19 d 18:35 ago)

@ dennis001
Posting: # 23179
Views: 318

Non-Superiority

Hi Dennis,

» » You hope that the complication rates in group A (with medicine) are lower than in group B (without medicine), right?
»
» However, we try to show the complication rates in group B (without medicine) are lower (or not worse or non-inferiority) than in group A (with medicine), …

Essentially you are trying to show that the medicine doesn’t perform better than placebo (assuming the study is double-blind).

» … duo to the medicine is very expensive.

Likely you will face problems with the IEC/IRB:
• The – traditional* – medicine is approved (I guess). The fact that it is expensive is not relevant.
• While it would be acceptable to administer yet another treatment (hoping that it performs at least as good as the standard whilst still being cheaper), administering just nothing is unacceptable.
It is the job of the IEC/IRB to make a benefit/risk-assessment for the subjects in the study. To assess a potential socio-econommic advantage for the health system is not in the job description of the members.

» Is there any other probelm for the design?

IMHO, ethics.

• Can you elaborate a bit – without naming the good stuff – what ‘traditional’ means in this context? Any studies demonstrating its efficacy & safety? Or ‘traditional’ like ‘It’s used for ages and obviously [sic] it works’.

Dif-tor heh smusma 🖖
Helmut Schütz

The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes
dennis001
☆

Hong Kong,
2022-07-27 05:18
(19 d 04:21 ago)

@ Helmut
Posting: # 23180
Views: 281

Non-Superiority

Dear Helmut

Thanks!

» Essentially you are trying to show that the medicine doesn’t perform better than placebo (assuming the study is double-blind).

But I have to confess that I'm a little bit confused.
Dose the effect of placebo could be consided as the effect of patinets without medicine in the situation of our study?
that is, the effect of patinets with placebo equal to the effect of patinets without medicine in our study?

» IMHO, ethics.

It's very important for our study.

» Can you elaborate a bit – without naming the good stuff – what ‘traditional’ means in this context? Any studies demonstrating its efficacy & safety? Or ‘traditional’ like ‘It’s used for ages and obviously [sic] it works’.

Yes.
The medicine is a common prescription medications.
Helmut
★★★

Vienna, Austria,
2022-07-27 12:17
(18 d 21:22 ago)

@ dennis001
Posting: # 23184
Views: 271

Placebo effect

Hi Dennis,

» » Essentially you are trying to show that the medicine doesn’t perform better than placebo (assuming the study is double-blind).
»
» But I have to confess that I'm a little bit confused.
» Dose the effect of placebo could be consided as the effect of patinets without medicine in the situation of our study?
» that is, the effect of patinets with placebo equal to the effect of patinets without medicine in our study?

Oh dear! Depending on the drug, the disease, and the setting (administration at home or by an affectionate and trustworthy ‘God’ with horn-rimmed glasses wearing a white coat with a stethoscope around the neck), the placebo effect amounts up to 30%.1,2 Studies to assess the placebo effect are difficult.3

When we assess the effect of a drug, it’s always against the placebo effect and not against ‘nothing’.

» » Can you elaborate a bit – without naming the good stuff – what ‘traditional’ means in this context? Any studies demonstrating its efficacy & safety? Or ‘traditional’ like ‘It’s used for ages and obviously [sic] it works’.
»
» The medicine is a common prescription medications.

OK, that helps.

1. Seemingly there is an ‘optimal’ size of a placebo in adults. If the pill is relatively small → little ‘content’, low effect. If it is larger → difficult to swallow, must be great stuff, higher effect. If is really large → distrust, low effect.

2. Once researchers were interested whether the color (‼) of a sham ‘sleeping pill’ influenced the placebo effect in European patients. It turned out the blue one ‘worked’ best in almost all countries (blue associated with calmness, dusk, …).
There was one exception – Italy. ⚽ Football/soccer is a kind of religion in Italy. The national team is called the Squadra Azzurra (the azure team). Hence the associations were awareness, speed, sta­mina, you name it. No way to fall asleep.

Obviously Pfizer didn’t know this study. Why is ➡️ this goody blue in all countries? Cause it was originally approved for males (boys = blue, girls = pink)? Should consider a modification.

3. One option is to compare a treatment vs placebo in two studies. One is double-blind and the other open-label. There are even ones where the sub­jects are told:
• This is a placebo. It’s just a pill made from starch and sugar. It does not contain any active ingredient at all. You shouldn’t expect to experience an effect.
Then compare the placebo-arms of both studies. Power of such a comparison is poor. Com­monly a meta-analysis of many studies is performed.
Images from Wikimedia Commons. First © 2013 Michael J. Ermarth @FDA, second © 2006 user SElefant.

Dif-tor heh smusma 🖖
Helmut Schütz

The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes
ElMaestro
★★★

Denmark,
2022-07-26 12:44
(19 d 20:55 ago)

@ dennis001
Posting: # 23178
Views: 332

Could we design the RCT as a non-inferiority trial at this situation?

Hi dennis001

» So, could we design the RCT as a non-inferiority trial?

Clearly a case for an NI trial one way or another.

» Moreover, how to defining the non-inferiority margin at the situation?

Clearly a case for a scientific advice where you need the authority's nod for margin and design before you kick off.

Pass or fail!
ElMaestro
dennis001
☆

Hong Kong,
2022-07-27 05:21
(19 d 04:17 ago)

@ ElMaestro
Posting: # 23181
Views: 276

Could we design the RCT as a non-inferiority trial at this situation?

Dear ElMaestro
Thanks!

» Clearly a case for an NI trial one way or another.

Thank you for your comment that helped to assuage my doubts.

» Clearly a case for a scientific advice where you need the authority's nod for margin and design before you kick off.

Ok. we will discuss with them.