Feeds:
Posts
Comments

Posts Tagged ‘non inferiority trials’

Why Superior Scientists, will never rely on Non-Inferiority trials?

Posted in Uncategorized, tagged , , , , , , , , , , , on April 4, 2026|

Statistics is the most advanced form of mathematics by which predictions about the future can be made with some degree of surety.  It is a vital cog linking biology with  maths. However, since the last few decades, an abnormal man made bio-mathematical mutation is being spotted in the field of medical statistics.

Why superior scientists avoid Non Inferiority (NI)trials?

They avoid relying NI -trials due to fundamental statistical, interpretive, and ethical weaknesses that undermine rigorous evidence and patient benefit.

Superiority trial models , reduce bias  towards the null hypothesis and proactively protect against false claims. NI trials just do the opposite. The same issues dilute differences, biasing toward falsely concluding the new treatment is “not unacceptably worse” than the active control by a pre-specified margin . Without a placebo arm, sensitivity cannot be confirmed, and the assumption remains tentative.

Key vulnerabilities In NI trials include arbitrary, manipulable margins that may permit clinically meaningful inferiority. High success rates (>85–95%, especially industry-sponsored) suggesting bias, and risks ( Biocreep) where successive approvals erode standards. “Non-inferior” does not mean equivalent .It can mask statistically significant inferiority. Reporting often deviates from guidelines, with inconsistent margin justification and analyses (ITT vs. per-protocol).

Ethically, NI designs accept potential efficacy loss for unproven gains (convenience, cost, safety) without direct proof, exposing patients unjustifiably in high-stakes areas. They place disproportionate proof burdens on de-escalation or alternatives while entrenching suboptimal standards.

A Lancet Oncology piece highlights this as “the tyranny of non-inferiority trials”:. The authors propose abandoning superiority/NI distinctions for simple “comparative” trials.

Final message

Superior scientists prefer superiority or hybrid designs, or direct comparisons of net patient-centered outcomes .They never go for the NI shortcuts driven by regulatory or commercial pragmatism.

Reference

1.Tannock IF, Buyse M, De Backer M, et al. The tyranny of non-inferiority trials. Lancet Oncol. 2024;25(10):e520-e525. doi:10.1016/S1470-2045(24)00218-3.

2.Beryl P, Vach W. Is there a danger of “biocreep” with non-inferiority trials? Trials. 2011 Dec 13;12(Suppl 1):A29. doi: 10.1186/1745-6215-12-S1-A29. PMCID: PMC3287743.

Read Full Post »

“Non-superiority” trial of sub-optimal PCI vs OMT : A proposal for a study design

Posted in Uncategorized, tagged , , , , , , , , , , , , , , , , , , , , on February 3, 2025|

The therapeutics of coronary stenosis has become a technogical wonder, interwoven with statistical wordplay in the last few decades. PCI is sitting pretty at its peak glory.The term OMT or GDMT is a popular terminology, but realistically exist only in guidelines.

It is a strange academic habit among cardiologists, that they have subdivided medical management into optimal and suboptimal. Meanwhile, we haven’t seen any papers from cardiology forums that classify PCI according to its quality. How many of use a term like optimal PCI or guideline-directed PCI (O-PCI, GDPCI). Every PCI, by default, is perceived as good by our flawed coronary intellect.

A single patient experience

Let me share a patient consult from a remote town of north India. He is a STEMI patient (1 year old) with mild LV dysfunction and thinning of IVS and anterior wall. His CAG showed a significant looking, yet non-flow limiting LAD lesion without any troubling symptoms. I came to know he had consulted two institutions and was apparently not happy with their approach (In his own words, “They seem to be primarily interested in caging my LAD than listening to me”).

Somebody has suggested my name. He called me over the phone for a consult. I asked him remain there to follow his doctor’s advice. But, he flew some 2000 km to meet me. He was so knowledgeable and was aware of everything I wanted to tell. Like, viability, scars, futility, and benefits of revascularization, imaging-assisted PCI, impact of PCI on exercise capacity, importance of risk factor management, etc.

I told him, “In my opinion, you have technically a single vessel disease that can be managed well with drugs. But if PCI is to be done, it should be done in a proficient manner, as the lesion looked hard and was close to the LAD ostium, trespassing LCX as well.” I stressed the importance of a professionally done procedure with enough expertise and follow-up maintenance care.

He was not entirely satisfied with my response. He wanted a clear yes or no! . I told him, “If you have full trust, continue with the drugs at full intensity and do a stress test after 3 months. otherwise, if you keep getting even the slightest doubt and anxiety over the hidden blocks, go for a stent immediately at a good Institution. (My conscience said the latter half of my advice was unwarranted, but I had to; after all, me too need a protective mechanism)

He left my clinic profusley thanking me. I am not sure , how my consult was useful for him and what he is going to decide.

Academic lessons from this patient.

1.Patient fear factor over coronary blocks may be the ultimate game changer. Cardiologists should try to mitigate this fear and at the least should not be an amplifier to this emotion.

2.Leaving tricky profesionaly complex decisions to the patient, is an easy escape route for us, however it comes very close to professional incompetence. (Of course, we do this on a routine basis, approved by the modern medical guidelines, ethics, and legal system, in the name of patient empowerment)

3.Finally, we can grow a potential research hypothesis. A sub-optimal PCI is non-superior to OMT.It is curious there is no study available to compare sub-optimal PCI to OMT. We must also realize there is nothing called standalone PCI. Without concomitant OMT, PCI is a dud. Every young cardiology fellow need to etch this fact in their cortical cardiac memory. OMT often turns out to be the savior of stents, but the latter ruthlessly steals the credit.

Postamble

I could find one study analyzing suboptimal stenting (Ref 1), but it didn’t compare it with OMT. Suddenly, as I finish writing this, a big fact struck me hard, i.e., even a well-done PCI in sophisticated core labs with meticulous care struggled to beat OMT in a barrage of landmark trials (like COURAGE, ISCHEMIA, ORBITA). What is the big deal to analyze suboptimal PCI vs OMT?

Prati F, Romagnoli E, Gatto L, La Manna A, . Clinical Impact of Suboptimal Stenting and Residual Intrastent Plaque/Thrombus Protrusion in Patients With Acute Coronary Syndrome: The CLI-OPCI ACS Substudy Circ Cardiovasc Interv. 2016 Dec;9(12):e003726. .

Read Full Post »