Thread by @HadidiSamer

Analyzing this study of defibrotide in VOD after transplant from Spain @amarkelkar @AkshaySharmaMD @hmgcoa2
➡️Link:tandfonline.com/doi/full/10.10… (accepted, not final)
➡️ Price data from Spain with €2019
➡️ FDA approved (March 30, 2016): fda.gov/drugs/resource…

➡️ VOD (~15% ,all types): a bad complication with high mortality (severe type, historically >80%)
➡️ Diagnosis &grading(see tables👇) (nature.com/articles/bmt20…)
➡️Authors question: is defibrotide cost-effective using price input from Spain (note this is likely different from US)

➡️ Simplified definitions:
1⃣ Cost-effectiveness analysis(CEA) is usually measured by incremental cost-effectiveness ratio (ICER) and it establishes how much extra has to be paid for extra benefit.

2⃣ Cost-utility analysis, a subset of CEA, measures quality-adjusted life years (QALY’s) which is considered the standard measure in the U.S. For example, one year of life is valued less if a patient is severely disabled than if a patient is not disabled.

3⃣ Markov model: a type of decision tree model especially important in cancer where patients may initially get worse, then improve, then relapse, then improve again and be thought of as cured. Simply it assumes that the underlying decision tree has cyclic end points.

➡️A patient who is well at the beginning of the cycle can make the transition to any one of these states at the end of the cycle: Well(CR), Sick(VOD), Dead( absorbing state)
➡️See authors Markov model 👇 they assumed 2 phases: acute & long term based on survival.

➡️To make Markov model useful for economic evaluation, you need to add in costs or benefits or both for each state (what you put in the model from cost and benefit matter)
➡️ Tables they provide for their assumptions and values ( note utility values based on historical cohort)

➡️Note price differences between Spain and US ( in the US this also differs based on other variables)
➡️They report €27,829/QALY (good value for a bad disease and outcome) especially they have children included who are expected to live longer and with $$ of complications)

➡️ Standard acceptable limits (ceiling ratio) differ between countries ( authors said €30,000/QALY commonly used in Spain where in the US no limit but an upper max of $50,000-100,000/QALY)

Not sure if historical cohort got ursodeoxycholic acid

➡️Now for sensitivity analysis : it is an extra step to do if there is uncertainty about the assumptions that are appropriate for the model and the data used by the model( it illustrates the plausible range of outcomes).
➡️Multiple types:one-way, multi-way,statistically based

➡️There were previous efforts to look into this in the US:
pubmed.ncbi.nlm.nih.gov/28008770/

➡️Multiple limitations (most important)
1⃣relying on 2 sources of data for both arms: a phase III trial and a historical cohort (Outcomes are expected to be worse for historical cohort compared to a positive RCT, variation is a problem).

2⃣ Authors relied on experts who assumed shorter ttt duration than the phase III ( picking some things from the phase III vs relying on opinion in others, need to be careful when you see that😃)

The cost-effectiveness output is dependent on the input

Finally,check @ASCO #ASCOConnection on some definitions and suggested curriculum
connection.asco.org/blogs/should-w…
➡️ Education in cost-effectiveness is needed, many take results from abstract which may not be accurate
Hope this help @amarkelkar @AkshaySharmaMD @hmgcoa2