As a statistician I like to know the actual odds so I can make a rational decision. I have encounter situations where the operation was successful and there was a 40% risk of patient death within 6 months — and that was never disclosed to the patient or the family. Personally, I would love to see printed risk statement required to be given with each drug (and compulsory reporting of any or odd reaction — typically MDs will dismiss them and not report them). – the patient reports (giving the MD’s name) and the MD reports giving patient name.
“There is a low risk of ….” is misinformation.. numbers are needed!
” Idiosyncratic DILI is like other adverse effects of drugs underestimated and under reported in most epidemiological studies.” 
Case Study: Infliximab
Infliximab aka Remicade, Remsima, Inflectra has been cited in the immediate prior posts and is likely a good example because it has been heavily used — including for:
So what are the risks if you use it? A PubMed search returned over 3000 articles dealing with adverse effects.
- “Infliximab caused liver injury in 8.3% of treated patients” 
- “10~20 % of patients was Remicade-resistant.” 
- “primary-non response” 22% 
- “65.5%/69.4% for clinical response”
- high pre-treatment hemoglobin was also a predictor of good response 
- The sensitivity and specificity in predicting response to IFX based on this gene profiling was 95% and 85%, respectively.
- all of these studies more severe disease was associated to adverse outcomes and less favorable response to anti-TNF 
- “Chronic sinusitis.. approximately 2%”
- infusion reactions and delayed hypersensitivity 10%  27% 
- Mild and moderate IR occurred in 17% 
- Adverse drug reaction: 41.4%  35%
- Serious adverse events: 67.9% 
- 48.48 % of patients who received infliximab presented (ANA, ENA, anti-dsDNA) autoantibodies 
- Infections and elevation of transaminases occurred in 28.40% 
- severe infections 6% 
- Anaphylaxis: 6%
- mild acute infusion reaction in 6%
- hypotension in 6%
- respiratory distress 6%
- skin rash and eruptions 6%
- hypertension 3%
- tightness in the chest 3%
No statistics on, but..
- ” an increase in weight” 
- “an association between focal mucinosis and thyroid dysfunction, as well as possible adverse effects of biological therapy with TNF-α antagonists.”
- ” probably exert a direct, toxic effect on the bone marrow”
- Mycobacterium chelonae bacteremia 
- Hepatitis 
- Herpes zoster
- Aseptic meningitis  – 5 cases at least reported
- a significant association with lupus 
While the analysis is not complete, we may sum it up as: This drug as a 70-80% response (and for many diseases it is used for, less than 25% chance of remission) with probablity a 10% of having liver damage (regardless of outcome), 30% change of getting an infection (with the immune system depressed), 70% chance of adverse reaction. The odds of taking it and everything going “fine” is about 1 in 15. 14 in 15 people will have problems.
As a side note, I found two of the studies found success (or issues, such as developing APS) was predictable from DNA. I really doubt that more than 1% of patients prescribe this are testing for DNA, exposing patients to needless risks with little chance of success.
After the above article, someone who was about to take aka Remicade, contacted me and allow access to their DNA (from 23andMe.com ). I found a beautiful table at: http://www.nature.com/tpj/journal/v13/n4/fig_tab/tpj201253t2.html (Gene polymorphisms that can predict response to anti-TNF therapy in patients with psoriasis and related autoimmune diseases (Aug 2013) R Prieto-Pérez, T Cabaleiro, E Daudén and F Abad-Santos) and both created a LiveWello.com panel from it and walked their DNA SNPS. The result was a strong prediction that they would be a non-responder (i.e. 6x more ↓ than ↑). I wonder how the MD is going to respond to “According to the latest studies (2013), my DNA says that I am going to be a non-responder to Infliximab and other anti-TNF-Alpha drugs. What do you suggest next?”
The sweet thing would be for Insurance Companies to demand DNA testing before approving the use of Remicade — it will save them money (and cut premiums too!) and save patients side-effects!