loading words...

Dec 08, 2018 07:12:41

On probabilistic decision-making

by @vickenstein | 222 words | 🐣 | 218💌

Victoria Maung

Current day streak: 0🐣
Total posts: 218💌
Total words: 55041 (220 pages 📄)

Did you know that 90% of drugs only work in 30-50% of people? As an example, 300 people prescribed statins have to take it for a year before 1 heart attack, stroke, or other adverse event is prevented. However, 5% of patients, or 15 people, experience side effects, such as muscle and joint pain or gastrointestinal distress. This is not atypical.

Daniel Levitin, a neuroscientist, introduces the concept called number needed to treat (NNT), which measures the number of participants needed in an intervention to benefit one person. The NNT for the particular statin mentioned above is 300. NNT can inform doctors about the likelihood that a patient will be helped, harmed, or unaffected by a particular treatment. 

NNT demonstrates that even medical decision-making is more of a probability-driven process than absolute right or wrong. What's reasonable to one doctor might not be for another. 

If we framed our personal goals in a similarly probabilistic fashion, would that help us become more tactical decision-makers? Rather than expecting a prescription of "A" to be a guaranteed path to positive outcome "B," "A" merely represents an increase in the likelihood of outcome "B." 

With the limited number of days we posses, what would be a reasonable course of action beyond treatment "A" to maximize the probability of achieving this outcome "B"? 

contact: email - twitter / Terms / Privacy