Question by ~EmAd~ C: Determine the sensitivity and specificity of the drug and individuals being tested in the group?
You are working in a laboratory investigating a new drug intervention to
assist in the reduction of withdrawal symptoms of addicts wishing to come off
chronic methamphetamine abuse and are presently in the phase II portion of the study. In your group, you have 750 subjects who are chronically addicted to methamphetamine and you randomly assign these individuals to a control group and a treatment group. Determine the sensitivity and specificity of this drug on the individuals you are testing given this number of individuals in this group

Best answer:

Answer by checkmate!
For the question, you can put 375 in each of the groups but statistically they don’t need to have the same number. You can find the sensitivity of the diagnostic test, observation, or self-report from the number of true positives it finds. Then divide that by the total number of people randomly assigned to the treatment group, which is likely to be 375. Specificity of the same test is from the number of true negatives divided by the total number of 375 in the control group. The ratio or proportion formulas you have been taught will result in the same value as using one – p. For example, if n=250 for the treatment group and 200 are true positives then it is 200/(200+50 false negatives) = 0.80. That’s the same as 200/250 = 0.80 and 1-0.20=0.80 if the proportion of false negatives is 50/250 = 0.20. Probability is excellent to use regardless of how many subjects you have. So, that means it is a robust test even if you have unequal numbers in the control and treatment groups because proportion is unaffected by the number of people. But “they” will get upset if you don’t split the group into two groups of the same number because they will criticise that unequal numbers might bias the results. The thing is, if the new drug really works it will work regardless of whether one group has 300 and the other 450, as long as the difference isn’t too extreme like more than double the size of 200 with 550 in the treatment group. It’s the idea that testing two different things on the same number of people gives each a fair go. But really it makes little difference because equality doesn’t always come with a fair chance when individual humans don’t all react in the same way. So unlike coin tosses, equal numbers aren’t going to change that. You haven’t yet been taught probability, which is harder than how it looks, so you probably won’t understand my detailed answer.

It seems the question should have been better worded like so: “How do you calculate the specificity and the sensitivity of the new treatment?” OR “How would you design a clinical trial to find the specificity and sensitivity of a new drug?” So simple without all the irrelevant info in the question above. But I know it’s not your doing. Let me know when I get full marks for you : )

Know better? Leave your own answer in the comments!

 

Why America's Addiction Treatment System Is Broken & What You Can Do

Filed under: drug addict treatment

Why America's Addiction Treatment System Is Broken & What You Can Do We know a lot about addiction. We know that addiction is a complex brain disease – not a weakness or lack of willpower. We know its risk factors include “genetic predisposition …
Read more on PsychCentral.com (blog)

 

Is There An Addiction Treatment For Everyone?

Filed under: drug addict treatment

With growing evidence that the 12-step program doesn't work for everyone, can we create a universal addiction-treatment system? Hosted by: … introllectual July 19, 2013 17:15 If you don't want to be an addict, there's only one solution, STOP USING DRUGS.
Read more on Huffington Post

 

Paula Deen & Brother Lose Bid To Seal Court Records Detailing 'Drug And …

Filed under: drug addict treatment

Paula Deen and her brother have lost a bid to seal court documents that reveal his “past treatment for drug and alcohol addiction,” RadarOnline.com has exclusively learned. The Queen of Southern Cuisine had appealed a Georgia judge's earlier ruling in …
Read more on Radar Online