Talk:Number needed to treat

Reversion of Psarka's Reversion

Listen up Psarka; I have no doubts you have heard these words before, but I will say them anyways. In this instance at least, you are arrogant and lack both manners and the expertise sufficient to say what you have.

Your nearly 900 word post here is a testament to just how extreme your need to criticize my work is. Instead of spending such an absurd amount of time on explaining why you will revert my edits, did you perhaps think you might instead spend such time on simply making the edits you feel are appropriate? The only thing you have done here is waste your time to justify wasting mine, all while never improving the quality of this article. You seem to have been so incensed by my request not to delete my work without reason that you set out to prove my incompetence just to spite me for making such a bold request.

Your Complaint	Response
"I get the feeling that you are very emotional about your additions, but also that you had a hunch that they will get reverted. What made you think that? Did you saw some weakness in your text when you wrote it?"	Drop your god complex for a moment, would you? Would you like to know why I said what I did? Sure, I'll tell you, even though you deserve nothing of the sort. I would refer you to the specific words I used - "DO NOT REVERT...". I'm afraid I'm going to have to oppose your absolutely asinine theory that I somehow put hours of effort into improving in article all the while somehow retaining the knowledge that I was simply too incompetent to possibly have anything to contribute. In fact, I added such a statement exactly for pedants like you. As I've highlighted, my comment does not say "DON'T YOU DARE EDIT MY REVISION", it says revert for a reason, which is that, unless one has a FACT-BASED issue with my edit, he should focus his efforts on ADDING to the article instead of DELETING my work. Your pathetically patronizing explanation of my statement is a delusion in your thick skull only.
"Or maybe you thought that inevitably some evil guy will show up to destroy your valuable effort? If so, I'm afraid that no matter what I write, you already made up your mind and will curse me to hell. Well do that, while I will try to do my best explaining my point of view in the politest and most constructive way I can. Here I go."	Nice save, buddy. Say one pseudo-polite thing to give the appearance that your previous statements weren't thoroughly insulting. Your baseless idea that I would oppose you no matter what you said is utter stupidity. Why, if I was so unable to hear any criticism, would I specify "WITHOUT THOROUGH EXPLANATION OF REASONING"? Do you truly think, not knowing me in the slightest, that my request to not delete my hours of work is an act of absolute arrogance? How exactly would that fit with your previous theory that I am somehow fundamentally confident in my own incompetence? You can say you're being "polite" all you want. It does not alter how polite your words are. It is laughable that you would call what you decided to do the "most constructive" way you could possibly address your concerns. As I have mentioned previously, you seem to be missing the entire idea of Wikipedia, which is not one of academic sparring, but of cooperative contribution.
First you did not correct many inaccuracies. If you read the old version of that section generously, you will see that all the numbers and notions are accurate. The text is quite poor, I agree, but accurate.	Here is where I begin to question what article you could have possibly been reading. For someone supposedly so knowledgeable in epidemiology, you have missed many glaring false statements present in the original. Allow me to address your largely invalid, wholly unhelpful points now, in no particular order (if it does not harm your sense of self too severely).
"Your other additions are valuable, but they should find their place in different articles. Why? Take a look at the table at the bottom of the article. NNT is one epidemiological measure out of 25(!). All the discussion in this article should be focused on the NNT itself, and not the broad context of epidemiology."	You know, I am not an idiot. Despite your implication, yes, I am aware that NNT is not the only important term in epidemiology. I have taken epidemiology classes. Unlike mine, your negative contribution, in the form of reverting mine, is not valuable. Believe it or not, expounding on one topic will inevitably contain some information on other topics. If it didn't, it would be a dictionary, not an encyclopedia. An article should contain material that is pertinent to the understanding of the article's topic. That requires explaining some things which would otherwise, by your idea of what articles should contain, be scattered broadly over up to 25(!) different articles. Despite the absurdity of your claim, I will address each individual point.
"When I think about the article such as NNT, I imagine a reader that wants to quickly become familiar with the terminology of epidemiology measurements. So they click around and try to understand the differences between the measurements and their usage. My goal is to provide them this understanding as painlessly as I can."	Yeah... no. That is not the goal of wikipedia. Clearly, the specific wording of articles should be as simple and digestible as possible, but, beyond this, it would be a disservice to the reader to leave out important aspects of the specific thing they are reading about simply in the name of "painlessness". In every situation, it is better to err on the side of too much detail than too little, as readers are capable of sorting through more information than they wanted, but are not capable of synthesizing from nothing information they do not have.
"That includes not burdening their brain with extraneous information that they could easily find in a dedicated article."	Please explain how you expect readers to find information in a "dedicated article" that is relevant to the information about to be presented to them if you do not provide the reader with any inkling that there is indeed an additional concept to understand before they ought to continue reading? You could certainly argue an internal link to a specific subsection of a different page on the subject could replace this, but you did not edit the text to include that, you instead REVERTED the entire thing.
'Sentence "The treatment may be a drug in the form of a pill or injection, a surgical procedure, or many other possibilities" should find its place in Treatment and control groups or Clinical trial'	HAVE YOU READ the original article??? It uses EXCLUSIVELY the example of a "pill" being tested in a trial. If I didn't know better, reading this entire section I'd think pills are the only treatment relevant to the NNT measure. So, in fact, the sentence in question is acutely relevant to a proper understanding of the section. If you wanted to say it a different way, perhaps you should have done so instead of REVERTING the entire thing.
Chunk of text "This is typically a type of study <...> but are much more difficult and time-consuming to perform" should find its place in Clinical trial	I was actually going to edit this bit again as it's not quite correct, but you seem to have missed that and locked in on its relevance. If you'll turn your attention to the section just below you will see the real-life example is a study that involves a group of participants selected, in part, on the basis that they do not have heart disease. In this case, when the primary outcome measure is development of heart disease, the entire design of the study (starting with all sick or all healthy participants) is wildly different, which warrant a short explanation. Unless you disagree with how I did this, in which case you could edit it instead of REVERTING the entire thing.
Chunk of text "The control group may receive a placebo <...> quality of the placebo" should find its place in Clinical trial or Clinical endpoint or Treatment and control groups	Again, no it shouldn't, because it's extremely important to understand in order to make good use of the table it precedes. Someone without a comprehensive understanding of the term NNT would certainly be at high risk of thinking the only kind of control is a placebo, which is also exclusively what the section said before you reverted my work. Even readers who have read a number of academic papers documenting clinical trials would probably have seen the term "placebo controlled" in nearly every one. It is absolutely critical to the understanding of NNT to know that an NNT of 10 does not indicate a treatment is ineffective in cases where the control is an existing, effective treatment. AND it just so happens this EXACT scenario plays out in the table in this section. Not every reader has the incredible mastery of all epidemiological knowledge that you do, and not every reader want to open 6 different articles to understand one concept.
Repeating the same information again and again and again does not make things clear. It makes reader annoyed that they have to constantly skip over the same ramblings.	You are wrong. Please learn literally anything about the science behind education or memory formation then get back to me. Repetition, especially when given in the form of similar information written in different ways, is without a doubt the best way to ensure understanding and retention of a concept. You just don't know what you're talking about.
Finally, if you forgive me, your contribution is not of high quality.	No, I will not forgive that. There are a hundred much less rude ways of saying the same thing. If you were correct I'd be fine with it, but you aren't.
You write a sentence 'There are a number of factors that can affect the meaning of the NNT". What would you expect in a following sentence? I would expect an elaboration on these factors. Instead, you continue with "The treatment may be a drug in the form of a pill or injection, a surgical procedure". Why? How pill vs injection vs surgical procedure alter the meaning of NNT?	I'm sorry, do you know what NNT is? Did you read what I wrote literally right after the item you quoted? The route of administration, which is what "pill vs. injection" describes, IS A FACTOR affecting the meaning of the NNT. For example, the standard treatment is a pill consisting of a racemic formulation of a drug, and the experimental treatment is a subdermal implant that slowly releases the same drug, but in the form of the pure dextrorotatory enantiomer. Should the NNT be very high or negative, it is vital to know by what method the drug is delivered, as without this knowledge one might think the pure enantiomer is either not much more effective than the racemic mixture or less effective than the racemic mixture, when in fact it may be that the pure isomer is significantly more effective but the subdermal implant is not an effective mode of drug delivery. You are absolutely dead wrong on this which you would have realized if you weren't so focused on proving me wrong and instead used that energy to determine the truth and relevance of the actual information.
You write "The following examples", and follow up with "In this example".	Wow, well there it is. There's the thing that justified REVERTING the edit instead of changing or deleting ONE SENTENCE. It's pathetic how little you care about actually improving the quality of the article compared to making sure you control every detail of all epidemiology articles.
Your text contains too much boilerplate such as "depending on the situation", "or many other possibilities", "The following examples", "In this example", "it is important to understand", "This is typically a type", "which may be a reduction or halt in worsening of the condition, an improvement in the condition, or an outright cure of the condition.".	I fail to see how you can possibly call your last quote "boilerplate". Please point me toward the numerous uses of that exact wording elsewhere on wikipedia. No, nevermind, why don't you instead EDIT the phrases you don't like INSTEAD OF REVERTING THE ENTIRE THING?
You make strong claims without any citations, such as "This <...> type of study that would occur only if <...>", "Prospective studies produce much higher quality evidence <...>"	First, your convienient omission of THREE words in the first part of the quote via "<...>" provides a pretty important qualifier. The full quote reads "typically a type of study that would occur only if". As far as the other claim I made, if your competence in study design was as great as your arrogance, you'd know a prospective study is factually a source of higher quality evidence. It's called the scientific method. A study is only a true experiment if the scientist actually manipulates a variable, which is impossible in a retrospective study. If you think I'm wrong please tell me why, but I don't think you will because you have no idea what you are talking about.
In fact, you did not add a single citation	Oh BOO-HOO cry me a river. If every edit that did not include additional citations was reverted wikipedia would be a wasteland. Do you know what you do if something is missing a citation? YOU ADD ONE. You do not REVERT the entire thing.
You redefine $I_{e}$ and $I_{u}$ differently from how they were defined in the article introduction.	No kidding? Well, maybe that's because the definitions given in this section were TOTALLY WRONG. Why don't you tell me this original definition of $I_{e}$ is correct, cause I'd love to see if you'll straight up lie: Original: $I_{e}$ is the probability of still having the disease after taking the pill (i.e. complement of the probability of being cured after taking the pill). Wrong parts: $I_{e}$ is the probability of *still having the disease* after *taking the pill* (i.e. complement of the probability of *being cured* after *taking the pill*). Seriously, do you understand these concepts? Because your idea that the article should be using the word "cure" in the definition of the term is absolutely ludicrous and wrong.
"You write "$I_{e}$ is the probability <...> this is the inverse of the probability <...>". This is incorrect. Inverse of a probability is never a probability."	No, you are incorrect. Here is the first definition of the term "inverse" from wiktionary: "Opposite in effect, nature or order." Not satisfied? Want to use the technical definition? Sure, here it is: "(mathematics) Having the properties of an inverse; said with reference to any two operations, which, when both are performed in succession upon any quantity, reproduce that quantity." An action which may produce multiple events is repeated 20 times. Event x occurs 6 times, and other events occur 14 times. Probability of event x: 0.3. Method of calculation: 1.0 minus probability of "not x". Calculation: 1.0 - (14/20). Inverse: 1.0 - 0.3 = 0.7. "Inverse of 'not x'" calculation: 1.0 - 0.7 = 0.3. Repeat the two operations ad infinitum. Hey genius, why don't you explain this to me, because clearly I am too simple minded to understand your advanced logic which goes beyond both the technical and non-technical definitions of the word.
Sentence "For simplicity, a low number of participants will be used, thought scientific studies almost always require many more." just appears out of nowhere in the second row of the table.	Out of nowhere, huh? Cause it would seem the me that the first mention of "ten" participants might warrant a one sentence explanation. Don't like how I did it? EDIT IT DON'T REVERT IT.
You write "Notice that, even though the treatment was effective in eight of ten cases (only one less than the previous example)", but the previous example had two out of ten.	No, actually, you are wrong YET AGAIN. The incompetence is only surpassed by the arrogance. Previous example: "...Nine improve with the treatment and one does not."
You write "This is because" and then give a definition of NNT. That's not helpful.	You again have no idea how to write for an audience other than yourself. The repetition of the definition is very helpful to convey the meaning behind an NNT of -10, which counter-intuitively does not mean "really bad", as many readers might assume without the clarification.
If you decide to give another shot to this section, go ahead. But don't feel obliged, there are plenty of articles that would benefit much more from your attention than this one. I made an effort to clean up epidemiology measurements corner of the wikipedia (see my notes here https://en.wikipedia.org/wiki/User:Psarka/What_a_mess!), but ran out of steam. One page that is a particularly in need of love is Population Impact Measures. If you could try to sort that one out, it would be wonderful.	Listen close you disrespectful, incompetent, patronizing quack: why don't you give it ONE shot by EDITING the page rather than DELETING my hours of work. I don't care how much you want to rule over all epidemiology articles; that does not give you the unilateral authority to revert others' good faith work.

Sincerely,
Lumentex

--Lumentex (talk) 04:28, 12 November 2018 (UTC)

I am sorry you feel this way. Do as you wish. Psarka (talk) 05:27, 12 November 2018 (UTC)

Reasons for reverting edits by Lumentex

Hi Lumentex,

I'm sorry, but I'm reverting your edits. Reading your message "Completely rewrote this section to correct many inaccuracies and greatly increased clarity of explanations. DO NOT REVERT WITHOUT THOROUGH EXPLANATION OF REASONING." I get the feeling that you are very emotional about your additions, but also that you had a hunch that they will get reverted. What made you think that? Did you saw some weakness in your text when you wrote it?

Or maybe you thought that inevitably some evil guy will show up to destroy your valuable effort? If so, I'm afraid that no matter what I write, you already made up your mind and will curse me to hell.

Well do that, while I will try to do my best explaining my point of view in the politest and most constructive way I can. Here I go.

First you did not correct many inaccuracies. If you read the old version of that section generously, you will see that all the numbers and notions are accurate. The text is quite poor, I agree, but accurate.

Second, your explanations made the 'Interpretation' column less tricky to understand, but are very verbose. I would like to try to find a middle ground between the two versions.

Your other additions are valuable, but they should find their place in different articles. Why? Take a look at the table at the bottom of the article. NNT is one epidemiological measure out of 25(!). All the discussion in this article should be focused on the NNT itself, and not the broad context of epidemiology.

When I think about the article such as NNT, I imagine a reader that wants to quickly become familiar with the terminology of epidemiology measurements. So they click around and try to understand the differences between the measurements and their usage. My goal is to provide them this understanding as painlessly as I can.

That includes not burdening their brain with extraneous information that they could easily find in a dedicated article. In particular:

• Sentence "The treatment may be a drug in the form of a pill or injection, a surgical procedure, or many other possibilities" should find its place in Treatment and control groups or Clinical trial

• Chunk of text "This is typically a type of study <...> but are much more difficult and time-consuming to perform" should find its place in Clinical trial

• Chunk of text "The control group may receive a placebo <...> quality of the placebo" should find its place in Clinical trial or Clinical endpoint or Treatment and control groups

Repeating the same information again and again and again does not make things clear. It makes reader annoyed that they have to constantly skip over the same ramblings.

Finally, if you forgive me, your contribution is not of high quality. Consider the following things that caught my eye:

• You write a sentence 'There are a number of factors that can affect the meaning of the NNT". What would you expect in a following sentence? I would expect an elaboration on these factors. Instead, you continue with "The treatment may be a drug in the form of a pill or injection, a surgical procedure". Why? How pill vs injection vs surgical procedure alter the meaning of NNT?

• You write "The following examples", and follow up with "In this example".

• Your text contains too much boilerplate such as "depending on the situation", "or many other possibilities", "The following examples", "In this example", "it is important to understand", "This is typically a type", "which may be a reduction or halt in worsening of the condition, an improvement in the condition, or an outright cure of the condition.".

• You make strong claims without any citations, such as "This <...> type of study that would occur only if <...>", "Prospective studies produce much higher quality evidence <...>"

• In fact, you did not add a single citation

• You redefine $I_{e}$ and $I_{u}$ differently from how they were defined in the article introduction.

• You write "$I_{e}$ is the probability <...> this is the inverse of the probability <...>". This is incorrect. Inverse of a probability is never a probability.

• Sentence "For simplicity, a low number of participants will be used, thought scientific studies almost always require many more." just appears out of nowhere in the second row of the table.

• You write "Notice that, even though the treatment was effective in eight of ten cases (only one less than the previous example)", but the previous example had two out of ten.

• You write "This is because" and then give a definition of NNT. That's not helpful.

If you decide to give another shot to this section, go ahead. But don't feel obliged, there are plenty of articles that would benefit much more from your attention than this one. I made an effort to clean up epidemiology measurements corner of the wikipedia (see my notes here https://en.wikipedia.org/wiki/User:Psarka/What_a_mess!), but ran out of steam. One page that is a particularly in need of love is Population Impact Measures. If you could try to sort that one out, it would be wonderful.

Sincerely, Paulius Psarka (talk) 22:02, 11 November 2018 (UTC)

Asprin example

How about another example with aspirin. The New York Times explains it in easy to understand terms. The Wikipedia article is written by statisticians for other statisticians. --Richard Arthur Norton (1958- ) (talk) 18:03, 6 February 2015 (UTC) http://www.nytimes.com/2015/02/03/upshot/how-to-measure-a-medical-treatments-potential-for-harm.html?

Y I think one example is enough. Psarka (talk) 23:04, 8 September 2018 (UTC)

Caption wrong?

Right now, the caption for the first image is "Group exposed to a treatment (left) has reduced risk of an adverse outcome (grey) compared to the unexposed group (right). 4 individuals need to be treated to prevent 1 adverse outcome (NNT = 4)." Is that correct? From my understand, the adverse outcome would be colored in red as there are more "greys" in the (effective) treatment group. Or am I missing something? 2A02:908:D78:D1A0:3058:DE3D:230B:77EF (talk) 18:15, 23 February 2023 (UTC)

I think you are correct. Even if the infographic is accurate, it does not enhance understanding. In fact, I suspect it confuses more than clarifies. I removed it.

If another editor disagrees and wants to restore the image, please imporve the infographic so that the average reader can understand it. Mark D Worthen PsyD (talk) [he/him] 14:08, 5 November 2024 (UTC)