Home / Healthcare & Medicine / Pharmacy / Evidence-Based Medicine in Clinical Pharmacy Practice / Interpretation of OR, RR, HR, and NNT/NNH Analysis
Interpretation of OR, RR, HR, and NNT/NNH Analysis
Interpretation of OR, RR, HR, and NNT/NNH Analysis
So here let’s summarize again how to interpret an OR RR and an HR so the outcome is negative or a bad outcome you want the results of the treatment group to be less than one because that would be considered protective. So you want less of the patients to experience this bad risk and some examples that you’ll see in the literature of negative outcomes are things like death MI, hospitalization strokes things like that. Less often and very rarely do you see a good outcome and if you do have a good outcome then you want the number to be RR or to be geater than one.
Because you want to see more people experience that good effect and some examples are vomit free period LIBOR smoking cessation. So I said from the majority of the time you’re gonna have negative outcomes but just want to make sure that you’re prepared and can interpret positive outcomes correctly and so I just wanted to provide you with one example in the literature of a good outcome there was a study that looked at drunk combinations in the treatment of rheumatoid arthritis. They considered their outcome a good response to the treatment and so the RR was 1.59 which again remember because it’s a good response it’s a good outcome. You want that number to be greater than 1.
And you can see the confidence interval does not touch nor cross 1 and the p-value is similar to what our consistent with the clinically relevant confidence interval that you have. So these often instead of caught being called a risk ratio as the other ones all are they’re often called a benefit ratio. So let’s now move on and talk about NNT and NNH statement. As I said just giving the number doesn’t tell you anything. If you just say the n NT is 24, looks like 24 what and these are the six key things that really need to be in NNT and NNH statement to make sure you can clearly identify what’s going on.
Obviously the one is the first thing you want to have is the number of patients. But you also want to have the treatment intervention with the dose, because if you don’t have the dose then you’re not sure how you’re going to extrapolate that information to the general population. You also want to have the duration. Because NNT of 24 irrespective of the situation if it happens over six months or six years will be vastly different. And then you want to make sure you list it correctly is it causing a benefit is it caused or is it preventing or is it harming one additional patient above what was seen in the study.
Then you also want to make sure it’s tagged to the primary outcome. Again because of clinical relevance although you also do these for the ad ADR often times for the NNH and then you want the comparator with dose. Again making sure that you’re comparing apples to apples with this. So here’s just the example of what you would want it to say I know it seems long but if you think about it from a clinical standpoint you want all this information to be able to determine if that treatment might be beneficial for your patients.
So you listed as the number of patients who need to be treated with a specified treatment with the dose over a specified time in order for one additional patient to benefit from that treatment with the dose or to prevent one additional person from having an event or cause one additional negative effect compared with the control or comparative group with the dose. So we’ll obviously go through some examples, Now just in case you like algorithms I’ve provided this one here for you, I will not go through the non-inferiority one right now because we’re going to be doing that one next session.
But we’ll look at this one and I’m sorry the algorithm wouldn’t fit on one slides are gonna have to have it on two slides. So you can see you want to start with your primary outcome, and if it’s a nominal type data then obviously you’re gonna be looking at measures of Association CIs. If it’s not then you’re going to be looking at general CIs. So you can see if the primary outcome is bad and then you also have the flow section that says if the outcome is good and will this depend on how what you state there.
If the studied you grow is significantly better than control, then you do the NNT statement for the study, if the safety drug was not then you’re gonna have NNH step statement for the study group. And then you could see that if the drug was significantly better than control at NNH statement for study drug and then same thing for the other ones and you can see at the bottom it’s similar statements either talking about preventing or harm depending on what was seen in the results.
Prof. Mary Ferrill summarizes how to interpret an OR, RR, and an HR.
First, if the outcome is negative, we want to see that the results of the treatment group should be less than one because that would be considered protective and make fewer patients to experience this bad risk.
On the contrary, if the outcome is positive, we want the number to be greater than one since more people experience that good effect.
Besides, she took a good MOA outcome for an example. We should consider a benefit ratio rather than a risk ratio in a good outcome.
Next, we need to understand the 6 steps of an NNT/NNH statement.
Finally, we can learn how to assess NNT/NNH possibilities with the algorithmic diagram.
Do you have any questions about this section? Please leave them below.
This article is from the online course:
Evidence-Based Medicine in Clinical Pharmacy Practice
Created by
This article is from the free online
Evidence-Based Medicine in Clinical Pharmacy Practice
Reach your personal and professional goals
Unlock access to hundreds of expert online courses and degrees from top universities and educators to gain accredited qualifications and professional CV-building certificates.
Join over 18 million learners to launch, switch or build upon your career, all at your own pace, across a wide range of topic areas.
