Calculating Number Needed To Treat: Answering Biostatistics Questions For IFOM And The USMLE Part 2

Calculating Number Needed To Treat: Answering Biostatistics Questions For IFOM And The USMLE Part 2


Calculating Number Needed To Treat: Answering Biostatistics Questions For IFOM And The USMLE Part 2 Peter     665  

2 years ago

~1.9 mins read

Let us look at the sample NBME question below:


A randomized controlled trial is conducted to compare a new oral proton pump inhibitor with oral omeprazole for the prevention of recurrent duodenal ulcer bleeding. A total of 2000 patients with newly diagnosed duodenal ulcer who have had recent bleeding are randomly assigned to receive the new drug or omeprazole. After 1 month, results show that 4% of patients treated with the new drug have recurrent ulcer bleeding compared with 5% of patients treated with Omeprazole (p<0.05). Based on these results, how many patients would need to be treated with the new drug to prevent one recurrent duodenal ulcer bleed?

 

A.    A. 10

B.    B. 40

C.    C. 50

D.    D.100

E.     E. 200

Dissecting the question

The question targets number needed to treat.

To answer this type of question asking “Number Needed to Treat” you need to remember Absolute Risk Reduction [ARR] (That is how much risk is taken away after exposure or treatment, ARR is gotten when the exposed group is subtracted from the unexposed group)

The formula for Absolute Risk Reduction (ARR): Unexposed group – Exposed group = ARR

To get absolute risk reduction from the question, we need the number of patients who were treated with a new drug (4%) and the number of patients treated with Omeprazole (5%).

ARR = 5 – 4 = 1%

Convert percentage to number = 1/100 =  0.01

The formula for Number Needed to Treat: 1 / ARR

Number needed to treat = 1/0.01 = 100


The answer is D.100

 

.

Was my post useful? Support me to keep creating useful content

Disclaimer If this post is your copyrighted property, please message this user or email us your request at team@pejoweb.com with a link to this post




23 likes
 

Advertisement

";

Advertisement