Download Absolute Risk

FAST STATS A 5-minute Review for the Evidence-Based Practitioner

When evaluating medical literature, there are two important concepts

absolute risk vs. relative risk number needed to treat

Absolute Risk & Relative Risk Relative Risk (RR) compares the probability of an event occurring in the exposed group vs. the non-exposed group.

Absolute Risk (AR) describes the incidence of a condition in a population.

Absolute Risk & Relative Risk Example:

Drug A prevents heart attacks (MIs) but also increases the risk of lung cancer. A placebo-controlled trial in 200 people (100 in each arm) found that there were three heart attacks in the treated group and six heart attacks in the placebo group. There were three cases of lung cancer in the treated group, and one in the placebo group.

MIs Lung Ca

Drug A

Placebo

3 3

6 1

Absolute Risk & Relative Risk We might say that Drug A reduces heart attack risk by 50% or cuts heart attack rate in half.

We could also say that the heart attack risk is reduced by 3%.

Absolute Risk & Relative Risk We could say that the lung cancer risk increases by 200%. We could also say that lung cancer risk increases by 2%.

MIs Lung Ca

Drug A

Placebo

3 3

6 1

Absolute Risk & Relative Risk Absolute risk makes risks or benefits look smaller.

Relative risk makes risks or benefits look bigger.

Absolute Risk & Relative Risk To be fair, both harms and benefits should be presented in absolute terms or relative terms.

Number Needed to Treat Number Needed to Treat (NNT) is an important means of evaluating data. NNT is calculated by taking the reciprocal of the absolute difference between experimental groups. Another way is by dividing 100 by absolute risk.

Number Needed to Treat 3% of the treatment group and 6% of the control group had heart attacks. The absolute difference is 3% because 6%-3%= 3% = 0.03

-ORThat means that 33 people must be treated with this drug to prevent one heart attack.

Number Needed to Harm Number Needed to Harm (NNH) is also calculated by taking the reciprocal of the absolute difference between experimental groups. Another way is by dividing 100 by absolute risk.

Number Needed to Harm 3% of the treatment group developed lung cancer, compared to 1% of the placebo group. The absolute difference is 2% or 0.02.

-ORSo the NNH is 50, meaning that we would expect one additional case of lung cancer for every 50 people treated with Drug A.

NNT vs. NNH One out of every 33 people treated with Drug A will be spared a heart attack; however, one out of every 50 people will develop lung cancer caused by Drug A.

MIs Lung Ca

Drug A

Placebo

3 3

6 1

Congratulations! Now you understand the difference between absolute and relative risk, and you can calculate NNT and NNH. This information will help you make evidence-based decisions for your practice.

PharmedOut is a Georgetown University Medical Center project that advances evidence-based prescribing, and educates healthcare professionals about pharmaceutical marketing practices. Learn more about us at http://pharmedout.org/aboutus.htm. Cover art by Andrea Sikora, inspired by Wojtek Kozak Presentation layout by Andrea Sikora and Grace Doan

Nov 2011 version 2.4