• Users Online: 59
  • Home
  • Print this page
  • Email this page
Home About us Editorial board Ahead of print Current issue Search Archives Submit article Instructions Subscribe Contacts Login 


 
 Table of Contents  
CURRICULUM IN CARDIOLOGY - STATISTICAL PAGES
Year : 2017  |  Volume : 3  |  Issue : 2  |  Page : 106-108

Number needed to treat


Division of Forensic Science, School of Basic and Applied Sciences, Galgotias University, Greater Noida, Uttar Pradesh, India

Date of Web Publication20-Nov-2017

Correspondence Address:
Amitabh Biswas
Division of Forensic Science, School of Basic and Applied Sciences, Galgotias University, Greater Noida, Uttar Pradesh
India
Login to access the Email id

Source of Support: None, Conflict of Interest: None


DOI: 10.4103/jpcs.jpcs_31_17

Rights and Permissions
  Abstract 


The Number Needed To Treat (NNT) is a measure used in epidemiology to convey the effectiveness of an intervention. It is the average number of patients who need to be treated to prevent one bad outcome. It is the reverse of the Absolute Risk Reduction. The lower the NNT, the more effective the intervention. In this article we discuss the concept and limitations of this measure.

Keywords: Number needed to treat, odd's ratio, randomized clinical trials, relative risk reduction


How to cite this article:
Biswas A. Number needed to treat. J Pract Cardiovasc Sci 2017;3:106-8

How to cite this URL:
Biswas A. Number needed to treat. J Pract Cardiovasc Sci [serial online] 2017 [cited 2019 May 20];3:106-8. Available from: http://www.j-pcs.org/text.asp?2017/3/2/106/218806




  Introduction Top


Number needed to treat (NNT) is a simple measure depicting the effectiveness of a health-care intervention. The NNT presents a measure of the effect of a medicine or therapy by estimating the number of patients who need to be treated to benefit one person. This is a statistical concept, but intuitive, for we know that everyone is not benefited by a medicine or intervention – some benefit, some are harmed, and some are unaffected. The NNT tells us how many of each.[1],[2]

In any comparison of a previous treatment with a new drug treatment, the clinical question is whether this new treatment benefit the patients, has no effect or is harming patients. The researchers are focused on whether the difference is small or big. For example, if the new drug improved the left ventricular ejection fraction by 15% with respect to placebo, is that 15% improvement meaningful? To answer these questions, there are calculations which measure the effect size. These measures include the standardized mean difference, relative risk (RR), odd's ratio (OR), NNT, and number needed to harm.


  Definition of Terms Top


Absolute risk (AR) = The number of events in treated or control groups, divided by the number of people in that group.

  1. ARC = the AR of events in the control group
  2. ART = the AR of events in the treatment group
  3. AR reduction (ARR) = ARC − ART
  4. RR = ART/ARC. RR is the ratio of the probability of an event occurring in an exposed/treatment group to the probability of the event occurring in a control group
  5. The NNT is the inverse of the ARR
  6. An OR is a measure of association between an exposure and an outcome. It represents the odds that an outcome will occur given a particular exposure, compared to the odds of the outcome occurring in the absence of that exposure.


Odds ratio = (Rate group 1/[1 − Rate group 1]) ÷ (Rate group 2/[1 − Rate group 2]).


  Understanding the Calculation for Number Needed to Treat Top


Let us suppose, two treatments are administered in a randomized control trial. Rate or proportions of events for both drugs are found to be p1 and p2, respectively. Difference in the probability/rate (p1-p2) of both drugs is defined as ARR which expresses in absolute terms the benefit or harm of the drug tested over the control treatment. RR, OD and RR reduction are other measures to stipulate the benefit or harm of the new drug treatment. At the same, NNT is suggested as the measure to express the benefit of the new treatment.[3] The inverse of AR reduction is defined as NNT, that is, NNT = 1/ARR.


  An Example Top


Let us suppose, in a double-blind, randomized, placebo-controlled clinical trial, Valsartan 160 mg/day treatment was randomly given to 35 hypertensive patients and placebo treatment was given to 37 patients. The effect of the drug was measured through ambulatory blood pressure (BP) monitoring. It is found that three patients in the placebo arm showed lowering of BP as compared to 11 patients on Valsartan (P< 0.05, Fischer's exact test).

If we calculate the proportion of patients with lowering BP: Proportions are p1 = 11/35 = 0.314 for Valsartan and p2 = 3/37 = 0.081 for placebo. The AR reduction, ARR = p1-p2 = 0.233, which measures the absolute gain due to the new drug. We can calculate OR (0.314/0.686)/(0.081/0.919) = 5.19.[4] With a RR of 3.87 (p1/p2 = 0.31/0.80), probability of lowering of BP is about four times more than that of control treatment. NNT calculated as 1/0.233 = 4.29, means that four patients must be treated with Valsartan to benefit one extra patient as compared to control treatment.


  Another Example Top


Let us assume that in a cohort of patients with heart failure, a new drug has to be tested called TEST DRUG. We do a study where all the patients get the best medical care, including a PLACEBO in group 1 and get the best medical care and TEST DRUG in group 2. After 6 months, the mortality in the PLACEBO arm is 5.9% and in the TEST DRUG arm is 2.6%. Hence, the TEST DRUG reduced Deaths by 3.3%. The numbers needed to treat (ARR = 3.3% =3.3/100 = 1/30> NNT = 1/ARR = 30) is 30. So you have to treat 30 additional patients to prevent one death.


  Drawbacks of Number Needed to Treat Top


  • It is described as a single number; and therefore, some investigators have started describing a confidence interval (e. g., NNT 5 (95% confidence interval 3–9)
  • It does not account for a patient's baseline risk
  • The NNT of a given treatment will be different when it is compared to different drugs or placebo and so it cannot be used as a single value
  • The time frame of a given study should also be mentioned. For example, if a treatment was conducted over a mean of 4 years, its NNT should be expressed with the same time component.[5],[6]



  Online Calculator for Number Needed to Treat Top


There are many online calculators available in the market, GraphPad software is one of them providing all sorts of statistical calculators for the investigators. Simply insert the number of patients benefited for control and new treatment with their total numbers of trials for each.[7] For the abovementioned example, you may see the result in [Figure 1].
Figure 1: GraphPad NNT calculator overview for the example used in this article.

Click here to view



  Conclusion Top


The “NNT” is a simple concept which is a measure of the impact of an intervention or a medicine. It shows the impact by calculating the number of patients, which need to be treated to affect the outcome of one patient. The advantage is that it is simple to understand, can be easily calculated from the ARR, and gives a clear idea of the impact of an intervention. NNT is the inverse of the ARR. ARR is the Control Event Rate – Experimental Arm Event Rate.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.



 
  References Top

1.
Suissa S. Calculation of number needed to treat. N Engl J Med 2009;361:424-5.  Back to cited text no. 1
[PUBMED]    
2.
Church EW, Gundersen A, Glantz MJ, Simon SD. Number needed to treat for stroke thrombectomy based on a systematic review and meta-analysis. Clin Neurol Neurosurg 2017;156:83-8.  Back to cited text no. 2
[PUBMED]    
3.
Laupacis A, Sackett DL, Roberts RS. An assessment of clinically useful measures of the consequences of treatment. N Engl J Med 1988;318:1728-33.  Back to cited text no. 3
[PUBMED]    
4.
Hutton J. Number needed to treat: Properties and problems. J R Stat Soc Ser A 2000;163:403-19.  Back to cited text no. 4
    
5.
McQuay HJ, Moore RA. Using numerical results from systematic reviews in clinical practice. Ann Intern Med 1997;126:712-20.  Back to cited text no. 5
[PUBMED]    
6.
Altman DG. Confidence intervals for the number needed to treat. BMJ 1998;317:1309-12.  Back to cited text no. 6
[PUBMED]    
7.
Available from: https://www.graphpad.com/quickcalcs/NNT2/. [Last accessed on 2017 Sep 02].  Back to cited text no. 7
    


    Figures

  [Figure 1]


This article has been cited by
1 Novel Mineral–Vitamin Treatment for Reduction in Cigarette Smoking: A Fully Blinded Randomized Placebo-Controlled Trial
Phillipa K Reihana,Neville M Blampied,Julia J Rucklidge
Nicotine & Tobacco Research. 2018;
[Pubmed] | [DOI]



 

Top
 
 
  Search
 
Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
 Related articles
Access Statistics
Email Alert *
Add to My List *
* Registration required (free)

 
  In this article
   Abstract
  Introduction
  Definition of Terms
   Understanding th...
  An Example
  Another Example
   Drawbacks of Num...
   Online Calculato...
  Conclusion
   References
   Article Figures

 Article Access Statistics
    Viewed967    
    Printed41    
    Emailed0    
    PDF Downloaded181    
    Comments [Add]    
    Cited by others 1    

Recommend this journal


[TAG2]
[TAG3]
[TAG4]