Appendix C: Developing and Maintaining Numbers Needed to Treat (NNT) – Standard Operating Procedure (SOP)


This SOP is intended to describe the roles of the Association of Scottish Medicines Information Practitioners (ASMIP) in the development and maintenance of the NNTs and to describe a systematic approach to their calculation.

Originally described by Laupacis et al (1988)31 and cited in Cook (1995)32, NNT was introduced as an approach to summarise the effect of treatment in terms of the number of patients a clinician needs to treat with a particular therapy to expect to prevent one adverse event over a specified time period.


The Scottish Government Polypharmacy Guidance 2017 is intended as a practical tool to help prescribers decide when it is appropriate to initiate and continue long-term medicines, especially in the management of long term conditions. In some circumstances, and in consultation with the patient, it may be appropriate to discontinue treatments.  Presentation of numbers needed to treat (NNT) for a range of medicines is one tool that prescribers may use to aid discussions with patients about the likely benefit.

The NNT is defined as the expected number of people who need to receive the experimental rather than the comparator intervention for one additional person to either incur or avoid an event in a given time frame. An NNT of 10 can be interpreted that one additional (or less) person will incur an event for every 10 participants receiving the experimental intervention rather than control over a given time frame.

1. Defining NNTs

1.1 Defining the medicine, intervention and the clinical outcome of relevance

The medicines used in previous editions of the Scottish Government Polypharmacy Guidance 2012 and 2015 should be included. These will be reviewed to ensure that they are both specific and measurable. Consideration also needs to be given to their relevance to clinical practice, e.g. is the medicine likely to be used in this clinical context and is the comparator described the most relevant to clinical practice?


1.2 Identifying relevant medical literature

The following principles should be applied:

  • Cochrane reviews where available should be used
  • Systematic reviews should generally be used in preference to individual randomised controlled clinical trials (RCT), unless the RCT includes a greater number of patients
  • Where systematic reviews are not available individual randomised controlled trials may be used

The MI pharmacist should carry out a standard Medline® or Embase® search using relevant Medical Subject Heading (MeSH) terms and Boolean operators.  In particular the following should be identified:

  • Cochrane systematic reviews
  • Other high quality systematic reviews
  • Pivotal trials for the medicine in the relevant indication

Ideally  studies should be identified from the previous five years, but in exceptional circumstances, e.g. where only a single pivotal trial has been published, or no newer systematic reviews have been published, older clinical trials or systematic reviews may be used.


1.3 Dealing with multiple trials and meta-analyses

Where more than one review or trial is identified for the relevant indication and intervention the following criteria should be assessed:

  • Relevance to the defined medicine and intervention
  • Size of the study or review
  • Similarity of review and study cohort to the Scottish population

A judgement can then be made, using the criteria above to identify the most relevant trial or review from which the NNT can be calculated. Where the studies are very similar, the NNT should be calculated for each individual study and the mean taken for inclusion in the table.

2. Calculating NNTs

The NNT can be calculated from the absolute risk reduction (ARR) taken from a clinical trial or systematic review.  ARR = p1- p2 , where p1 is the baseline or placebo rate and p2 is response rate in the intervention group in a clinical trial. 

The NNT can be calculated as 1/ (p1-p2).33 Where the benefit is accrued over a number of years, the annual NNT can be calculated by multiplying the NNT by the number of years over which the study was conducted.  



3. Recording research

The MiDatabank® project management function should be used to record all research.  The following information should be recorded:

  • The literature search
  • Trials/ reviews identified
  • Absolute risk reduction figures taken from the study(ies)
  • The calculation used to define the NNT
4. Presenting the NNT data

All NNT data should be tabulated to include the following:

  • Intervention - the medicine or other intervention of interest
  • The comparator
  • Outcome - the desired outcome from the proposed treatment
  • NNT - calculated using standard methodology
  • Duration of study/ intervention
  • Demographics of population - age, sex (where relevant), co-morbidities
  • Reference - main reference used to calculate the NNT
5. Referencing

Vancouver style should be used to reference all trials/ reviews used in the calculation of NNTs. Where data has been taken from websites, the web address and the date accessed should be recorded.

6. Checking/ Peer Review

A peer check should be undertaken by another MI or clinical pharmacist prior to publication.  The check should include:

  • Clarity and completeness
  • Any obvious gaps in the information concerning the patient demographics
  • A calculation check for the NNT
Drug Efficacy Chart

31. Laupacis A, Sackett D, Roberts R. An Assessment of Clinically Useful Measures of the Consequences of Treatment. New England Journal of Medicine. 1988;318(26):1728-1733

32. Cook R, Sackett D. The number needed to treat: a clinically useful measure of treatment effect. BMJ. 1995;310(6977):452-454


33. Hutton J. Number needed to treat and number needed to harm are not the best way to report and assess the results of randomised clinical trials. British Journal of Haematology. 2009;146(1):27-30