Article Text


SAT0229 Number needed to treat (nnt): implication in rheumatology clinical practice
  1. M Osiri1,
  2. ME Suarez-Almazor2,
  3. GA Wells3,
  4. P Tugwell3
  1. 1Department of Medicine, Chulalongkorn University Hospital, Bangkok, Thailand
  2. 2Health Services Research, Baylor College of Medicine, Houston, USA
  3. 3Department of Medicine, University of Ottawa, Ottawa, Canada


Background A major challenge for rheumatologists is to translate the research data into treatment decision for their patients in clinical practice. One approach that is becoming increasingly used is the ?number needed to treat? (NNT), the reciprocal of absolute risk reduction, which is the number of patients who must be treated in order to obtain the benefit of interest in one new patient.

Objectives To calculate the NNT from rheumatology systematic reviews and clinical trials.

Methods We calculated the NNT from clinically important outcomes in rheumatology from the Cochrane Database of Systematic Reviews in the Cochrane Library Issue 2, 2000 or from the absolute risk differences of the primary outcomes from the original RCTs. If these outcomes were not available, the NNT was calculated from mean improvement of tender joint counts. Key words included musculoskeletal, rheumatology, and systematic reviews.

Results 15 systematic reviews and 11 original articles contributed the data for NNT calculation. The smallest NNT for the patients meeting the ACR20 criteria was from etanercept treatment compared to placebo (NNT 1.8;95% CI 1.1,4.7). The NNT and 95% CI for the patients treated with MTX, SSZ and leflunomide was 5.0 (3.3,11), 3.6 (2.5,6.7), and 3.9 (2.7,6.6), respectively. The NNT for combination treatment (cyclosporin + MTX, infliximab + MTX, and prednisolone + MTX + SSZ) ranged from 2.6 to 4.8. The NNT for azathioprine (2.4; 95% CI 1.8, 8.9) and high dose D-penicillamine (5.2; 95% CI 2.8,34.2) was calculated from mean change of the number of tender joints. The NNT for meeting the Paulus criteria in antimalarial-treated patients was 4.6 (95% CI 2.5,25.6) and in injectable gold group was 4.0 (2.3,14.3). Among different NSAIDs used in hip OA, only etodolac (NNT 4.4; 95% CI 2.4,24.4) and tenoxicam (NNT 3.8, 95% CI 2.5,7.3) showed a significant NNT and 95% CI for relieving pain over placebo.

Conclusion The NNT is a term translated from the less understandable research data to help clinicians in routine practice decision-making. Comparisons of the NNTs can be done among different treatment options for the same disease and outcome. The NNT should be adjusted for the baseline risk and treatment duration in individual patients.

Statistics from

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.