Sensitive and Personalized Determinations of Likelihood of Being Helped or Harmed

This work may not be copied, distributed, displayed, published, reproduced, transmitted, modified, posted, sold, licensed, or used for commercial purposes. By downloading this file, you are agreeing to the publisher’s Terms & Conditions.

See article by Andrade

Sensitive and Personalized Determinations of Likelihood of Being Helped or Harmed

To the Editor: We read Dr Andrade’s recent article on likelihood of being helped or harmed (LHH) with interest.1 In our view, this work raised important issues related to the need to make adequately sensitive and personalized number needed to harm (NNH) (thus LHH) determinations, as was originally intended by Dr Sharon Straus when she originated the idea in her evidence-based medicine classes some years ago.2,3 There are many considerations regarding proximal and distal benefits and harms.4

Reasons for dropping out (or remaining in) clinical trials are complex, making the selection for individual patients of the optimal NNH for calculation of LHH complex. This is true particularly if the selected harm involves trial discontinuation (eg, all-cause discontinuation or adverse effect discontinuation), as patients may be more reluctant to discontinue participation in a clinical trial than to discontinue a clinically administered medication. This could undermine the sensitivities of harms involving trial discontinuation.

Spontaneously reported adverse events for common problems such as sedation, weight gain, or akathisia (eg, for second-generation antipsychotics) may be more sensitive harms but may be too sensitive (ie, reflecting harms with insufficient severity to be actionable) or, once again, susceptible to complex causes. Adverse events leading to discontinuation may address the former sensitivity (but not the latter complexity) challenge. Adverse events thresholded for potential clinical significance (eg, ≥ 7% weight gain) may address both challenges, but may not be relevant for all patients. Thus, it appears to these authors that it is particularly important to make adequately sensitive and personalized NNH (thus LHH) determinations when using these tools.5

References

1. Andrade C. Likelihood of being helped or harmed as a measure of clinical outcomes in psychopharmacology. J Clin Psychiatry. 2017;78(1):e73-e75. PubMed doi:10.4088/JCP.16f11380

2. McAlister FA, Straus SE, Guyatt GH, et al; Evidence-Based Medicine Working Group. Users’ guides to the medical literature, XX: integrating research evidence with the care of the individual patient. JAMA. 2000;283(21):2829-2836. PubMed doi:10.1001/jama.283.21.2829

3. Straus SE. Individualizing treatment decisions: the likelihood of being helped or harmed. Eval Health Prof. 2002;25(2):210-224. PubMed

4. Citrome L, Kantrowitz J. Antipsychotics for the treatment of schizophrenia: likelihood to be helped or harmed, understanding proximal and distal benefits and risks. Expert Rev Neurother. 2008;8(7):1079-1091. PubMed doi:10.1586/14737175.8.7.1079

5. Citrome L, Ketter TA. When does a difference make a difference? interpretation of number needed to treat, number needed to harm, and likelihood to be helped or harmed. Int J Clin Pract. 2013;67(5):407-411. PubMed doi:10.1111/ijcp.12142

Terence A. Ketter, MDa

tketter@stanford.edu

Leslie Citrome, MD, MPHb

aDepartment of Psychiatry and Behavioral Sciences, Stanford University School of Medicine, Stanford, California

bDepartment of Psychiatry and Behavioral Sciences, New York Medical College, Valhalla, New York

Potential conflicts of interest: None.

Funding/support: None.

J Clin Psychiatry 2017;78(5):e554

https://doi.org/10.4088/JCP.17lr11508

© Copyright 2017 Physicians Postgraduate Press, Inc.