ClinicSpeak: risk benefit, risk harm analyses

Do we need a new way to communicate benefits and risks? #ClinicSpeak #MSBlog #MSResearch
'Lies, damned lies, and statistics', Mark Twain

"The analysis below looks at the benefit and risk of DMTs from a different statistical perspective, i.e. how many patients do you need to treat with DMT to benefit, or harm, one patient? 

NNTB = number-needed-to-treat to benefit
NNTH = number-needed-to-treat to harm
LHH = likelihood to be helped-or-harmed
LHH = NNTH/NNTB ratio
AAR = annualized relapse rate
PPR-F = proportion of relapse-free patients
PP-F-CDPS3M = proportion of disability progression free patients confirmed or sustained at 3 months

When analysed like this it is remarkable how well subcutaneous interferon-beta-1a (Rebif) does compared to the newer DMTs. Of the newer agents Natalizumab fares the best. Is this a compelling way to present data to people with MS? If I presented data in this way I suspect many of my patients may select interferon-beta as their agent of choice. The problem with this analysis is that it classifies harm very broadly and not all 'harmful events' are necessarily serious or life threatening. For example, raised liver enzyme that are asymptomatic and reversible is very different to PML or a life-threatening infusion reaction. In addition, this type of analysis does not take into account individual patient characteristics (prognostic profile and baseline disease activity) nor does it take into account lifestyle issues (family planning, cosmetic issues, etc.). I may be wrong but I am not sure it will be easy to present the relative risks and benefits of each of these DMTs to patients in this way without getting bogged down in the small print about rare adverse events. However, I may be wrong. What do you think?"


Mendes et al. Benefit-Risk of Therapies for Relapsing-Remitting Multiple Sclerosis: Testing the Number Needed to Treat to Benefit (NNTB), Number Needed to Treat to Harm (NNTH) and the Likelihood to be Helped or Harmed (LHH): A Systematic Review and Meta-Analysis. CNS Drugs. 2016 Aug 18. [Epub]

OBJECTIVE: This study aimed to test the number needed to treat to benefit (NNTB) and to harm (NNTH), and the likelihood to be helped or harmed (LHH) when assessing benefits, risks, and benefit-risk ratios of disease-modifying treatments (DMTs) approved for relapsing-remitting multiple sclerosis (RRMS).

METHODS: In May 2016, we conducted a systematic review using the PubMed and Cochrane Central Register of Controlled Trials databases to identify phase III, randomized controlled trials with a duration of ≥2 years that assessed first-line (dimethyl fumarate [DMF], glatiramer acetate [GA], β-interferons [IFN], and teriflunomide) or second-line (alemtuzumab, fingolimod, and natalizumab) DMTs in patients with RRMS. Meta-analyses were performed to estimate relative risks (RRs) on annualized relapse rate (ARR), proportion of relapse-free patients (PPR-F), disability progression (PP-F-CDPS3M), and safety outcomes. NNTB and NNTH values were calculated applying RRs to control event rates. LHH was calculated as NNTH/NNTB ratio.

RESULTS: The lowest NNTBs on ARR, PPR-F, and PP-F-CDPS3M were found with IFN-β-1a-SC (NNTB 3, 95 % CI 2-4; NNTB 7, 95 % CI 4-18; NNTB 4, 95 % CI 3-7, respectively) and natalizumab (NNTB 2, 95 % CI 2-3; NNTB 4, 95 % CI 3-6; NNTB 9, 95 % CI 6-19, respectively). The lowest NNTH on adverse events leading to treatment discontinuation was found with IFN-β-1b (NNTH 14, 95 % 2-426) versus placebo; a protective effect was noted with alemtuzumab versus IFN-β-1a-SC (NNTB 22, 95 % 17-41). LHHs >1 were more frequent with IFN-β-1a-SC and natalizumab.

CONCLUSIONS: These metrics may be valuable for benefit-risk assessments, as they reflect baseline risks and are easily interpreted. Before making treatment decisions, clinicians must acknowledge that a higher RR reduction with drug A as compared with drug B (versus a common comparator in trial A and trial B, respectively) does not necessarily mean that the number of patients needed to be treated for one patient to encounter one additional outcome of interest over a defined period of time is lower with drug A than with drug B. Overall, IFN-β-1a-SC and natalizumab seem to have the most favourable benefit-risk ratios among first- and second-line DMTs, respectively.

CoI: multiple

Labels: , , , , ,