☀️     🌓

Prescribing Advice for GPs

An NHS Prescribing Advisers' Blog

Post hoc analysis of TNT

The Lancet has made a post hoc analysis of the Treating to New Targets (TNT) study available as an early online publication.

This analysis examines the effects that aggressive lipid lowering therapy with atorvastatin 80mg compared to regular lipid lowering with atorvastatin 10mg had in patients with metabolic syndrome. Metabolic syndrome was defined by the NCEP ATP III criteria. The analysis included 5,584 of the original 10,001 patients recruited to the TNT study.

Efficacy assessment was still based upon the primary outcome measure of a major cardiovascular event, defined as death from CHD, non-fatal MI, resuscitated cardiac arrest and fatal or non-fatal stroke.

Over the average follow up of 4.9 years 367 events occurred in the 2,820 patients assigned to the atorvastatin 10mg arm and 262 events occurred in the 2,764 patients assigned to the atorvastatin 80mg arm. This equates to an absolute risk reduction of 3.5% and a Number Needed to Treat (NNT) of 28 over 4.9 years.

Throughout this analysis no mention is made of adverse events. In the original study, patients in the 80mg arm were more statistically likely to experience a treatment-related adverse event (8.1% vs. 5.8%). Persistent liver enzyme elevation (1.2% vs. 0.2%) and treatment discontinuation (7.2% vs. 5.3%) were also more likely.

The original study reported an NNT of 45 to prevent one event over 4.9 years but against this there is a Number Needed to Harm (NNH) of 43 for an adverse event and NNH of 53 for a discontinuation. Aggressive treatment will only work if it can be tolerated and therefore taken by patients.

Action: Considering the limitations of post hoc analyses and the additional adverse events experienced by patients in the aggressive arm of the study it is inappropriate to change practice based on these findings. The study may formulate a hypothesis about targeting aggressive treatment to those at highest risk that should be tested under controlled trial conditions.

Share 'Post hoc analysis of TNT' by emailShare 'Post hoc analysis of TNT' on FacebookShare 'Post hoc analysis of TNT' on TwitterShare 'Post hoc analysis of TNT' on LinkedInShare 'Post hoc analysis of TNT' on reddit


No Comments to “Post hoc analysis of TNT”

Leave a Comment

Your email address will not be published. Required fields are marked *

Please be aware that you comment is subject to our Privacy Policy.

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Prescribing Advice for GPs is powered by ClassicPress.
Connect to our RSS or Atom Feeds.