Letters HRT and cardiovascular events

Was the composite outcome specified in the original protocol?

BMJ 2012; 345 doi: http://dx.doi.org/10.1136/bmj.e8144 (Published 03 December 2012) Cite this as: BMJ 2012;345:e8144
  1. Jeppe Schroll, researcher1,
  2. Andreas Lundh, researcher1
  1. 1Nordic Cochrane Centre, 2200 Copenhagen N, Denmark
  1. js{at}cochrane.dk

It is unclear whether Schierbeck and colleagues’ primary composite outcome was defined in the original protocol.1 In other words, is their analysis exploratory or confirmatory? The main question is whether the results should lead to changes in clinical practice. Schierbeck has advocated for the safety of hormone therapy,2, but we believe that the study results should be interpreted with caution and put in the context of previous findings.

In their response Schierbeck and colleagues state that “the outcomes in the present study were secondary outcome measures in the original protocol” and “the primary endpoint of the present study was defined in the protocol as safety measures for cardiovascular events.”3

A previous publication on the design of the Danish Osteoporosis Prevention Study mentions coronary heart disease (defined as angina and myocardial infarction) but not the composite outcome.4 The published trial of fracture outcomes uses the term “cardiovascular incidents” but does not describe what this means.5 Nothing is stated on clinicaltrials.gov about cardiovascular outcomes. The protocol we received from Schierbeck (1995 version, patients enrolled 1990-93) states that myocardial infarction would be assessed as a side effect but not specified as a primary or secondary outcome in the data analysis. The statistical analysis section states that “incidence of ischaemic heart disease” will be evaluated, without describing what this means.

We found no information on the outcome, admission to hospital for heart failure, or the composite outcome in previous publications or the provided protocol. Can the authors please clarify whether the composite outcome of death, admission to hospital for heart failure, and myocardial infarction was defined before the trial started enrolling in 1990 or whether it was defined afterwards?


Cite this as: BMJ 2012;345:e8144


  • Competing interests: None declared.