Article Text

other Versions

Download PDFPDF
Selection of the primary end point in an observational cohort study
  1. Guadalupe Gómez1,
  2. Oleguer Plana-Ripoll2,
  3. Urania Dafni3,4
  1. 1Departament d'Estadística i Investigació Operativa, Universitat Politècnica de Catalunya, Barcelona, Spain
  2. 2Section of Epidemiology, Department of Public Health, Aarhus University, Aarhus, Denmark
  3. 3Frontier Science Foundation-Hellas, Athens, Greece
  4. 4Laboratory of Biostatistics, Department of Nursing, School of Health Sciences National and Kapodistrian University of Athens, Athens, Greece
  1. Correspondence to Dr Guadalupe Gómez Melis, Departament d'Estadística i Investigació Operativa, Universitat Politècnica de Catalunya, Carrer Jordi Girona 1-3, Barcelona 08034, Spain; lupe.gomez{at}

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.


Any study exploring specific research questions requires to test well-defined hypotheses that efficiently translate these questions into measurable quantities. To attribute differences in outcomes between two or more groups to their respective different exposure levels, the flagship comparative approach is the randomised clinical trial (RCT).

An RCT is not always feasible, as is the case if the exposure of interest cannot be reasonably randomised, such as smoking, age, family or personal history of a disease. A cohort study is the next most rigorous approach to answer cause and effect questions, sharing with an RCT the advantage of prospective follow-up. In addition, observational studies are more suitable to answer certain important questions for an intervention such as detecting rare or late adverse effects of treatments.1

A comparative study that lacks the randomisation component should not also lag behind in other rigorous requirements that are customarily expected of randomised studies. The value of the scientific evidence produced by an observational study lies equally heavily on how explicit its design characteristics are in the protocol, and how faithfully the study implementation followed the study protocol design. Guidelines to that effect are available and should be respected.2

One of the most crucial parameters in a study is the choice of the end point that would best translate the objective and capture the effect of interest. It is often the case that several relevant end points are of comparable importance and it might be difficult to select the most appropriate one. In those situations, the union of several end points, a composite end point (CE), is used as the primary end point (PE) of interest. A key advantage of a CE is that it provides a better description of a disease process.3 For example, in the cardiovascular literature, efficacy of interventions is …

View Full Text


  • Contributors GG is responsible for the overall content of the article, as well as for submitting and acting as the correspondent. GG, OP-R and UD are equal contributors to the planning, conduct, and reporting of the work described in the article, as well as to the manuscript writing.

  • Funding The study sponsors have not had any involvement in the study design; in the collection, analysis and interpretation of the data; in the writing of the report; and in the decision to submit the paper for publication.

  • Competing interests None declared.

  • Provenance and peer review Commissioned; externally peer reviewed.