Evidence-based recommendations or Show me the patients selected and I will tell you the results

Modern treatment decision making in medicine is recommended to be evidence based. In order to have a high grade of evidence, the studies should have sufficient power, be randomized and double blinded. Moreover, the evaluation parameters should be solid and reproducible. While there is a lot focus on primary study design much less is focused on patient eligibility and to what extent the patients included in a clinical trial are representative for the patients treated in real life. By knowing the profile and action of the drug it is thus often possible to design inclusion criteria's that already from the start, with high probability, predict the results. If you wish to prove that the addition of long-acting beta-2 agonist is better that increasing the anti-inflammatory treatment you should select mild stable and highly reversible patients. On the other hand, if you wish to prove that increasing anti-inflammatory treatment is beneficial, you choose patients proven to be steroid responsive and slightly under treated. Applying common inclusion and exclusion criteria's often render only a few percent of the patient population eligible. This is often forgot, when the results from these strictly selected patient populations are extrapolated into evidence based treatment recommendations directed towards a much larger and less selected patient population. Thus when evidence are graded, it is important also to consider to what extent the results are extendable to a much larger real life patient population. Modern asthma management must consider pathophysiological mechanisms that not necessarily are reflected by lung function parameters. Demands from medical authorities as the medical authorities in Europe (EMEA) asking for 15% reversibility of asthma study patients, helps to conserve the imbalance between clinical trials and real life. (c) 2006 Published by Elsevier Ltd.