The goal was to arrive at a workable number of indicators, based on the recommendations most relevant to improving the quality of care. An iterative consensus procedure was followed for this. The
First step in developing the set of indicators was to select those recommendations that were based on the highest level of scientific evidence. This meant that only recommendations supported by the results of a systematic review, or by the results of at least two independently performed randomized clinical trials with sufficient sample size, were selected. Excluding four recommendations that related to the timing of a specific intervention, the number of selected recommendations was reduced to 38. These 38 recommendations were then reformulated into a preliminary set of indicators.