The presence of information on the selected indicators (e.g.,BMI) was defined as an indicator of good quality, and treated as the primary outcome in the statistical analysis. The effect of participation in KOLIBRI was measured in terms of change in presence of information from the 1st to the 2nd audit on the selected indicators. Since the data in the analysis came from hospital (strata) of different sizes, we estimated overall change in presence of information by weighted estimation, i.e., each observation being weighted by the inverse of its sampling probability.