Disparities, variations, inequalities or inequities: whatever you call them, we need data to monitor them.

Health inequalities are a problem in high, middle and low income countries. Most are unfair ('inequities') and could be minimised but primarily through policies outside the health service.In the US, the Center for Diseases Control has used high quality, nationally-available data to monitor conditions and determinants of health among different groups (by sex, disability, race, ethnicity, and language) to motivate action to reduce inequalities. In the UK, the 10 top level 'health' indicators in London at the turn of the millennium included unemployment, education, housing quality, crime, air pollution, road travel injuries, as well as traditional health measures. Most of these affect mental and physical health through social determinants or adverse environmental exposures. Current inequalities monitoring in England includes a Local Basket of Inequalities Indicators focusing on a wide range of determinants of health as well as traditional health metrics.Israel, like the US, has above average socio-economic inequalities but has universal healthcare. Health inequalities in Israel occur within different Jewish groups and by income, education, ethnicity, and religion, with disadvantages often clustering. Current monitoring in Israel focuses on health outcomes and 'midstream' healthcare-related provision. I agree with Abu-Saad and her colleagues that including monitoring of social determinants of health is crucial to identify and tackle health inequalities in Israel.National, 'upstream', interventions are the most effective ways to reduce inequalities and improve the population's health. High-level political support is crucial for this. While a 'Health in all Policies' approach combined with political will to 'leave no one behind' can lead to great improvements, regular monitoring is essential, to: identify the inequities; plan appropriate and effective, targeted interventions; implement and evaluate them; and change them where needed. All of this requires adequate and timely data on health and its determinants, including information about undiagnosed and poorly controlled disease, obtained from the general population not just those attending for healthcare, analysed for each population sub-group at risk of experiencing inequalities.This is a commentary on https://doi.org/10.1186/s13584-018-0208-1.