We maximally protected participants as all data were fully anonymized before accessing them: a) data in the study database did not include identifying variables, such as names or addresses, b) the study database was encoded by staff of the routine programme.