[Correspondence] Cancer survival rates: the CONCORD-2 study – Authors’ reply
Ruffo Freitas-Junior and colleagues ask if the data from Goiânia, Brazil, are reliable. We paid obsessive attention to documenting the quality of data and survival estimates. The estimates for Goiânia were flagged as less reliable than would be preferred for breast cancer and for eight of the other nine cancers examined in adults (see p 84 and p 99 in the appendix of our Article1). We agree that the flagged estimates for Goiânia are not optimal; improvement in data quality is in progress. Data from flagged survival estimates were not included in pooled national estimates.