YugabyteDB has a feature for strongly consistent secondary indexes, but modifying existing indexes is not currently possible through the Yugabyte CQL API. However, there are ways to achieve this without taking downtime or restarting applications. One approach is to drop and recreate the index with the desired columns, which can take time and potentially affect application performance. Another approach is to create a new index first, including additional columns using the INCLUDE clause, and then dropping the old index once it has backfilled. This second approach avoids potential issues with disk space usage and write penalties, but still requires testing in a lower environment before production deployment.