Dealing with missing data in terms of the dates | Adding a single imputation of quarters if only the quarters are missing. Or impute quarters (or directly reporting delay) in the general imputation model Possible ways to do it: - imputing the outcome (so imputing the reporting delay) in the reporting delay model. - in jomo constraint imputation, or impute RD in a scale that guarantees that is will be positive on the original scale - Poisson model in MICE Adding an error message for the reporting delay if the dates necessary for the delay calculation (notification and /or diagnosis dates) are completely missing or missing above a threshold (of 50%).
Dealing with missing data in terms of the dates | Adding a single imputation of quarters if only the quarters are missing. Or impute quarters (or directly reporting delay) in the general imputation model Possible ways to do it: - imputing the outcome (so imputing the reporting delay) in the reporting delay model. - in jomo constraint imputation, or impute RD in a scale that guarantees that is will be positive on the original scale - Poisson model in MICE Adding an error message for the reporting delay if the dates necessary for the delay calculation (notification and /or diagnosis dates) are completely missing or missing above a threshold (of 50%).