We need to quantify our outcome labels for out-of-hospital mortality. As discussed in the weekly meeting (2019-06-17), we will make predictions at the end of every admission event. Thus, the outcome label will be binary mortality at discrete time points (e.g. 72hrs, 1 wk, 1 month, 6 months, 1 year). This data can be found in the PATIENTS table under the dod_ssn field.
The result should be an m-by-n matrix with m admissions and n rows corresponding to the mortality at these timepoints.
Additionally, this should be repeated with the inclusion of hospital re-admission.
Recall that patients who died in the hospital should be omitted from the out-of-hospital mortality cohort.
We need to quantify our outcome labels for out-of-hospital mortality. As discussed in the weekly meeting (2019-06-17), we will make predictions at the end of every admission event. Thus, the outcome label will be binary mortality at discrete time points (e.g. 72hrs, 1 wk, 1 month, 6 months, 1 year). This data can be found in the
PATIENTStable under thedod_ssnfield.The result should be an m-by-n matrix with m admissions and n rows corresponding to the mortality at these timepoints.
Additionally, this should be repeated with the inclusion of hospital re-admission.
Recall that patients who died in the hospital should be omitted from the out-of-hospital mortality cohort.