2024‐02‐01 - aasiyahrashan/PhD-Aasiyah GitHub Wiki

Meeting with Abi and Matt about the prolonged MV model

• Actions are on the document. https://docs.google.com/document/d/1L6F5gK6I0ihrMBOutkTZxfVCGbzy1CWSHZ4Apx8hSds/edit#heading=h.2owqv0fyypeq
• Remove APACHE II from the model - there's no real justification for including it, and it causes collinearity.
• Instead, include SOFA or its components. No real substitute for platelets though.
• Use Charlson for comorbidities
• Split tables so there's one with all possible variables in the appendix
• Add histograms for distribution of variables
• Write out justifications for everything in the methods section
• Use complete case analysis - if have time, do normal imputation