proportional hazards 에 대한 supremum test ?
-
게시물 수정 , 삭제는 로그인 필요
mortality에 대한 논문을 읽던 중 진단과 사망 사이 기간에 대한 분석을 위해 cox proportional hazard model 을 사용한다는 부분을 읽게 되었습니다. 해당 모델에 대한 이해도가 부족해서 도움을 요청드립니다.
1) hazard assumption 3.5 -> 어떤 의미를 가지고 있는 것인지?
2) proportioanl hazard assumption이 hold 되지 못했다는 게 무슨 뜻인지?
(해당 모델을 적용하는 것이 부적절하다는 뜻인지? -> 그래서 5 yr fu 으로 대체하겠다는 것인지?)
궁금합니다. 읽어주셔서 감사합니다.
The time to event data was analysed using the Cox proportional hazard models and the results are pre- sented as hazard ratios (HRs) and their 95% confidence intervals. We used the PHREG procedure to estimate the HRs. The propor- tional hazards assumption was tested using the SAS ‘assess’ state- ment within the PHREG procedure, which performs graphical and quantitative methods for checking the adequacy of the model.21 Based on this model, the maximum absolute value from the supremum test for proportional hazards assumption was 3.5, and the test had a statistically significant P-value of <0.001, meaning that the proportional hazards assumption did not hold. We observed a statistically significant quantitative (non-crossover) interaction between exposure (non-psychotic depression versus psychotic depression) and time until death. Because the propor- tional hazards assumption was violated, we present some of the results stratified by follow-up time and our main analyses are based on the 5-year follow-up period. Causes of death were included as competing risks in the mortality models. This resulted in more conservative effect estimates compared with models where compet- ing causes of death were not censored. All individuals remained in the at-risk population for mortality until death or end of follow-up period, whichever came first, i.e. potential subsequent change of diagnosis did not change their at-risk status.
#proportional hazards assumption #proportional hazards tests and diagnostics based on weighted residuals #proportional hazards model #proportional hazards assumption test #proportional hazards regression analysis #proportional hazards assumption in r #proportional hazards model python #proportional hazards assumption kaplan meier #proportional hazards assumption not met #proportional hazards assumption explained