APLIKASI MODEL HAZARD ADDITIVE LIN & YING SEBAGAI PENDUGA KEKAMBUHAN PADA PENDERITA ENDOMETRIOSIS

Survival analysis is a statistical procedure related to the time from the start to the time of occurrence of the event. Survival analysis is usually used to analyze risk factors associated with clinical events such as patient survival time, recovery time, relapse time or death at a certain time. There are two types of hazard regression models that can be used, namely the multiplicative hazard model and the additive hazard model. The multiplicative hazard model is a cox proportional hazard model, with a proportional hazard assumption test using the graph method or Goodness Of Fit. If a cox model does not meet the proportional assumptions, then the conclusions in the cox proportional hazard model can potentially be biased and the alternative can use an additive hazard model. The advantages of using the additive hazard model are that there are no proportional assumptions and the Lin & Ying additive hazard model uses a method similar to the maximum partial likelihood as in the Cox model where the regression coefficients can be searched directly so that it is easier to interpret. This study uses recurrence data experienced by endometriosis patients who have undergone surgical treatment. There are several factors that can influence recurrence, namely fruit consumption, vegetable consumption, gluten consumption, soy consumption, milk consumption, fat consumption and cheese consumption. Based on these variables, analysis was carried out using the Lin & Ying additive semiparametric hazard method. The results and discussion showed that the variable consumption of milk consumed 4-7 times/week and the variable consumption of cheese consumed 1-4 times/week were factors that significantly influenced the relapse time of endometriosis patients. Keywords: Endometriosis, Lin & Ying, Additive hazard.