生存时间的cutoff天数是2000天，如何得出time-varying interaction term 的CI？在网上查了要用hazardratio command 但是出来的interaction 的HR怎么不对

以下是code：
proc surveyphreg data=final;
class smoking_status(ref='0') household_income(ref='0') cultural_background(ref='0') education(ref='0') multimorbidity_index(ref='0') Sex(ref='0') BMI(ref='0') alcohol(ref='0') stress(ref='0') martial_status(ref='0') provinces(ref='1') phs_act(ref='0') asthma(ref='0') arthritis(ref='0') blood_pressure(ref='0') diabetes(ref='0') heart_disease(ref='0') cancer(ref='0') stroke(ref='0') mood_disorder(ref='0') copd(ref='0');
model survitime*outcome(0)=smoking_status household_income cultural_background education alcohol BMI stress martial_status provinces phs_act multimorbidity_index multi1/rl;
multi1=multimorbidity_index*(survitime>2000);
hazardratio multimorbidity_index;
hazardratio multi1;
weight wts_sdle;
run;

[url=][/url]

谢谢指点！