Test of equivalence for five groups.
##
## Kruskal-Wallis rank sum test
##
## data: siglec by cirrhosis_status
## Kruskal-Wallis chi-squared = 15.327, df = 4, p-value = 0.004068
If we group Healthy Controls and Sepsis together.
##
## Wilcoxon rank sum test with continuity correction
##
## data: siglec by group
## W = 1508.5, p-value = 0.0001088
## alternative hypothesis: true location shift is not equal to 0
Sick patient only.
## Call:
## coxph(formula = Surv(fucensor_days, death_90_admission) ~ siglec_bin,
## data = subset(dat3, cirrhosis_status != "Healthy Controls"))
##
##
## coef exp(coef) se(coef) z p
## siglec_bin 0.741 2.098 0.281 2.64 0.0083
##
## Likelihood ratio test=7.29 on 1 df, p=0.00693
## n= 128, number of events= 55
All patient. I added this in case reviewer request it.
## Call:
## coxph(formula = Surv(fucensor_days, death_90_admission) ~ siglec_bin,
## data = dat3)
##
##
## coef exp(coef) se(coef) z p
## siglec_bin 0.808 2.242 0.281 2.88 0.004
##
## Likelihood ratio test=8.67 on 1 df, p=0.00323
## n= 134, number of events= 55
Corrlation between siglec (original scale) and AST, ALT, APRI (after log-transformation).
## siglec log.ast log.alt log.apri
## siglec 1.0000000 0.3438546 0.2895265 0.5079886
## log.ast 0.3438546 1.0000000 0.8524903 0.8282806
## log.alt 0.2895265 0.8524903 1.0000000 0.7003898
## log.apri 0.5079886 0.8282806 0.7003898 1.0000000