Serum indoxyl sulfate is associated with mortality in hospital-acquired acute kidney injury: a prospective cohort study

Table 4 Multivariate Cox proportional hazard model of mortality during 90-day follow-up in AKI patients (serum creatinine entered as a continuous or dichotomous variable)

Model	HR	95% CI	P
Serum creatinine (continuous variable)
Unadjusted	1.00	1.00–1.00	0.002
Model 1	1.00	1.00–1.00	0.002
Model 2	1.00	1.00–1.00	0.696
Model 3	1.00	1.00–1.00	0.597
Model 4	1.00	1.00–1.00	0.924
Model 5	1.00	1.00–1.00	0.360
Serum creatinine (dichotomous variable)
Unadjusted	2.02	1.32–3.07	0.001
Model 1	1.96	1.28–2.99	0.002
Model 2	1.07	0.68–1.68	0.762
Model 3	0.87	0.54–1.42	0.583
Model 4	1.14	0.69–1.88	0.608
Model 5	0.93	0.55–1.57	0.791

Variables of model 1 include age and gender
Variables of model 2 include age, gender, non-renal APACHE II score, sepsis, RRT, and surgery
Variables of model 3 include age, gender, non-renal APACHE II score, sepsis, RRT, surgery, serum albumin, IS, BUN and β₂-microglobulin
Variables of model 4 include age, gender, non-renal APACHE II score, sepsis, RRT, surgery, serum albumin, IS, BUN, β₂-microglobulin, ALT, neutrophilic granulocyte and platelet counts
Variables of model 5 include age, gender, non-renal APACHE II score, sepsis, RRT, surgery, serum albumin, IS, BUN, β2-microglobulin, ALT, neutrophilic granulocyte, platelet counts and interaction terms (gender*platelet counts, surgery*serum albumin, surgery*creatinine, and ALT*platelet counts)

ISSN: 1471-2369