Table 4.
Variable | Univariable analysis | Multivariable analysis* | ||
---|---|---|---|---|
sHR (95% CI) | p | sHR (95% CI) | p | |
Age in years | 1.03 (1.00–1.06) | 0.092 | 1.02 (0.99–1.06) | 0.17 |
Male gender | 0.95 (0.49–1.87) | 0.89 | ||
Diabetes mellitus | 1.18 (0.59–2.36) | 0.64 | ||
Chronic renal failure | 0.79 (0.32–1.99) | 0.62 | ||
Severe hepatic failure | 0.70 (0.17–2.97) | 0.63 | ||
Solid neoplasms | 1.28 (0.58–2.82) | 0.54 | ||
Hematological malignancies | 0.71 (0.26–1.91) | 0.49 | ||
Causative agent of BSI | 0.67 | |||
Enterobacteriaceae | 1 (Reference) | |||
Pseudomonas spp. | 1.15 (0.43–3.05) | |||
Acinetobacter spp. | 1.72 (0.82–3.62) | |||
Enterobacteriaceae plus Pseudomonas spp. | 1.80 (0.37–8.91) | |||
Enterobacteriaceae plus Acinetobacter spp. | 1.66 (0.27–10.36) | |||
Colistin susceptibility test method | 0.13 | |||
Vitek 2 | 1 (Reference) | |||
Etest | 1.69 (0.85–3.35) | |||
Polymicrobial | 0.55 (0.21–1.43) | 0.22 | ||
ICU stay | 1.63 (0.83–3.2) | 0.16 | ||
Presence of central venous catheter | 0.47 (0.23–0.95) | 0.035 | 0.56 (0.26–1.23) | 0.15 |
Presence of septic shock | 1.17 (0.59–2.29) | 0.66 | ||
Neutropenia | 0.28 (0.04–1.93) | 0.2 | ||
Serum hemoglobin in g/dl | 1.05 (0.83–1.35) | 0.67 | ||
Serum total bilirubin in mg/dl | 0.97 (0.90–1.05) | 0.48 | ||
Serum creatinine in mg/dl | 0.73 (0.38–1.42) | 0.35 | ||
Serum albumin < 2.5 g/dl | 2.55 (1.27–5.15) | 0.009 | 2.37 (1.15–4.87) | 0.019 |
Type of adequate therapy | 0.69 | |||
Colistin monotherapy | 1 (Reference) | |||
2 active agents including colistin | 0.75 (0.38–1.51) | |||
3 active agents including colistin | 0.90 (0.26–4.72) | |||
Time to adequate therapy in days | 1.02 (0.91–1.15) | 0.7 | ||
9 MU colistimethate loading-dose | 1.82 (0.92–3.6) | 0.086 | 1.83 (0.88–3.82) | 0.11 |
Use of other nephrotoxic agents | 0.87 (0.20–3.77) | 0.85 |
KDIGO, Kidney Disease: Improving Global Outcomes31; sHR, subdistribution hazard ratio; CI, confidence intervals; BSI, bloodstream infection; ICU, intensive care unit; MU, million units.
*Only results for variables retained in the final multivariable model are presented.