Table 2.
Four Models Describing the Association of AKI Severity with Postoperative Outcome
| Mortality
|
Morbidity
|
|||||||
|---|---|---|---|---|---|---|---|---|
| In-hospital Death
|
Survival*
|
Hospital Length of Stay†
|
Discharge‡
|
|||||
| P Value | OR (95% CI) | P Value | HR (95% CI)§ | P Value | β (95% CI) | P Value | HR (95% CI)|| | |
| Constant# | <0.001 | — | — | — | <0.001 | 9.75 (9.21–10.30) | — | — |
| Age, yr | 0.002 | 0.99 (0.99 to 1.00) | 0.39 | 1.00 (0.99 to 1.01) | 0.30 | −0.00 (−0.01 to 0.00) | <0.001 | 1.00 (1.00 to 1.00) |
| Sex, female vs. male | 0.19 | 0.90 (0.77 to 1.05) | 0.67 | 0.97 (0.84 to 1.12) | 0.08 | 0.09 (−0.01 to 0.20) | 0.27 | 0.99 (0.97 to 1.01) |
| Charlson Comorbidity Index | <0.001 | 1.18 (1.16 to 1.21) | <0.001 | 1.13 (1.11 to 1.15) | <0.001 | 0.17 (0.16 to 0.19) | <0.001 | 0.96 (0.96 to 0.96) |
| Congestive heart failure | <0.001 | 2.15 (1.77 to 2.61) | <0.001 | 1.61 (1.36 to 1.89) | <0.001 | 1.12 (0.91 to 1.33) | <0.001 | 0.85 (0.81 to 0.89) |
| Preoperative hemoglobin, mg/dl | <0.001 | 0.79 (0.76 to 0.86) | <0.001 | 0.87 (0.84 to 0.90) | <0.001 | −0.38 (−0.40 to −0.35) | <0.001 | 1.09 (1.08 to 1.10) |
| Preoperative creatinine, mg/dl** | 0.08 | 0.90 (0.80 to 1.01) | 0.55 | 0.97 (0.88 to 1.07) | <0.001 | −1.11 (−1.24 to −0.98) | <0.001 | 1.20 (1.17 to 1.24) |
| Radiocontrast agent†† | <0.001 | 1.90 (1.61 to 2.25) | <0.001 | 1.48 (1.28 to 1.72) | <0.001 | 0.72 (0.55 to 0.89) | <0.001 | 0.87 (0.84 to 0.90) |
| Noncardiac vs. cardiac surgery | <0.001 | 1.94 (1.48 to 2.55) | 0.002 | 1.43 (1.14 to 1.78) | <0.001 | 2.27 (1.99 to 2.54) | 0.003 | 0.92 (0.87 to 0.97) |
| Risk factors of surgical AKI‡‡ | ||||||||
| 1 risk factor | 0.93 | 1.01 (0.83 to 1.23) | 0.69 | 1.04 (0.97 to 1.49) | 0.86 | −0.01 (−0.13 to 0.11) | 0.14 | 1.02 (0.99 to 1.05) |
| 2 risk factors | 0.09 | 1.23 (0.97 to 1.58) | 0.10 | 1.20 (0.97 to 1.49) | <0.001 | 1.04 (0.79 to 1.28) | 0.15 | 0.96 (0.92 to 1.01) |
| ≥ 3 risk factors | <0.001 | 3.27 (2.28 to 4.69) | <0.001 | 1.98 (1.49 to 2.63) | <0.001 | 4.25 (3.55 to 4.96) | <0.001 | 0.65 (0.56 to 0.74) |
| Postoperative ICU admission | <0.001 | 2.54 (2.09 to 3.09) | 0.004 | 1.31 (1.09 to 1.58) | <0.001 | 4.83 (4.69 to 4.98) | <0.001 | 0.57 (0.56 to 0.59) |
| Severity of AKI§§ | ||||||||
| No F/U | 0.001 | 0.60 (0.45 to 0.81) | 0.35 | 0.87 (0.65 to 1.16) | <0.001 | −2.27 (−2.42 to −2.13) | <0.001 | 1.70 (1.65 to 1.75) |
| ΔCr 1–24% | 0.49 | 0.92 (0.71 to 1.17) | 0.19 | 0.85 (0.67 to 1.08) | 0.001 | 0.28 (0.11 to 0.45) | 0.81 | 1.00 (0.97 to 1.04) |
| ΔCr 25–49% | <0.001 | 1.74 (1.28 to 2.38) | 0.04 | 1.36 (1.02 to 1.82) | <0.001 | 2.44 (2.12 to 2.75) | <0.001 | 0.80 (0.75 to 0.86) |
| KDIGO stage 1 | <0.001 | 2.21 (1.69 to 2.87) | <0.001 | 1.55 (1.22 to 1.98) | <0.001 | 2.88 (2.58 to 3.17) | <0.001 | 0.77 (0.72 to 0.82) |
| KDIGO stage 2 | <0.001 | 3.82 (2.65 to 5.52) | <0.001 | 2.24 (1.62 to 3.09) | <0.001 | 5.12 (4.52 to 5.71) | <0.001 | 0.65 (0.58 to 0.73) |
| KDIGO stage 3 | <0.001 | 15.64 (12.07 to 20.28) | <0.001 | 4.46 (3.57 to 5.56) | <0.001 | 10.87 (10.36 to 11.37) | <0.001 | 0.38 (0.34 to 0.42) |
On the left, a binary logistic regression model estimating the OR and a corresponding Cox proportional hazard regression model estimating the HR for all-cause in-hospital death; on the right, a robust regression model estimating the coefficients for hospital length of stay and a corresponding Cox proportional hazard model calculating the HR for hospital discharge.
A total of 37,772 cases were censored due to hospital discharge;
R2 = 0.35;
842 cases were censored due to in-hospital death;
HR for in-hospital death;
HR for discharge from hospital;
not applicable for Cox proportional hazard regression models;
International Systems of Units conversion factor: to convert creatinine to μM, multiply values by 76.3;
administered within 1 week before surgery;
as described by Kheterpal et al.3;
mutually exclusive categories of AKI severity by defnition of the KDIGO group.27
AKI = acute kidney injury; Cr = creatinine; HR = hazard ratio; ICU = intensive care unit; KDIGO = Kidney Disease: Improving Global Outcome; no F/U = no follow-up; OR = odds ratio; SE = standard error.