Skip to main content
. Author manuscript; available in PMC: 2017 Feb 1.
Published in final edited form as: Anesthesiology. 2016 Feb;124(2):339–352. doi: 10.1097/ALN.0000000000000959

Table 3.

Multivariable predictors of severity of postoperative acute kidney injury*

Predictors Proportional Odds Ratio (95% Confidence Interval) P value
Thrombocytopenia, platelet count ≤ 74 × 109/L 3.04 (2.26–4.07) < 0.0001
Duration of cardiopulmonary bypass, per 30-minute increase 1.19 (1.11–1.28) < 0.0001
Postoperative nadir hemoglobin, per 1 g/dL increase 0.83 (0.74–0.93) 0.001
*

Multivariable proportional odds model based on n = 2,267 patients who have postoperative acute kidney injury stage I, II or III. Clinical variables with P < 0.05 from the final multivariable regression model for predicting postoperative acute kidney injury as a binary outcome variable were selected and tested in this model using a stepwise selection method. This model excluded cases without postoperative acute kidney injury because the score test (see Methods) indicated that testing for these patients violated the proportional odds assumption. This model gives odds ratios for increase in risk for worsening stages of postoperative acute kidney injury, ie, from stage I to II, or from stage II to III, as defined by the Kidney Disease: Improving Global Outcomes (KDIGO) Clinical Practice Guidelines.21