Skip to main content
. 2022 Apr 26;17(4):e0266737. doi: 10.1371/journal.pone.0266737

Table 3. Univariate (crude) and adjusted association of different outcomes among patients with versus without AKI.

Univariate Association
Continuous outcome
Outcomes AKI Non-AKI p value
N Median (IQR) N Median (IQR)
Hospital LOS (days) 270 9.04 (5.11–16.66) 2,228 5.09 (2.58–9.94) <0.001
PIM 2 Probability of Death (%) 274 1.20 (0.86–3.83) 2,272 0.96 (0.79–1.72) <0.001
PIM 3 Probability of Death (%) 274 0.98 (0.72–2.93) 2,272 0.78 (0.69–1.26) <0.001
Categorical outcome
Outcomes AKI [N (%)] Non-AKI [N (%)] p value Crude Odds ratio (95% CI)
Mortality 21 (7.7%) 37 (1.6%) <0.001 5.01 (2.89–8.70)
Airway / Respiratory support 152 (55.5%) 983 (43.3%) <0.001 1.63 (1.27–2.10)
Cardio-respiratory support 8 (2.9%) 19 (0.8%) 0.006 3.57 (1.55–8.23)
Kidney support 13 (4.7%) 9 (0.4%) <0.001 12.52 (5.30–29.58)
Vascular access 184 (67.2%) 675 (29.7%) <0.001 4.84 (3.70–6.32)
Adjusted Association *
Continuous outcome
Outcomes Un-standardized Coefficients (95% CI)^ p value
Hospital LOS (days) 6.29 (3.95–8.64) <0.001
PIM 2 Probability of Death (%) 0.81 (-0.11–1.72) 0.083
PIM 3 Probability of Death (%) 0.61 (-0.30–1.51) 0.187
Categorical outcome
Outcomes Adjusted odds ratio (95% CI)^ p value
Mortality 2.69 (1.48–4.88) 0.001
Airway / Respiratory support 1.61 (1.16–2.24) 0.005
Cardio-respiratory support 1.88 (0.77–4.56) 0.165
Kidney support 5.34 (2.15–13.25) <0.001
Vascular access 3.51 (2.63–4.70) <0.001

AKI: Acute Kidney Injury; LOS: Length of stay; PIM: Pediatric index of mortality; IQR: Interquartile range; CI: Confidence Interval

^AKI versus non-AKI

*Adjusted for variables significant in the uni-variate analysis “Table 1” such as age (continuous in years), presence of cardiovascular, endocrine, hematology, and respiratory co-morbidities (categorical). The clinical and laboratory variables significant in the uni-variate analysis “Table 2” were not considered because for those variables the data was available for <25% of the children [which was leading to decrease in the sample size for adjusted analysis to a greater extent]