Table 2.
Primary, Secondary, and Safety Outcomes in cluster randomized analyses of the POINCARE-2 trial
| Control group % [95% CI]b | Strategy group % [95% CI]b | Crude effect of strategy | Adjusted effect of strategya | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Median (IQR)c | Median (IQR)c | Model (part) | Exp(parameter) [95% CI] | p-value | Model (part) | Exp(parameter) [95% CI] | p-value | |||||
| Primary outcome | ||||||||||||
| 60-day mortality rateb | 33.9 [29.6–38.2] | 30.5 [26.2–34.8] | MP | RR | 0.90 [0.75–1.08] | 0.26 | MP | RR | 0.90 [0.75–1.08] | 0.99 | ||
| Secondary outcomes | ||||||||||||
| 28-day mortality rateb | 26.7 [22.7–30.7] | 21.7 [17.8–25.6] | MP | RR | 0.77 [0.56–1.06]d | 0.26d | MP | RR | 0.77 [0.56–1.06]d | > 0.99d | ||
| In-hospital mortality rateb | 32.6 [28.4–36.9] | 29.6 [25.3–33.9] | MP | RR | 0.91 [0.68–1.21]d | > 0.99d | MP | RR | 0.91 [0.68–1.21]d | > 0.99d | ||
| Mechanical ventilator-free daysc | 9.0 [0.0–18.0] | 9.0 [0.0–17.0] | ZINB | (NB) | MF | 1.01 [0.91–1.12]d | > 0.99d | ZINB | (NB) | MF | 1.01 [0.91–1.12]d | > 0.99d |
| (ZI) | OR | 0.91 [0.57–1.46]d | > 0.99d | (ZI) | OR | 0.91 [0.57–1.46]d | > 0.99d | |||||
| Vasopressor-free daysc | 17.0 [7.0–24.0] | 17.0 [8.0–26.0] | ZINB | (NB) | MF | 1.03 [0.96–1.10]d | > 0.99d | ZINB | (NB) | MF | 1.03 [0.96–1.10]d | > 0.99d |
| (ZI) | OR | 0.84 [0.42–1.68]d | > 0.99d | (ZI) | OR | 0.84 [0.42–1.68]d | > 0.99d | |||||
| Renal replacement therapy-free daysc | 23.0 [11.0–47.5] | 23.0 [11.0–49.0] | ZIPoi | (Poi) | MF | 0.996 [0.97–1.03]d | > 0.99d | ZIPoi | (Poi) | MF | 0.996 [0.97–1.03]d | > 0.99d |
| (ZI) | OR | 1.11 [0.40–3.11]d | > 0.99d | (ZI) | OR | 1.11 [0.40–3.11]d | > 0.99d | |||||
| Safety outcomes | ||||||||||||
| Arterial hypotensionb | 77.8 [74.0–81.5] | 71.6 [67.4–75.9] | MP | RR | 0.94 [0.85–1.03] | 0.16 | MP | RR | 0.94 [0.85–1.03] | 0.39 | ||
| At least one episode of hypernatremia> 155 mmol/lb | 2.3 [1.0–3.7] | 5.8 [3.6–8.0] | LB | RR | 2.72 [1.23–6.00] | 0.01 | LB | RR | 2.72 [1.23–6.00] | 0.02 | ||
| At least one episode of hypokalemia < 2.8 mmol/lb | 7.6 [5.2–10.0] | 7.2 [4.8–9.6] | LB | RR | 1.02 [0.61–1.72] | 0.93 | LB | RR | 1.02 [0.61–1.72] | 0.60 | ||
| Renal damage b,e | 20.3 [16.6–24.0] | 23.5 [19.5–27.6] | LB | RR | 1.16 [0.90–1.50] | 0.25 | LB | RR | 1.16 [0.90–1.50] | 0.24 | ||
| Myocardial infarctionb | 0.4 [0.0–1.0] | 1.15 [0.2–2.2] | LB | RR | 1.82 [0.23–14.15] | 0.57 | LB | RR | 1.82 [0.23–14.15] | 0.61 | ||
| Mesenteric ischemiab | 0.6 [0.0–1.4] | 0.2 [0.0–0.7] | LB | RR | 0.36 [0.04–3.49] | 0.38 | LB | RR | 0.36 [0.04–3.49] | 0.45 | ||
CI Confidence interval, IQR Interquartile range, MP Mixed modified Poisson, RR Risk ratio, ZINB Zero-inflated negative binomial mixed, NB Negative binomial part, MF Multiplicative factor, ZI Zero-inflated part, ZIPoi Zero-inflated Poisson mixed, Poi Poisson part, LB Log-binomial mixed
aAdjusted for the following baseline characteristics: age, main cause of admission (recoded as CNS injury vs. other causes of admission), McCabe score, SAPS II at day 1
bBinary outcomes: log-binomial (modified Poisson) mixed models were used and “exp(parameter)” are estimations of RR
cNumeric outcomes: modeled with zero-inflated negative binomial (Poisson) mixed models which combine a logistic regression and a negative binomial (Poisson respectively) model and “exp(parameter)” are odds ratios of excess of zeros for the zero-inflated part (logistic) and multiplicative factor for the counting part (negative binomial/Poisson)
dBonferroni correction was used: Confidence intervals according to α = 5/22 = 0.23% and p-values by a multiplying a coefficient of 22
eRenal damage was defined by worsening RIFLE between Day 3 and Day 14 as compared with the higher RIFLE observed on Day 1 or Day 2. For renal damage, analyses were conducted within the subsample of at-risk patients (i.e., in 883/905 patients not classified as RIFLE “End Stage Kidney Disease” at Day 2)