Prognostic value of early vs late acute renal dysfunction in the critically ill obstetric patient

Introduction

Acute renal dysfunction/failure (ARF) is newly considered as an independent mortality risk factor in the ICU [1]. The time of onset of ARF could have a prognostic impact.

Objective

To determine prognosis of acquired vs not acquired renal dysfunction.

Methods

Retrospective analysis of prospectively collected data as part of the APRiMo study [2]. We defined ARF as serum creatinine level >100 μmol/l. We defined the early ARF group (E) as renal dysfunction that occurred during maternity management (less than 24 hours before transfer to the ICU) or at day 1 of ICU admission. The late ARF group (L) was defined as renal dysfunction occurring from day 2 of ICU admission. Inclusion criteria: critically ill obstetric patients presenting with one or more of the presenting conditions: severe pre-eclampsia/eclampsia, postpartum haemorrhage, stroke, HELLP syndrome, hemolytic uremic syndrome, and so on. Exclusion criteria: length of stay in the ICU <2 calendar days, chronic renal failure. Setting: patients first managed in a tertiary referral maternity for high-risk pregnancies, then transferred to our independent multidisciplinary ICU. Collected data: demographic, obstetric management, daily SOFA score. Main outcome of interest: vital status at ICU discharge. Adequate statistical tests were used (t test, chi-square test, etc.). P < 0.05 was considered significant.

Results

Six hundred and forty patients, overall mortality 13.3%, included n = 223: 193 in group E and 30 in group L. There was no difference in demographic data, main admission diagnosis between E and L groups. Mean maximum creatinine level: group E = 270 μmol/l, group L = 220 μmol/l (P = 0.15). Renal replacement therapy: 30/193 and 7/30 (P = 0.3). Comparing group L vs group E: L had a higher mortality rate (73% vs 20.7%; P < 0.001), duration of mechanical ventilation (5.8 ± 1 vs 3.5 ± 0,26; P = 0,001), rate of massive transfusion in the ICU (13/30 vs 34/193; P = 0.003) and length of stay (8.9 vs 6.2, P = 0.03). Mean day 1 MOD score summing organ dysfunctions without renal dysfunction: 7.6 ± 3.3 vs 5.5 ± 4 (P = 0.45). Mean total maximum MOD score (TM_MODS):11 ± 6 vs 6 ± 5, P = 0.001. For group E, nonsurvivors (NS) showed significantly more unstable haemo-dynamic state with a lower diastolic arterial pressure (P = 0.007), uterine atonia (P = 0.001), transfusion rate in the labor ward/ operating room and ICU stay than survivors (S). For group L, NS vs S: mean MODS at day 1 ICU admission 8.1 ± 3.7 vs 6.2 ± 2.7, P = 0.15), mean TM_MODS (16.2 ± 3.3 vs 7.8 ± 3.1, P < 0.001). NS were paradoxically younger than the S (29 years vs 35 years, P < 0.01). Duration of mechanical ventilation was higher among NS (P = 0.001). The main diagnosis of admission was postpartum hemorrhage complicating HELLP syndrome or sepsis. Three organ failures lasting >2 days (renal failure not counted) is associated with 100% mortality (n = 17).

Conclusion

Acquired ARF is an important prognostic factor. It is more serious than early ARF for the same level of serum creatinine.