Epidemiology of acute kidney injury in hospitalized pregnant women in China

EACH study (Epidemiology of AKI in Chinese Hospitalized patients) was a multi-centered retrospective cohort study conducted in 25 regional medical centers from 15 provinces across China. The study cohort included 3,044,224 patients admitted from January 1, 2013 to December 31, 2015. We selected from the cohort women of childbearing age (14–50 years) who had at least two SCr tests within any 7-day window during their first 30 days of hospitalization as analysis set. We only included SCr tests that used an enzymatic assay and excluded patients with end-stage renal disease (ESRD), receiving maintenance dialysis or renal transplantation. For patients with multiple hospitalizations, we only included the first hospitalization in the analysis set. Pregnancy and its related comorbidities were identified by the diagnostic code. Delivery were identified by the operation procedures.

We obtained patient-level data from the electronic hospitalization databases and laboratory databases in the participating centers. The hospitalization records consisted of patients’ age, sex, date and diagnostic code at admission and discharge, operation procedures and dates, need for intensive care, in-hospital death, and total hospitalization cost. The laboratory data included time of patients’ SCr tests and value of SCr. The Medical Ethics Committee of Nanfang Hospital approved the study protocol and waived patient consent. While the approval number is NFEC-2014-098.

AKI was defined as an increase in SCr by 26.5 μmol/L (0.3 mg/dl) within 48 h or a 50% increase in SCr from the baseline within 7 days according to the Kidney Disease Improving Global Outcomes (KDIGO) criteria [21]. We screened patients’ SCr for the onset of AKI using an algorithm described previously [22]. At any time point t, a baseline SCr was dynamically defined as the mean of SCr levels within 30 days prior to t, and each of the available SCr data within 7 days after t was compared with this baseline. The earliest day when the SCr change met the KDIGO criteria was defined as the date of AKI onset. Patients who met at least one of the following criteria were classified as having community acquired AKI [22]: (a) admitted with AKI according to diagnostic code; (b) having multiple creatinine tests in the clinic, and comparing to the previous measurement, the increase in creatinine on the first day of hospitalization met the KDIGO definition; and (c) SCr on admission ≥1.5 fold of standardized SCr reference value and ≥ 1.5 fold of the minimal SCr level during hospitalization. In these cases, the lowest SCr during hospitalization was used as the baseline creatinine level for patients with community acquired AKI, and the mean SCr during the first 30 days of hospitalization for those without AKI. Cases that met the KDIGO creatinine criteria but not the criteria for the community acquired AKI were identified as having hospital acquired AKI.

The primary outcome of AKI was in-hospital death. Other outcomes included adverse perinatal outcomes (containing preterm delivery, abortion, dystocia and stillbirth), length of stay in hospital and daily cost of hospitalization. Primary diagnosis and coexisting comorbidities (including chronic kidney disease, convulsion, chronic obstructive pulmonary disease, diabetes mellitus, diarrhea, gastrointestinal bleeding, peptic ulcer disease, heart failure, hematological tumor, hypertension, liver disease, pulmonary infection, myocardial infarction, glomerulonephritis, pregnancy-induced hypertension syndrome, acute fatty liver, sepsis, shock, systemic lupus erythematosus, stroke, thrombotic microangiopathy, trauma, hyperlipidemia)were identified by the diagnostic codes (ICD10-CM codes, see Additional file 1).

Among 110,873 hospitalized women of childbearing age (14–50 years old), a total of 10,920 pregnant women were included as analysis set (Fig. 1). We identified 795 AKI events from the pregnant women (in Additional file 2: Table S1). The characteristics of pregnant women stratified by AKI status were shown in Table 1. Compared with the non-AKI group, AKI group had a higher percentage of patients requiring intensive care and patients with preexisting chronic kidney disease (CKD), hypertension and systemic lupus erythematosus. Women who developed community acquired AKI had a higher baseline SCr compared to those without AKI. Considering the pregnancy-related complications, pregnancy-induced hypertension (PIH) syndrome (including gestational hypertension, preeclampsia, eclampsia) was more prevalent in AKI group (41.0% in community acquired AKI, 23.1% in hospital acquired AKI). The proportion of PIH and acute fatty liver (AFL) in women with community acquired AKI were significantly higher than that in women without AKI and women with hospital acquired AKI.

Of 10,920 pregnant women, community acquired AKI and hospital acquired AKI occurred in 3.6% (n = 393) and 3.7% (n = 402) of the population in the analysis set, respectively, giving rise to an overall incidence of 7.3%. The incidence of community acquired and hospital acquired AKI in various clinical settings was depicted in Fig. 2. The three clinical settings with the highest incidence of community acquired AKI were acute fatty liver (55.9%), sepsis (20.0%) and shock (18.9%). For hospital acquired AKI, the top three clinical settings were respiratory failure (29.6%), acute fatty liver (10.2%), and heart failure (9.4%). Only 4.0% (n = 32) of AKI events, which were identified by SCr change, were diagnosed on the discharge record.

Among 99,953 non-pregnant women of childbearing age (14–50), we calculated the incidence of AKI at 6.0%. Compared with the non-pregnant group, pregnant women had a 51% increased risk of AKI (Odds Ratio [OR] 1.51, 95% Confidence Interval [CI] 1.35 to 1.69), after adjusting for age, baseline SCr, length of stay in hospital, division, hospital, need for intensive care, and clinical comorbidities (in Additional file 2: Table S1).

The risk profile of AKI during pregnancy is presented in Table 2. Advanced maternal age was not a significant risk factor for pregnancy-related AKI. We estimated the population attributable fractions (PAF) of the risk factors to assess their contribution to AKI during pregnancy. Overall, the top three risk factors, ranked in order of decreasing PAF, were PIH (21.2%), AFL (13.5%), and CKD (6.2%) for AKI during pregnancy.

In-hospital outcomes and adverse perinatal outcomes of pregnant women were illustrated in Table 3. The incidence of in-hospital death was 0.1, 1.0, 2.1, and 7.4% in non-AKI, patients with AKI stage 1, 2, and 3, respectively (in Additional file 2: Table S2). The Hazard Ratio (HR) of in-hospital death adjusted for age, baseline SCr, length of stay in hospital, division, hospital, need for intensive care, and clinical comorbidities was 6.8 (95%CI, 2.8 to 16.2) for patients with AKI. Pregnancy-related AKI was associated with longer length of stay in hospital and higher daily cost during hospitalization. The average cost in pregnant women with AKI is 83% higher than that in patients without AKI. The average length of stay was 11 days in AKI patients and 9 days in non-AKI patients. The incidence of adverse perinatal outcomes was 23.9% (n = 190) and 22.8% (n = 2313) in hospitalized pregnant women with and without AKI, respectively (P = 0.485). Compared to patients without AKI, pregnant women with AKI are more likely to have stillbirth (1.0% in AKI group versus 0.2% in non-AKI group, P < 0.05), while the risk of preterm-delivery, abortion and dystocia showed no significant difference.