The data for our study were derived from a prospective multicentre cohort study conducted by the China Critical Care Sepsis Trial (CCCST) working group. The study was carried out in 18 Chinese ICUs. The period was from 1 January 2014 to 31 August 2015. The CCCST study aimed to investigate the epidemiology and characteristics of critically ill patients with sepsis. Our study specifically targeted patients who met the sepsis 3.0 criteria [1] after ICU admission, thus being conclusively diagnosed with sepsis. The first day of diagnosis of sepsis was defined as the onset of sepsis. Following this diagnosis, patients were closely monitored for any signs of AKI that occurred within the first seven days. The exclusion criteria included: (1) patients already suffered from AKI of ICU admission; (2) AKI occurred before sepsis; (3) AKI occurred seven days later after the diagnosis of sepsis; (4) patients suffered chronic kidney disease (CKD), operated with nephrectomy or kidney transplantation; (5) patients received renal replacement therapy (RRT) for non-renal conditions; (6) missing data; (7) patients younger than 18 years of age. All participants or their close relatives had signed an informed consent form.

The included patients were categorized into two groups: those who developed AKI (AKI group) within seven days following a sepsis diagnosis and those who did not develop AKI (non-AKI group).

The definition of sepsis 3.0 was used to define sepsis [1]. For cases before 2016, this definition was applied retrospectively. Septic shock was defined as sepsis requiring vasopressors after adequate fluid resuscitation to maintain mean arterial pressure (MAP) ≥ 65 mmHg and blood lactate concentration ≥ 2 mmol/L [1, 12]. The diagnosis and stage of AKI were based on the serum creatinine and urine output criteria proposed by Kidney Disease: Improving Global Outcomes (KDIGO) [13]. The definition of baseline creatinine was as follows: if at least five values were available, the median of all values from 6 months to 6 days before hospitalization was used; otherwise, the minimum value was five days before hospitalization [14]. If no creatinine was available before hospitalization or the emergency patient’s serum creatinine was abnormal at admission, baseline creatinine was estimated using the Modification of Diet in Renal Disease (MDRD) equation [15]. Attributable mortality of AKI was defined as the proportion of deaths that can be statistically attributed to AKI [10, 11, 16, 17]. It was calculated by subtracting the mortality of matched septic patients without AKI from the mortality of matched septic patients with AKI (i.e., attributable mortality of AKI = Mortality in matched AKI patients–Mortality in matched non-AKI patients).

The primary endpoint was 30-day mortality after sepsis diagnosis. The secondary endpoint was the ICU length of stay (LOS), hospital LOS, ICU mortality, and hospital mortality.

Data information included demographic characteristics [gender, age, and body mass index (BMI)], comorbidities [chronic obstructive pulmonary disease (COPD) or asthma, cardiovascular disease, hypertension, diabetes, cancer, and chronic liver disease], admission type (medical, surgical or emergency), and ICU diagnosis. The Acute Physiology and Chronic Health Evaluation II (APACHE II) score and the non-renal Sequential Organ Failure Assessment (SOFA) score were recorded on the sepsis diagnosis day. We collected clinical management data, including mechanical ventilation and renal replacement therapy (RRT), on the day of sepsis diagnosis. The mean arterial pressure (MAP) and the presence of septic shock were recorded daily within seven days of the definitive diagnosis of sepsis. We also recorded the baseline creatinine, the use of nephrotoxic drugs (angiotensin-converting enzyme inhibitors, aminoglycosides, nonsteroidal anti-inflammatory drugs, and vancomycin, et al.) [18, 19], ICU LOS, hospital LOS, ICU mortality, 30-day mortality, and hospital mortality.

Statistics were analyzed using R 4.2.3. Continuous variables were presented as medians (interquartile range, IQR). Categorical variables are expressed as percentages. For comparing continuous data, we used the Mann-Whitney U test, and for categorical variables, the Chi-square test was utilized. A significance level of p < 0.05 was considered to indicate statistical significance. Applying PSM between septic patients with and without AKI to estimate the excess mortality attributed to AKI (Fig. 1A). PSM was performed using the “matchit” package to balance non-renal variables on outcomes. This study used the nearest neighbor method with a caliper value of 0.06 of the standard deviation of the logit of the propensity score. The propensity score, which was based on baseline characteristics and clinical covariates, was used to adjust the differences between matched patients with and without AKI. The variables analyzed for PSM included age, gender, BMI, comorbidities, APACHE II score, non-renal SOFA score, baseline creatinine, mechanical ventilation, MAP, septic shock, and the use of nephrotoxic drugs. The variables used for propensity score matching models did not exhibit multicollinearity. The multicollinearity tests were presented in Supplemental Table S2. We checked the covariate balance using standardized differences and considered values of over 25% indicative of meaningful differences. McNemar’s test was applied to sensitivity analysis [20]. Then, the 30-day mortality in matched septic patients with and without AKI was calculated separately. Subgroup analysis of matched septic patients with and without AKI was based on the AKI stage (Fig. 1B). Attributable mortality was calculated separately for total and different stages of AKI. A 95% confidence interval (CI) for the attributable mortality difference was estimated by Newcombe’s method [21].

Study flow diagram (A) Flowchart of all the participants (B) Flowchart of the subgroup analysis. Abbreviations: ICU intensive care unit; AKI, acute kidney injury

Survival analysis was conducted on matched AKI and matched non-AKI septic patients. We used the Kaplan-Meier survival curve to compare the survival status of both groups and the log-rank test to compare their survival time and rate differences. We also used the multivariate Cox proportional hazard model to estimate the hazard ratio (HR) and 95% confidence interval (CI) for 30-day mortality after adjusting for age, gender, BMI, comorbidities, APACHE II score, non-renal SOFA score, baseline creatinine, mechanical ventilation, RRT, MAP, septic shock, and the use of nephrotoxic drugs.