Abstract
Introduction
Guidelines recommend oral hydration (OH) to prevent contrast-associated acute kidney injury (CA-AKI) after coronary angiography or intervention, but quantitative protocols are lacking, and practice varies. Shorter hospital stays limit opportunities for post-procedure creatinine monitoring, and real-world OH patterns and their early renal effects remain poorly described. This study aimed to characterize periprocedural weight-adjusted OH trajectories and to assess whether trajectory membership was associated with a change in serum creatinine within 24 h.
Methods
This single-center prospective cohort study enrolled 192 inpatients undergoing coronary angiography or intervention between November 2024 and May 2025. We recorded weight-adjusted oral intake in four windows: pre-12 h, post-0–6 h, 6–12 h, and 12–24 h. We computed partial-overlap dynamic time warping distances and clustered trajectories using partitioning around medoids. Group comparisons used nonparametric tests. The primary outcome was percent change in serum creatinine, and dose-response was assessed with generalized additive models (GAMs) adjusted for baseline creatinine, estimated glomerular filtration rate, comorbidities, procedural indication, contrast volume, and intravenous hydration.
Results
After excluding 18 patients for missing periprocedural oral intake or post-procedure creatinine, 174 patients were analyzed (median age 65.0 years, IQR 57.2 to 72.0; 71.8% male). Median weight-adjusted cumulative oral intake was 10.9, 18.6, 23.9, and 33.3 mL/kg for the pre-12 h, post-0–6 h, 0–12 h, and 0–24 h windows. Time-series clustering produced two stable groups (low-OH n = 85, high-OH n = 89; average silhouette = 0.393; bootstrap adjusted Rand index = 0.845). The largest between-group difference occurred in the 0–6 h window (median 4.12 vs. 2.00 mL/kg/h, p < 0.001). No patient met CA-AKI criteria. In adjusted GAMs, the 0–6 h OH rate was not a significant smooth term (edf = 3.79, p = 0.795), and cluster membership was not associated with creatinine change. However, the high-OH cluster had substantially greater 6-h urine output and a larger positive fluid balance (p < 0.001), while serum potassium was similar between clusters.
Conclusion
Periprocedural OH was generally ample and formed two reproducible patterns. Concentrating oral intake in the first 6 h increased early urine output and positive fluid balance but did not show a significant dose-response association with early serum creatinine change. This null result should be interpreted cautiously because the cohort was generally low-risk, follow-up was short, and the study had limited power. Future studies should validate whether early concentrated OH protects renal function using earlier sensitive biomarkers and by targeting high-risk patients.
Keywords: Oral hydration, Contrast-associated acute kidney injury, Coronary angiography and intervention, Time-series clustering, Prospective real-world cohort
Introduction
Contrast-associated acute kidney injury (CA-AKI) remains a common and clinically important complication of coronary angiography (CAG) and intervention, arising from direct nephrotoxicity of contrast media or from other precipitating factors [1]. Reported CA-AKI incidence varies widely across series, typically 1–20% in unselected cohorts and markedly higher in patients with diabetes, preexisting renal impairment, or other risk factors, reaching 40–50% in some high-risk groups [2, 3]. CA-AKI is associated with prolonged hospital stay, higher short- and long-term mortality, and increased medical costs. Effective prevention is a central concern in contemporary practice [4–6].
Limiting intravascular contrast exposure and ensuring adequate periprocedural hydration are the principal preventive measures recommended in guidelines [2, 6, 7]. Both intravenous and oral routes can preserve intravascular volume and renal perfusion and may reduce tubular exposure to contrast media by dilution and faster renal clearance [8]. Although some studies suggest that oral hydration (OH) may be non-inferior to intravenous protocols in selected settings, evidence is heterogeneous, and many studies combine oral and intravenous hydration or apply varied regimens, limiting direct generalization to real-world, short-stay practice [9, 10]. Current guidelines endorse OH but do not specify concrete protocols for rate, timing, or total volume [2, 6, 7]. In practice and in the published literature, OH regimens vary widely. Some studies report preprocedural fluid boluses and postoperative regimens ranging from several hundred milliliters to a few liters over 3–24 h [9–12], and some trials define adequate OH as approximately 12–15.6 mL/kg over 24 h [13–15].
Existing investigations generally summarize cumulative oral intake over 12 or 24 h and do not capture dynamic differences in intake rate across early post-procedural windows [9–15]. Shorter hospital stays mean that many patients are discharged within 24 h after coronary procedures, so early intervals may represent a more practical and modifiable window for hydration interventions [16, 17]. Evidence also supports that individualized, physiology-guided hydration may outperform fixed-rate protocols, but data describing real-world OH time series and their immediate renal and fluid-balance consequences remain scarce [7]. The diagnostic criteria of Kidney Disease: Improving Global Outcomes (KDIGO) for acute kidney injury(AKI) rely on changes in serum creatinine observed over 48 h or up to 7 days [18, 19]. However, short hospital stays after coronary procedures mean that many patients get blood tests within the first 24 h or are discharged before creatinine peaks [19]. Studies reported that about half of patients missed their post-procedure serum creatinine (SCr) measurement [20, 21], hampering reliable CA-AKI surveillance.
With shorter hospital stays, periprocedural OH patterns and their effects on early renal function are unclear, and guidance on optimal timing and dose is lacking. In this single-center prospective cohort study, we recorded patients' weight-adjusted OH volume during four time segments (pre-12 h, post-0–6 h, 6–12 h, and 12–24 h) and applied time-series clustering to identify real-world periprocedural OH trajectories. The primary aim was to describe these trajectories and test whether higher trajectory membership was associated with different changes in serum creatinine within 24 h as well as to compare short-term urine output and electrolyte changes across trajectories. Because clustering showed maximal separation in the immediate post-procedure period, we performed an exploratory analysis of the dose-response between the post-0–6 h weight-adjusted OH rate and percent change in serum creatinine within 24 h.
Methods
Study Design and Participants
We conducted a single-center prospective observational cohort study at Peking University First Hospital (Beijing) from November 2024 to May 2025. We screened 192 adult inpatients scheduled for elective CAG with or without percutaneous coronary intervention (PCI). The observation window ran from 12 h before the procedure to 24 h after the procedure or until hospital discharge, whichever came first. Eligible patients were aged 18 years or older, scheduled for an elective coronary procedure, and able to provide written informed consent. We excluded patients with emergency procedures, urinary tract obstruction, and ongoing maintenance hemodialysis. Reporting follows the Reporting of Observational Studies in Epidemiology (STROBE) recommendations.
Data Collection and Measurements
Baseline demographic, body weight, clinical history, and medication data were collected at enrollment. At the study center, all patients undergoing CAG or intervention were encouraged to maintain OH throughout the periprocedural period unless there was an explicit clinical restriction, such as decompensated heart failure. No fixed OH prescription (timing or volume) was mandated. Intravenous hydration was provided at the attending physician’s discretion according to baseline renal function, comorbidities, contrast volume, and other risk factors for AKI. The usual institutional intravenous regimen, when used, was isotonic saline at 1 mL/kg/h starting at 22:00 the night before the procedure and continuing until 24 h after the procedure; the infusion rate was reduced for patients with heart failure at the treating physician’s judgement. Participants received a standardized diary and recorded all oral fluids from 20:00 the evening before the procedure until 24 h after; recorded fluids included water, juice, milk, soup, and other oral liquids, together with the exact time of ingestion. The study personnel will measure and mark the capacity of the patient’s drinking glass, helping the patient accurately record fluid intake. Urine output over the first 6 h after the procedure was collected in a dedicated container and measured by study personnel. The post-0–6 h cumulative volume was converted to an hourly, weight-standardized urine flow rate (mL/kg/h) for analysis. Positive fluid balance during the same interval was calculated as oral plus intravenous fluid intake minus urine output. All procedures used the same nonionic, low-osmolar iodinated contrast agent, iomeprol. Procedural data (including contrast volume and details of the coronary intervention) and in-hospital laboratory results (including baseline and post-procedure serum creatinine within 24 h) were abstracted from the hospital information system, and baseline serum creatinine was defined as the most recent preprocedural measurement obtained during the current hospital admission. Participants with neither any recorded oral intake during the observation window nor a post-procedure creatinine within 24 h were treated as lost to follow-up for the primary analysis.
Exposures
The primary exposures were periprocedural weight-adjusted OH trajectories during four prespecified windows (pre-12 h, post-0–6 h, 6–12 h, and 12–24 h), which were identified by time-series clustering. OH rates were calculated as the total oral volume in each interval divided by the interval duration (hours) and by body weight (mL/kg/h).
Outcomes
Because of routine post-procedure sampling and short hospital stays, the prespecified primary endpoint was the occurrence of AKI within the first 24 h after the procedure, defined using KDIGO thresholds [19]. Given limited events within 24 h, the primary outcome was analyzed as early post-procedural serum creatinine and estimated glomerular filtration rate (eGFR) change, reported both as absolute change and as percent change from baseline. The prespecified short-term physiologic and safety outcomes included the post-0–6-h urine flow rate, positive fluid balance, and early serum electrolytes (including serum potassium) within 24 h.
Covariates
A comprehensive set of prespecified covariates was included to control for confounding in multivariable analyses. Sociodemographic variables comprised age, sex, body mass index, smoking status, and alcohol use. Relevant clinical history that reflects patients’ cardiovascular risk profile and comorbidity burden included the indication for the coronary procedure, prior PCI, prior myocardial infarction, heart failure, hypertension, hyperlipidemia, diabetes mellitus, chronic kidney disease, hyperuricemia, and cancer. Medication use at baseline was recorded to reflect background therapy and potential effect modifiers; these variables included use of an angiotensin-converting enzyme inhibitor (ACEI) or an angiotensin receptor blocker (ARB), use of diuretics, and use of metformin. Preprocedural laboratory and imaging measurements comprised left ventricular ejection fraction, baseline serum creatinine, baseline eGFR, blood urea nitrogen, serum potassium, N-terminal pro B-type natriuretic peptide, high-sensitivity C-reactive protein, triglycerides, total cholesterol, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, hemoglobin, and hemoglobin A1c. Procedural covariates included intravascular contrast volume and whether intracoronary imaging had been performed in the preceding week. In addition, the Mehran score was calculated for each participant and included as an overall measure of CA-AKI.
Statistical Analysis
Analyses were performed in R version 4.4.1. For each patient, we built a time series matrix of weight-adjusted OH rates in mL/kg/h for four prespecified intervals: pre-12 h, post-0–6 h, 6–12 h, and 12–24 h. Pairwise sequence distances were computed using partial-overlap dynamic time warping (DTW) to allow comparison when some intervals were missing, and partitioning around medoids was used to derive trajectory clusters. Candidate cluster numbers were selected based on average silhouette width, clinical interpretability, cluster sample size, and clustering stability as measured by the bootstrap adjusted Rand index (ARI). Stability was assessed using 200 bootstrap subsamples, and ARI values were summarized across repeats.
Continuous variables are reported as median (25th–75th percentile) and categorical variables as n (%). Between-cluster comparisons used the Mann-Whitney U test for two-group continuous comparisons, the Kruskal-Wallis test for multigroup continuous comparisons, and the chi-square or Fisher’s exact test for categorical variables as appropriate. All tests were two-sided, and p < 0.05 was considered statistically significant.
Prompted by the clustering results, we explored the association of the post-0–6-h OH rate with 24-h percent change in serum creatinine using generalized additive models (GAMs) with penalized splines to allow for nonlinearity. Multivariable GAMs were adjusted for prespecified covariates, including age, baseline serum creatinine and eGFR, comorbidities, indication for the coronary procedure, intravascular contrast volume, and intravenous hydration status. Model goodness-of-fit was summarized using adjusted R2.
For the two core variables, post-0–6-h OH and post-procedure serum creatinine, we did not impute missing values and excluded participants missing either (n = 18). Post-12 h and 24 h oral volumes, which had substantial missingness (35 [20.1%] and 65 [37.4%], respectively), were left missing for clustering because the partial-overlap DTW method compares sequences with partially missing intervals without imputation. Other variables followed pragmatic rules for missingness: single median imputation for minimal missingness (<5%), and median imputation plus a missingness indicator for moderate missingness (5–15%) [22]. The observed missing counts were height 1 (0.5%), left ventricular ejection fraction 2 (1.1%), contrast volume 1 (0.5%), 6-h urine 3 (1.7%), N-terminal pro B-type natriuretic peptide 7 (4.0%), high-sensitivity C-reactive protein 15 (8.6%), lipids 9 (5.2%), and hemoglobin A1c 22 (12.6%).
Two prespecified sensitivity analyses were performed. First, we compared the 18 participants excluded for missing post-0–6-h OH or post-procedure serum creatinine with the analytic cohort on baseline demographic, clinical, and procedural variables and, where appropriate, modelled missingness by logistic regression to assess systematic differences. Second, we repeated clustering and outcome analyses using k = 3 to test robustness. Sensitivity results are reported alongside the primary findings.
Results
Baseline Characteristics
Of 192 patients enrolled and followed for about 24 h after the procedure, 18 were excluded for missing data (12 [6.3%] lacked post-procedure serum creatinine, and 6 [3.1%] lacked OH records), leaving 174 patients for analysis (Fig. 1). The median age was 65.0 years (IQR 57.2 to 72.0), and 71.8% were men. Unstable angina was the leading indication (111, 63.8%). The most common comorbidities were hyperlipidemia (158, 90.8%), hypertension (117, 67.2%), and diabetes mellitus (74, 42.5%). Median baseline serum creatinine was 84.8 µmol/L (IQR 76.8 to 99.8), and eGFR was 77.6 mL/min/1.73 m2 (IQR 64.8 to 88.9). Eight participants had a prior charted diagnosis of chronic kidney disease; among these, the median preprocedural eGFR was 23.7 (IQR 15.4 to 71.8) mL/min/1.73 m2, and four had dipstick-positive proteinuria. Median iodinated contrast volume was 120 mL (IQR 60 to 150). Vascular access was predominantly radial (165, 94.8%). A total of 114 patients (65.5%) underwent PCI, and 25 (14.4%) received intravenous hydration. Recipients of intravenous hydration had lower baseline renal function than the overall cohort (median preprocedural eGFR 47.6 mL/min/1.73 m2, IQR 33.3 to 58.9). Mehran risk categories were low 117 (67.2%), moderate 42 (24.1%), high 14 (8.0%), and very high 1 (0.6%). Additional baseline variables are summarized in Table 1. We present core demographic, comorbidity, renal function, medication, and procedural variables in Table 1. Less directly relevant laboratory measures, such as the lipid profile, are provided in online supplementary Table S1 (for all online suppl. material, see https://doi.org/10.1159/000550260).
Fig. 1.
Flowchart of enrollment, exclusions, and the analytic population in this study. CAG, coronary angiography; PCI, percutaneous coronary intervention; OH, oral hydration.
Table 1.
Baseline characteristics of participants overall and by hydration pattern
| Characteristic | Overall (n = 174) | Low-OH cluster (n = 85) | High-OH cluster (n = 89) | p value |
|---|---|---|---|---|
| Age, years | 65.0 (57.2, 72.0) | 66.0 (56.0, 72.0) | 64.0 (58.0, 72.0) | 0.960 |
| Male gender, n (%) | 125 (71.8) | 63 (74.1) | 62 (69.7) | 0.514 |
| BMI, kg/m2 | 25.1 (23.3, 27.7) | 25.9 (23.9, 28.1) | 24.5 (23.2, 26.6) | 0.016 |
| Indication for CAG, n (%) | ||||
| STEMI | 4 (2.3) | 3 (3.5) | 1 (1.1) | 0.006 |
| NSTEMI | 9 (5.2) | 9 (10.6) | 0 (0.0) | |
| Unstable angina | 111 (63.8) | 48 (56.5) | 63 (70.8) | |
| Stable angina | 31 (17.8) | 14 (16.5) | 17 (19.1) | |
| Other | 19 (10.9) | 11 (12.9) | 8 (9.0) | |
| Previous PCI, n (%) | 68 (39.1) | 36 (42.4) | 32 (36.0) | 0.387 |
| Previous MI, n (%) | 21 (12.1) | 12 (14.1) | 9 (10.1) | 0.418 |
| HF, n (%) | 11 (6.3) | 9 (10.6) | 2 (2.2) | 0.024 |
| Hypertension, n (%) | 117 (67.2) | 62 (72.9) | 55 (61.8) | 0.117 |
| Hyperlipidemia, n (%) | 158 (90.8) | 80 (94.1) | 78 (87.6) | 0.139 |
| DM, n (%) | 74 (42.5) | 41 (48.2) | 33 (37.1) | 0.137 |
| CKD, n (%) | 8 (4.6) | 6 (7.1) | 2 (2.2) | 0.161 |
| Hyperuricemia, n (%) | 13 (7.5) | 9 (10.6) | 4 (4.5) | 0.126 |
| Cancer, n (%) | 15 (8.6) | 6 (7.1) | 9 (10.1) | 0.473 |
| ARB/ACEI, n (%) | 98(56.3) | 54 (63.5) | 43 (48.3) | 0.043 |
| Diuretics, n (%) | 13 (7.5) | 10 (11.8) | 3 (3.4) | 0.035 |
| Metformin, n (%) | 45 (25.9) | 31 (36.5) | 14 (15.7) | 0.002 |
| Statin, n (%) | 167 (96.0) | 80 (94.1) | 87 (97.8) | 0.216 |
| LVEF, % | 68.8 (63.0, 73.9) | 67.8 (61.5, 73.8) | 69.8 (65.8, 74.0) | 0.198 |
| ICM within the past week, n (%) | 12 (6.9) | 4 (4.7) | 8 (9.0) | 0.265 |
| SCr, μmol/L | 84.8 (76.8, 99.8) | 83.7 (77.4, 102.4) | 85.0 (73.9, 95.0) | 0.281 |
| eGFR, mL/min/1.73 m2 | 77.6 (64.8, 88.9) | 75.7 (61.5, 87.4) | 79.3 (68.7, 90.4) | 0.223 |
| K, mmol/L | 3.9 (3.7, 4.1) | 3.9 (3.7, 4.2) | 3.9 (3.6, 4.1) | 0.045 |
| hs-CRP, mg/L | 1.2 (0.6, 2.3) | 1.2 (0.6, 3.0) | 1.2 (0.6, 1.9) | 0.148 |
| Hb, g/L | 139.0 (128.0, 149.0) | 138.0 (128.0, 148.0) | 141.0 (128.0, 152.0) | 0.298 |
| HbA1c, % | 6.3 (5.9, 6.9) | 6.3 (5.9, 7.2) | 6.3 (5.9, 6.6) | 0.157 |
| Vascular access, n (%) | ||||
| Radial | 165 (94.8) | 80 (94.1) | 85 (95.5) | 0.485 |
| Femoral | 8 (4.6) | 4 (4.7) | 4 (4.5) | |
| Radial+femoral | 1 (0.6) | 1 (1.2) | 0 (0.0) | |
| ICM volume, mL | 120.0 (60.0, 150.0) | 140.0 (80.0, 150.0) | 120.0 (60.0, 150.0) | 0.151 |
| Stent implantation, n (%) | 79 (45.4) | 41 (48.2) | 38 (42.7) | 0.463 |
| DCB implantation, n (%) | 61 (35.1) | 34 (40.0) | 27 (30.3) | 0.182 |
| Mehran grade, n (%) | ||||
| Low risk | 117 (67.2) | 53 (62.4) | 64 (71.9) | 0.175 |
| Moderate risk | 42 (24.1) | 21 (24.7) | 21 (23.6) | |
| High risk | 14 (8.0) | 10 (11.8) | 4 (4.5) | |
| Very high risk | 1 (0.6) | 1 (1.2) | 0 (0.0) | |
| IVH, n (%) | 25 (14.4) | 14 (16.5) | 11 (12.4) | 0.440 |
Continuous variables were expressed as interquartile range and categorical variables as frequencies (n) and percentages (%).
BMI, body mass index; CAG, coronary angiography; STEMI, ST-elevation myocardial infarction; NSTEMI, non-ST-elevation myocardial infarction; PCI, percutaneous coronary intervention; MI, myocardial infarction; HF, heart failure; DM, diabetes mellitus; CKD, chronic kidney disease; ARB/ACEI, angiotensin-receptor blocker/angiotensin-converting enzyme inhibitor; LVEF, left ventricular ejection fraction; SCr, serum creatinine; eGFR, estimated glomerular filtration rate; K, serum potassium; hs-CRP, high-sensitivity C-reactive protein; Hb, hemoglobin; HbA1c, hemoglobin A1c; ICM, iodinated contrast medium; DCB, drug-coated balloon; IVH, intravenous hydration; OH, oral hydration.
OH Volumes by Time Window
OH volumes by time window are shown in Table 2. A complete four-time point record was available for 108 of 174 patients (62.1%). Median intake was greatest in the post-0–6 h window and lowest in the post-6–12 h window. By 6 h after the procedure, 106 of 174 patients (60.9%) had reached the previously proposed target of ≥15.6 mL/kg. Median weight-adjusted oral intake was 18.6 mL/kg in the 0–6 h window, 23.9 mL/kg in the 0–12 h window and 33.3 mL/kg cumulatively at 0–24 h.
Table 2.
Periprocedural OH volumes
| Variable | Pre-12 h (n = 174) | Post-0–6 h (n = 174) | Post-6–12 h (n = 139) | Post-12–24 h (n = 108) | Post-0–24 h (n = 108) |
|---|---|---|---|---|---|
| OHV, mL | 800.0 (500.0, 1127.5) | 1300.0 (900.0, 1800.0) | 300.0 (115.0, 600.0) | 745.0 (400.0, 1200.0) | 2375.0 (1750.0, 3170.0) |
| OHV/W, mL/kg | 10.9 (7.4, 16.2) | 18.6 (12.2, 25.0) | 4.2 (1.8, 8.0) | 11.1 (6.1, 16.4) | 33.3 (24.0, 44.8) |
Continuous variables were expressed as interquartile range.
OHV, oral hydration volume; OHV/W, oral hydration volume per body weight.
Temporal Clustering of OH and Cluster-Specific Characteristics
Among 174 participants, partial-overlap DTW clustering showed the highest average silhouette at k = 2 (0.393); bootstrap stability (ARI mean 0.845, SD 0.159) supported a two-cluster solution. Average silhouette coefficient and per-sample silhouette widths are provided in online supplementary Figures S1 and S2. The cohort was split into a low-OH cluster (n = 85) and a high-OH cluster (n = 89). Cluster trajectories are plotted in Figure 2 and summarized in Table 3. Both clusters followed the same temporal pattern, with the highest OH rate in the immediate post-0–6 h window, the lowest rate in the post-6–12 h window, and intermediate rates in the pre-12 h and post-12–24 h windows. The largest between-group separation occurred in the post-0–6 h window (median 4.1 vs. 2.0 mL/kg/h, p < 0.001), meaning the high-OH cluster drank more than twice as fast in this interval. Outside the 0–6 h window median, OH rates were similar between clusters or showed only small differences: pre-12 h, 1.0 vs. 0.9 mL/kg/h (p = 0.299); post-6–12 h, 0.9 vs 0.6 mL/kg/h (p = 0.208); post-12–24 h, 1.1 vs. 0.8 mL/kg/h (p = 0.038). Among participants with complete 24-h records (n = 54 per cluster), cumulative 24-h OH was substantially greater in the high-OH cluster (43.6 [33.3, 52.3] vs. 27.0 [19.4, 33.0] mL/kg, p < 0.001). Baseline comparisons (Table 1) showed higher BMI in the low-OH cluster (25.9 [23.9, 28.1] vs. 24.5 [23.2, 26.6]; p = 0.016) and cluster differences in indication for angiography (p = 0.006): non-ST-elevation myocardial infarction (NSTEMI) occurred only in the low-OH group (10.6% vs. 0.0%), while unstable angina was more frequent in the high-OH group (70.8% vs. 56.5%). Heart failure (10.6% vs. 2.2%), ACEI/ARB use (63.5% vs. 48.3%), diuretic use (11.8% vs. 3.4%), and metformin use (36.5% vs. 15.7%) were all more common in the low-OH cluster (p < 0.05). Preprocedural potassium was slightly higher in low-OH (3.93 vs. 3.87; p = 0.045). Other comorbidities, medications, laboratory, and procedural factors did not differ significantly (p > 0.05). These differences likely reflect distinct clinical presentations and background therapy rather than causal effects of hydration patterns; therefore, cluster membership was treated as a descriptive grouping, and outcome analyses were adjusted for key baseline covariates.
Fig. 2.
Individual and cluster-mean oral hydration rate trajectories for k = 2. Clustering by partial-overlap dynamic time warping and partitioning around medoids. Thin lines represent individual trajectories and thick lines show cluster means. Oral hydration rates are weight adjusted and expressed in mL/kg/h for four prespecified intervals. OH, oral hydration.
Table 3.
Cluster-specific OH rates at 4 time segments
| Cluster | OH rates, mL/kg/h | |||
|---|---|---|---|---|
| pre-12 h | post-0–6 h | post-6–12 h | post-12–24 h | |
| Low-OH (n = 85) | 0.9 (0.5, 1.4) | 2.0 (1.4, 2.5) | 0.6 (0.3, 1.2), n = 68 | 0.8 (0.5, 1.2), n = 54 |
| High-OH (n = 89) | 1.0 (0.6, 1.4) | 4.1 (3.6, 5.1) | 0.9 (0.2, 1.5), n = 71 | 1.1 (0.5, 1.6), n = 54 |
| p value | 0.299 | <0.001 | 0.208 | 0.038 |
Continuous variables were expressed as interquartile range.
OH, oral hydration.
Renal Function Outcomes and Association with OH Post-6 h
Post-procedural renal indices are shown in Table 4. Median serum creatinine change was −2.2 µmol/L (IQR −7.6 to 2.9), with individual changes ranging from −38.9 to +19.8 µmol/L. The median percent change was −2.3% (IQR −8.5% to 3.6%). No patient met KDIGO criteria for AKI within the 24-h observation window.
Table 4.
Renal function outcomes overall and by hydration pattern
| Characteristic | Overall (n = 174) | Low-OH cluster (n = 85) | High-OH cluster (n = 89) | p value |
|---|---|---|---|---|
| SCr | ||||
| Post, μmol/L | 82.2 (71.4, 93.5) | 81.9 (71.6, 94.2) | 82.7 (69.8, 92.3) | 0.659 |
| Change, μmol/L | −2.2 (−7.6, 2.9) | −2.3 (−10.3, 3.5) | −1.8 (−5.8, 2.5) | 0.555 |
| Change proportion, % | 2.3 (−8.5, 3.6) | −2.2 (−10.0, 3.4) | −2.9 (−6.2, 3.2) | 0.589 |
| eGFR | ||||
| Post, mL/min/1.73 m2 | 81.7 (68.2, 90.6) | 78.8 (65.0, 91.4) | 82.7 (70.9, 89.9) | 0.354 |
| Change, mL/min/1.73 m2 | 1.6 (−2.4, 6.4) | 1.5 (−2.8, 6.8) | 1.8 (−2.5, 5.5) | 0.771 |
| Change proportion, % | 2.26 (−3.0, 8.5) | 2.6 (−3.2, 8.8) | 2.0 (−2.6, 7.5) | 0.774 |
Continuous variables were expressed as interquartile range.
OH, oral hydration; SCr, serum creatinine; eGFR, estimated glomerular filtration rate.
In adjusted GAMs, the 0–6-h OH rate was the primary exposure, and models controlled for baseline eGFR, age, comorbidities, indication for CAG, contrast volume, and intravenous hydration. The 0–6-h OH rate was not a significant smooth term (edf = 3.79, p = 0.795), indicating no dose-response relationship with early serum creatinine change (online suppl. Fig. S3, S4). The final model explained 38.4% of the deviance (adjusted R2 = 0.272), consistent with the conclusion that OH volume does not independently determine early postoperative creatinine change after adjustment for clinical and procedural factors.
Short-Term Physiologic Response and Safety by Hydration Cluster
Post-procedural fluid metrics differed by cluster (Table 5). Patients in the high-OH cluster had a markedly greater urine flow rate within 6 h and a larger positive fluid balance than those in the low-OH cluster (both p < 0.001). Serum potassium and its absolute and relative changes were similar between clusters (p > 0.05). These findings are compatible with higher oral intake and greater renal excretion in the high-OH cluster but do not establish a causal effect of hydration pattern on renal or electrolyte outcomes.
Table 5.
Short-term physiologic response and safety by hydration cluster: post-0–6-h urine flow rate, positive fluid balance, and electrolyte
| Characteristic | Overall (n = 174) | Low-OH cluster (n = 85) | High-OH cluster (n = 89) | p value |
|---|---|---|---|---|
| Post-0–6-h urine flow rate, mL/kg/h | 2.9 (1.9, 3.9) | 2.1 (1.5, 2.9) | 3.6 (2.7, 4.7) | <0.001 |
| Positive fluid balance, L | 0.1 (−0.2, 0.4) | −0.1 (−0.4, 0.2) | 0.3 (0.0, 0.6) | <0.001 |
| K | ||||
| Post, mmol/L | 3.6 (3.5, 3.9) | 3.69 (3.4, 4.0) | 3.61 (3.5, 3.8) | 0.133 |
| Change, mmol/L | −0.2 (−0.5, 0.0) | −0.3 (−0.6, 0.0) | −0.2 (−0.4, −0.0) | 0.219 |
| Change proportion, % | −6.2 (−12.0, 0.0) | −6.8 (−13.6, 0.8) | −4.6% (−10.6, −0.3) | 0.269 |
Continuous variables were expressed as interquartile range.
OH, oral hydration; K, serum potassium.
Sensitivity Analyses
The 18 patients excluded for missing 6-h OH data or post-procedure serum creatinine did not differ from the analytic cohort (n = 174) in age, sex, indication for the coronary procedure, baseline serum creatinine, baseline eGFR, contrast volume, stent implantation, or Mehran score, supporting a missing-at-random mechanism (online suppl. Table S2). Repeating clustering and outcome analyses with k = 3 produced similar results: the higher hydration cluster again showed greater 6-h urine output and larger positive fluid balance, while early serum-creatinine change remained unrelated to OH volume (online suppl. Tables S3, S4).
Discussion
In this single-center prospective cohort of 174 patients undergoing CAG or intervention, we identified two reproducible periprocedural OH trajectories that diverged mainly in the immediate post-procedure window (0–6 h). However, neither cluster membership nor the 0–6-h OH rate showed a statistically significant association with early postoperative serum-creatinine change. Early drinking behavior altered short-term fluid balance and diuresis but did not produce detectable differences in creatinine-based renal indices within our 24 h, underscoring limitations of short-interval creatinine sampling for detecting early renal effects.
Compared with prior reports, OH in our cohort was substantial. Median intake was 18.6 mL/kg in the first 6 h and about 33 mL/kg by 24 h. These values exceed thresholds proposed in earlier studies. Xie et al. [15] observed benefit when 24-h OH exceeded 15.6 mL/kg, and Song et al. [14] used a ≥12 mL/kg 24-h target in their trial design. Many studies combined oral intake with routine intravenous hydration, typically around 1 mL/kg/h [6, 14, 15], whereas only 14.4% of our cohort received intravenous hydration. Large randomized trials such as PRESERVE and AMACING have not reported a substantial benefit of routine prophylactic intravenous hydration over simpler standard care in reducing CA-AKI, using creatinine-based definitions that differ from KDIGO criteria [23, 24]. However, a recent systematic review showed that intravenous fluids can reduce contrast-induced nephropathy and that OH may be similarly effective, although heterogeneity in AKI definitions and study designs should be acknowledged [10]. In line with this mixed evidence, the updated NICE guidance [6] encourages OH in at-risk adults, reserving intravenous volume expansion for inpatients at particularly high risk. For low-risk patients, hypovolemia should still be avoided. OH is feasible and cost-effective. In short-stay practice, intravenous hydration is therefore often reserved for high-risk patients, and supervised 24-h OH is often impractical [16, 17]. These observations support focusing pragmatic interventions on the immediate post-procedural window, which is both biologically plausible for accelerating contrast clearance and feasible in fast-turnover care pathways.
We extended prior work by characterizing periprocedural OH as a time-resolved pattern instead of a single 12- or 24-h summary. Oral intake was summarized in four consecutive windows and clustered by partial-overlap DTW with partitioning around medoids. The two-cluster solution showed acceptable separation (mean silhouette 0.393) and high bootstrap stability (mean ARI 0.845). We identified two reproducible trajectories that diverged mainly in the immediate post-procedure period, with median 0–6 h rates of 4.1 versus 2.0 mL/kg/h in the high-OH and low-OH clusters, respectively. Early concentrated intake is clinically plausible because patients are more attentive to instructions and nursing staff perform frequent rounds that encourage drinking and monitoring of urine output, whereas the post-6–12 h window often overlaps with sleep and lower oral intake. Mechanistically, iodinated contrast is principally renally excreted, and a large fraction of the injected dose is eliminated during the first hours after administration, thus supporting intravascular volume and urine flow immediately after exposure; therefore, it can accelerate contrast clearance and reduce tubular exposure [16, 25]. Prior studies that reported protective associations used 24-h cumulative targets and frequently combined oral fluids with routine intravenous saline [14, 15], which limits direct comparison with our time-resolved, largely OH data. Early concentrated oral intake in the 0–6 h window may be both a feasible strategy in short-stay care and a rational target for randomized evaluation. Technical advances and reductions in contrast volume have lowered the overall risk of contrast-associated AKI, and same-day discharge pathways mean that an increasing proportion of patients stay in the hospital for less than 6 h. In these settings, structured OH strategies need to be specifically tested.
The cluster separation can be explained by clinical and physiological differences between groups. The low-OH cluster had higher prevalences of heart failure and diuretic use, conditions that prompt clinicians and patients to restrict oral fluids to avoid volume overload and thereby reduce observed intake [26, 27]. Higher body mass index in some participants reduces weight-adjusted hydration rates even when absolute intake is similar, amplifying apparent between-group differences. Acute presentations such as NSTEMI and background exposures, including ACEI or ARB therapy and metformin, reflect more complex clinical status and closer periprocedural management, which can both limit oral intake and alter renal hemodynamics and contrast handling [8, 28]. For these reasons, cluster membership likely reflects a composite of clinical presentation and background therapy rather than a pure causal exposure. Accordingly, we treated clusters as descriptive phenotypes, adjusted outcome models for key baseline covariates, and conducted sensitivity checks, including a k = 3 partition; these steps reduce but do not eliminate residual confounding, and randomized trials will be needed to establish causality.
Kidney injury is often multifactorial and may reflect illness severity as well as direct contrast nephrotoxicity. Hemodynamic instability, nephrotoxic drugs, atheroembolism, and reduced renal reserve can affect post-procedural creatinine [1]. We adjusted models for baseline comorbidities, concurrent medications, and procedural factors. In adjusted GAMs, the 0–6-h OH rate was not significantly associated with early post-procedural serum creatinine change (edf = 3.79, p = 0.795; adjusted R2 = 0.272). No patient met KDIGO criteria for AKI during the 24-h observation window. The modest and uniform decrease in serum creatinine observed across the cohort should not be interpreted as improved renal function. It is probably because serum creatinine is a delayed marker that often does not peak within 24 h and may rise more clearly at 48–72 h or later, so early sampling can miss true injury [19, 29, 30]. Second, substantial periprocedural volume expansion can transiently dilute serum creatinine and blunt short-term rises, masking small degrees of injury [28, 31, 32]. Additionally, the cohort was relatively low risk for CA-AKI, for more than 75% of participants had preprocedural eGFR ≥60 mL/min/1.73 m2, and nearly 70% had low Mehran risk, which may partly explain the absence of KDIGO-defined AKI within 24 h and limits generalizability to higher-risk populations. For these reasons, a single creatinine measurement within 24 h is an imperfect sentinel of early kidney injury after contrast exposure in short-stay practice. Future studies may include earlier or more sensitive biomarkers, such as plasma or urinary neutrophil gelatinase-associated lipocalin(NGAL) and cystatin C, to distinguish hemodilution from genuine renal injury and to more sensitively evaluate hydration interventions [33–35].
Urine flow rate, positive fluid balance, and serum potassium were included to complement creatinine and eGFR as indicators of the physiological impact and safety of differing OH patterns. The high-OH cluster showed substantially greater early urine flow rate (3.6 vs. 2.1 mL/kg/h, p < 0.001) and a larger short-interval positive fluid balance (+0.3 vs. −0.1 L, p < 0.001), while serum potassium remained clinically stable across clusters. Increased immediate oral intake increases intravascular volume and urine flow, promoting renal clearance and transiently raising urine output without large acute electrolyte shifts in most patients [26, 36]. From a pragmatic viewpoint, short-interval process measures such as 0–6-h urine volume and net fluid balance are useful for implementing and auditing OH protocols, but patients with heart failure, ongoing diuretics, or other volume-sensitive conditions require individualized targets and closer monitoring [27]. Future trials should test early concentrated OH strategies while measuring sensitive renal biomarkers to separate true injury from dilutional effects.
This study has several strengths and some important limitations. We prospectively quantified oral intake in four prespecified time windows and measured 6-h urine output, yielding time-resolved data that improve on prior reports using only 12- or 24-h aggregates. The cohort represents contemporary short-stay practice in which routine intravenous hydration is uncommon, so our observation that the first 6 h is a key and practicable window is directly relevant to current care pathways. We used a partial-overlap DTW clustering approach with bootstrap validation to identify reproducible trajectory phenotypes and performed sensitivity checks comparing excluded and included patients, which argues against major selection bias from the 18 cases missing core variables. Limitations include a single-center and relatively low-risk cohort, which may limit generalizability. Adequately powered subgroup analyses in high-risk groups were not performed because of the limited sample size. Because follow-up and in-hospital sampling were limited to the early post-procedure period, this study cannot estimate the full KDIGO incidence of CA-AKI. The serum creatinine often peaks later and can be transiently diluted by substantial fluid loading, both factors that reduce the sensitivity of short-interval creatinine endpoints. We therefore recommend that future studies measure earlier kidney injury biomarkers such as cystatin C or NGAL and extend follow-up beyond 24 h. In addition, the applicability of our findings to same-day discharge settings with very short (<6-h) post-procedural observation is limited. Finally, preprocedural fluid intake was self-reported. Despite measuring and labelling drinking cups and instructing patients to record volumes, some incomplete or inaccurate recording was unavoidable. Although we applied pragmatic missing-data rules, residual confounding and unmeasured factors that influence both drinking patterns and renal outcomes cannot be excluded, and randomized trials will be required to establish causality.
Conclusion
In this cohort of patients undergoing CAG or intervention, periprocedural OH formed two reproducible trajectories that diverged mainly in the immediate post-procedural period. Early concentrated oral intake markedly increased urine output and short-term positive fluid balance but did not show a significant association with early creatinine change within the first 24 h. This null finding should be interpreted with caution, given the cohort’s generally low AKI risk, the short follow-up interval, and limited statistical power to detect small effects. The first 6 h post-procedure may be a feasible target for pragmatic hydration strategies in short-stay care. Because creatinine is a delayed marker and hemodilution may obscure small changes, future randomized studies that include earlier biomarkers such as cystatin C or NGAL and that focus on high-risk patients are required to determine whether early concentrated OH confers renal protection.
Acknowledgments
We thank the nurses and ward staff of the three cardiovascular wards at Peking University First Hospital for their assistance with patient recruitment and real-time data collection. Their support was essential to the successful completion of this study.
Statement of Ethics
The study protocol was approved by the Institutional Review Board of Peking University First Hospital (Approval No.: 2024 Yan 418-002, 23 August 2024), and the first patient was enrolled on 13 November 2024. Written informed consent was obtained from all participants prior to study procedures.
Conflict of Interest Statement
The authors have no conflicts of interest to declare.
Funding Sources
The project was funded by the National High Level Hospital Clinical Research Funding (Scientific Research Seed Fund of Peking University First Hospital) (2024SF69). The funder had no role in the design, data collection, data analysis, and reporting of this study.
Author Contributions
Study conception, design, and statistical analysis: Dan Zhu and Lingling Gao. Data collection and interpretation: Dan Zhu, Nanhui Zhuang, Yuan Ma, Lijing Fan, and Haiyan Ma. Drafting of the manuscript: Dan Zhu. Critical revision of the manuscript: Dan Zhu, Lingling Gao, and Yimei Zheng. Supervision: Yimei Zheng. All authors read and approved the final manuscript and agree to be accountable for all aspects of the work, ensuring the integrity and accuracy of the study.
Funding Statement
The project was funded by the National High Level Hospital Clinical Research Funding (Scientific Research Seed Fund of Peking University First Hospital) (2024SF69). The funder had no role in the design, data collection, data analysis, and reporting of this study.
Data Availability Statement
The data that support the findings of this study are not publicly available due to their containing information that could compromise the privacy of research participants but are available from the corresponding author upon reasonable request.
Supplementary Material.
Supplementary Material.
Supplementary Material.
Supplementary Material.
Supplementary Material.
Supplementary Material.
Supplementary Material.
Supplementary Material.
Supplementary Material.
References
- 1. Mehran R, Dangas GD, Weisbord SD. Contrast-associated acute kidney injury. N Engl J Med. 2019;380(22):2146–55. [DOI] [PubMed] [Google Scholar]
- 2. Chinese Society of Clinical Pharmacy; Hospital Pharmacy Professional Committee of Chinese Pharmaceutical Association; Chinese Society of Nephrology . Expert consensus on prevention and treatment of iodine contrast media-induced acute kidney injury. Chin J Nephrol. 2022;38(3):265–88. [Google Scholar]
- 3. Azzalini L, Kalra S. Contrast-induced acute kidney injury-definitions, epidemiology, and implications. Interv Cardiol Clin. 2020;9(3):299–309. [DOI] [PubMed] [Google Scholar]
- 4. Azzalini L, Candilio L, McCullough PA, Colombo A. Current risk of contrast-induced acute kidney injury after coronary angiography and intervention: a reappraisal of the literature. Can J Cardiol. 2017;33(10):1225–8. [DOI] [PubMed] [Google Scholar]
- 5. McCullough PA, Choi JP, Feghali GA, Schussler JM, Stoler RM, Vallabahn RC, et al. Contrast-induced acute kidney injury. J Am Coll Cardiol. 2016;68(13):1465–73. [DOI] [PubMed] [Google Scholar]
- 6. NICE Acute kidney injury: prevention, detection and management National Institute for Health and Care Excellence (NICE); 2023. http://www.ncbi.nlm.nih.gov/books/NBK552160/(Accessed November 12, 2023). [PubMed] [Google Scholar]
- 7. Moroni F, Baldetti L, Kabali C, Briguori C, Maioli M, Toso A, et al. Tailored versus standard hydration to prevent acute kidney injury after percutaneous coronary intervention: network meta-analysis. J Am Heart Assoc. 2021;10(13):e021342. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Almendarez M, Gurm HS, Mariani J, Montorfano M, Brilakis ES, Mehran R, et al. Procedural strategies to reduce the incidence of contrast-induced acute kidney injury during percutaneous coronary intervention. JACC Cardiovasc Interv. 2019;12(19):1877–88. [DOI] [PubMed] [Google Scholar]
- 9. Kong DG, Hou YF, Ma LL, Yao DK, Wang LX. Comparison of oral and intravenous hydration strategies for the prevention of contrast-induced nephropathy in patients undergoing coronary angiography or angioplasty: a randomized clinical trial. Acta Cardiol. 2012;67(5):565–9. [DOI] [PubMed] [Google Scholar]
- 10. Zaki HA, Bashir K, Iftikhar H, Alhatemi M, Elmoheen A. Evaluating the effectiveness of pretreatment with intravenous fluid in reducing the risk of developing contrast-induced nephropathy: a systematic review and meta-analysis. Cureus. 2022;14(5):e24825. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Angoulvant D, Cucherat M, Rioufol G, Finet G, Beaune J, Revel D, et al. Preventing acute decrease in renal function induced by coronary angiography (PRECORD): a prospective randomized trial. Arch Cardiovasc Dis. 2009;102(11):761–7. [DOI] [PubMed] [Google Scholar]
- 12. Chen YQ, Xiong Y, Li GQ, et al. Meta-analysis of effect of oral hydration for preventing contrast-induced nephropathy after percutaneous coronary intervention. Chin Evid-based Nurs. 2018;4(9):780–6. [Google Scholar]
- 13. Akyuz S, Karaca M, Kemaloglu OT, Altay S, Gungor B, Yaylak B, et al. Efficacy of oral hydration in the prevention of contrast-induced acute kidney injury in patients undergoing coronary angiography or intervention. Nephron Clin Pr. 2014;128(1–2):95–100. [DOI] [PubMed] [Google Scholar]
- 14. Song F, Sun G, Liu J, Chen JY, He Y, Liu L, et al. Efficacy of post-procedural oral hydration volume on risk of contrast-induced acute kidney injury following primary percutaneous coronary intervention: study protocol for a randomized controlled trial. Trials. 2019;20(1):290. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Xie W, Zhou Y, Liao Z, Lin B. Effect of oral hydration on contrast-induced acute kidney injury among patients after primary percutaneous coronary intervention. Cardiorenal Med. 2021;11(5–6):243–51. [DOI] [PubMed] [Google Scholar]
- 16. van der Molen AJ, Dekkers IA, Geenen RWF, Bellin MF, Bertolotto M, Brismar TB, et al. Waiting times between examinations with intravascularly administered contrast media: a review of contrast media pharmacokinetics and updated ESUR contrast media safety committee guidelines. Eur Radiol. 2024;34(4):2512–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Madan M, Bagai A, Overgaard CB, Fang J, Koh M, Cantor WJ, et al. Same-day discharge after elective percutaneous coronary interventions in Ontario, Canada. J Am Heart Assoc. 2019;8(13):e012131. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Li Y, Wang J. Contrast-induced acute kidney injury: a review of definition, pathogenesis, risk factors, prevention and treatment. BMC Nephrol. 2024;25(1):140. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Khwaja A. KDIGO clinical practice guidelines for acute kidney injury. Nephron Clin Pract. 2012;120(4):c179–184. [DOI] [PubMed] [Google Scholar]
- 20. Matheny ME, Carpenter-Song E, Ricket IM, Solomon RJ, Stabler ME, Davis SE, et al. Sustained improvements after intervention to prevent contrast‐associated acute kidney injury: a randomized controlled trial. J Am Heart Assoc. 2025;14(10):e038920. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Lu X, Li Q, Chen W, Deng J, Shi S, Huang H, et al. Effect of missed post-procedure creatinine measurement on sub-acute kidney injury following coronary angiography. Angiology. 2025;76(7):681–9. [DOI] [PubMed] [Google Scholar]
- 22. Sterne JAC, White IR, Carlin JB, Spratt M, Royston P, Kenward MG, et al. Multiple imputation for missing data in epidemiological and clinical research: potential and pitfalls. BMJ. 2009;338:b2393. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Weisbord SD, Gallagher M, Jneid H, Garcia S, Cass A, Thwin SS, et al. Outcomes after angiography with sodium bicarbonate and acetylcysteine. N Engl J Med. 2018;378(7):603–14. [DOI] [PubMed] [Google Scholar]
- 24. Nijssen EC, Rennenberg RJ, Nelemans PJ, Essers BA, Janssen MM, Vermeeren MA, et al. Prophylactic hydration to protect renal function from intravascular iodinated contrast material in patients at high risk of contrast-induced nephropathy (AMACING): a prospective, randomised, phase 3, controlled, open-label, non-inferiority trial. Lancet Lond Engl. 2017;389(10076):1312–22. [DOI] [PubMed] [Google Scholar]
- 25. Pistolesi V, Regolisti G, Morabito S, Gandolfini I, Corrado S, Piotti G, et al. Contrast medium induced acute kidney injury: a narrative review. J Nephrol. 2018;31(6):797–812. [DOI] [PubMed] [Google Scholar]
- 26. Pickering JW, Ralib AM, Endre ZH. Combining creatinine and volume kinetics identifies missed cases of acute kidney injury following cardiac arrest. Crit Care. 2013;17(1):R7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Bei WJ, Wang K, Li HL, Guo XS, Guo W, Abuduaini T, et al. Safe hydration to prevent contrast-induced acute kidney injury and worsening heart failure in patients with renal insufficiency and heart failure undergoing coronary angiography or percutaneous coronary intervention. Int Heart J. 2019;60(2):247–54. [DOI] [PubMed] [Google Scholar]
- 28. Jin J, Chang SC, Xu S, Xu J, Jiang W, Shen B, et al. Early postoperative serum creatinine adjusted for fluid balance precisely predicts subsequent acute kidney injury after cardiac surgery. J Cardiothorac Vasc Anesth. 2019;33(10):2695–702. [DOI] [PubMed] [Google Scholar]
- 29. Liu Y, Duan C, Wang K, Bei WJ, Guo XS, Li HL, et al. Could late measurement of serum creatinine be missed for patients without early increase in serum creatinine following coronary angiography? Medicine. 2017;96(50):e8460. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Modi K, Padala SA, Gupta M. Contrast-induced nephropathy. StatPearls StatPearls Publishing; 2025. http://www.ncbi.nlm.nih.gov/books/NBK448066/(Accessed July 8, 2025). [PubMed] [Google Scholar]
- 31. Macedo E, Bouchard J, Soroko SH, Chertow GM, Himmelfarb J, Ikizler TA, et al. Fluid accumulation, recognition and staging of acute kidney injury in critically-ill patients. Crit Care. 2010;14(3):R82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Jin J, Xu J, Xu S, Hu J, Jiang W, Shen B, et al. Hemodilution is associated with underestimation of serum creatinine in cardiac surgery patients: a retrospective analysis. BMC Cardiovasc Disord. 2021;21(1):61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Devarajan P. Neutrophil gelatinase-associated lipocalin: a promising biomarker for human acute kidney injury. Biomark Med. 2010;4(2):265–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Quintavalle C, Anselmi CV, De Micco F, Roscigno G, Visconti G, Golia B, et al. Neutrophil gelatinase-associated lipocalin and contrast-induced acute kidney injury. Circ Cardiovasc Interv. 2015;8(9):e002673. [DOI] [PubMed] [Google Scholar]
- 35. Briguori C, Visconti G, Rivera NV, Focaccio A, Golia B, Giannone R, et al. Cystatin C and contrast-induced acute kidney injury. Circulation. 2010;121(19):2117–22. [DOI] [PubMed] [Google Scholar]
- 36. Andreucci M, Solomon R, Tasanarong A. Side effects of radiographic contrast media: pathogenesis, risk factors, and prevention. Biomed Res Int. 2014;2014:741018. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are not publicly available due to their containing information that could compromise the privacy of research participants but are available from the corresponding author upon reasonable request.
