CARDIOVASCULAR JOURNAL OF AFRICA • Volume 35, No 3, September – October 2024 AFRICA 161 As a known determinant of AKI, patients with major bleeding were excluded as well. Major bleeding was defined as > 3 g/l drop in haemoglobin level, and bleeding requiring transfusion or surgery. Patients who died immediately after (< 48 hours) PCI was not included either. Patients with missing baseline or follow-up laboratory or procedural data were excluded. Hospital records were used to detect AKI after the index procedure. Follow-up laboratory was defined as either in-hospital (generally for ACS patients) or short-term visit (generally for elective procedures). The estimated glomerular filtration rate (eGFR) was calculated according to the simplified modification of diet in renal disease formula.15 Lastly, patients on long-term dialysis were excluded as the diagnosis of AKI is not applicable. As per institutional protocol, patients with chronic renal failure (eGFR < 60 ml/h/1.73 m2), ACS, and over 70 years old were hydrated with normal saline, starting before the procedure. The usual protocol was 1 ml/kg/h of normal saline infused for six hours before, and 0.5 ml/kg/h for 12 hours after a procedure. Access site (radial or femoral) and size of catheters (6F or 7F) were at the operators’ discretion. Low osmolar, non-ionic iohexol was used exclusively at the time of the study. Ad hoc interventions were considered as one procedure. Contrast volume and procedure time were calculated from the beginning of diagnostic angiography. Procedural anticoagulation was done with unfractionated heparin 50–70 U/ kg, as per institutional protocol. Target vessels, procedure time, contrast volume, procedural complications and post-procedural deaths were recorded. The primary endpoint of the study was the development of AKI after PCI, which was defined as an increase in serum creatinine (SCr) level of 0.5 mg/dl or 25% from the baseline. Secondary endpoints were absolute (mg/dl) or proportional (%) changes in SCr level, an increase in SCr level of > 0.3 and > 0.5 mg/dl, and an increase in SCr level of > 25 and > 50%. Statistical analysis SPSS Statistics for Windows, version 22.0 (Armonk, NY: IBM Corp) was used for the statistical analysis. Continuous variables are presented as means and standard deviations (SD) or medians and interquartile ranges (IQR), according to the normality of distribution. Comparison for continuous variables between the radial and femoral groups were performed either with the t-test or Mann–Whitney U-test, as appropriate. Categorical data are presented as numbers and percentages, and compared using chi-squared or Fisher’s exact test, as appropriate. A propensity score (PS) analysis was performed to alleviate the selection bias and other clinical imbalances between the radial and femoral groups due to demographic, laboratory, clinical and procedural characteristics. PS matching was performed using the R-essentials plugin in SPSS Statistics for Windows. The nearest neighbour 1:1 matching with a caliper of 0.1 was performed. The radial and femoral groups were matched for these variables: age, gender, hypertension, diabetes, history of coronary intervention (PCI and/or coronary artery bypass graft), baseline serum eGFR, white blood cells, haemoglobin, blood cholesterol, clinical presentation (elective or ACS), contrast volume and treated vessel [left anterior descending artery (LAD), circumflex artery (CX), right coronary artery (RCA)]. In addition, independent predictors of AKI following PCI were determined using logistic regression analysis; the model included vascular access site in addition to the aforementioned variables. Odds ratios (OR) and 95% confidence intervals (CI) were reported. Two-sided p < 0.05 was considered significant. Results A total of 463 patients were identified. We excluded 124 patients with missing laboratory or procedural data (n = 96), acute/ decompensated heart failure (n = 8), major bleeding (n = 5), haemodynamic instability (n = 3), long-term dialysis (n = 6) and mortality (n = 6), and the remaining 339 patients made up our study population. The radial group included 134 patients (40%) and the femoral group comprised 205 patients (60%). A total of seven (2.1%) access site crossovers was observed. The reasons for a radial-to-femoral crossover were failure to access the radial artery in three patients (2.2%) and severe tortuosity of the brachiocephalic artery in two patients (1.5%). A femoral-to-radial crossover was needed in severe iliac tortuosity in two patients (1%). Procedures included in the radial or femoral groups according to the access site procedure were successfully carried out. Demographic, laboratory and procedural data are presented in Table 1. There was no difference between the two groups in terms of baseline demographic characteristics. There was also no difference between the two groups in the laboratory data, except for white blood cell count, which was higher in the femoral group. There were 110 patients (32%) undergoing elective PCI, 137 patients (40%) with non-ST segment elevation myocardial infarction (NSTEMI), and 92 patients (27%) with ST-segment elevation myocardial infarction (STEMI). The operators more frequently chose the TRA in elective procedures, whereas TFA was more common in STEMI. In NSTEMI, the rate of TRA and TFA were similar. Ad hoc and/or primary PCI were performed more commonly through the femoral route. The distribution of target vessels was comparable between the groups. Procedure time and contrast volume used were also similar. Total failed procedures were five (1.5%) overall, one (0.7%) in the TRA and four (2.0%) in the TFA (p = 0.652). After PS matching, we had a well-balanced population of 182 patients (Table 1). SCr level was controlled after the procedure at a median of three days (range two to seven days). According to our definition, AKI developed in 35 patients (10.3%) in the overall study population. The differences between the TRA and TFA in AKI incidence were not significant in both the overall (9.0 vs 11.2%, p = 0.503) and PS-matched (9.9 vs 7.7%, p = 0.601) study population (Fig. 1). Multivariate logistic regression analysis of AKI following PCI is presented in Table 2. The overall success of the models for unmatched and matched patients was 90.2 and 93.2%, respectively. Age, female gender, baseline SCr level, baseline eGFR and contrast volume were independent predictors of AKI in all patients. The same predictors were found in the matched patients; the only exception was baseline SCr level.
RkJQdWJsaXNoZXIy NDIzNzc=