Radial Versus Femoral Access for Diagnostic Coronary Angiography: Insights From a Tertiary Academic Center in Johannesburg Authors: Khulasande Liso Ntaka, Dineo Mpanya, Umar G Adamu, Thomas Kalk, Nqoba Tsabedze Published: 2026 ABSTRACT ABSTRACTBackgroundThe radial artery is the preferred access for performing diagnostic coronary angiograms (DCA). However, data on the safety and efficacy of radial compared to femoral access in low‐ and middle‐income countries is limited.AimsThis study aims to evaluate the safety and efficacy of radial versus femoral access in patients referred for a DCA.MethodsWe conducted a cross‐sectional, retrospective study on patients referred for DCA to evaluate the safety and efficacy of radial versus femoral access routes.ResultsThe study population comprised 653 patients with a mean age of 58.2 ± 12.7 years, of which 408 (62.5%) were males. Radial access was utilized in 318 (48.7%) patients. The median duration of the radial access was 50 min (Interquartile range [IQR]: 40–60) and was longer than that of the femoral artery access, which had a median duration of 45 min (IQR: 35–60) (p = 0.010). The median total radiation dose emitted during the DCA in the femoral access group was 3511 milligray per square meter (mGy/m²), (IQR: 2154–5821), and patients in the radial access group received a median radiation dose of 4011 mGy/m² (IQR: 2298–6411) (p = 0.0661). The mean contrast volume administered in the radial and femoral artery access group did not differ significantly, measuring 83.7 ± 36.1 milliliters (mL) and 88.9 ± 37.1 mL, respectively (p = 0.073).ConclusionsRadial access was used in nearly half of all DCA, reflecting its growing adoption in clinical practice. These findings support the use of radial access as a safe and viable alternative to femoral access, even in resource‐limited settings.Keywords: contrast media, coronary angiography, femoral access, radial access, radiation exposure, vascular access Introduction In the recent decade, interventional cardiology centers have adopted a radial access‐first policy internationally. In some countries, the radial artery has replaced the femoral artery as the primary vascular access route for percutaneous coronary intervention (PCI) [1]. This is mainly due to the evident advantages of radial artery access over femoral artery access. These advantages include fewer vascular complications such as hemorrhage, pseudoaneurysms, arteriovenous fistula, arterial dissection, and major adverse cardiovascular events (MACE) [2, 3]. Despite the increasing prevalence of cardiovascular disease (CVD) and access to catheterization suites in low‐ and middle‐income countries (LMIC), there is a paucity of data on the commonly used access during diagnostic coronary angiograms (DCA). Ekou et al. conducted a study evaluating the role of PCI in managing ST‐segment elevation myocardial infarction (STEMI) in sub‐Saharan Africa (SSA). Femoral access was the most commonly used vascular access route in SSA when performing coronary angiograms. Also, in their study, femoral access accounted for most non‐fatal vascular complications [4]. Similarly, most international studies associate femoral access with more vascular complications when compared to radial access [5, 6]. The ease of post‐procedural care and shorter hospital stays are additional advantages associated with radial access [7, 8]. Although there are several studies on the safety and efficacy of radial versus femoral access, only a few have been published in LMIC. Therefore, this study aims to describe the clinical and procedural factors of the transradial compared to the transfemoral approach for coronary angiography in the Division of Cardiology at the Charlotte Maxeke Johannesburg Academic Hospital (CMJAH) in Johannesburg, South Africa. Methods Study Design, Setting and Population We conducted a cross‐sectional, single‐center, observational study and retrospectively reviewed inpatient medical records and DCA reports of consecutive patients referred to the CMJAH between January 1, 2018 and December 31, 2019. The CMJAH is one of the state‐owned facilities in Johannesburg, and it is equipped with an intensive care unit, cardiac admission wards, outpatient services, and electrophysiology and catheterization laboratories. The catheterization laboratory at the CMJAH is considered a high‐volume center and is accredited for training cardiology fellows. All patients aged 18 years and older who underwent DCA were included in the study. Patients referred for physiological or haemodynamic studies, those who had PCI performed during the same session as the DCA, and those with a history of previous coronary artery bypass graft (CABG) surgery were excluded. The rationale for excluding patients with prior CABG surgery lies in the prolonged or unpredictable procedural times observed when performing DCA in these individuals. Data Collection We documented demographic data, comorbidities, biochemical laboratory data, concomitant anticoagulant therapy, the left ventricular ejection fraction (LVEF), procedural findings. Biochemical data was extracted from the National Health Laboratory Services website. The Modification of Diet in Renal Disease (MDRD) equation was used to calculate the estimated glomerular filtration rate (eGFR) in mL/min/1.73m2. Angiographic data were obtained from the angiography report registry on the Research Electronic Data Capture (REDCap) database, the procedure book in the catheterization laboratory, filed hard copies of angiogram reports, and archived cine images. Variables such as vascular access route, procedural duration, contrast volume, total radiation dose and reported complications were also extracted. Procedural Factors Routinely, when performing DCA, lignocaine is used to anaesthetize the site of injection. After the puncture, verapamil and nitrocine are administered into the radial artery as prophylaxis against spasm. Heparin was also used to prevent vessel occlusion. After the procedure, the radial artery access was closed using an inflatable radial band, whereas femoral artery closure was achieved with manual compression followed by a compressive bandage. Since our institution is a public hospital, efforts are generally made to reduce costs without compromising patient care. Accordingly, vascular closure devices (VCD) were used for femoral access during the study period on a case‐by‐case basis rather than routinely for all patients. The procedural time was defined as the period from the patient's entry into the catheterization laboratory until their exit after the procedure. As such, the procedural time also incorporated the time required to apply the radial artery band, which was routinely applied while the patient was still in the catheterization laboratory. However, manual compression after femoral access was routinely done in the recovery room and was omitted when estimating the procedural time. The radiation dose was defined as the total amount of radiation emitted in the interventional lab during each procedure in milligray per square meter (mGy/m2). This information was obtained from the theater registry book. Study variables were captured in the REDCap database hosted by the University of the Witwatersrand. Study Outcomes The safety outcomes of interest included the total radiation dose, the total volume of contrast media used, and the prevalence of local vascular complications. Ethics Clearance Permission to conduct the study was obtained from the University of the Witwatersrand Human Research Ethics Committee (Clearance certificate number: M2011118) and relevant hospital authorities. Statistical Analysis The statistical analysis was conducted using Stata/SE version 17.0 (Stata CORP LLC, Texas). Categorical data is expressed using frequencies and percentages. Normally distributed continuous variables are expressed as the mean and standard deviation (±), while variables with a non‐normal distribution are presented using the median and the interquartile range (IQR). The student's t‐test was used to compare means of normal data, and the Pearson's Chi‐square test was used to compare categorical data. The Wilcoxon rank‐sum test was used to compare non‐normal data. Confidence intervals were set at 95% and a p‐value less than 0.05 was considered as a threshold for statistical significance. Results Demographic and Clinical Characteristics Between January 1, 2018 and December 31, 2019, 1189 DCA were performed (Figure 1). After excluding ineligible patients, the final study cohort comprised 653 patients, of which 408 (62.5%) were males. The mean age was 58.2 ± 12.7 years. Hypertension was the most common comorbidity, reported in 299 (45.8%) patients, followed by dyslipidaemia in 157 (24.0%), diabetes mellitus in 146 (22.4%) patients, and chronic kidney disease in 62 (9.5%). Study flow chart showing vascular access route. PCI = percutaneous coronary intervention. Diagnostic coronary angiography was performed as an elective procedure in 457 (81.6%) patients and as an emergency procedure in 103 (15.8%) patients. The remaining 96 patients had insufficient data to discern whether the DCA was scheduled as an elective or an emergency procedure. The femoral access route was used in 335 (51.3%) patients. Among the 157 patients with dyslipidaemia, 93 (59.2%) had transfemoral access (p = 0.022), and 18 of the 25 (72.0%) patients with atrial fibrillation had transradial access (p = 0.017). Among the patients with atrial fibrillation, 96% had an elective DCA. A total of 55 (8.4%) patients were prescribed warfarin, and 38 (69.0%) of these patients had radial access (p = 0.002). The median INR among patients on warfarin and those not taking warfarin was 1.84 (IQR: 1.22–2.45) versus 1.07 (IQR: 1.01–1.16), p < 0.001. There were no bleeding complications among patients on warfarin. In the entire study cohort, 202 (30.9%) patients were taking oral nonsteroidal anti‐inflammatory drugs (NSAID), of which 110 (54.5%) had radial access (p = 0.049). Among the 653 patients referred for DCA, 80 (12.2%) had underlying valvular heart disease, and 48 (60.0%) of these patients had radial access (p = 0.031). The rest of the baseline demographic and clinical characteristics did not differ between patients subjected to femoral and radial access during DCA (Table 1). Table 1Baseline demographic and clinical characteristics of the study cohort stratified according to the site of coronary angiography vascular access.All patients (n = 653)Transradial access (n = 318) 48.7%Transfemoral access (n = 335) 51.3% p value Clinical Parameters Age, years (mean, SD)58.2 ± 12.758.7 ± 12.557.7 ± 12.70.3231Males (%)408 (62.5)198 (62.3)210 (62.9)0.872Ethnicity (%)0.221Black African245 (37.6)127 (39.9)118 (35.3)Mixed Ancestry58 (8.9)33 (10.4)25(7.5)Indian104 (16)45 (14.2)59 (17.7)Caucasian245 (37.6)113 (35.5)132 (39.5)Hypertension (%)299 (45.8)141 (44.3)158 (47.2)0.469Diabetes Mellitus (%)146 (22.4)68 (21.4)78 (23.3)0.560HIV infection (%)27 (4.1)12 (3.8)15 (4.5)0.652Dyslipidaemia (%)157 (24)64 (20.1)93 (27.8)0.022Smoker (%)131 (20.1)60 (18.9)71 (21.2)0.458Previous smoker (%)77 (11.8)40(12.6)37 (11.0)0.544CKD (%)62 (9.5)26(8.2)36 (10.8)0.263Atrial Fibrillation (%)25 (3.8)7 (2.1)18 (5.7)0.017Previous CAD (%)83 (12.7)40 (12.6)43 (12.8)0.921No comorbidities (%)60 (9.2)29 (9.1)31 (9.3)0.953Warfarin (%)55 (8.4)38 (12)17 (5.1)0.002NSAID (%)202 (30.9)110 (34.6)92 (27.5)0.049Systolic BP, mmHg127.4 ± 26.4126.4 ± 27128.5 ± 25.90.4262Diastolic BP, mmHg75 ± 14.576 ± 1474 ± 150.1472eGFR, mL/min/1.73m2 (mean, SD)68.3 ± 26.768.1 ± 24.168.5 ± 290.8455Hb, g/dl (mean, SD)13.9 ± 2.214 ± 2.213.7 ± 2.10.1330Platelet count, 109/L (mean, SD)261.7 ± 92.8260.1 ± 88.5263.1 ± 96.70.6779INR, median [IQR]1.08 (1.01–1.2)1.09 (1.0–1.2)1.07 (1.02–1.16)0.2755LVEF, % (mean, SD)46.7 ± 15.746.9 ± 16.046.4 ± 15.50.7775 Note: Categorical data are represented as frequencies and percentages. Numerical data that is normally distributed is summarized as the mean and standard deviation (SD), while the median and the interquartile range (IQR) were used to summarize numerical data with a non‐normal distribution.Abbreviations: BP, blood pressure; CAD, coronary artery disease; CKD, chronic kidney disease; eGFR, estimated glomerular filtration rate; Hb, hemoglobin; HIV, human immunodeficiency virus; INR, International normalized ratio; LVEF, left ventricular ejection fraction; NSAID, Nonsteroidal anti‐inflammatory drugs. Baseline demographic and clinical characteristics of the study cohort stratified according to the site of coronary angiography vascular access. Note: Categorical data are represented as frequencies and percentages. Numerical data that is normally distributed is summarized as the mean and standard deviation (SD), while the median and the interquartile range (IQR) were used to summarize numerical data with a non‐normal distribution. Abbreviations: BP, blood pressure; CAD, coronary artery disease; CKD, chronic kidney disease; eGFR, estimated glomerular filtration rate; Hb, hemoglobin; HIV, human immunodeficiency virus; INR, International normalized ratio; LVEF, left ventricular ejection fraction; NSAID, Nonsteroidal anti‐inflammatory drugs. Procedural Findings In 2018, 297 diagnostic coronary angiograms were performed, with a 19.9% increase in 2019. Also, 112 (37.7%) of all angiograms done in 2018 were conducted through the radial artery. By 2019, this proportion rose significantly to 206 (57.9%) (p < 0.001). The coronary angiograms demonstrated normal epicardial coronary vessels in 229 (50.3%), single vessel disease in 95 (20.9%), double vessel disease in 65 (14.3%) and triple vessel disease in 66 (14.5%). The median duration of the procedure was longer for the radial access group compared to the femoral access group at (50 min (IQR: 40.0–60.0) vs. 45 min (IQR: 35.0–60.0), p = 0.011) (Figure 2). The mean contrast volume administered in the radial and femoral artery access group did not differ significantly, measuring 83.7 ± 36.1 milliliters (mL) and 88.9 ± 37.1 mL, respectively (p = 0.073). The median radiation dose was 3726 mGy/m2 (IQR: 2181.7–6205.0), and it did not differ significantly between the radial and femoral access group (4011.0 mGy/m2 (IQR: 2298.0 – 6411.4) vs. 3511.0 mGy/m2 (IQR: 2154.0–5821.5), p = 0.066) (Table 2). During the study period, 13 operators performed DCA using radial artery access. Of these, 11 were fellows in cardiology at various stages of training, each with experience of ≤ 50 transradial DCA procedures. The most experienced operator had performed 70 cases. Procedural factors grouped according to vascular access (A) radiation dose (mGy/m2), (B) contrast volume (mL), (C) duration of procedure (minutes) and (D) estimated glomerular filtration rate (mL/min/1.73m2). [Color figure can be viewed at wileyonlinelibrary.com] Table 2Procedural factors for diagnostic coronary angiography stratified according to access route.VariableTotal (N = 653)Transradial (N = 318) 48.7%Transfemoral (N = 335) 51.3% p value Indications for DCA, % Known CAD99 (15.2)44 (13.9)55 (16.4)0.358Suspected CAD484 (74.1)225 (70.8)259 (77.3)0.056Valvular disease80 (12.3)48 (15.1)32 (9.6)0.031Perioperative9 (1.4)7 (2.2)2 (0.6)0.079Duration of procedure, min, IQR46 (40–60)50 (40–60)45 (35.0–60)0.011 Urgency, % 0.010Elective457 (81.6)231(87.2)226 (76.6)Emergency103 (18.4)34 (12.8)69 (23.4)Total radiation dose, mGy/m2, IQR3726 (2181–6205)4011 (2298–6411)3511 (2154–5822)0.066Total contrast volume, mL (mean, SD)86.4 ± 36.783.8 ± 36.189 ± 37.20.073 Diseased vessels on angiography, % Normal angiogram229 (50.3)118(55.1)111(46.1)0.053Single vessel disease95 (20.9)40 (18.7)55 (22.8)0.279Double vessel disease65 (14.3)26 (12.1)39 (16.2)0.220Triple vessel disease66 (14.5)30 (14.0)36 (14.9)0.781 Note: Categorical data are represented as frequencies and percentages. Numerical data that are normally distributed is summarized as the mean and standard deviation (SD), while the median and the interquartile range (IQR) were used to summarize numerical data with a non‐normal distribution.Abbreviations: CAD, coronary artery disease; DCA, diagnostic coronary angiography; mGy/m2, milligray per square meter; IQR, interquartile range; mL, milliliters; min, minutes. Procedural factors for diagnostic coronary angiography stratified according to access route. Note: Categorical data are represented as frequencies and percentages. Numerical data that are normally distributed is summarized as the mean and standard deviation (SD), while the median and the interquartile range (IQR) were used to summarize numerical data with a non‐normal distribution. Abbreviations: CAD, coronary artery disease; DCA, diagnostic coronary angiography; mGy/m2, milligray per square meter; IQR, interquartile range; mL, milliliters; min, minutes. Outcomes In total, 653 DCA were performed, and 613 (94.2%) were completed without the need to repeat the procedure. Failed access was mainly documented during the transfemoral approach, accounting for 24 (61.5%) failed procedures. Clinically significant procedure‐related complications were uncommon. Vascular spasms were encountered in eight patients subjected to radial artery access. Other documented complications were femoral artery dissection (n = 2), hematoma (n = 2), aortic dissection (n = 1), and femoral artery hemorrhage (n = 1) (Table 3). Table 3Complications associated with diagnostic coronary angiography stratified according to access route.VariableTotal (N = 653)Transradial (N = 318) 48.7%Transfemoral (N = 335) 51.3%Dissection3 (0.46)0 (0.0)3 (0.9)Radial artery occlusion1 (0.15)1 (0.31)0 (0.0)Hematoma2 (0.31)0 (0.0)2 (0.6)Vascular spasm8 (1.23)6 (1.89)2 (0.6)Hemorrhage1 (0.15)0 (0.0)1 (0.3)Ventricular fibrillation1 (0.15)1 (0.31)0 (0.0) Complications associated with diagnostic coronary angiography stratified according to access route. Discussion We retrospectively reviewed medical records of 653 patients referred for DCA in Johannesburg and assessed the efficacy and safety of radial versus femoral access. Transradial access was used in 48.7% of patients, and the longer procedural duration was in the transradial access group. Radial access has been reported to be associated with longer procedural times and higher radiation exposure than femoral access [9, 10]. Additionally, the longer procedural duration in our study was probably due to the inherent complexity of catheterizing the radial artery with its smaller diameter and higher propensity for spasm [11]. We did not find significant differences in the absorbed radiation dose or volume of contrast media administered between patients in the radial and femoral artery access groups. This is consistent with findings from a study by Feldkemp et al., who investigated whether radial access was protective against contrast‐induced nephropathy compared to femoral access [12]. They showed that the contrast volume was comparable in each vascular access group. Surprisingly, they concluded that despite the volume of contrast media administered being similar in both access arms, radial access was associated with significantly fewer acute kidney injuries, particularly in patients with acute coronary syndromes. The mean age in our study was 58 years, and almost half (46%) of patients had hypertension. In more than half of patients with dyslipidaemia, DCA were performed via the femoral artery. We hypothesize that perhaps patients with dyslipidaemia presented with higher low‐density lipoprotein cholesterol levels and critical disease and, as such, were subjected to femoral artery access in anticipation of potential coronary interventions. Our finding is similar to Alkatiri et al., who reported that the transfemoral approach tends to be the preferred access route in patients with multiple comorbidities [13]. However, there is limited evidence in the literature showing a similar association [5, 14]. In the Radial Versus Femoral Randomized Investigation in ST‐Elevation Acute Coronary Syndrome (RIFLE‐STEACS), there was no association between dyslipidaemia and vascular access route, and the prevalence of hypercholesterolemia was comparable in both vascular access groups [5]. In our study, 6% of patients with atrial fibrillation had transfemoral access compared to 2% subjected to transradial access. These findings are similar to a study by Sherwood et al. investigating outcomes of cardiac catheterization in patients with atrial fibrillation on anticoagulation therapy in contemporary practice [15]. Similar to their cohort, 96% of patients with atrial fibrillation had their procedure done electively via the femoral artery, with the interruption of oral anticoagulation [15]. Also, in our center, radial access was used more frequently than the transfemoral access route in patients prescribed warfarin or NSAID. This supports the evidence that radial access has fewer bleeding complications when compared to transfemoral access [16]. Our institutional strategy for managing ACS has not undergone significant changes over the study period, and the same access protocols and procedural approaches were consistently applied throughout the 5 years. However, in select cases of high‐risk non‐ST segment elevation myocardial infarction (NSTEMI) and STEMI, femoral arterial access was preferred. There was a low rate of institutional volume in the first year. Our study findings showed that the rate of transradial access increased from 38% in 2018 to 58% in 2019. This indicates a low institutional volume of transradial access in 2018. Jolly et al. demonstrated an association between improved outcomes with transradial access and high procedural volume centers [17]. As such, failure to achieve improved outcomes in LMIC could be explained by low volume and a lack of experience among operators. Given the variability in operator experience and the small number of high‐volume operators, it was challenging to reliably assess and compare learning curves across operators with respect to procedural characteristics. We acknowledge that our institutional experience during the study period, characterized by a transition from femoral to radial access and a small number of high‐volume operators, mostly cardiology fellows, may not reflect the “steady state” seen in high‐volume expert radial centers described in high‐income countries. The challenges highlighted above, including fellows at various stages of training, a steep learning curve, and evolving institutional protocols, are not unique to our center but are common across most academic and public‐sector hospitals in SSA and other resource‐limited settings, which serve as vital training hubs. Therefore, we believe that this context makes our findings a highly relevant reflection of real‐world practice in many LMIC at the time the study was undertaken. Although our study does not reflect the ideal transradial access experience, it captures real‐world practice and enthusiasm for adopting a new technique in a LMIC setting. Despite these limitations, transradial access was successfully performed in nearly half of our cohort, with no significant differences in major safety outcomes, including radiation exposure, contrast volume, or major complications, compared to femoral access. These results demonstrate that TRA can be safely and feasibly implemented, even during the early stages of institutional adoption, within a training center environment. The operator profile and institutional learning curve we described may serve as an impetus for other centers in LMIC settings seeking to transition to radial‐first practices. Our data may provide a realistic benchmark and offer reassurance that the benefits of TRA can be achieved, even in the absence of high‐volume expert operators. The radial artery was the preferred vascular access in patients with valvular heart disease who were being worked up for open heart surgery. This was an expected finding, as most of these elective procedures were done as a screening investigation. There was little or no foreseen need for intervention. The operators preferably selected this route for its perceived lower vascular complication risk and shorter hospital stays [5]. Conversely, femoral access is more commonly used in emergency settings. Firstly, the patient's clinical presentation may require assistive devices such as an intra‐aortic balloon pump, which cannot be inserted via radial access. Secondly, the probability of coronary artery intervention is higher in emergency situations; therefore, the operator may need to use a larger cannula for larger intervention equipment. However, an overwhelming body of evidence suggests that radial artery access in these emergency cases is just as efficient at providing emergency interventions and may have an additional advantage of fewer MACE and bleeding complications [16, 18, 19]. Vascular complications have been the crux of discussion between transradial and transfemoral access, particularly bleeding, which is associated with poorer outcomes and higher rates of MACE [16, 20]. Upon reviewing the medical records of patients included in our study, complications were rarely documented. The rates were much lower than most published studies [21, 22]. This could be explained by the exclusion of patients requiring PCI from our study cohort. Limitations This was a single‐center retrospective study with a relatively small sample size. We relied on information documented in medical records. As this was a retrospective review, the specific breakdown of urgent cases, including those performed for ACS, was not recorded in this study. As a result, we were unable to report the exact proportion of urgent cases presenting with ACS. Furthermore, due to the low rate of documented complications, we were also unable to determine the variables associated with the occurrence of complications. Additionally, the long‐term sequelae resulting from vascular complications were not assessed. Regarding access route selection, there may have been a clinical selection bias, as the access route was left to the operator's discretion. Moreover, during the study period, our center was still transitioning from femoral access to radial access. The primary operators in our institution are typically fellows in training, and they may have been more comfortable using femoral access instead of the newly introduced radial access. Though this adoption was due to the growing evidence favoring radial artery access, the vascular access used by the operator was still largely based on their discretion. No set standards of operation were set during that time. Given the relatively small sample size, low rate of institutional volume, and event rates, our study was not powered to detect slight differences in uncommon outcomes. Our study was not meant to provide definitive evidence on rare events, but rather to provide real‐world experience and offer hypothesis‐generating insights. Despite these limitations, this study sheds some light on the safety and efficacy of DCA, as well as the extent to which competency is achieved when adopting newer interventional techniques in a high‐volume center in South Africa. Larger prospective studies and meta‐analyses are required to validate these findings further (Central illustration 1). Clinical characteristics and procedural factors among patients referred for diagnostic coronary angiography. [Color figure can be viewed at wileyonlinelibrary.com] Conclusions The radial artery was the preferred vascular access in 49% of coronary angiograms. These findings support the continued use of radial access as a safe and viable alternative to femoral access, even in resource‐limited settings. Funding The authors have nothing to report. Ethics Statement Permission to conduct the study was obtained from the University of the Witwatersrand Human Research Ethics Committee (Clearance certificate number: M2011118) and relevant hospital authorities. Consent Informed consent was waived as the study was a retrospective review of patient records. Conflicts of Interest N.T. is a cardiologist and has received consultation fees from Pfizer, Novartis Pharmaceuticals, Novo Nordisk, Boston Scientific, Servier, Phillips, Takeda, AstraZeneca, Acino Health Care Group, Sanofi and Merck. NT has also received educational and research grants from Medtronic, Biotronik, Boston Scientific and Vertice Health Care Group. K.L.N., D.M., U.G.A. and T.K. declare no conflicts of interest. Acknowledgments The authors would like to thank colleagues in the Division of Cardiology for assisting with retrieving patient data. Data Availability Statement The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions. References Associated Data Data Availability Statement The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.