NDT Advance Access originally published online on December 29, 2005
Nephrology Dialysis Transplantation 2006 21(4):1006-1012; doi:10.1093/ndt/gfk007
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
© The Author [2005]. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org
Original Articles: Dialysis and Transplantation
Predictors and outcome of cardiopulmonary resuscitation (CPR) calls in a large haemodialysis unit over a seven-year period
Hemodialysis Unit, Division of Nephrology, Maisonneuve-Rosemont Hospital, Affiliated to University of Montreal, QC, Canada
Correspondence and offprint requests to: Martine Leblanc, MD, FRCPc, Nephrology and Critical Care, Maisonneuve-Rosemont Hospital, 5415 de l’Assomption, Montreal, QC, Canada, H1T 2M4. Email: martine.leblanc{at}sympatico.ca
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and methodsResultsDiscussionReferences |
|---|
Background. Cardiac mortality is the leading cause of death in dialysis patients, with cardiac arrests being most frequent. Our purpose was to determine the epidemiology, predictors and outcomes of calls for cardiopulmonary resuscitation (CPR) occurring in our haemodialysis unit.
Methods. We reviewed retrospectively all calls for CPR occurring in our unit between August 1997 and December 2004 and compared data to a cohort of chronic haemodialysis patients from our unit. Dialysis sessions performed in the ICUs were not included.
Results. A total of 38 calls occurred over 307 553 sessions, corresponding to an incidence of 0.012%. In a multivariate logistic regression model, statistically significant predictors to have a call for CPR were ischaemic heart disease (OR: 3.93; 95% CI: 1.70–9.07), heart failure (OR: 2.74; 95% CI: 1.12–6.74) and female gender (OR: 2.96; 95% CI: 1.37–6.43). Patients who had a call for CPR had a lower dialysis vintage than control patients (OR: 0.98; 95% CI: 0.965–0.996). Twenty of the 38 events presented on Mondays or Tuesdays (P = 0.012); 78% occurred during haemodialysis, vs 14 and 8% immediately after and immediately before dialysis but still on the unit, respectively. Of the 38 events, 24 were true cardiopulmonary arrests. Cardiac etiology was the most frequent (34%) and only 4 events were attributed to potassium disorders. One quarter of patients were dialyzed against a dialysate potassium concentration of 1 mmol/l or below. An arrhythmia was identified in 19 patients; a malignant ventricular fibrillation or ventricular tachycardia was most frequently found (32%), followed by severe bradycardia (26%). For the whole group, there were 6 deaths (16%) within 48 h; 30 patients (79%) were alive at 30 days and discharged from the hospital. Among the 24 cardiopulmonary arrests, there were 4 deaths (17%) within 48 h; 18 patients (75%) were alive at 30 days and discharged from the hospital. There was a trend for worse prognosis at 60 days when related to cardiopulmonary etiology (P = 0.054) and when a true cardiopulmonary arrest occurred (P = 0.134).
Conclusions. This study confirms that arrest codes occur more frequently on Mondays and Tuesdays in a haemodialysis unit. Survival after an arrest code appears to be better than in certain other circumstances, probably in part because of the presence of witness, physician and equipment, and vascular access being readily available.
Keywords: cardiac arrest; cardiopulmonary resuscitation; haemodialysis; mortality; predictors
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionReferences |
|---|
Haemodialysis patients have an annual age-adjusted mortality rate above 20%, nearly 25 times higher than the general population [1]. Chronic kidney diseases (CKD) and end-stage renal disease (ESRD) are now clearly established as major cardiovascular risk factors [2,3]. Approximately 45% of dialysis patients’ mortality is from cardiac causes. Of these deaths, around 50% are from cardiac arrest [1].
Left ventricular hypertrophy (LVH), autonomic dysfunction, ischaemic heart disease, hypertension, left-ventricular dysfunction, diabetes, male gender and electrolytes disturbances are all risk factors for ventricular arrhythmias in haemodialysis patients [4,5]. Furthermore, the dialysis treatment itself may induce arrhythmias.
Survival after a cardiac arrest is usually poor [6]. Few studies have been conducted to evaluate predictors and outcome of cardiac arrest occurring during haemodialysis. We reviewed epidemiology, predictors and outcomes of calls for cardiopulmonary resuscitation (CPR) over the last seven years in our haemodialysis unit, which is hospital-based.
|
Subjects and methods |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionReferences |
|---|
We reviewed the medical records of patients haemodialyzed at the haemodialysis unit of Maisonneuve-Rosemont hospital between August 1997 and December 2004. Calls for CPR were identified and considered if they occurred within the haemodialysis unit. Patients who had CPR during haemodialysis but were dialyzed in the intensive care units of Maisonneuve-Rosemont hospital were excluded.
Before calling for CPR, our dialysis nurses are trained to stop dialysis, retransfuse, keep the catheter open or the needles in the access, infuse saline and wait for team response. Our dialysis nurses have received CPR training and are able to initiate resuscitation when needed. For less critical cases, they would call for medical assistance but would not call for CPR. For example, there would not be a CPR call for a breathing patient having a systolic blood pressure of 60 mmHg and appearing very ill.
Data was extracted from medical records, cardiac arrest data forms and dialysis run sheets. Ischaemic heart disease was defined as angina, previous myocardial infarction, positive results for ischaemia at investigational tests, previous coronary angioplasty or coronary artery bypass surgery. Heart failure was defined as a left ventricular ejection fraction below 40%. Peripheral vascular disease was defined as significant carotid stenosis (symptomatic or above 70%), history of stroke or previous carotid surgery, peripheral vascular disease, previous peripheral bypass or amputation for ischemia, and history of an aortic aneurysm. Evolution of identified patients was followed until death or for 60 days after the CPR event.
Data from a cohort of chronic haemodialysis patients from the same unit during the same period was available for comparison. This cohort included all remaining patients in the dialysis unit in year 2000. Similar definitions were applied and similar characteristics were available for both.
Statistical analysis
Continuous variables were described as means±SD and analysed with a two-tailed Student t-test or a Mann–Whitney test (for parametric and non-parametric variables respectively). Categorical variables were analysed using the 
2 test. Significant variables (P<0.05) were included in a logistic regression model, and adjusted odds ratios were determined. A two-tailed z-approximation test was used to test if days were significant predictors of CPR calls. Kaplan–Meier curves were plotted for survival after the event. A Cox proportional model was used to find independent predictors of survival at 60 days.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionReferences |
|---|
Thirty-eight calls for CPR were recorded during the 89 month observation period of the study. A total of 307 553 haemodialysis treatments were delivered in the dialysis unit during this period. Thus, the incidence of CPR events was 12.4 per 100 000 haemodialysis treatments or 1 event per 8094 sessions.
Demography of our patients with and without CPR events is presented in Table 1. Mean age was 64.5 years and was not significantly different from the haemodialysis cohort. The majority of patients with CPR events were females (60.5%) and this was significantly higher than for our cohort (P = 0.022). Most patients were Caucasians (92%), with only 1 African-American, 1 Asian and 1 Hispanic. The median dialysis vintage was 21.8 months (interquartile range of 8.1–53.6 months), and this was significantly lower than for our cohort (median 36 months, interquartile range of 23–60 months). The prevalence of diabetes was 36.8% (2.6% type 1 and 34.2% type 2). Most patients had hypertension (81.6%).
|
Patients who had a CPR event were more likely to have ischaemic heart disease (73.4 vs 40.1%, P<0.001), and/or to have a heart failure (31.6 vs 11.3%, P = 0.001). Within the patients with ischaemic heart disease and who had a CPR event (n = 28), 19 had a history of myocardial infarction and 4 had a previous coronary artery bypass graft (CABG). More than a third of patients with a CPR event had a history of arrhythmia, and 18.4% had a permanent pacemaker previously inserted. A significant proportion had peripheral vascular disease (23.7%) and cerebrovascular disease (21.1%). Patients who had a CPR event were more likely to be dialyzed via a central catheter (42.1% compared to 32.6% for our overall haemodialysis population), although not statistically significant (P = 0.264).
Significant variables for having a CPR event (gender, ischaemic heart disease, heart failure and dialysis vintage) were entered into a logistic regression model; they remained independently significant. Odds ratios (OR) with 95% confidence intervals (95% CI) and P-values are shown in Table 2.
|
Most CPR events occurred during the haemodialysis treatment (78%); only 8 and 14% occurred either just before or just after the session, but when patients were still on the haemodialysis unit. Twenty calls for CPR occurred on Mondays or Tuesdays (P = 0.012), as compared with 8 on Wednesdays or Thursdays and 10 on Fridays or Saturdays (Figure 1). Among all the identified CPR calls, 24 (63.2%) were considered ‘real’ cardiopulmonary arrests, but almost all (97.4%) needed emergency medical support and rapid intervention.
|
We classified CPR events in categories, as presented in Table 3. If no abnormality in electrolytes was noted and no other causes suspected by the CPR team, and no acute myocardial infarction occurred, events were classified as from a cardiac etiology. However, even if the patient was known for heart disease, if a predisposing factor was found to explain the malignant arrhythmia (i.e. hyperkalaemia), it was then classified in that last category. As expected, the most frequent etiology was cardiac (34.2%), whereas pulmonary (15.8%), seizures (10.5%), bleeding (10.5%) and electrolytes (10.5%) were less common categories. The category called miscellaneous includes 5 severe hypovolemic hypotension episodes or vagal shock, 1 anaphylactic shock after the administration of dextran iron and 1 narcosis. A cardiac arrest occurred in 92.3% of the cardiac causes, and a cardiac event was the etiology in half of the cardiac arrests. The category called electrolytes included 2 hyperkalaemias, 1 hypokalaemia and 1 hypokalaemia combined with hypomagnesemia. Both cases of hypokalaemia had a recent serum potassium value above 4.5 mmol/l and were dialyzed against a dialysate bath containing a potassium concentration between 1 and 1.5 mmol/l. Noteworthy, both CPR calls for what turned to be hyperkalaemia occurred just before starting dialysis, one on Monday and one on Tuesday.
|
One patient with hyperkalaemia had a recent serum value of 6 mmol/l and was dialyzed against a 2 mmol/l potassium bath. However, his previous serum potassium values were quite fluctuant over time and the episode was considered not predictable. The other case of hyperkalaemia had a recent serum potassium value of 7.3 mmol/l five days previously; this patient was usually dialyzed in a 2.5 mmol/l potassium bath, suggesting that a lower bath could have been a preferable choice.
Nearly two-thirds of patients with a CPR event (63.9%) were dialyzed against a dialysate bath containing a potassium concentrations between 1.5 and 3 mmol/l. Dialysate potassium concentrations of 1 mmol/l or below were used in 25.0% of them and dialysate potassium concentration of 3.5 mmol/l or above in 11.1%. Among the subgroup of patients dialyzed in a low potassium bath, 22.2% had a recent serum potassium above 5 mmol/l. One particular patient dialyzed against a potassium bath of 1 mmol/l had a recent serum potassium below 4 mmol/l, suggestive of an inappropriate choice. However, although the serum potassium of this patient during the CPR event was low, it was not judged to be causal when considering the overall episode and after reading the medical records. Recent serum potassium values (mean time 13.5 days) related to dialysate potassium concentrations are presented in Figure 2.
|
Without considering the primary cause, arrhythmia was documented in half of the CPR events (the first arrhythmia encountered when the code team arrived is presented in Table 4). Ventricular tachycardia (VF) and ventricular fibrillation (VF) were most frequent (31.6%), followed by severe bradycardia (26.3%). Other documented arrhythmias were asystole (15.8%), electromechanical dissociation (15.8%) and atrial fibrillation (10.5%). However, the first arrhythmia encountered, when a monitor was initially installed, may not be the causing arrhythmia of the cardiac arrest and can be followed by other malignant rhythms. Within these 19 documented arrhythmias, 17 occurred in patients with cardiac arrest. For the two patients without cardiac arrest, one had an unstable atrial fibrillation (that needed electrical cardioversion) and the other had a severe bradycardia secondary to hyperkalaemia.
|
Early death (less than 48 h) occurred in 15.8% of cases. If they survived the first two days after the CPR event, most patients were discharged and alive at 30 days (78.9% survival rate at 30 days). More than two-thirds (68.4%) were still alive at 60 days. When considering cardiac arrests only, the early survival rate was quite similar at 48 h (16.7%), and 75% were alive at 30 days (Table 5). Figure 3 shows the Kaplan–Meier curves for survival of all cases, and for those who had a cardiac arrest or not. The survival curve after a cardiac arrest was not statistically significantly different from the other survival curve according to a univarate Cox regression model (P = 0.134). Other characteristics (shown in Table 1) were tested in the model, and none independently predicted survival. However, the most significant variable to predict survival at 60 days was a cardiopulmonary etiology compared to other causes (P = 0.054), as shown in Figure 4. Figure 5 shows more details on cardiac rhythms and outcomes of the subgroup of cardiac arrest patients, presented in a Utstein style diagram [7].
|
|
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionReferences |
|---|
Haemodialysis patients have approximately a 10-fold higher risk of dying from cardiac arrest than the general Medicare population [1]. In the USA in 2002, a dialyzed white man of 65-year-old had an expected remaining lifetime of 3.3 years compared to 16.5 years if not dialyzed. The rate of cardiac arrest in the dialysis population has remained relatively stable between 1994 and 2002.
Haemodialysis may have an arrhythmogenic effect by itself, but controversial results were found by different groups. Cardiopulmonary resuscitation in haemodialysis was reported to occur in the dialysis unit in 1.58 and 6.8% of cases, which is less than what is expected for the time spent on the unit by these patients [8,9]. However, a possible deleterious effect of haemodialysis may last longer than that during the treatment alone. Several factors were reported to increase the likelihood of ventricular arrhythmias in haemodialysis patients: LVH and/or dysfunction, autonomic dysfunction, ischaemic heart disease, hypertension, diabetes, male gender and electrolytes disturbances [4,5].
Karnik et al. [10] reported 400 cardiac arrests in dialysis units. Compared to the Fresenius Medical Care North America (FMCNA) cohort, older age, diabetes and catheter use for vascular access were all risk factors for cardiac arrest in the unit. Age was not found to be predictor of cardiac arrest in our series; indeed, it is not always reported as a factor for sudden death in dialysis patients [11]. Neither diabetes nor catheter use were significant predictors in our cases. Surprisingly, ischaemic heart disease and heart failure were not associated with a higher cardiac arrest rate, although this is often reported as being the case. Female gender is usually not associated with a higher cardiovascular risk; in our group, it may relate to an older age, as observed in postmenopausal women for whom the cardiovascular risk becomes almost equal to men. Dialysis vintage seems to protect against a CPR event, despite the fact that mortality rates increase with time on haemodialysis [1].
Bleyer et al. [11] found an increased incidence of cardiac and sudden deaths on Mondays and Tuesdays in the USRDS cohort. They hypothesized that greater fluid overload, electrolytes and toxins fluctuations on the three-day-dialysis-free intervals may contribute to an enhanced risk. Karnik et al. [10] also reported more cardiac arrests on Mondays compared to Wednesdays and Fridays. However, this pattern was not reproduced on the Tuesday-Thursday-Saturday schedule. For statistical power purposes, we combined the two schedules together. The higher incidence of events on Mondays and Tuesdays is present in our series as well, supporting the hypothesis that a longer dialysis-free interval may enhance the risk.
We voluntarily included all the identified CPR calls in our analysis (among which 24 were considered ‘real’ cardiopulmonary arrests while the other 14 required emergency medical support and rapid intervention) since they happen in real life and mobilize the CPR team. Furthermore, they could potentially be more easily reversible, influencing the overall prognostic of having a CPR call in a dialysis unit. This may also impact on the decision to have a DNR order or not for a given dialysis patient.
Cardiovascular disease accounts for approximately 45% of deaths in dialysis patients. Cardiac cause is the most frequent etiology of cardiac arrest in the general population [12,13]. In our cases, cardiac problem was also the leading cause of arrest, followed by pulmonary disease.
In their meta-analysis, Ebell et al. [6] reported a 40.7% chance of immediate efficiency of CPR and a 13.4% rate of survival until discharge after in-hospital CPR in the general population. In dialysis patients, such rates appear even worse. In 74 dialyzed patients who underwent CPR, Moss et al. [9] found a 36% rate of effective CPR; only 8% were discharged from the hospital, and 3% alive at 6 months. In another series, a 20% hospital discharge was observed [14]. These results contrast with our survival rates, that are at least twice as high. Maybe survival after CPR has improved in current years compared to studies from the early 1990s, although a recent study argues that survival has not changed since 40 years ago [13].
A poor prognosis is associated with asystole and pulseless electrical activity as initial rhythms [6–13]. In our cases, these two rhythms were less frequent than those previously reported in the general population [13,15,16]. This may have contributed to the better outcome observed in our patients. Moreover, witnessed cardiac arrests have a better chance of effective resuscitation until hospital discharge than unwitnessed ones [13]. By definition, all our events were witnessed and, this may at least in part explain the higher survival rate of our patients. Nevertheless, in the series of Karnik et al. [10], where all events were also witnessed, 60% of patients died within 48 h. In the study by Lai et al. [8], witnessed events resulted in a 54% mortality within 24 h; only 8.3% of cases were alive at 30 days.
Furthermore, we reviewed outcome of all calls for CPR on the unit, and not exclusively for cardiac arrest per se. The survival after a cardiac arrest alone was not statistically different from the survival of patients without frank arrest. The initial survival curves are similar in both subgroups. However, these are small numbers and a trend for a higher mortality at 30 days is observed after cardiac arrest. Nonetheless, survival of our dialysis patients after a cardiac arrest remains higher than numbers previously released.
On the other hand, even if our dialysis unit is ambulatory, it is hospital-based and there is always a physician on site. All equipment for advanced CPR is available on the unit. The CPR team in the adjacent building usually arrives within a delay of 3 min to our dialysis unit. Blood or vascular access is readily available for most dialysis patients who have an event; thus, fluid and intravenous drugs can be delivered rapidly if needed. It is well known that early CPR and early defibrillation are associated with improved survival [15,16]. For this reason and since the incidence of cardiac arrests appears higher in a dialysis unit than in certain other medical fields, automated external defibrillators (AED) are now recommended by many authors and by new guidelines on cardiovascular disease in dialysis patients from the National Kidney Foundation [17,18]. Nonetheless, defibrillation alone was needed only for a minority of our patients. However, the small numbers in this series may mask the potential benefit of an earlier rapid defibrillation. Even if the addition of a prehospital advanced life-support system programme for rapid defibrillation was not found to improve survival after out-of-hospital cardiac arrests, such a conclusion probably does not apply to events occurring in dialysis units, even if adjacent to a hospital [15].
Knowing outcomes of CPR in haemodialysis patients is essential to the discussion of advanced directives with them. The proportion of dialysis patients requesting CPR in case of an event ranges from 42 to about 80% [19]. In a recent survey, 87% of haemodialyzed patients requested CPR if a cardiac arrest occurred during haemodialysis treatment. Of the remaining 13%, 71% states that they would refuse CPR, despite a high chance of reversibility and survival [20]. Discussion with patients and their representatives should be held in advance and if possible systematically to assure that their wishes and directives are respected.
In summary, cardiovascular mortality remains a major problem in the dialysis population. In the present study, we found three main factors associated with an increased risk for a call for CPR in the dialysis unit: female gender, ischaemic heart disease, heart failure and a short dialysis vintage. Mondays and Tuesdays carry a higher incidence risk than other days of the week, probably in relation to the three-day-dialysis-free interval. Finally, in our unit, survival after CPR was found to be better than in other situations; this may also apply to other dialysis units with similar settings. However, these findings are limited by the small number of patients and further studies are awaited.
Conflict of interest statement. None declared.
|
References |
|---|
|
TopAbstractIntroductionSubjects and methodsResultsDiscussionReferences |
|---|
- U.S. Renal Data System, USRDS 2004 Annual Data Report: Atlas of End-Stage Renal Disease in the United States, National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD: 2004
- Go AS, Chertow GM, Fan D, McCulloch CE, Hsu C. Chronic kidney disease and the risks of death, cardiovascular events, and hospitalization. N Engl J Med 2004; 351: 1296–1305
[Abstract/Free Full Text] - Collins AJ, Li S, Ma JZ, Herzog C. Cardiovascular disease in end-stage renal disease patients. Am J Kidney Dis 2001; 38: S26–S29[ISI][Medline]
- McCullough PA, Sandberg KR. Chronic kidney disease and sudden death: strategies for prevention. Blood Purif 2004; 22: 136–142[CrossRef][Medline]
- Herzog CA. Cardiac arrest in dialysis patients: approaches to alter an abysmal outcome. Kidney Int 2003; [Suppl 84]: S197–S200[CrossRef]
- Ebell MH, Becker LA, Barry HC, Hagen M. Survival after in-hospital cardiopulmonary resuscitation: a meta-analysis. J Gen Intern Med 1998; 13: 805–816[CrossRef][ISI][Medline]
- Cummins RO, Chamberlain DA, Abramson NS et al. Recommended guidelines for uniform reporting of data from out-of-hospital cardiac arrest: the Utstein style. Ann Emerg Med 1991; 20: 861–874[CrossRef][ISI][Medline]
- Lai MN, Hung KY, Huang JW, Tsai TJ. Clinical findings and outcomes of intra-hemodialysis cardiopulmonary resuscitation. Am J Nephrol 1999; 19: 468–473[CrossRef][Medline]
- Moss AH, Holley JL, Upton MB. Outcomes of cardiopulmonary resuscitation in dialysis patients. J Am Soc Nephrol 1992; 3: 1238–1243[Abstract]
- Karnik JA, Young BS, Lew NL et al. Cardiac arrest and sudden death in dialysis units. Kidney Int 2001; 60: 350–357[CrossRef][ISI][Medline]
- Bleyer AJ, Russell GB, Satko SG. Sudden and cardiac death rates in hemodialysis patients. Kidney Int 1999; 55: 1553–1559[CrossRef][ISI][Medline]
- Horsted TI, Rasmussen LS, Lippert FK, Nielsen SL. Outcome of out-of-hospital cardiac arrest – why do physicians withhold resuscitation attempts? Resuscitation 2004; 63: 287–293[CrossRef][ISI][Medline]
- Brindley PG, Markland DM, Mayers I, Kutsogiannis DJ. Predictors of survival following in-hospital adult cardiopulmonary resuscitation. CMAJ 2002; 167: 343–348
[Abstract/Free Full Text] - Tzamaloukas AH, Murata GH, Avasthi PS. Outcome of cardiopulmonary resuscitation in patients on chronic dialysis. ASAIO Trans 1991; 37: M369–M370[Medline]
- Stiell IG, Wells GA, Field B et al. Advanced cardiac life support in out-of-hospital cardiac arrest. N Engl J Med 2004; 351: 647–656
[Abstract/Free Full Text] - Bunch TJ, White RD, Gersh BJ et al. Long-term outcomes of out-of-hospital cardiac arrest after successful early defribrillation. N Engl J Med 2003; 348: 2626–2633
[Abstract/Free Full Text] - Becker L, Eisenberg M, Fahrenbruch C, Cobb L. Cardiac arrest in medical and dental practices. Arch Intern Med 2001; 161: 1509–1512
[Abstract/Free Full Text] - K/DOQI Workgroup. K/DOQI clinical practice guidelines for cardiovascular disease in dialysis patients. Am J Kidney Dis 2005; 45 4 [Suppl 3]: S16–S153[CrossRef]
- Hijazi F, Holley JL. Cardiopulmonary resuscitation and dialysis: outcome and patients’ views. Semin Dial 2003; 16: 51–53[Medline]
- Moss AH, Hozayen O, King K, Holley JL, Schmidt RJ. Attitudes of patients toward cardiopulmonary resuscitation in the dialysis unit. Am J Kidney Dis 2001; 38: 847–852[ISI][Medline]
Accepted in revised form: 28.11.05
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  B. Kestenbaum, K. D. Rudser, M. G. Shlipak, L. F. Fried, A. B. Newman, R. Katz, M. J. Sarnak, S. Seliger, C. Stehman-Breen, R. Prineas, et al. Kidney Function, Electrocardiographic Findings, and Cardiovascular Events among Older Adults Clin. J. Am. Soc. Nephrol., May 1, 2007; 2(3): 501 - 508. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||