| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Eur J Cardiothorac Surg 2006;29:964-970
© 2006 Elsevier Science NL
a Department of Cardiothoracic Surgery, Blackpool Victoria Hospital, United Kingdom
b Department of Cardiothoracic Surgery and Clinical Governance, The Cardiothoracic Centre-Liverpool, United Kingdom
c Department of Cardiothoracic Surgery, Manchester Royal Infirmary, United Kingdom
d Department of Cardiothoracic Surgery, South Manchester University Hospital, United Kingdom
Received 18 September 2005; received in revised form 14 March 2006; accepted 20 March 2006.
* Corresponding author. Tel.: +44 151 293 2336; fax: +44 151 288 2371. (Email: tony.grayson{at}ctc.nhs.uk).
 |
Abstract |
|---|
 Top Abstract 1. Introduction2. Methods3. Results4. DiscussionAppendix AReferences |
|---|
Key Words: Renal disease • Non-dialysis support • CABG • Mortality • Morbidity
|
1. Introduction |
|---|
|
TopAbstract1. Introduction2. Methods3. Results4. DiscussionAppendix AReferences |
|---|
Renal dysfunction is a well-recognised risk factor for post-operative morbidity and mortality after CABG [4,5], although the emphasis has dwelt on dialysis-dependent renal dysfunction. The European System for Cardiac Operative Risk Evaluation (EuroSCORE) allocates two points to patients with non-dialysis-dependent renal dysfunction (NDRD) with serum creatinine levels >200 µmol/L. In clinical practice it is felt that NDRD confers a significantly higher detrimental effect on post-operative outcome.
We conducted a large contemporary multi-centre study, using the North West Quality Improvement Programme in Cardiac Interventions (NWQIP) Cardiac Surgery Registry, to compare the outcomes after CABG of NDRD patients and patients with no history of renal dysfunction.
|
2. Methods |
|---|
|
TopAbstract1. Introduction2. Methods3. Results4. DiscussionAppendix AReferences |
|---|
Data was collected on a total of 19,625 consecutive patients undergoing isolated CABG between 1 April 1997 and 31 March 2003 in the northwest of England. Data collection methods and definitions have been described in detail previously [6] and are available from http://www.nwheartaudit.nhs.uk/. Data was collected prospectively and included patient characteristics (Table 1 ) and in-hospital outcome (Table 2 ) variables. Validation of data was conducted in each centre, which involved checking each record for completeness and flagging back to the relevant surgical team any erroneous data. All records entered onto the databases were also cross-checked against finance activity lists and theatre log-books to ensure capture of all cases. Data was collected in each hospital and returned to a central source for analysis on a 6-monthly basis. Data would be returned to the providing hospital if data completeness did not achieve a rate of 98% or above. Any missing risk factor data after acceptance into the central registry was treated as absent and this occurred in less than 2%. Outcome variables of in-hospital mortality and length of post-operative hospital stay were available for all patients in our study. Data on post-operative complications (acute renal failure, stroke, atrial arrhythmia, re-exploration for bleeding, deep sternal wound infection, and duration of ventilation) were only available for 15,427 patients.
|
|
Different techniques of coronary revascularisation were used in our study. These included operations done with and without cardiopulmonary bypass (CPB). The technique of myocardial protection for patients done on CPB again varied according to operator preference. Although blood cardioplegia was the favoured choice, cold crystalloid cardioplegia and intermittent cross-clamp and fibrillation were also used by some surgeons. Mammary arteries were harvested as a pedicle.
2.2 Patient follow-up
Patient records were linked to the National Strategic Tracing Service (NSTS), which records all cause mortality in the United Kingdom, to establish current vital status, and were matched to the NSTS based on the patients National Health Service number and date of birth.
2.3 Statistical analysis
Categorical variables are shown as a percentage. Comparisons were made with chi-square tests as appropriate. Standard statistical tests were used to calculate odds ratios and 95% confidence intervals (CI). Continuous variables are shown as median values with 25th and 75th percentiles, due to non-normality of data. Comparisons between continuous data were made with the Wilcoxon rank sum test. The additive and logistic versions of the EuroSCORE [8] were both calculated to assess differences in patient case-mix. Logistic regression was used to examine the impact of NDRD on in-hospital mortality and morbidity, while adjusting for differences in patient characteristics (treatment selection bias). Deaths occurring as a function of time were described using the product-limit methodology of Kaplan–Meier. To control for treatment selection bias, we used Cox proportional hazards analysis to calculate adjusted hazard ratios (HR) with 95% CI and to risk adjust the Kaplan–Meier survival curves.
Treatment selection bias was controlled for by constructing a propensity score [9]. The propensity score was the probability of a patient having NDRD, and was constructed using the variables listed in Table 1 (C statistic = 0.83). Once the propensity score is constructed for each patient, there are three ways of using the score for comparisons: matching, stratification, and multivariable adjustment. Due to the small sample size available to us in the main study group, we decided to use multivariable adjustment because matching would have reduced the study size even further and stratification can be difficult to interpret. The propensity score is then included along with the comparison variable (NDRD vs no history of renal dysfunction) in the multivariable analyses of outcome producing adjusted odds ratios and adjusted Kaplan–Meier survival curves. The propensity score adjusts for the treatment selection bias, which is evident in Table 1, between the two groups. In all cases a p-value < 0.05 was considered significant. All statistical analysis was performed retrospectively with SAS for Windows Version 8.2 (SAS Institute, Cary, NC).
|
3. Results |
|---|
|
TopAbstract1. Introduction2. Methods3. Results4. DiscussionAppendix AReferences |
|---|
There were 19,172 patients (98%) with preoperative serum creatinine < 200 µmol/L and 386 patients (2%) with creatinine > 200 µmol/L. Table 1 lists patient characteristics based on the presence of NDRD. Patients with NDRD were more likely to have severe angina, shortness of breath, diabetes, hypertension, peripheral vascular disease, cerebrovascular disease, respiratory disease, myocardial infarction, poor ejection fraction, triple-vessel disease, emergent procedures, and to be older. These patients were also more likely to undergo off-pump CABG surgery. However, NDRD patients were also more likely to have a lower body mass index and less use of the left internal mammary artery.
In the operations performed on cardiopulmonary bypass, patients with NDRD had significantly longer bypass [99 min (25th and 75th percentiles: 75–127) vs 90 min (25th and 75th percentiles: 62–115); p < 0.001] and cross-clamp times [53 min (25th and 75th percentiles: 34–74) vs 47 min (25th and 75th percentiles: 29–65); p = 0.002] compared to other patients.
3.2 Crude and risk-adjusted in-hospital outcomes
Table 2 shows that NDRD patients had significantly higher in-hospital mortality. Table 3
shows the variables identified as predictors of NDRD group membership, which contributed to the propensity score. After adjusting for the propensity score, NDRD patients still had a significantly higher risk of in-hospital mortality (Table 4
). The post-operative lengths of stay were significantly longer for NDRD patients compared to other patients [9 days (25th and 75th percentiles: 7–14) vs 7 days (25th and 75th percentiles: 6–9); p
< 0.001].
|
|
3.3 Crude and risk-adjusted mid-term mortality
One thousand one hundred and eighty-three (6.1%) deaths occurred during the study with a total follow-up period of 58,062 patient-years (mean follow-up of 3.0 years). The number of patients at risk of death during the follow-up period for both study groups is shown in Fig. 1
. The crude hazard ratio of mid-term mortality for NDRD patients was 3.6 (95% CI 2.8–4.6; p
< 0.001). Freedom from death in NDRD patients at 1, 2, 3, 4, and 5 years was 90.3%, 86.2%, 83.2%, 74.9%, and 71.4%, respectively, compared with 97.4%, 96.0%, 94.6%, 92.9%, and 91.0% for patients with no history of renal dysfunction (Fig. 1).
|
|
|
4. Discussion |
|---|
|
TopAbstract1. Introduction2. Methods3. Results4. DiscussionAppendix AReferences |
|---|
NDRD affects the function of every organ system and several abnormalities may be present in asymptomatic patients with advanced impairment of kidney function. Large alterations in the composition of intracellular and extra-cellular fluids, the effects of retained toxic metabolites and many other abnormalities may be tolerated by patients with preserved renal function but may prove to be detrimental to those with little reserve. The insult of surgery may further upset the delicate balance of homeostatic mechanisms resulting in high mortality and morbidity.
The present study and previous publications have confirmed a higher incidence of hypertension, diabetes, prior myocardial infarction, left ventricular dysfunction, peripheral vascular disease and dyslipidemia in patients with NDRD, which are co morbidities, widely accepted to be cardiovascular risk factors. The association of NDRD with recurrent cardiovascular disease outcomes, stroke and all-cause mortality has been published by previous workers [1,13]. Our study concurs with recent studies, which showed an independent effect of renal dysfunction on patient outcomes in terms of cardiovascular events and mortality following acute myocardial infarction, acute coronary syndromes and revascularisation by surgical or percutaneous intervention [2]. A number of possible explanations have been put forward by these authors. Both kidney disease and cardiovascular disease may be sequelae of atherosclerosis implying greater burden of disease in patients with renal dysfunction [1,14]. Angiographic evidence of greater severity of atherosclerosis and calcification of coronary vessels in NDRD patients has been published previously. Patients with kidney disease may have accelerated cardiovascular disease and conversely cardiovascular disease may perpetuate kidney disease.
NDRD patients also have a higher prevalence of non-traditional risk factors such as hyper-homocystinemia, inflammation, and oxidative stress, which may result in acceleration of atherosclerosis [2,15]. Derangement of calcium-phosphate homeostasis, endothelial dysfunction, conditions promoting coagulation, anemia, albuminuria and elevated uric acid levels have all been noted in the published literature [14].
Previous investigators have also highlighted that NDRD patients do not receive the optimal risk factor modification and intervention with cardioprotective medications such as aspirin, beta-blockers, angiotensin converting enzyme inhibitors and statins, due to fear of complications, a concept called ‘therapeutic nihilism’ [13]. Patients in the lowest tier of renal function are noted to be least likely to receive these risk-modifying medications. Other possible explanation is the well-known adverse impact of cardiac surgery on postoperative renal function [16]. In the presence of NDRD, cardiac surgery has synergistic effect in producing a more pronounced post-operative renal impairment, which is partly contributes to high morbidity and mortality.
NDRD results in prolonged hospital stay, increased morbidity and reduced long-term survival.
NDRD patients spend longer time in hospital and critical care area due to increased need of postoperative renal support and additional time waiting for the biochemistry to revert back to base line. They tend to have more advanced degrees of atherosclerosis and calcification of coronary vessels and the aorta so they are not only technically more challenging but also stand a higher risk of stroke with aortic manipulations.
Large community based longitudinal studies as well as cardiac surgical follow-up studies have shown that although patients with NDRD are at risk of both cardiovascular and renal adverse outcomes, the former is far more predominant and is deemed to be the cause of reduced mid-term survival compared to the healthy population [17,18]. In a recently reported series the actuarial 1, 2, 3 and 4-year survival rates in NDRD group were 88.2%, 81.4%, 75.3% and 71.8% compared to 99.4%, 98.7%, 96.9% and 94.5% in the control group [19]. Mid-term survival data from our study has reinforced the findings from these studies.
The management of renal failure patients requires careful management of fluid retention, hypertension, electrolyte imbalance, metabolic acidosis, anemia, bleeding disorders and glucose intolerance. Use of arterial grafts, pulsatile flow on bypass, preoperative [20] and intra-operative hemodialysis have been suggested to improve the risk in CABG [21]. Off-pump coronary bypass grafting has also been shown to be beneficial in some studies [22] while others did not support this opinion. Other strategies aimed at preservation of renal function have focused on relatively higher perfusion pressure, use of frusemide, mannitol and dopamine infusions and maintenance of overall hemodynamic stability. Close cooperation between the surgeon, anesthetist, intensivist and nephrologist is required to optimise the care. In the postoperative ward regular estimation of acid base balance and electrolytes, hemodynamic parameters, daily weight are essential. Potassium levels should be monitored closely to avoid the dangers of arrhythmia associated with high or low potassium.
Previous studies on this issue are limited by a variety of factors including small sample sizes, use of different cut off values of serum creatinine and heterogeneity of case mix. They focused mainly on operative mortality and there is little reported morbidity and longer term results [10–12,23,24]. The design of this study and large number of patients involved, has given us enough statistical power to come to a firm conclusion. Like others we have observed an increased association of NDRD with known risk factors for cardiovascular events. This study has, however, shown conclusively that even after adjustment for these risk factors, there is a strong and independent association of NDRD with significantly increased perioperative death and complications as well as a reduced mid-term survival after surgical coronary revascularisation.
There are some limitations, which may affect the conclusions drawn from this study. Firstly this is an observational study and by its retrospective nature cannot account for the unknown variables affecting the outcomes that are not correlated strongly with the variables used in the risk adjustment. However, retrospective comparisons with propensity score adjustment are recognised as highly robust and may in some cases be more widely acceptable than randomised control trials. This is particularly true in this case, as it is not possible to randomise a patient to either NDRD or no history of renal dysfunction. A further limitation of the study is the fact that we have not assessed the impact of cardiac related deaths or other mid-term outcomes, such as graft patency. Also, we do not have data on the duration and etiology of renal dysfunction and hence we cannot comment on the effect of either factor on the risk of adverse outcome.
Glomerular filtration rate (GFR) is believed to be a more reliable indicator of renal function than serum creatinine estimation alone [2,25]. The limitation of serum creatinine is its nonlinear association with GFR which varies according to age, sex, race and lean body mass. However, serum creatinine is the biochemical marker of renal function, most readily available to the clinicians. It has been accepted by learned societies as part of minimum dataset on the patients [8]. The incremental risk of increasing creatinine levels on the morbidity and mortality is well recognised [4,15].
This study carries valuable information for the treatment of NDRD patients with surgical revascularisation. Surgeons are often asked to provide short- and long-term mortality rates, along with morbidity outcomes, for possible CABG surgery to patients and families. This study provides contemporary and accurate data to surgical clinicians for the purposes of patient consent.
In conclusion our study suggests that preoperative renal dysfunction is a powerful, independent and easily identifiable predictor of perioperative morbidity and mortality as well as reduced mid-term survival after CABG surgery.
|
Appendix A |
|---|
|
TopAbstract1. Introduction2. Methods3. Results4. DiscussionAppendix AReferences |
|---|
Dr H.-J. Schaefers (Homburg/Saar, Germany): How many of those patients were on dialysis at some point between the operation and the 5-year follow-up? Did you have any or did you exclude such patients?
Mr Devbhandari : When we undertook this study, we excluded 68 patients already on dialysis. And among the patients, which were followed up, in the non-dialysis-dependent renal dysfunction group, 13% required dialysis at some stage compared to 0.7% in the control group.
Dr Schaefers : So, in other words, secondary need for dialysis at one point during follow-up was not high enough to explain this long-term difference in long-term survival, correct?
Mr Devbhandari : That's correct, I agree with you.
Dr J. Bachet (Paris, France): Just a small comment. It always seems to me a little artificial to separate renal function from the rest of the patient's pathology. You showed a very complicated slide with all the co-morbidities. Don’t you think that people with renal failure or renal insufficiency preoperatively have worse results not only because they have a superior creatinine level but because in general they have bad arteries, a bad heart or bad lungs or whatsoever? Intellectually it is a little difficult to accept that it is only renal failure that makes the difference.
Mr Devbhandari : A very important question, and this is the reason why I am presenting this paper. We see that the renal dysfunction group has all the higher comorbidities in terms of more angina, more shortness of breath, more peripheral vascular disease, more incidence of myocardial infarction, and more left ventricular dysfunction and so on and so forth. But what we are emphasizing in the paper is that when we balance differences the in case mix between the two groups by means of propensity scores, we can still see that the difference in outcome is much more than expected from the distribution of risk factors. So it appears that non-dialysis-dependent renal dysfunction has a powerful and independent effect over and above the differences in the comorbidities.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Presented at the joint 19th Annual Meeting of the European Association for Cardio-thoracic Surgery and the 13th Annual Meeting of the European Society of Thoracic Surgeons, Barcelona, Spain, 25–28 September 2005. |
References |
|---|
|
TopAbstract1. Introduction2. Methods3. Results4. DiscussionAppendix AReferences |
|---|
2.0 mg/dl). Eur J Cardiothorac Surg 2001;20:565-572.This article has been cited by other articles:
|
|
 |
|
  K. I. Paraskevas Applications of statins in cardiothoracic surgery: more than just lipid-lowering Eur. J. Cardiothorac. Surg., March 1, 2008; 33(3): 377 - 390. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Ibanez, M. Riera, J. I. Saez de Ibarra, A. Carrillo, R. Fernandez, J. Herrero, M. Fiol, and O. Bonnin Effect of preoperative mild renal dysfunction on mortality and morbidity following valve cardiac surgery Interactive CardioVascular and Thoracic Surgery, December 1, 2007; 6(6): 748 - 752. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
