Abstract
Cardiac troponin T (cTnT), even at low concentrations, is a risk factor for 30-day mortality in patients undergoing noncardiac surgery, but it is uncertain whether that risk is generalizable to patients with poor kidney function. We, therefore, evaluated the relationship between cTnT concentration and kidney function on the outcome of 30-day mortality in a post hoc analysis of a prospective cohort study of patients undergoing noncardiac surgery. cTnT was measured for 3 days after surgery and considered abnormal if the peak was ≥0.02 ng/ml. Of the included 14,037 patients, 267 (1.9%) patients died within 30 days of surgery. The adjusted hazard ratios for death with an abnormal cTnT concentration were 4.37 (95% confidence intervals [95% CI], 3.21 to 6.22), 6.15 (95% CI, 2.95 to 140.9), 6.30 (95% CI, 3.12 to 21.23), 1.33 (95% CI, 0.56 to 4.85), and 1.46 (95% CI, 0.46 to 9.21) for eGFR≥60, 45 to <60, 30 to <45, 15 to <30, and <15 ml/min per 1.73 m2 or on dialysis, respectively. Compared with patients with eGFR≥60 ml/min per 1.73 m2, the adjusted hazard ratio was significantly lower for patients with eGFR=15 to <30 ml/min per 1.73 m2 (interaction P value=0.02). Redefining abnormal cTnT concentration as ≥0.03 ng/ml or a change of ≥0.02 ng/ml did not alter results. Because the risk associated with postoperative cTnT levels may be different for patients with eGFR<30 ml/min per 1.73 m2, additional research is required to determine how to interpret perioperative cTnT values for patients with low kidney function.
Keywords: cardiovascular disease, CKD, survival, risk factors, epidemiology and outcomes
Over 200 million noncardiac surgeries are performed every year globally.1 Although these procedures can be lifesaving and significantly improve quality of life for patients, >1 million individuals will die each year within 30 days after surgery.1 Cardiovascular complications are the most important contributor to 30-day mortality in patients undergoing noncardiac surgery.2
CKD, largely characterized by a reduced eGFR, affects over 30% of older adults and is a strong independent risk factor for cardiovascular events in both patients undergoing and not undergoing surgery.3–5 Furthermore, increasing numbers of patients with CKD are undergoing surgery because of its increasing prevalence, longer survival, and improved surgical and anesthetic safety.
We previously showed that >11% of major noncardiac surgeries in patients ≥45 years old are complicated by an elevation in cardiac troponin T (cTnT), a commonly used marker of myocardial injury, to a concentration of ≥0.02 ng/ml.6 This concentration of cTnT was independently associated with a >2-fold increased risk of death within 30 days of surgery. We also identified a borderline statistical interaction between kidney function (on the basis of the following categories: eGFR<30 or on dialysis, 30–44, 45–59, and ≥60 ml/min per 1.73 m2) and abnormal cTnT (overall interaction P value =0.05). This did not meet our a priori threshold for a statistically significant interaction. However, uncertainty remained about whether an elevated cTnT has the same meaning in patients with and without reduced eGFR, particularly for those with severely reduced eGFR.
To address the possibility that renal function modifies the prognostic properties of postoperative cTnT, we performed a post hoc analysis exploring the interaction between several strata of preoperative kidney function and previously defined prognostically important postoperative cTnT concentrations.
Results
There were 15,133 patients potentially eligible for these analyses. Of these, 1096 patients were missing preoperative creatinine, leaving 14,037 (92.8%) patients. Of these, 26 (0.2%) patients did not complete 30-day follow-up and were censored at last follow-up (Figure 1).
Figure 1.
Flow diagram of patients included in the cohort.
Table 1 reports the preoperative characteristics of patients by eGFR strata. Patients in lower eGFR strata more frequently had a history of diabetes mellitus, congestive heart failure, and high–risk coronary artery disease, whereas malignancy was less common.
Table 1.
Preoperative characteristics by eGFR
| Characteristics | All Patients (n=14,037) | eGFRa | P Value | ||||
|---|---|---|---|---|---|---|---|
| ≥60 (n=11,266) | 45 to <60 (n=1488) | 30 to <45 (n=763) | 15 to <30 (n=274) | <15 or on dialysis (n=246) | |||
| Age (yr) | |||||||
| Median | 65.2 | 62.6 | 74.9 | 77.7 | 77.8 | 66.5 | <0.001 |
| 25th, 75th percentiles | 56.0, 75.1 | 54.4, 72.4 | 67.8, 80.6 | 70.1, 83.6 | 68.9, 84.2 | 55.8, 75.4 | |
| Women, n (%) | 7169 (51.1) | 5735 (50.9) | 781 (52.5) | 393 (51.5) | 148 (54.0) | 112 (45.5) | 0.25 |
| Current, n (%) | |||||||
| Atrial fibrillation | 497 (3.5) | 317 (2.8) | 71 (4.8) | 66 (8.7) | 28 (10.2) | 15 (6.1) | <0.001 |
| Diabetes | 2833 (20.2) | 1978 (17.6) | 384 (25.8) | 266 (34.9) | 97 (35.4) | 108 (43.9) | <0.001 |
| History of, n (%) | |||||||
| Congestive heart failure | 694 (4.9) | 392 (3.5) | 100 (6.7) | 101 (13.2) | 56 (20.4) | 45 (18.3) | <0.001 |
| CAD | 1766 (12.6) | 1133 (10.1) | 291 (19.6) | 201 (26.3) | 76 (27.7) | 65 (26.4) | <0.001 |
| High-risk CAD | 172 (1.2) | 101 (0.9) | 25 (1.7) | 24 (3.2) | 8 (2.9) | 14 (5.7) | <0.001 |
| Cardiac arrest | 62 (0.4) | 44 (0.4) | 8 (0.5) | 6 (0.8) | 1 (0.4) | 3 (1.2) | 0.11 |
| DVT/PE | 454 (3.2) | 317 (2.8) | 65 (4.4) | 46 (6.0) | 12 (4.4) | 14 (5.7) | <0.001 |
| Stroke | 670 (4.8) | 420 (3.7) | 120 (8.1) | 81 (10.6) | 28 (10.2) | 21 (8.5) | <0.001 |
| Transient ischemic attack | 361 (2.6) | 236 (2.1) | 57 (3.8) | 39 (5.1) | 17 (6.2) | 12 (4.9) | <0.001 |
| Peripheral vascular disease | 788 (5.6) | 466 (4.1) | 143 (9.6) | 89 (11.7) | 37 (13.5) | 53 (21.5) | <0.001 |
| Sleep apnea | 734 (5.6) | 604 (5.4) | 70 (4.7) | 39 (5.1) | 9 (3.3) | 12 (4.9) | 0.49 |
| Hypertension | 7309 (52.1) | 5180 (46.0) | 1073 (72.1) | 627 (82.2) | 227 (82.9) | 202 (82.1) | <0.001 |
| COPD | 1252 (8.9) | 919 (8.2) | 165 (11.1) | 101 (13.2) | 43 (15.7) | 24 (9.8) | <0.001 |
| Active cancer, n (%) | 3865 (27.5) | 3196 (28.4) | 381 (25.6) | 181 (23.7) | 69 (25.2) | 38 (15.5) | <0.001 |
| eGFR (ml/min per 1.73 m2) | |||||||
| Medianb | 82.6 | 87.3 | 53.8 | 38.3 | 24.7 | 9.4 | <0.001 |
| 25th, 75th percentiles | 66.1, 95.0 | 76.1, 97.3 | 49.7, 57.2 | 34.5, 41.8 | 20.3, 27.6 | 6.7, 12.3 | |
| Urgent surgery, n (%) | 1902 (13.6) | 1428 (12.7) | 205 (13.8) | 145 (19.0) | 69 (25.1) | 55 (22.4) | <0.001 |
| Type of surgery, n (%) | |||||||
| Vascular | 480 (3.4) | 295 (2.6) | 91 (6.1) | 63 (8.3) | 15 (5.5) | 16 (6.5) | <0.001 |
| General | 2877 (20.5) | 2294 (20.4) | 313 (21.0) | 142 (18.6) | 76 (27.7) | 52 (21.1) | 0.03 |
| Thoracic | 366 (2.6) | 326 (2.9) | 27 (1.8) | 9 (1.2) | 1 (0.4) | 3 (1.2) | 0.004 |
| Major urology/gynecologic | 1732 (12.3) | 1386 (12.3) | 187 (12.6) | 91 (11.9) | 27 (9.9) | 41 (16.7) | 0.20 |
| Major orthopedic | 3013 (21.5) | 2242 (19.9) | 405 (27.2) | 245 (32.1) | 69 (25.2) | 52 (21.1) | <0.001 |
| Major neurologic | 796 (5.7) | 710 (6.3) | 53 (3.6) | 27 (3.5) | 3 (1.1) | 3 (1.2) | <0.001 |
| Low-risk surgeries | 5374 (38.3) | 4507 (40.0) | 476 (32.0) | 207 (27.1) | 96 (35.0) | 88 (35.8) | <0.001 |
CAD, coronary artery disease; DVT/PE, deep venous thrombosis or pulmonary embolus; COPD, chronic obstructive pulmonary disease.
eGFR calculated using the Chronic Kidney Disease Epidemiology Collaboration equation.
Patients on dialysis not included in the median eGFR calculation.
A median of three (two to four) cTnTs were drawn in each patient. The proportion of patients who died within 30 days after surgery was greater for patients with an elevated cTnT than for those without an elevated cTnT in every strata of eGFR (Table 2). For patients with an eGFR≥30 ml/min per 1.73 m2, a cTnT≥0.02 ng/ml was associated with a 4- to 6-fold increase in the adjusted risk of 30-day mortality compared with a 1.3- to 1.5-fold increase in the adjusted risk in patients with an eGFR<30 ml/min per 1.73 m2 or on dialysis (Figure 2). A statistically significant difference in the effect was only seen for the 15 to <30 ml/min per 1.73 m2 group (interaction P value=0.02).
Table 2.
Frequency of an abnormal cTnT defined as ≥0.02 ng/ml and 30-day all-cause mortality by eGFR strata
| eGFR Strata (ml/min per 1.73 m2) and cTnT Value (ng/ml) | n (%) | Died (95% Confidence Interval) | HR (95% Confidence Interval) |
|---|---|---|---|
| ≥60 | |||
| <0.02 | 10,418 (92.5) | 0.9 (0.8 to 1.1) | Referent |
| ≥0.02 | 848 (7.5) | 7.3 (5.7 to 9.3) | 4.37 (3.13 to 6.11) |
| 45 to <60 | |||
| <0.02 | 1214 (81.6) | 1.0 (0.6 to 1.7) | 0.77 (0.42 to 1.42) |
| ≥0.02 | 274 (18.4) | 8.4 (5.7 to 12.3) | 4.74 (2.94 to 7.65) |
| 30 to <45 | |||
| <0.02 | 510 (66.8) | 1.4 (0.7 to 2.8) | 0.93 (0.43 to 2.03) |
| ≥0.02 | 253 (33.2) | 10.3 (7.1 to 14.6) | 5.87 (3.69 to 9.33) |
| 15 to <30 | |||
| <0.02 | 127 (46.4) | 4.7 (2.2 to 9.9) | 3.46 (1.51 to 7.95) |
| ≥0.02 | 147 (53.6) | 10.2 (6.3 to 16.2) | 4.60 (2.62 to 8.09) |
| <15 or on dialysis | |||
| <0.02 | 75 (30.5) | 4.0 (1.4 to 11.1) | 3.96 (1.25 to 12.54) |
| ≥0.02 | 171 (69.5) | 8.8 (5.4 to 14.0) | 5.77 (3.28 to 10.15) |
Figure 2.
Forest plot demonstrating attenuated adjusted hazard ratios for death within 30 days of noncardiac surgery associated with an abnormal cTnT in patients with an eGFR<30 ml/min per 1.73 m2 compared to patients with an eGFR 30 ml/min per 1.73 m2 or greater. An abnormal cTnT was defined as cTnt ≥0.02. P values test the hypothesis that the HR for an abnormal TnT in an eGFR strata differs from the HR associated with an abnormal TnT in eGFR≥60 ml/min per 1.73 m2. 95% CI, 95% confidence interval.
The addition of postoperative cTnT to the clinical model using preoperative characteristics did not improve risk classification in patients with an eGFR≤30 ml/min per 1.73 m2. The net reclassification improvement (NRI) for a model using cTnT compared with a model without cTnT was 1.5% (Supplemental Table 1).
Using an alternative definition of an abnormal cTnT as ≥0.03 ng/ml did not materially alter these results. The adjusted hazard ratio (aHR) for patients with an eGFR≥30 ml/min per 1.73 m2 was between 5.14 (95% CI, 3.57 to 7.10) and 7.23 (95% CI, 3.61 to 18.3), whereas the aHR for patients with an eGFR<30 ml/min per 1.73 m2 or on dialysis ranged from 1.18 (95% CI, 0.44 to 3.73) to 1.31 (0.47 to 8.54) (Supplemental Table 2). Similarly, the alternative definition of an abnormal cTnT as a change of ≥0.02 ng/ml did not materially alter the primary results, with an aHR for patients with an eGFR<30 ml/min per 1.73 m2 of 1.03 (95% CI, 0.27 to 2.72) (Supplemental Table 3).
Discussion
In patients with an eGFR>30 ml/min per 1.73 m2, a postoperative cTnT ≥0.02 ng/ml predicts a 4- to 6-fold higher risk of death within 30 days compared with similar patients with postoperative cTnT concentrations <0.02 ng/ml. Patients with an eGFR≤30 ml/min per 1.73 m2 are at substantial risk of death within 30 days of surgery, but a postoperative cTnT ≥0.02 ng/ml was associated with only an approximately 1.5-fold increased risk of death and did not improve risk discrimination compared with clinical characteristics.
To our knowledge, we are the first to assess the risk associated with postoperative cTnT in patients with severely impaired kidney function. Other studies showed that older generations of cTnT assays were prognostic of mortality in patients with reduced eGFR or ESRD.7–14 Studies of patients with reduced eGFR and an acute coronary syndrome or presenting to an emergency room with chest pain also showed an increased risk of death in those with an elevated cTnT compared with a cTnT below the established necrosis limit.15–17 Older studies typically defined an elevated cTnT as ≥0.09 ng/ml and found roughly a 2.6-fold increased risk of death in those with an elevated cTnT. We found an approximately 1.5-fold increase in the risk of death in patients with an eGFR<30 ml/min per 1.73 m2. This difference may be because of the lower cTnT threshold that we used to define an abnormal concentration, characteristics of the assay, a truly lesser associated risk in patients undergoing noncardiac surgery, greater residual confounding in older studies, or random error. In particular, if circulating cTnT is cleared by the kidneys, patients with reduced kidney function may develop measurable concentrations with less cardiac injury than patients with normal kidney function. Small increases in cTnT would be expected to incur less risk in patients with reduced eGFR in this case.
Also, up to 43% of patients without symptoms of myocardial ischemia but with reduced eGFR not requiring dialysis were found to have an abnormal cTnT after surgery. The elevated cTnT in patients with poor kidney function in our study may be chronic and not have the same prognostic (or pathophysiologic) implications as an acute rise brought on by surgery. However, in a substudy of the Vascular Events in Noncardiac Surgery Patients Cohort Evaluation (VISION), we found that 21% of patients had an elevated high–sensitivity cTnT before surgery, regardless of kidney function or urgency of the surgery, suggesting that measurable quantities of cTnT are also present in many patients without severely impaired kidney function.18 Whether this means that substantial numbers of patients have myocardial injury before undergoing surgery is unclear.
It is important to recognize that patients that undergo surgery are selected to do so. Patients with reduced eGFR may be more rigorously screened before elective surgery and may be more commonly rejected. The population that is actually operated on may, therefore, represent a special subgroup of patients with reduced eGFR in whom the risk of cardiovascular events is not the primary risk factor for early death after surgery. Some evidence of this is seen by virtue that 24% of the group with an eGFR<30 ml/min per 1.73 m2 underwent urgent/emergent procedures compared with 13% with an eGFR≥30 ml/min per 1.73 m2. Patients with poor kidney function may either be operated on under more emergent circumstances, in which case their risk of death may be dominated by the urgency of the surgery, and have nonvascular complications after surgery that are more relevant.
It bears mention that there was a statistically significant interaction for the eGFR strata 15 to <30 ml/min per 1.73 m2 but not for the eGFR strata <15 ml/min per 1.73 m2 or the on dialysis strata. This may be because of lack of power to detect the interaction in the lowest kidney function strata, which is supported by the consistency of the effect sizes between these two strata. However, patients on dialysis may also be qualitatively different than patients with low eGFR. However, our sample was too small to make a valid comparison of urgent/emergent surgeries and nonurgent emergent surgeries.
Our study has several notable strengths. The study sample was large and representative of patients undergoing major noncardiac surgery from several hospitals in several countries, and patients were well characterized at baseline. We lost very few patients to follow-up and had excellent adherence to the cTnT measurement in the first 3 days after surgery. Our cTnT threshold was empirically determined using an objective method and shown to be the most important risk factor for 30-day mortality in patients undergoing noncardiac surgery. Finally, there were 269 deaths within 30 days, allowing us to fit a stable regression model.
The limitations of our study must also be considered. There were still relatively few patients with very poor kidney function. Thus, only large effects could be reliably detected by a test of interaction. We are unable to determine how stable kidney function was before surgery. Patients may be misclassified because of acute changes in kidney function before surgery (i.e., AKI). Whether the interaction that we detected applies equally to patients with acutely deteriorating kidney function and patients with chronically poor kidney function is unclear. Our results are on the basis of only one assay (the fourth generation cTnT assay). Direct extrapolation of our results to other assays and particularly, the newly available high–sensitivity assays is problematic. One might expect that highly sensitive assays will reinforce the issue that we identified. Our results may be more important, because highly sensitive assays will be even more difficult to interpret in the setting of poor kidney function. Additional studies in both surgical and nonsurgical settings are needed urgently given that patients with reduced eGFR are at high risk of cardiac events. Finally, it is important to highlight that our study was on the basis of post hoc analyses, and our findings require confirmation in additional studies.
Monitoring postoperative cTnT remains an important instrument to identify patients at high risk of postoperative death. However, the interpretation of postoperative cTnTs in patients with very poor kidney function is unclear. Given the high prevalence of CKD and that >200 million individuals undergo surgery each year, additional studies to determine how to use these important tests for cardiac injury to identify patients at high risk of death and other adverse events are required urgently.
Concise Methods
The VISION Cohort Methods
The VISION (clinicaltrials.gov identifier NCT00512109) is an international prospective cohort study evaluating major complications after noncardiac surgery. The VISION recruited patients undergoing noncardiac surgery at 12 centers in eight countries between August of 2007 and December of 2011. Participants were followed during their hospitalization and up to 30 days after surgery.
Participants
Participants in the VISION were 45 years old or older and undergoing an operation that required a general or regional anesthetic and at least an overnight hospital stay. Participants provided consent before surgery, or for those for whom we could not obtain consent preoperatively because of emergency surgery, consent was obtained within 24 hours after surgery. Furthermore, eight centers used a deferred consent process for patients unable to provide consent (e.g., patients sedated and mechanically ventilated) for whom no legal substitute decision maker was available. This allowed collection of cTnT measurements while awaiting patient’s or legal substitute decision maker’s consent. Patients were excluded if they had participated in the VISION for a previous surgery. Potential participants were identified by screening daily patient lists in the preoperative assessment clinics, surgical wards, intensive care units, surgical lists, and patients in preoperative holding areas. In centers where the surgical volume exceeded the research staff’s capacity to enroll at least 80% of all patients, the centers were assigned random weeks for recruitment of all or randomly selected surgical services.
Outcomes and Exposures
The outcome of interest was mortality from any cause up to 30 days after surgery. The main exposures of interest were preoperative kidney function and postoperative cTnT. Preoperative kidney function was estimated using the serum creatinine concentration measured most recently before surgery as part of routine clinical care. The serum creatinine concentration was converted to an eGFR using the Chronic Kidney Disease Epidemiology Collaboration equation.19 Each participant’s eGFR was then categorized as ≥60, 45 to <60, 30 to <45, 15 to <30, or <15 ml/min per 1.73 m2 or on dialysis. Therefore, in this study, we further categorized low eGFRs compared with our prior study, which categorized a low eGFR as <30 ml/min per 1.73 m2 or on dialysis.
cTnT was measured prospectively using the Roche Fourth Generation Elecsys Assay at 6–12 hours postoperatively and then, days 1, 2, and 3 postoperatively. Patients enrolled between 12 and 24 hours after surgery had cTnT measured immediately and then daily up to the third postoperative day. All cTnT measurements were analyzed at the participating center and reported to the attending physicians. For our primary analysis, participants were classified by whether their cTnT was <0.02 or ≥0.02 ng/ml, because this was the first threshold shown to have prognostic importance in our previous study.6 In our exploratory analyses, participants were classified as having a cTnT either <0.03 or ≥0.03 ng/ml or a change in cTnT by at least 0.02 ng/ml over the course of the first 3 postoperative days. For the change in cTnT analysis, patients with a lowest cTnT value of ≤0.01 ng/ml met the abnormal cTnT threshold if the highest cTnT value was at least 0.03 ng/ml, whereas patients with a lowest value >0.01 ng/ml were required to have a value at least 0.02 ng/ml higher to be considered abnormal.
Other potential confounders were assessed before surgery or as soon as possible after surgery for patients enrolled using deferred consent (i.e., consented within 24 hours of surgery). A complete list of the potential confounders and their definitions is included in Supplemental Material.
Statistical Analyses
Patient characteristics are described as means (SDs) or medians (25th, 75th percentiles) as appropriate for continuous data and numbers (%) for frequency data. A two–sided P value <0.05 was regarded as statistically significant for all analyses without adjustment for multiple comparisons.
For the primary objective, we performed Cox proportional hazards regression analysis, in which the independent variables were the baseline characteristics determined to be of prognostic importance in our previous study.6 These variables included age (categorized as 45–65 [referent], >65–75, and >75 years old), history of high–risk coronary artery disease, history of peripheral vascular disease, history of stroke, chronic obstructive pulmonary disease, active cancer, urgent or emergent surgery (versus other surgery), major general surgery, and major neurosurgery. The presence of an abnormal cTnT (i.e., ≥0.02 ng/ml) was used as a dichotomous independent variable. Also, kidney function was included as an independent categorical variable with five levels (i.e., ≥60, 45 to <60, 30 to <45, 15 to <30, and <15 ml/min per 1.73 m2 or on dialysis) using the ≥60-ml/min per 1.73 m2 category as the reference category. Included in the model were the interaction terms between abnormal cTnT and each kidney function category.
We calculated the effect estimates from the model as an aHR with a corresponding 95% confidence interval computed by bootstrapping the model 500 times. We assessed model predictive discrimination through evaluation of the NRI in patients with severely impaired kidney function (i.e., eGFR<30 ml/min per 1.73 m2).20 The NRI is defined as the probability that an event is predicted with a new model but not with an old model. Our old model included all of the predictors of 30-day postoperative mortality above (age, high–risk coronary artery disease, peripheral vascular disease, stroke, chronic obstructive pulmonary disease, active cancer, urgent/emergent surgery, and type of surgery), and the new model included the same predictors plus postoperative cTnT. For our analyses, the old model was on the basis of preoperative clinical variables only (age, high–risk coronary artery disease, peripheral vascular disease, stroke, chronic obstructive pulmonary disease, active cancer, urgent/emergent surgery, and type of surgery), and our new model included cTnT in addition to the preoperative clinical variables. We categorized the predicted risk for each patient with each model as <1%, 1%–5%, 5.1%–10%, and >10%. The NRI represents the proportion of patients with predicted risk that more closely approximates their actual 30-day event status when cTnT is used in addition to the preoperative clinical variables.
Exploratory analyses were conducted to determine if alternative definitions of an abnormal cTnT ameliorated any effect modification of eGFR. The alternative definitions of an abnormal cTnT included a peak cTnT of ≥0.03 ng/ml or a change in cTnT of ≥0.02 ng/ml as described above.
All analyses were performed with SAS v9.2 (Cary, NC).
Disclosures
Roche Diagnostics Global Office provided cardiac troponin T assays and some financial support for the VISION study. P.J.D. received research grants from Roche Diagnostics (makers of the troponin T assay) and Abbott Diagnostics (makers of a troponin I assay).
Supplementary Material
Acknowledgments
The authors acknowledge W.A.W. Ahmad and C.Y. Chong from the University of Malaysia for their efforts in data collection.
Funding for this study comes from >50 grants for the Vascular Events in Noncardiac Surgery Patients Cohort Evaluation (VISION) and its substudies. M.W. is supported by a New Investigator Award from the Kidney Research Scientist Core Education National Training Program. A.X.G. is supported by a Clinician Scientist Award from the Canadian Institutes of Health Science. P.J.D. is supported by a Career Investigator Award from the Heart and Stroke Foundation of Ontario.
The VISION study funding sources had no role in the design and conduct of the study; the collection, management, analysis, and interpretation of the data; or the preparation, review, or approval of the manuscript.
Footnotes
Published online ahead of print. Publication date available at www.jasn.org.
This article contains supplemental material online at http://jasn.asnjournals.org/lookup/suppl/doi:10.1681/ASN.2014060536/-/DCSupplemental.
References
- 1.Weiser TG, Regenbogen SE, Thompson KD, Haynes AB, Lipsitz SR, Berry WR, Gawande AA: An estimation of the global volume of surgery: A modelling strategy based on available data. Lancet 372: 139–144, 2008 [DOI] [PubMed] [Google Scholar]
- 2.Devereaux PJ, Goldman L, Cook DJ, Gilbert K, Leslie K, Guyatt GH: Perioperative cardiac events in patients undergoing noncardiac surgery: A review of the magnitude of the problem, the pathophysiology of the events and methods to estimate and communicate risk. CMAJ 173: 627–634, 2005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Coresh J, Selvin E, Stevens LA, Manzi J, Kusek JW, Eggers P, Van Lente F, Levey AS: Prevalence of chronic kidney disease in the United States. JAMA 298: 2038–2047, 2007 [DOI] [PubMed] [Google Scholar]
- 4.Mathew A, Devereaux PJ, O’Hare A, Tonelli M, Thiessen-Philbrook H, Nevis IF, Iansavichus AV, Garg AX: Chronic kidney disease and postoperative mortality: A systematic review and meta-analysis. Kidney Int 73: 1069–1081, 2008 [DOI] [PubMed] [Google Scholar]
- 5.Matsushita K, van der Velde M, Astor BC, Woodward M, Levey AS, de Jong PE, Coresh J, Gansevoort RT, Chronic Kidney Disease Prognosis Consortium : Association of estimated glomerular filtration rate and albuminuria with all-cause and cardiovascular mortality in general population cohorts: A collaborative meta-analysis. Lancet 375: 2073–2081, 2010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Devereaux PJ, Chan MT, Alonso-Coello P, Walsh M, Berwanger O, Villar JC, Wang CY, Garutti RI, Jacka MJ, Sigamani A, Srinathan S, Biccard BM, Chow CK, Abraham V, Tiboni M, Pettit S, Szczeklik W, Lurati Buse G, Botto F, Guyatt G, Heels-Ansdell D, Sessler DI, Thorlund K, Garg AX, Mrkobrada M, Thomas S, Rodseth RN, Pearse RM, Thabane L, McQueen MJ, VanHelder T, Bhandari M, Bosch J, Kurz A, Polanczyk C, Malaga G, Nagele P, Le Manach Y, Leuwer M, Yusuf S, Vascular Events In Noncardiac Surgery Patients Cohort Evaluation (VISION) Study Investigators : Association between postoperative troponin levels and 30-day mortality among patients undergoing noncardiac surgery. JAMA 307: 2295–2304, 2012 [DOI] [PubMed] [Google Scholar]
- 7.Khan NA, Hemmelgarn BR, Tonelli M, Thompson CR, Levin A: Prognostic value of troponin T and I among asymptomatic patients with end-stage renal disease: A meta-analysis. Circulation 112: 3088–3096, 2005 [DOI] [PubMed] [Google Scholar]
- 8.Havekes B, van Manen JG, Krediet RT, Boeschoten EW, Vandenbroucke JP, Dekker FW, NECOSAD Study Group : Serum troponin T concentration as a predictor of mortality in hemodialysis and peritoneal dialysis patients. Am J Kidney Dis 47: 823–829, 2006 [DOI] [PubMed] [Google Scholar]
- 9.Roberts MA, MacMillan N, Hare DL, Ratnaike S, Sikaris K, Fraenkel MB, Ierino FL: Cardiac troponin levels in asymptomatic patients on the renal transplant waiting list. Nephrology (Carlton) 11: 471–476, 2006 [DOI] [PubMed] [Google Scholar]
- 10.Wang AY, Lam CW, Wang M, Chan IH, Goggins WB, Yu CM, Lui SF, Sanderson JE: Prognostic value of cardiac troponin T is independent of inflammation, residual renal function, and cardiac hypertrophy and dysfunction in peritoneal dialysis patients. Clin Chem 53: 882–889, 2007 [DOI] [PubMed] [Google Scholar]
- 11.Helleskov Madsen L, Ladefoged S, Hildebrandt P, Atar D: Comparison of four different cardiac troponin assays in patients with end-stage renal disease on chronic haemodialysis. Acute Card Care 10: 173–180, 2008 [DOI] [PubMed] [Google Scholar]
- 12.Roberts MA, Hare DL, Macmillan N, Ratnaike S, Sikaris K, Ierino FL: Serial increased cardiac troponin T predicts mortality in asymptomatic patients treated with chronic haemodialysis. Ann Clin Biochem 46: 291–295, 2009 [DOI] [PubMed] [Google Scholar]
- 13.Orea-Tejeda A, Sánchez-González LR, Castillo-Martínez L, Valdespino-Trejo A, Sánchez-Santillán RN, Keirns-Davies C, Colín-Ramírez E, Montańo-Hernández P, Dorantes-García J: Prognostic value of cardiac troponin T elevation is independent of renal function and clinical findings in heart failure patients. Cardiol J 17: 42–48, 2010 [PubMed] [Google Scholar]
- 14.Apple FS, Pearce LA, Doyle PJ, Otto AP, Murakami MM: Cardiac troponin risk stratification based on 99th percentile reference cutoffs in patients with ischemic symptoms suggestive of acute coronary syndrome: Influence of estimated glomerular filtration rates. Am J Clin Pathol 127: 598–603, 2007 [DOI] [PubMed] [Google Scholar]
- 15.Aviles RJ, Askari AT, Lindahl B, Wallentin L, Jia G, Ohman EM, Mahaffey KW, Newby LK, Califf RM, Simoons ML, Topol EJ, Berger P, Lauer MS: Troponin T levels in patients with acute coronary syndromes, with or without renal dysfunction. N Engl J Med 346: 2047–2052, 2002 [DOI] [PubMed] [Google Scholar]
- 16.Abbas NA, John RI, Webb MC, Kempson ME, Potter AN, Price CP, Vickery S, Lamb EJ: Cardiac troponins and renal function in nondialysis patients with chronic kidney disease. Clin Chem 51: 2059–2066, 2005 [DOI] [PubMed] [Google Scholar]
- 17.McCullough PA, Nowak RM, Foreback C, Tokarski G, Tomlanovich MC, Khoury NE, Weaver WD, Sandberg KR, McCord J: Performance of multiple cardiac biomarkers measured in the emergency department in patients with chronic kidney disease and chest pain. Acad Emerg Med 9: 1389–1396, 2002 [DOI] [PubMed] [Google Scholar]
- 18.Kavsak PA, Walsh M, Srinathan S, Thorlacius L, Buse GL, Botto F, Pettit S, McQueen MJ, Hill SA, Thomas S, Mrkobrada M, Alonso-Coello P, Berwanger O, Biccard BM, Cembrowski G, Chan MT, Chow CK, de Miguel A, Garcia M, Graham MM, Jacka MJ, Kueh JH, Li SC, Lit LC, Martínez-Brú C, Naidoo P, Nagele P, Pearse RM, Rodseth RN, Sessler DI, Sigamani A, Szczeklik W, Tiboni M, Villar JC, Wang CY, Xavier D, Devereaux PJ: High sensitivity troponin T concentrations in patients undergoing noncardiac surgery: A prospective cohort study. Clin Biochem 44: 1021–1024, 2011 [DOI] [PubMed] [Google Scholar]
- 19.Levey AS, Stevens LA: Estimating GFR using the CKD Epidemiology Collaboration (CKD-EPI) creatinine equation: More accurate GFR estimates, lower CKD prevalence estimates, and better risk predictions. Am J Kidney Dis 55: 622–627, 2010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Pencina MJ, D'Agostino RB, Sr., D'Agostino RB, Jr., Vasan RS: Evaluating the added predictive ability of a new marker: From area under the ROC curve to reclassification and beyond. Stat Med 27: 157–172, 2008 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.