The ability of intensive care unit physicians to estimate long-term prognosis in survivors of critical illness

PURPOSE

To assess the reliability of physicians' prognoses for intensive care unit (ICU) survivors with respect to long-term survival and health related quality of life (HRQoL).

METHODS

We performed an observational cohort-study in a single mixed tertiary ICU in The Netherlands. ICU survivors with a length of stay >48h were included. At ICU discharge, one-year prognosis was estimated by physicians using the four-option Sabadell score to record their expectations. The outcome of interest was poor outcome, which was defined as dying within one-year follow-up, or surviving with an EuroQoL5D-3L index <0.4.

RESULTS

Among 1399 ICU survivors, 1068 (76%) subjects were expected to have a good outcome; 243 (18%) a poor long-term prognosis; 43 (3%) a poor short-term prognosis, and 45 (3%) to die in hospital (i.e. Sabadell score levels). Poor outcome was observed in 38%, 55%, 86%, and 100% of these groups respectively (concomitant c-index: 0.61). The expected prognosis did not match observed outcome in 365 (36%) patients. This was almost exclusively (99%) due to overoptimism. Physician experience did not affect results.

CONCLUSIONS

Prognoses estimated by physicians incorrectly predicted long-term survival and HRQoL in one-third of ICU survivors. Moreover, inaccurate prognoses were generally the result of overoptimistic expectations of outcome.

Accounting for treatment use when validating a prognostic model: a simulation study

Background

Prognostic models often show poor performance when applied to independent validation data sets. We illustrate how treatment use in a validation set can affect measures of model performance and present the uses and limitations of available analytical methods to account for this using simulated data.

Methods

We outline how the use of risk-lowering treatments in a validation set can lead to an apparent overestimation of risk by a prognostic model that was developed in a treatment-naïve cohort to make predictions of risk without treatment. Potential methods to correct for the effects of treatment use when testing or validating a prognostic model are discussed from a theoretical perspective.. Subsequently, we assess, in simulated data sets, the impact of excluding treated individuals and the use of inverse probability weighting (IPW) on the estimated model discrimination (c-index) and calibration (observed:expected ratio and calibration plots) in scenarios with different patterns and effects of treatment use.

Results

Ignoring the use of effective treatments in a validation data set leads to poorer model discrimination and calibration than would be observed in the untreated target population for the model. Excluding treated individuals provided correct estimates of model performance only when treatment was randomly allocated, although this reduced the precision of the estimates. IPW followed by exclusion of the treated individuals provided correct estimates of model performance in data sets where treatment use was either random or moderately associated with an individual's risk when the assumptions of IPW were met, but yielded incorrect estimates in the presence of non-positivity or an unobserved confounder.

Conclusions

When validating a prognostic model developed to make predictions of risk without treatment, treatment use in the validation set can bias estimates of the performance of the model in future targeted individuals, and should not be ignored. When treatment use is random, treated individuals can be excluded from the analysis. When treatment use is non-random, IPW followed by the exclusion of treated individuals is recommended, however, this method is sensitive to violations of its assumptions.

Electronic supplementary material

The online version of this article (doi:10.1186/s12874-017-0375-8) contains supplementary material, which is available to authorized users.

One-Year Mortality, Causes of Death, and Cardiac Interventions in Patients with Postoperative Myocardial Injury

BACKGROUND

To evaluate the role of routine troponin surveillance in patients undergoing major noncardiac surgery, unblinded screening with cardiac consultation per protocol was implemented at a tertiary care center. In this study, we evaluated 1-year mortality, causes of death, and consequences of cardiac consultation of this protocol.

METHODS

This observational cohort included 3224 patients ≥60 years old undergoing major noncardiac surgery. Troponin I was measured routinely on the first 3 postoperative days. Myocardial injury was defined as troponin I >0.06 μg/L. Regression analysis was used to determine the association between myocardial injury and 1-year mortality. The causes of death, the diagnoses of the cardiologists, and interventions were determined for different levels of troponin elevation.

RESULTS

Postoperative myocardial injury was detected in 715 patients (22%) and was associated with 1-year all-cause mortality (relative risk [RR] 1.4, P = 0.004; RR 1.6, P < 0.001; and RR 2.2, P < 0.001 for minor, moderate, and major troponin elevation, respectively). Cardiac death within 1 year occurred in 3%, 5%, and 11% of patients, respectively, in comparison with 3% of the patients without myocardial injury (P = 0.059). A cardiac consultation was obtained in 290 of the 715 patients (41%). In 119 (41%) of these patients, the myocardial injury was considered to be attributable to a predisposing cardiac condition, and in 111 patients (38%), an intervention was initiated.

CONCLUSIONS

Postoperative myocardial injury was associated with an increased risk of 1-year all-cause but not cardiac mortality. A cardiac consultation with intervention was performed in less than half of these patients. The small number of interventions may be explained by a low suspicion of a cardiac etiology in most patients and lack of consensus for standardized treatment in these patients.

Erratum to: Methods for evaluating medical tests and biomarkers

[This corrects the article DOI: 10.1186/s41512-016-0001-y.].

Kinetics of troponin I in patients with myocardial injury after noncardiac surgery

BACKGROUND

Myocardial injury after noncardiac surgery, as measured by troponin elevation, is strongly associated with mortality. However, it is unknown in which patients prognosis can be improved. The presence of kinetic changes of troponin may be associated with a worse prognosis and warrant more aggressive management. Therefore, we aimed to study the kinetics of troponin in patients with postoperative myocardial injury, and to determine the added predictive value of kinetic changes of troponin on mortality.

METHODS

This cohort study included patients with myocardial injury after noncardiac surgery. Troponin I (TnI) was measured on the first three postoperative days. The primary outcome was all-cause 1-year mortality. We studied both absolute and relative TnI changes, and determined the delta TnI that was associated with mortality to distinguish a rise-and-fall TnI pattern from a stable TnI pattern. Next, we determined the added predictive value of a rise-and-fall TnI pattern for mortality.

RESULTS

In total, 634 patients were included. The risk ratio (RR) for mortality increased significantly with an absolute delta TnI of ≥200 ng/L (RR 1.5, 99.4% CI 1.0-2.2, p=0.003). Using this delta TnI to define a rise-and-fall pattern, 459 patients (72%) had a stable TnI pattern and 175 patients (28%) had a rise-and-fall pattern. When added to a model including the highest TnI value and variables from the revised cardiac risk index (RCRI), the TnI pattern did not increase the predictive value for mortality.

CONCLUSIONS

A postoperative TnI rise-and-fall pattern was associated with 1-year mortality, but had no added value in addition to the highest TnI level to predict 1-year mortality. Therefore, postoperative TnI kinetics are not useful for further mortality risk stratification in patients with myocardial injury after noncardiac surgery.

Treatment use in prognostic model research: a systematic review of cardiovascular prognostic studies

BACKGROUND

Ignoring treatments in prognostic model development or validation can affect the accuracy and transportability of models. We aim to quantify the extent to which the effects of treatment have been addressed in existing prognostic model research and provide recommendations for the handling and reporting of treatment use in future studies.

METHODS

We first describe how and when the use of treatments by individuals in a prognostic study can influence the development or validation of a prognostic model. We subsequently conducted a systematic review of the handling and reporting of treatment use in prognostic model studies in cardiovascular medicine. Data on treatment use (e.g. medications, surgeries, lifestyle interventions), the timing of their use, and the handling of such treatment use in the analyses were extracted and summarised.

RESULTS

Three hundred two articles were included in the review. Treatment use was not mentioned in 91 (30%) articles. One hundred forty-six (48%) reported specific information about treatment use in their studies; 78 (26%) provided information about multiple treatments. Three articles (1%) reported changes in medication use ("treatment drop-in") during follow-up. Seventy-nine articles (26%) excluded treated individuals from their analysis, 80 articles (26%) modelled treatment as an outcome, and of the 155 articles that developed a model, 86 (55%) modelled treatment use, almost exclusively at baseline, as a predictor.

CONCLUSIONS

The use of treatments has been partly considered by the majority of CVD prognostic model studies. Detailed accounts including, for example, information on treatment drop-in were rare. Where relevant, the use of treatments should be considered in the analysis of prognostic model studies, particularly when a prognostic model is designed to guide the use of certain treatments and these treatments have been used by the study participants. Future prognostic model studies should clearly report the use of treatments by study participants and consider the potential impact of treatment use on the study findings.

[Recognising postoperative delirium in the elderly]

OBJECTIVE

To determine the degree of agreement between delirium experts on the diagnosis of delirium based on exactly the same information, and to assess the sensitivity of delirium screening methods used by clinical nurses.

DESIGN

Prospective observational longitudinal study.

METHOD

Older patients (≥ 60 years) who underwent major surgery were included. During the first three days after surgery they had a standardised cognitive screening test which was recorded on video. Two delirium experts independently evaluated these videos and the information from the patient records. They classified the patients as having 'no delirium', 'possible delirium' or 'delirium'. If there was disagreement, a third expert was consulted. The final classification, based on consensus of two or three delirium experts, was compared with the result of the delirium screening carried out by the clinical nurses.

RESULTS

A total of 167 patients were included and 424 postoperative classifications were obtained. The agreement between the experts was 0.61 (95% confidence interval (CI): 0.53-0.68), based on Cohen's kappa. In 89 (21.0%) of the postoperative classifications there was no agreement between the experts and a third expert was consulted. The nurses using the delirium screening tools recognised 32% of the cases that had been classified as delirium by the experts.

CONCLUSION

There was considerable disagreement between the classifications of individual delirium experts, based on exactly the same information, indicating the difficulty of the diagnosis. Furthermore, the sensitivity of the delirium screening tools used by the clinical nurses was poor. Further research should focus on the development of objective methods for recognising delirium.

Clinical recognition of acute aortic dissections: insights from a large single-centre cohort study

Aims

Acute aortic dissection (AD) requires immediate treatment, but is a diagnostic challenge. We studied how often AD was missed initially, which patients were more likely to be missed and how this influenced patient management and outcomes.

Methods

A retrospective cohort study including 200 consecutive patients with AD as the final diagnosis, admitted to a tertiary hospital between 1998 and 2008. The first differential diagnosis was identified and patients with and without AD included were compared. Characteristics associated with a lower level of suspicion were identified using multivariable logistic regression, and Cox regression was used for survival analyses. Missing data were imputed.

Results

Mean age was 63 years, 39% were female and 76% had Stanford type A dissection. In 69% of patients, AD was included in the first differential diagnosis; this was less likely in women (adjusted relative risk [aRR]: 0.66, 95% CI: 0.44–0.99), in the absence of back pain (aRR: 0.51, 95% CI: 0.30–0.84), and in patients with extracardiac atherosclerosis (aRR: 0.64, 95% CI: 0.43–0.96). Absence of AD in the differential diagnosis was associated with the use of more imaging tests (1.8 vs. 2.3, p = 0.01) and increased time from admission to surgery (1.8 vs. 10.1 h, p < 0.01), but not with a difference in the adjusted long-term all-cause mortality (hazard ratio: 0.76, 95% CI: 0.46–1.27).

Conclusion

Acute aortic dissection was initially not suspected in almost one-third of patients, this was more likely in women, in the absence of back pain and in patients with extracardiac atherosclerosis. Although the number of imaging tests was higher and time to surgery longer, patient outcomes were similar in both groups.

Electronic supplementary material

The online version of this article (doi:10.1007/s12471-016-0921-8) contains supplementary material, which is available to authorized users.

Explicit inclusion of treatment in prognostic modeling was recommended in observational and randomized settings

OBJECTIVES

To compare different methods to handle treatment when developing a prognostic model that aims to produce accurate probabilities of the outcome of individuals if left untreated.

STUDY DESIGN AND SETTING

Simulations were performed based on two normally distributed predictors, a binary outcome, and a binary treatment, mimicking a randomized trial or an observational study. Comparison was made between simply ignoring treatment (SIT), restricting the analytical data set to untreated individuals (AUT), inverse probability weighting (IPW), and explicit modeling of treatment (MT). Methods were compared in terms of predictive performance of the model and the proportion of incorrect treatment decisions.

RESULTS

Omitting a genuine predictor of the outcome from the prognostic model decreased model performance, in both an observational study and a randomized trial. In randomized trials, the proportion of incorrect treatment decisions was smaller when applying AUT or MT, compared to SIT and IPW. In observational studies, MT was superior to all other methods regarding the proportion of incorrect treatment decisions.

CONCLUSION

If a prognostic model aims to produce correct probabilities of the outcome in the absence of treatment, ignoring treatments that affect that outcome can lead to suboptimal model performance and incorrect treatment decisions. Explicitly, modeling treatment is recommended.

The predictive value of early acute kidney injury for long-term survival and quality of life of critically ill patients

BACKGROUND

Prognostic factors for the combination of long-term survival and health-related quality of life (HRQoL) after intensive care unit (ICU) stay have not yet been studied. Our aim was to assess whether early acute kidney injury (eAKI), AKI occurring on the first day of ICU admission, is an independent predictor of this combined one-year outcome.

METHODS

We included all patients admitted to the mixed ICU of the University Medical Centre Utrecht between July 2009 and April 2013, excluding patients with chronic dialysis, cardiac surgery, and length of stay shorter than 24 hours. eAKI was defined using the risk, injury, failure, loss, end-stage renal failure (RIFLE) classification, using a newly developed algorithm to classify AKI based on routinely collected patient data. In one-year survivors, HRQoL was measured using the EuroQoL 5D-3L™ (EQ-5D) questionnaire. The primary outcome measure was "poor outcome", defined as an EQ-5D index score <0.4 or death after one year follow up. A multivariable Poisson regression model was performed to adjust for age, comorbidities, admission type and severity of disease factors.

RESULTS

We enrolled 2,420 patients, of whom 871 (36.0 %) died within one year. An additional 286 of 1549 one-year survivors (11.8 %) experienced low HRQoL. The respective incidence of the RIFLE classes, risk, injury and failure, were 456 (18.8 %), 253 (10.5 %) and 123 (5.1 %). After adjustment for other covariates, the RIFLE classes, injury and failure, were independently associated with poor outcome (adjusted relative risk 1.14, 95 % CI 1.01, 1.29; p = 0.03, and 1.25, 95 % CI 1.01, 1.55; p = 0.04), when compared to no eAKI patients . The constituents of this composite outcome were also analysed separately. In a Cox regression model the RIFLE classes, injury and failure, were significantly associated with mortality (adjusted hazard ratio 1.35, 95 % CI 1.11, 1.65; p <0.01, and 1.78, 95 % CI 1.38, 2.30; p <0.01). In one-year survivors specifically, none of the RIFLE classes were significantly associated with low HRQoL.

CONCLUSIONS

ICU patients with moderate or severe AKI during the first 24 hours have a higher probability of mortality or low HRQoL (combined poor outcome), one year after ICU admission. Together with other available early prognostic factors, information on early acute kidney injury could improve informed decision-making on the continuation or withdrawal of treatment in ICU patients.

Diet Quality Scores and Prediction of All-Cause, Cardiovascular and Cancer Mortality in a Pan-European Cohort Study

Scores of overall diet quality have received increasing attention in relation to disease aetiology; however, their value in risk prediction has been little examined. The objective was to assess and compare the association and predictive performance of 10 diet quality scores on 10-year risk of all-cause, CVD and cancer mortality in 451,256 healthy participants to the European Prospective Investigation into Cancer and Nutrition, followed-up for a median of 12.8y. All dietary scores studied showed significant inverse associations with all outcomes. The range of HRs (95% CI) in the top vs. lowest quartile of dietary scores in a composite model including non-invasive factors (age, sex, smoking, body mass index, education, physical activity and study centre) was 0.75 (0.72-0.79) to 0.88 (0.84-0.92) for all-cause, 0.76 (0.69-0.83) to 0.84 (0.76-0.92) for CVD and 0.78 (0.73-0.83) to 0.91 (0.85-0.97) for cancer mortality. Models with dietary scores alone showed low discrimination, but composite models also including age, sex and other non-invasive factors showed good discrimination and calibration, which varied little between different diet scores examined. Mean C-statistic of full models was 0.73, 0.80 and 0.71 for all-cause, CVD and cancer mortality. Dietary scores have poor predictive performance for 10-year mortality risk when used in isolation but display good predictive ability in combination with other non-invasive common risk factors.

Novel Biomarkers to Improve the Prediction of Cardiovascular Event Risk in Type 2 Diabetes Mellitus

BACKGROUND

We evaluated the ability of 23 novel biomarkers representing several pathophysiological pathways to improve the prediction of cardiovascular event (CVE) risk in patients with type 2 diabetes mellitus beyond traditional risk factors.

METHODS AND RESULTS

We used data from 1002 patients with type 2 diabetes mellitus from the Second Manifestations of ARTertial disease (SMART) study and 288 patients from the European Prospective Investigation into Cancer and Nutrition-NL (EPIC-NL). The associations of 23 biomarkers (adiponectin, C-reactive protein, epidermal-type fatty acid binding protein, heart-type fatty acid binding protein, basic fibroblast growth factor, soluble FMS-like tyrosine kinase-1, soluble intercellular adhesion molecule-1 and -3, matrix metalloproteinase [MMP]-1, MMP-3, MMP-9, N-terminal prohormone of B-type natriuretic peptide, osteopontin, osteonectin, osteocalcin, placental growth factor, serum amyloid A, E-selectin, P-selectin, tissue inhibitor of MMP-1, thrombomodulin, soluble vascular cell adhesion molecule-1, and vascular endothelial growth factor) with CVE risk were evaluated by using Cox proportional hazards analysis adjusting for traditional risk factors. The incremental predictive performance was assessed with use of the c-statistic and net reclassification index (NRI; continuous and based on 10-year risk strata 0-10%, 10-20%, 20-30%, >30%). A multimarker model was constructed comprising those biomarkers that improved predictive performance in both cohorts. N-terminal prohormone of B-type natriuretic peptide, osteopontin, and MMP-3 were the only biomarkers significantly associated with an increased risk of CVE and improved predictive performance in both cohorts. In SMART, the combination of these biomarkers increased the c-statistic with 0.03 (95% CI 0.01-0.05), and the continuous NRI was 0.37 (95% CI 0.21-0.52). In EPIC-NL, the multimarker model increased the c-statistic with 0.03 (95% CI 0.00-0.03), and the continuous NRI was 0.44 (95% CI 0.23-0.66). Based on risk strata, the NRI was 0.12 (95% CI 0.03-0.21) in SMART and 0.07 (95% CI -0.04-0.17) in EPIC-NL.

CONCLUSIONS

Of the 23 evaluated biomarkers from different pathophysiological pathways, N-terminal prohormone of B-type natriuretic peptide, osteopontin, MMP-3, and their combination improved CVE risk prediction in 2 separate cohorts of patients with type 2 diabetes mellitus beyond traditional risk factors. However, the number of patients reclassified to a different risk stratum was limited.

Prediction models for cardiovascular disease risk in the general population: systematic review

OBJECTIVE

 To provide an overview of prediction models for risk of cardiovascular disease (CVD) in the general population.

DESIGN

 Systematic review.

DATA SOURCES

 Medline and Embase until June 2013.

ELIGIBILITY CRITERIA FOR STUDY SELECTION

 Studies describing the development or external validation of a multivariable model for predicting CVD risk in the general population.

RESULTS

 9965 references were screened, of which 212 articles were included in the review, describing the development of 363 prediction models and 473 external validations. Most models were developed in Europe (n=167, 46%), predicted risk of fatal or non-fatal coronary heart disease (n=118, 33%) over a 10 year period (n=209, 58%). The most common predictors were smoking (n=325, 90%) and age (n=321, 88%), and most models were sex specific (n=250, 69%). Substantial heterogeneity in predictor and outcome definitions was observed between models, and important clinical and methodological information were often missing. The prediction horizon was not specified for 49 models (13%), and for 92 (25%) crucial information was missing to enable the model to be used for individual risk prediction. Only 132 developed models (36%) were externally validated and only 70 (19%) by independent investigators. Model performance was heterogeneous and measures such as discrimination and calibration were reported for only 65% and 58% of the external validations, respectively.

CONCLUSIONS

 There is an excess of models predicting incident CVD in the general population. The usefulness of most of the models remains unclear owing to methodological shortcomings, incomplete presentation, and lack of external validation and model impact studies. Rather than developing yet another similar CVD risk prediction model, in this era of large datasets, future research should focus on externally validating and comparing head-to-head promising CVD risk models that already exist, on tailoring or even combining these models to local settings, and investigating whether these models can be extended by addition of new predictors.

Survival and quality of life after surgical aortic valve replacement in octogenarians

BACKGROUND

In patients with symptomatic severe aortic stenosis, advanced age is often a reason for a transcatheter rather than surgical aortic valve replacement. In this pre-transcatheter cohort we had the unique opportunity to study outcomes after surgical aortic valve replacement for severe aortic stenosis in patients who might currently be triaged to a percutaneous approach.

METHODS

In a prospective single-center cohort study we compared the incidence of peri-operative complications, mortality, and health-related quality of life in octogenarians versus patients aged <80 years. The quality of life was measured using the SF-36 questionnaire and expressed as a physical and mental component score (PCS and MCS respectively); a score of 50 equals the average score in the age-matched general population. The association between age and the component scores at one-year follow-up was studied with the use of linear regression, corrected for a set of confounding variables.

RESULTS

We included 762 patients, of whom 21.4 % was aged >80 and 49.0 % underwent concomitant revascularization. In octogenarians, the incidence of post-operative delirium was 11.0 %, which was higher than in patients aged below 80 (6.2 %, p = 0.034); the operative mortality (1.9 % vs. 2.9 %; p = 0.59) and long-term survival were not different however (log-rank p = 0.75). In octogenarians, the quality of life was impaired 30-days after surgery (PCS 45.01, p < 0.001; MCS 48.21, p = 0.04), which improved towards or above normal values at one-year follow-up (PCS: 49.92, p = 0.67, MCS: 52.55, p < 0.001). After correction for confounding, age was not significantly associated with the one-year PCS (β 0.08 per year, p = 0.34) or MCS (β 0.08 per year, p = 0.32).

CONCLUSIONS

This pre-transcatheter study showed that surgical aortic valve replacement in octogenarians could be performed with very low mortality, and with a relevant and significant increase of the quality of life towards normal values. Also, age was not associated with a lower PCS or MCS one-year after surgery.

Non-Invasive Continuous Respiratory Monitoring on General Hospital Wards: A Systematic Review

BACKGROUND

Failure to recognize acute deterioration in hospitalized patients may contribute to cardiopulmonary arrest, unscheduled intensive care unit admission and increased mortality.

PURPOSE

In this systematic review we aimed to determine whether continuous non-invasive respiratory monitoring improves early diagnosis of patient deterioration and reduces critical incidents on hospital wards.

DATA SOURCES

Studies were retrieved from Medline, Embase, CINAHL, and the Cochrane library, searched from 1970 till October 25, 2014.

STUDY SELECTION

Electronic databases were searched using keywords and corresponding synonyms 'ward', 'continuous', 'monitoring' and 'respiration'. Pediatric, fetal and animal studies were excluded.

DATA EXTRACTION

Since no validated tool is currently available for diagnostic or intervention studies with continuous monitoring, methodological quality was assessed with a modified tool based on modified STARD, CONSORT, and TREND statements.

DATA SYNTHESIS

Six intervention and five diagnostic studies were included, evaluating the use of eight different devices for continuous respiratory monitoring. Quantitative data synthesis was not possible because intervention, study design and outcomes differed considerably between studies. Outcomes estimates for the intervention studies ranged from RR 0.14 (0.03, 0.64) for cardiopulmonary resuscitation to RR 1.00 (0.41, 2.35) for unplanned ICU admission after introduction of continuous respiratory monitoring.

LIMITATIONS

The methodological quality of most studies was moderate, e.g. 'before-after' designs, incomplete reporting of primary outcomes, and incomplete clinical implementation of the monitoring system.

CONCLUSIONS

Based on the findings of this systematic review, implementation of routine continuous non-invasive respiratory monitoring on general hospital wards cannot yet be advocated as results are inconclusive, and methodological quality of the studies needs improvement. Future research in this area should focus on technology explicitly suitable for low care settings and tailored alarm and treatment algorithms.

Clinical Relevance of Cardiac Troponin Assessment in Patients Undergoing Carotid Endarterectomy

OBJECTIVE

Myocardial infarction (MI) is a frequent complication of carotid endarterectomy (CEA), yet most events are silent. Routine post-operative monitoring of cardiac troponin was implemented to facilitate timely recognition of MI and stratify high risk patients. The aim was to evaluate the incidence of troponin elevation after CEA and its association with adverse cardiovascular events.

METHODS

This analysis included patients ≥60 years old who underwent CEA, whose troponin-I levels were routinely monitored post-operatively and were included in a cohort study that assessed clinical outcomes. A clinical troponin cutoff of 60 ng/L was used. The primary endpoint was the composite of MI, stroke, and cardiovascular death. Secondary endpoints were MI, stroke, coronary intervention, cardiovascular death, and all cause death.

RESULTS

225 consecutive patients were included in the analysis. Troponin elevation occurred in 34 patients (15%) and a post-operative MI was diagnosed in eight patients. After a median follow up of 1.8 years (IQR 1.0-2.6), the primary endpoint occurred in 29% of patients with troponin elevation versus 6.3% without (HR 5.6, 95% CI 2.4-13), MI in 24% versus 1.6% (HR 18.0, 95% CI 4.7-68), stroke in 5.9% versus 4.2% (HR 1.4, 95% CI 0.3-6.7), coronary intervention in 5.9% versus 2.6% (HR 2.7, 95% CI 0.5-14), cardiovascular death in 5.9% versus 0.5% (HR 11.8, 95% CI 1.1-131), and all cause death in 15% versus 5.8% (HR 3.0, 95% CI 1.0-8.7), respectively. Incidences of the primary endpoint and all cause mortality in patients with a post-operative MI versus "troponin only" were 25% versus 7.7% and 25% versus 12%, respectively.

CONCLUSION

Troponin elevation after CEA occurred in 15% of patients. The incidence of adverse cardiovascular events was significantly higher in patients with troponin elevation, which was mainly attributable to silent non-ST segment elevation MIs that occurred in the early post-operative phase.