Search

Background: Postoperative hemodynamic deterioration among cardiac surgical patients can indicate or lead to adverse outcomes. Whereas prediction models for such events using electronic health records or physiologic waveform data are previously described, their combined value remains incompletely defined. The authors hypothesized that models incorporating electronic health record and processed waveform signal data (electrocardiogram lead II, pulse plethysmography, arterial catheter tracing) would yield improved performance versus either modality alone.Methods: Intensive care unit data were reviewed after elective adult cardiac surgical procedures at an academic center between 2013 and 2020. Model features included electronic health record features and physiologic waveforms. Tensor decomposition was used for waveform feature reduction. Machine learning-based prediction models included a 2013 to 2017 training set and a 2017 to 2020 temporal holdout test set. The primary outcome was a postoperative deterioration event, defined as a composite of low cardiac index of less than 2.0 ml min-1 m-2, mean arterial pressure of less than 55 mmHg sustained for 120 min or longer, new or escalated inotrope/vasopressor infusion, epinephrine bolus of 1 mg or more, or intensive care unit mortality. Prediction models analyzed data 8 h before events.Results: Among 1,555 cases, 185 (12%) experienced 276 deterioration events, most commonly including low cardiac index (7.0% of patients), new inotrope (1.9%), and sustained hypotension (1.4%). The best performing model on the 2013 to 2017 training set yielded a C-statistic of 0.803 (95% CI, 0.799 to 0.807), although performance was substantially lower in the 2017 to 2020 test set (0.709, 0.705 to 0.712). Test set performance of the combined model was greater than corresponding models limited to solely electronic health record features (0.641; 95% CI, 0.637 to 0.646) or waveform features (0.697; 95% CI, 0.693 to 0.701).Conclusions: Clinical deterioration prediction models combining electronic health record data and waveform data were superior to either modality alone, and performance of combined models was primarily driven by waveform data. Decreased performance of prediction models during temporal validation may be explained by data set shift, a core challenge of healthcare prediction modeling.Editor’s Perspective: [ABSTRACT FROM AUTHOR]

Introduction: Incomplete coronary revascularization is associated with suboptimal outcomes. We investigated the long-term effects of Incomplete, Complete, and Supra-complete revascularization and whether these effects differed in the setting of single-arterial and multi-arterial coronary artery bypass graft (CABG).Methods: We analyzed 15-year mortality in 7157 CABG patients (64.1 ± 10.5 years; 30% women). All patients received a left internal thoracic artery to left anterior descending coronary artery graft with additional venous grafts only (single-arterial) or with at least one additional arterial graft (multi-arterial) and were grouped based on a completeness of revascularization index (CRI = number of grafts minus the number of diseased principal coronary arteries): Incomplete (CRI ≤ -1 [N = 320;4.5%]); Complete (CRI = 0 [N = 2882;40.3%]; reference group); and two Supra-complete categories (CRI = +1[N = 3050; 42.6%]; CRI ≥ + 2 [N = 905; 12.6%]). Risk-adjusted mortality hazard ratios (AHR) were calculated using comprehensive propensity score adjustment by Cox regression.Results: Incomplete revascularization was rare (4.5%) but associated with increased mortality in all patients (AHR [95% confidence interval] = 1.53 [1.29-1.80]), those undergoing single-arterial CABG (AHR = 1.27 [1.04-1.54]) and multi-arterial CABG (AHR = 2.18 [1.60-2.99]), as well as in patients with 3-Vessel (AHR = 1.37 [1.16-1.62]) and, to a lesser degree, with 2-Vessel (AHR = 1.67 [0.53-5.23]) coronary disease. Supra-complete revascularization was generally associated with incrementally decreased mortality in all patients (AHR [CRI = +1] = 0.94 [0.87-1.03]); AHR [CRI ≥ +2] = 0.74 [0.64-0.85]), and was driven by a significantly decreased mortality risk in single-arterial CABG (AHR [CRI = +1] = 0.90 [0.81-0.99]; AHR [CRI ≥ +2] = 0.64 [0.53-0.78]); and 3-Vessel disease patients (AHR [CRI = +1] = 0.94 [0.86-1.04]; and AHR [CRI ≥ +2] = 0.75 [0.63-0.88]) with no impact in multi-arterial CABG (AHR [CRI = +1] = 1.07 [0.91-1.26]; AHR [CRI ≥ +2] = 0.93 [0.73-1.17]).Conclusions: Incomplete revascularization is associated with decreased late survival, irrespective of grafting strategy. Alternatively, supra-complete revascularization is associated with improved survival in patients with 3-Vessel CAD, and in single-arterial but not multi-arterial CABG. [ABSTRACT FROM AUTHOR]

Objective: This study was designed to determine the effect of ketorolac on mortality after cardiac surgery. Design: A retrospective multivariable analysis with propensity matching and propensity scoring. Setting: A tertiary care university-affiliated medical center. Participant: Eleven hundred eighty-six patients undergoing isolated coronary artery bypass surgery. Main Results: Between January 1, 2002, and November 1, 2004, 168 patients undergoing isolated coronary artery bypass surgery received ketorolac, whereas 1,018 patients did not. There were 2 deaths (1%) in the ketorolac group compared with 104 (10%) in the nonketorolac group (p < 0.001). Within 90 days of surgery, there was 1 death (1%) in the ketorolac group compared with 51 (5%) in the nonketorolac group (p = 0.01). By Cox modeling, ketorolac use was associated with a 7-fold lower risk of death (p = 0.02). In the patients who survived at least 90 days, there was 1 death (1%) in the ketorolac group compared with 53 (5%) in the nonketorolac group (p = 0.01). By Cox modeling, ketorolac use was associated with a 2.4-fold lower risk of death (p = 0.03) in the late hazard period. In the propensity-matched groups, Kaplan-Meier survival was better in patients who received ketorolac (p = 0.02). Conclusion: The use of ketorolac was associated with a statistically significant decrease in mortality at follow-up. [Copyright &y& Elsevier]

Background. Blood transfusions have been linked to increased morbidity and mortality. Bleeding during and after cardiac operations and the hemodilution effects of cardiopulmonary bypass commonly result in blood transfusions. Because we could not find any studies evaluating the effects of transfusion on long-term survival after cardiac operation, we sought to determine these effects.Methods. We studied 1,915 patients who underwent first-time isolated coronary artery bypass operations between July 6, 1994 and December 31, 1997 at our institution. Patients with transfusions were compared with those who had not been transfused. Long-term survival data were obtained from the United States Social Security Death Index. Groups were compared by Cox proportional hazard models, Kaplan-Meier survival plots, and hazard functions.Results. Six hundred forty-nine of 1,915 study patients (34%) received a transfusion during their hospitalization. Transfused patients were older, smaller, and more likely to be female, and had more comorbidity. Transfused patients also had twice the 5-year mortality (15% vs 7%) of nontransfused patients. After correction for comorbidities and other factors, transfusion was still associated with a 70% increase in mortality (risk ratio = 1.7; 95% confidence interval = 1.4 to 2.0; p = 0.001). By multivariate analysis, transfusion, peripheral vascular disease, chronic obstructive pulmonary disease, New York Heart Association functional class IV, and age were significant predictors of long-term mortality.Conclusions. We found that blood transfusions during or after coronary artery bypass operations were associated with increased long-term mortality. [Copyright &y& Elsevier]

Background Postoperative deep vein thrombosis (DVT) is associated with significant morbidity. Even with maximal thromboprophylaxis, postoperative DVT is present in 10% of cardiac surgery patients, and is linked to receiving transfusion. We hypothesized that the incidence of DVT varies with the transfused blood product type, and increases with transfusion dose. Study design and methods 139/1070 cardiac surgery patients have DVT despite maximal chemo and mechanical prophylaxis. DVTs were detected via serial perioperative duplex venous scans (DVS). Red blood cells (RBC), platelets (PLT), plasma (FFP) and cryoprecipitate transfusion data were collected. Results Transfusion was used in 506(47%) patients: RBC [468(44%); 4.0 ± 4.2 u]; FFP [155(14.5%); 3.5 ± 2.3 u]; PLT [185(17.3%); 2.2 ± 1.3 u] and Cryoprecipitate [51(4.8%); 1.3 ± 0.6 u]. Isolated RBC transfusion accounted for 92.6% patients receiving one product, and their DVT rate was increased considerably compared to no transfusion (16.7% versus 7.3%; P < 0.001). Incidence of DVT increased substantially for multiple product transfusions; particularly when both RBC and FFP are used (25%–40%). Relative to no RBC (n = 602), multivariate logistic regression analysis identified a significant RBC-DVT dose dependent relation (P < 0.001) with: 1–3 RBC units [n = 285, AOR = 1.95(1.23–3.07), adjusted odds ratio (95% confidence interval)]; 4–6 units [n = 117; AOR = 1.65(0.86–3.20)]; and ≥ 7 RBC units [n = 66; 3.19(1.52–6.70)]. This relation also increased according to an RBC ∗ FFP interaction term [AOR = 1.87(1.11–3.22); P = 0.022]. Conclusion RBC transfusion is associated with increased risk of DVT after cardiac surgery in a dose-dependent fashion that is exacerbated when accompanied with FFP. Postoperative screening diagnostic DVS are warranted in this transfused, high risk for DVT population to facilitate timely therapeutic intervention. [ABSTRACT FROM AUTHOR]

Objective: We sought to determine if decreased left ventricular systolic function was associated with an increased risk of postoperative infectious, respiratory, or renal complications in patients undergoing noncardiac surgery.Design: Retrospective cohort study.Setting: Single tertiary-care, university-based medical center.Patients: We studied individuals who participated in the American College of Surgeons National Quality Improvement Program and had a preoperative echocardiogram conducted at our institution.Interventions: None.Measurements: The incidences of postoperative respiratory (need for postoperative mechanical ventilation or unplanned intubation), renal (acute renal failure or renal insufficiency), and infectious (pneumonia, urinary tract infection, sepsis, or wound infection) complications were analyzed.Main Results: Postoperative infections (n=175, 10%) represented the most common postoperative complication seen in the study population of 1692 individuals. Respiratory complications occurred in 77 (5%) and renal complications occurred in 29 (2%) participants. The time between the echocardiogram and surgery ranged from 7 months (interquartile range [IQR] 1.8-15.7) for those with severely reduced left ventricular ejection fraction (LVEF) to 24 months (IQR 2.5-38.6) for those with a normal LVEF (P=.038). Univariate analysis demonstrated a relationship between decreased preoperative LVEF and infectious and renal complications, but not respiratory complications. After adjusting for preoperative characteristics, decreased preoperative LVEF was associated with infectious (odds ratio, 1.33; 95% confidence interval, 1.03-1.68; P=.0265) and renal (odds ratio, 1.69; 95% confidence interval, 1.12-2.48; P=.0142) complications.Conclusions: Decreased preoperative LVEF is associated with postoperative infections and renal complications. [ABSTRACT FROM AUTHOR]