Perioperative Risk of Noncardiac Surgery in Patients With Asymptomatic Significant Aortic Stenosis: A 10-Year Retrospective Study.

BACKGROUND: Aortic stenosis (AS) is a representative geriatric disease, and there is an anticipated rise in the number of patients requiring noncardiac surgeries in patients with AS. However, there is still a lack of research on the primary predictors of noncardiac perioperative complications in patients with asymptomatic significant AS. METHODS AND RESULTS: Among the cohort of noncardiac surgeries under general anesthesia, with an intermediate to high risk of surgery from 2011 to 2019, at Samsung Medical Center, 221 patients were identified to have asymptomatic significant AS. First, to examine the impact of significant AS on perioperative adverse events, the occurrences of major adverse cardiovascular events and perioperative adverse cardiovascular events were compared between patients with asymptomatic significant AS and the control group. Second, to identify the factors influencing the perioperative adverse events in patients with asymptomatic significant AS, a least absolute shrinkage and selection operator regression model was used. There was no significant difference between the control group and the asymptomatic significant AS group in the event rate of major adverse cardiovascular events (4.6% at control group versus 5.5% at asymptomatic significant AS group; P=0.608) and perioperative adverse cardiovascular events (13.8% at control group versus 18.3% at asymptomatic significant AS group; P=0.130). Cardiac damage stage was a significant risk factor of major adverse cardiovascular events and perioperative adverse cardiovascular events. CONCLUSIONS: There was no significant difference in major postoperative cardiovascular events between patients with asymptomatic significant AS and the control group. Advanced cardiac damage stage in significant AS is an important factor in perioperative risk of noncardiac surgery.

Risk of age older than 65 years for 30-day cardiac complication may be comparable to low-to-moderate risk according to revised cardiac risk index in non-cardiac surgery.

Revised cardiac risk index (RCRI) is widely used for surgical patients without containing age as a risk factor. We investigated age older than 65 years with respect to low-to-moderate risk of RCRI. From January 2011 to June 2019, a total of 203,787 consecutive adult patients underwent non-cardiac surgery at our institution. After excluding high-risk patients defined as RCRI score > 2, we stratified the patients into four groups according to RCRI and age (A: age < 65 with RCRI < 2, [n = 148,288], B: age ≥ 65 with RCRI < 2, [n = 42,841], C: age < 65 with RCRI = 2, [n = 5,271], and D: age ≥ 65 with RCRI = 2, [n = 5,698]). Incidence of major cardiac complication defined as a composite of cardiac death, cardiac arrest and myocardial infarction was compared. After excluding 1,689 patients with high risk (defined as RCRI score > 2), 202,098 patients were enrolled. The incidence with 95% confidence interval of major cardiac complication for A, B, C, and D groups was 0.3% (0.2-0.3), 1.1% (1.0-1.2), 1.8% (1.6-1.8), and 3.1% (2.6-3.6), respectively. In a direct comparison between B and C groups, old patients with RCRI < 2 showed a significantly lower risk compared to younger patients with RCRI = 2 (odd ratio, 0.62; 95% confidence interval, 0.50-0.78; p < 0.001). In non-cardiac surgery, the risk of age older than 65 years was shown to be comparable with low-to-moderate risk according to RCRI.

Antiplatelet therapy and long-term mortality in patients with myocardial injury after non-cardiac surgery.

BACKGROUNDS: Myocardial injury after non-cardiac surgery (MINS) has recently been accepted as a common complication associated with increased mortality. However, little is known about the treatment of MINS. The aim of this study was to investigate an association between antiplatelet therapy and long-term mortality after MINS. METHODS: From 2010 to 2019, patients with MINS, defined as having a peak high-sensitivity troponin I higher than 40 ng/L within 30 days after non-cardiac surgery, were screened at a tertiary centre. Patients were excluded if they had a history of coronary revascularisation before or during index hospitalisation. Clinical outcomes at 1 year were compared between patients with and without antiplatelet therapy at hospital discharge. The primary outcome was death, and the secondary outcome was major bleeding. RESULTS: Of the 3818 eligible patients with MINS, 940 (24.6%) received antiplatelet therapy at hospital discharge. Patients with antiplatelet therapy had a significantly lower mortality at 1 year than those without antiplatelet therapy (7.5% vs 15.9%, adjusted HR 0.60, 95% CI 0.45 to 0.79, p<0.001). A risk of major bleeding at 1 year was not significantly different between the patients with and without antiplatelet therapy (6.6% vs 7.6%, adjusted HR 0.85, 95% CI 0.62 to 1.17, p=0.324). In propensity score-matched analysis of 886 pairs, patients with antiplatelet therapy had a significantly lower risk of 1-year mortality (adjusted HR 0.53, 95% CI 0.39 to 0.73, p<0.001) than those without antiplatelet therapy. CONCLUSIONS: In patients with MINS, antiplatelet therapy at discharge was associated with decreased 1-year mortality.

Target and non-target analytical method for potential hazardous substances in livestock and pet hair using liquid- and gas chromatography quadrupole time-of-flight mass spectrometry.

Extraction using acetonitrile and water and quadrupole time-of-flight mass spectrometry (LC and GC-QTOF/MS) techniques were used to screen for potential hazardous substances in livestock and pet hair. In addition, LC-MS/MS and GC-MS/MS techniques were used for verification of the analytical method and quantitative analysis of pesticides, veterinary drugs, mycotoxins and antioxidants in hair. Optimized sample preparation involves extracting 0.05 g of sample with 0.6 mL of ACN and 0.4 mL of distilled water. In addition, the two layers were separated by adding 0.1 g of NaCl. Then, both the ACN and water layers were analyzed by LC-TOF/MS, and the ACN layer was analyzed by GC-TOF/MS. Most of the matrix effects of livestock and pet hair were less than 50%, but some matrices and components showed high results, so matrix matching correction was applied for more precise quantification. Method validation was performed for 394 constituents (293 pesticides, 93 veterinary drugs, 6 mycotoxins and 2 preservatives) in dog, cat, cow and pig hair and chicken and duck feathers. All components showed good linearity (r2 ≥0.98) in the developed assay. The quantification limit of all compounds was set at 0.02 mg/kg, which is the lowest level that satisfies the recovery rate standard. The recovery experiment was repeated 8 times at 3 concentrations. Most of the components were extracted with the ACN layer, and the recovery rate was 63.35-119.98%. In order to confirm the efficiency of extracting harmful substances from actual samples, 30 hairs of livestock and pets were screened.

Machine learning-based prediction model for postoperative delirium in non-cardiac surgery.

BACKGROUND: Postoperative delirium is a common complication that is distressing. This study aimed to demonstrate a prediction model for delirium. METHODS: Among 203,374undergoing non-cardiac surgery between January 2011 and June 2019 at Samsung Medical Center, 2,865 (1.4%) were diagnosed with postoperative delirium. After comparing performances of machine learning algorithms, we chose variables for a prediction model based on an extreme gradient boosting algorithm. Using the top five variables, we generated a prediction model for delirium and conducted an external validation. The Kaplan-Meier and Cox survival analyses were used to analyse the difference of delirium occurrence in patients classified as a prediction model. RESULTS: The top five variables selected for the postoperative delirium prediction model were age, operation duration, physical status classification, male sex, and surgical risk. An optimal probability threshold in this model was estimated to be 0.02. The area under the receiver operating characteristic (AUROC) curve was 0.870 with a 95% confidence interval of 0.855-0.885, and the sensitivity and specificity of the model were 0.76 and 0.84, respectively. In an external validation, the AUROC was 0.867 (0.845-0.877). In the survival analysis, delirium occurred more frequently in the group of patients predicted as delirium using an internal validation dataset (p < 0.001). CONCLUSION: Based on machine learning techniques, we analyzed a prediction model of delirium in patients who underwent non-cardiac surgery. Screening for delirium based on the prediction model could improve postoperative care. The working model is provided online and is available for further verification among other populations. TRIAL REGISTRATION: KCT 0006363.

Sex differences in myocardial injury after non-cardiac surgery and postoperative mortality.

BACKGROUND: Myocardial injury after non-cardiac surgery (MINS) has recently been accepted as a predictor of mortality. However, sex differences in the incidence of MINS and survival thereafter are not fully understood. This study aimed to compare the incidence of MINS and mortality among male and female patients. METHODS: This single-center study was conducted using the database of a large tertiary referral hospital. Consecutive patients with cardiac troponin (cTn) detected within 30 days after non-cardiac surgery performed between January 2010 and June 2019 were grouped according to sex. The incidence of MINS and mortality of patients with MINS were compared between men and women. RESULTS: Of the 33,311 patients, 18,546 (55.7%) were men and 14,765 (44.3%) were women. In a multivariable analysis, women showed a significantly lower incidence of MINS than did men (17.9% vs. 14.2%; odds ratio, 0.76; 95% confidence interval [CI], 0.71-0.81; P < 0.001). In patients with MINS, the propensity-score-matched analysis showed that 30-day mortality did not differ according to sex, but mortality in females was significantly lower than that in males during the overall follow-up (33.0% vs. 25.7%; hazard ratio, 0.75; 95% CI, 0.66-0.84; P < 0.001). CONCLUSION: The incidence of MINS was lower in women than in men. In patients with MINS, female sex may be associated with a survival benefit. Further studies are needed to confirm these findings.

Days alive and out of hospital at 30 days and outcomes of off-pump coronary artery bypass grafting.

Days alive and out of hospital (DAOH) is a simple estimator based on the number of days not in hospital within a defined period. In cases of mortality within the period, DAOH is regarded as zero. It has not been validated solely in off-pump coronary artery bypass grafting (OPCAB). This study aimed to demonstrate a correlation between DAOH and outcome of OPCAB. We identified 2211 OPCAB performed from January 2010 to August 2016. We calculated DAOH at 30 and 60 days. We generated a receiver-operating curve and compared outcomes. The median duration of hospital stay after OPCAB was 6 days. The median DAOH values at 30 and 60 days were 24 and 54 days. The estimated thresholds for 3-year mortality for DAOH at 30 and 60 days were 20 and 50 days. Three-year mortality was higher for short DAOH (1.2% vs. 5.7% and 1.1% vs. 5.6% DAOH at 30 and 60 days). After adjustment, the short DAOH 30 group showed significantly higher mortality during 3-year follow-up (hazard ratio 3.07; 95% confidence interval 1.45-6.52; p = 0.004). DAOH at 30 days after OPCAB showed a correlation with 3-year outcomes. DAOH 30 might be a reliable long-term outcome measure that can be obtained within 30 days after surgery.

Association between inflammation-based prognostic markers and mortality of non-cardiac surgery.

BACKGROUND: To evaluate the association between inflammation and nutrition-based biomarkers and postoperative outcomes after non-cardiac surgery. METHODS: Between January 2011 and June 2019, a total of 102,052 patients undergoing non-cardiac surgery were evaluated, with C-reactive protein (CRP), albumin, and complete blood count (CBC) measured within six months before surgery. We assessed their CRP-to-albumin ratio (CAR), neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), and modified Glasgow Prognostic Score (mGPS). We determined the best cut-off values by using the receiver operating characteristic (ROC) curves. Patients were divided into high and low groups according to the estimated threshold, and we compared the one-year mortality. RESULTS: The one-year mortality of the entire sample was 4.2%. ROC analysis revealed areas under the curve of 0.796, 0.743, 0.670, and 0.708 for CAR, NLR, PLR, and mGPS, respectively. According to the estimated threshold, high CAR, NLR, PLR, and mGPS were associated with increased one-year mortality (1.7% vs. 11.7%, hazard ratio [HR]: 2.38, 95% CI [2.05, 2.76], P < 0.001 for CAR; 2.2% vs. 10.3%, HR: 1.81, 95% CI [1.62, 2.03], P < 0.001 for NLR; 2.6% vs. 10.5%, HR: 1.86, 95% CI [1.73, 2.01], P < 0.001 for PLR; and 2.3% vs. 16.3%, HR: 2.37, 95% CI [2.07, 2.72], P < 0.001 for mGPS). CONCLUSIONS: Preoperative CAR, NRL, PLR, and mGPS were associated with postoperative mortality. Our findings may be helpful in predicting mortality after non-cardiac surgery.

Prediction model for myocardial injury after non-cardiac surgery using machine learning.

Myocardial injury after non-cardiac surgery (MINS) is strongly associated with postoperative outcomes. We developed a prediction model for MINS and have provided it online. Between January 2010 and June 2019, a total of 6811 patients underwent non-cardiac surgery with normal preoperative level of cardiac troponin (cTn). We used machine learning techniques with an extreme gradient boosting algorithm to evaluate the effects of variables on MINS development. We generated two prediction models based on the top 12 and 6 variables. MINS was observed in 1499 (22.0%) patients. The top 12 variables in descending order according to the effects on MINS are preoperative cTn level, intraoperative inotropic drug infusion, operation duration, emergency operation, operation type, age, high-risk surgery, body mass index, chronic kidney disease, coronary artery disease, intraoperative red blood cell transfusion, and current alcoholic use. The prediction models are available at https://sjshin.shinyapps.io/mins_occur_prediction/ . The estimated thresholds were 0.47 in 12-variable models and 0.53 in 6-variable models. The areas under the receiver operating characteristic curves are 0.78 (95% confidence interval [CI] 0.77-0.78) and 0.77 (95% CI 0.77-0.78), respectively, with an accuracy of 0.97 for both models. Using machine learning techniques, we demonstrated prediction models for MINS. These models require further verification in other populations.

Prediction Model for 30-Day Mortality after Non-Cardiac Surgery Using Machine-Learning Techniques Based on Preoperative Evaluation of Electronic Medical Records.

BACKGROUND: Machine-learning techniques are useful for creating prediction models in clinical practice. This study aimed to construct a prediction model of postoperative 30-day mortality based on an automatically extracted electronic preoperative evaluation sheet. METHODS: We used data from 276,341 consecutive adult patients who underwent non-cardiac surgery between January 2011 and December 2020 at a tertiary center for model development and internal validation, and another dataset from 63,384 patients between January 2011 and October 2021 at another center for external validation. Postoperative 30-day mortality was 0.16%. We developed an extreme gradient boosting (XGB) prediction model using only variables from preoperative evaluation sheets. RESULTS: The model yielded an area under the curve of 0.960 and an area under the precision and recall curve of 0.216, which were 0.932 and 0.122, respectively, in the external validation set. The optimal threshold calculated by Youden's J statistic had a sensitivity of 0.885 and specificity of 0.914. In an additional analysis with balanced distribution, the model showed a similar predictive value. CONCLUSION: We presented a machine-learning prediction model for 30-day mortality after non-cardiac surgery using preoperative variables automatically extracted from electronic medical records and validated the model in a multi-center setting. Our model may help clinicians predict postoperative outcomes.

Preoperative left atrial volume index may be associated with postoperative atrial fibrillation in non-cardiac surgery.

Background: Postoperative atrial fibrillation (POAF) is related to mortality after non-cardiac surgery. Left atrial volume index (LAVI) is known to be associated with prognosis and development of atrial fibrillation, but it has not been fully investigated in patients undergoing non-cardiac surgery. Materials and methods: A total of 203,787 consecutive adult patients underwent non-cardiac surgery at our institution between January 2011 and June 2019. After identifying those with available LAVI estimated during preoperative echocardiography, we divided them into those with LAVI higher and lower than 34 mL/m2. The primary outcome was incidence of POAF. Results: A total of 83,097 patients were enrolled in this study. The study patients were divided into the low (57,838 [69.6%]) and high (25,259 [30.4%]) LAVI groups. After an adjustment, higher LAVI was associated with increased incidence of POAF (5.1% vs. 8.1%; odds ratio [OR], 1.33; 95% confidence interval [CI], 1.25-1.41; p < 0.001). In 24,549 pairs of propensity-score-matched population, the result was similar (6.2% vs. 7.9%; OR, 1.30; 95% CI, 1.21-1.39; p < 0.001). The estimated threshold of LAVI associated with POAF was 36.4 mL/m2 with an area under the curve of 0.571. Subgroup analysis in non-thoracic and thoracic surgery showed that the association between preoperative LAVI and POAF significantly interacted with diastolic dysfunction (p for interaction < 0.001), and the observed association was valid in patients without diastolic dysfunction. Conclusion: Preoperative LAVI was shown to be associated with POAF in non-cardiac surgery. Our result needs verification in further studies.

Association between high preoperative body mass index and mortality after cancer surgery.

Despite an association between obesity and increased mortality in the general population, obesity has been paradoxically reported with improved mortality of surgery and some types of cancer. However, this has not been fully investigated in patients undergoing cancer surgery. Using a cohort consisting of mostly Asian population, we enrolled 87,567 adult patients who underwent cancer surgery from March 2010 to December 2019. They were divided into three groups according to body mass index (BMI): 53,980 (61.6%) in the normal (18.5-25 kg/m2), 2,787 (3.2%) in the low BMI (<18.5 kg/m2), and 30,800 (35.2%) in the high BMI (≥25 kg/m2) groups. The high BMI group was further stratified into overweight (25-30 kg/m2) and obese (≥30 kg/m2) groups. The primary outcome was mortality during three years after surgery. Following adjustment by inverse probability weighting, mortality during three years after surgery was significantly lower in the high BMI group than the normal (4.8% vs. 7.0%; hazard ratio [HR], 0.69; confidence interval [CI], 0.64-0.77; p < 0.001) and low BMI (4.8% vs. 13.0%; HR: 0.38; CI: 0.35-0.42; p < 0.001) groups. The mortalities of the overweight and obese groups were lower than that of the normal group (7.0% vs. 5.0%; HR: 0.72; CI: 0.67-0.77; p < 0.001 and 7.0% vs. 3.3%; HR: 0.57; CI: 0.50-0.65; p < 0.001, respectively). This association was not observed in female patients and those undergoing surgery for breast and gynecological cancers. High BMI may be associated with decreased mortality after cancer surgery. Further investigations are needed for clinical application of our finding.

Multi-residue analytical method for detecting pesticides, veterinary drugs, and mycotoxins in feed using liquid- and gas chromatography coupled with mass spectrometry.

Modified QuEChERS and triple quadrupole mass spectrometry (LC and GC-MS/MS) technology were used to sequentially analyze pesticides, veterinary drugs, and mycotoxins in feed. In order to analyze the harmful substances that may remain or occur in the feed, we performed optimization experiments for sample preparation and LC-MS/MS and GC-MS/MS conditions. Optimized sample preparation involves extracting 5 g of sample with 15 mL of 0.25 M EDTA and 10 mL of acetonitrile. And some extracts were diluted 10-fold with 100 mM ammonium formate aqueous solution and analyzed by LC-MS/MS, and some extracts were purified through 25 mg PSA and analyzed by GC-MS/MS by adding an analyte protectant. We confirmed the matrix effect of feed ingredients and compound feeds, and added a dilution process after extraction to increase on-site efficiency. Matrix-matched calibration was applied for quantification. Method validation was performed for 197 pesticides, 56 components for veterinary drugs, and 5 components for toxins. All the components showed good linearity (r2 ≥ 0.98) in the developed analytical method. For most compounds, the limit of quantitation was 0.05 mg/kg. The recovery rate experiment was repeated three times at three concentrations including LOQ in feed ingredient, compound feed for livestock, and compound feed for pets. The recovery rate was 70.09-119.76% and relative standard deviations were ≤ 18.91%. And the accuracy and precision were further verified through cross-validation between laboratories. The developed analytical method was used to monitor 414 domestically distributed and imported feeds.

The use of renin angiotensin aldosterone system inhibitors may be associated with decreased mortality after cancer surgery.

Renin-angiotensin-aldosterone system (RAAS) inhibitors are antihypertensive agents with conflicting results on protective effects against some types of cancer. In light of these controversies, we aimed to study the effects of RAAS inhibitors in patients undergoing cancer surgery. From March 2010 to December 2019, consecutive adult patients with antihypertensive drug prescription at discharge after cancer surgery were enrolled and divided into two groups according to RAAS inhibitors prescription. The primary outcome was 5-year mortality after surgery. Secondary outcomes included mortalities during 3-year and 1-year follow-ups and cancer-specific mortality and recurrence rates during 5-, 3-, and 1-year follow-ups. A total of 19,765 patients were divided into two groups according to RAAS inhibitor prescription at discharge: 8,374 (42.4%) patients in the no RAAS inhibitor group and 11,391 (57.6%) patients in the RAAS inhibitor group. In 5022 pairs of propensity-score matched population, 5-year mortality was significantly lower in the RAAS inhibitor group (11.4% vs. 7.4%, hazard ratio [HR] 0.73, 95% confidence interval [CI] 0.64-0.83, P < 0.001), and 5-year recurrence rate was also lower for the RAAS inhibitor group (5.3% vs. 3.7%, HR 0.82, 95% CI 0.68-0.99, P = 0.04). In our analysis, RAAS inhibitor was associated with decreased 5-year mortality in hypertensive patients who underwent cancer surgery. Prescription of RAAS inhibitor in accordance with current guidelines may be associated with improved mortality after cancer surgery.

Association Between Perioperative Adverse Cardiac Events and Mortality During One-Year Follow-Up After Noncardiac Surgery.

Background Cardiac complications are associated with perioperative mortality, but perioperative adverse cardiac events (PACEs) that are associated with long-term mortality have not been clearly defined. We identified PACE as a composite of myocardial infarction, coronary revascularization, congestive heart failure, arrhythmic attack, acute pulmonary embolism, cardiac arrest, or stroke during the 30-day postoperative period and we compared mortality according to PACE occurrence. Methods and Results From January 2011 to June 2019, a total of 203 787 consecutive adult patients underwent noncardiac surgery at our institution. After excluding those with 30-day mortality, mortality during a 1-year follow-up was compared. Machine learning with the extreme gradient boosting algorithm was also used to evaluate whether PACE was associated with 1-year mortality. After excluding 1203 patients with 30-day mortality, 202 584 patients were divided into 7994 (3.9%) patients with PACE and 194 590 (96.1%) without PACE. After an adjustment, the mortality was higher in the PACE group (2.1% versus 7.7%; hazard ratio [HR], 1.90; 95% CI, 1.74-2.09; P<0.001). Results were similar for 7839 pairs of propensity-score-matched patients (4.9% versus 7.9%; HR, 1.64; 95% CI, 1.44-1.87; P<0.001). PACE was significantly associated with mortality in the extreme gradient boostingmodel. Conclusions PACE as a composite outcome was associated with 1-year mortality. Further studies are needed for PACE to be accepted as an end point in clinical studies of noncardiac surgery.

Renin-Angiotensin-Aldosterone System Inhibitors and Risk of Cancer: A Population-Based Cohort Study Using a Common Data Model.

Studies have reported conflicting results on the association between the use of renin-angiotensin-aldosterone system (RAAS) inhibitors and cancer development. We compared the incidence of cancer between patients using RAAS inhibitors and other antihypertensive drugs. This retrospective observational cohort study used data from seven hospitals in Korea that were converted for use in the Observational Medical Outcomes Partnership Common Data Model. A total of 166,071 patients on antihypertensive therapy across the databases of the seven hospitals were divided into two groups according to the use of RAAS inhibitors. The primary outcome was the occurrence of cancer. A total of 166,071 patients across the databases of the seven hospitals was included in the final analysis; 26,650 (16%) were in the RAAS inhibitors group and 139,421 (84%) in the other antihypertensive drugs group. The meta-analysis of the whole cohort showed a lower incidence of cancer occurrence in the RAAS inhibitor group (9.90 vs. 13.28 per 1000 person years; HR, 0.81; 95% confidence interval [CI], 0.75-0.88). After propensity-score matching, the RAAS inhibitor group consistently showed a lower incidence of cancer (9.90 vs. 13.28 per 1000 person years; HR, 0.86; 95% CI, 0.81-0.91). The patients using RAAS inhibitors showed a lower incidence of cancer compared with those using other antihypertensive drugs. These findings support the association between the use of RAAS inhibitors and cancer occurrence.

Association between cardiologist evaluation and mortality in myocardial injury after non-cardiac surgery.

OBJECTIVE: Myocardial injury after non-cardiac surgery (MINS) is strongly associated with mortality, but few studies assessed treatment strategies. This study aimed to identify whether evaluation by cardiologists could reduce mortality in MINS patients. METHODS: From a single-centre retrospective cohort, we enrolled a total of 5633 adult patients diagnosed with MINS between January 2010 and June 2019. The patients were divided into two groups based on evaluation by cardiologist, which was defined as a cardiology consultation or transfer to the cardiology department. For the outcome, 30-day mortality was compared in crude and propensity-score matched populations. RESULTS: Of a total of 5633 patients, 2120 (37.6%) were evaluated by cardiologists and 3513 (62.4%) were not. Mortality during the first 30 days after surgery was significantly lower in MINS patients who were evaluated by cardiologists compared with those who were not (5.8% vs 8.3%; HR, 0.64; 95% CI 0.51 to 0.80; p<0.001 for all-cause mortality and 1.6% vs 2.0; HR 0.62; 95% CI 0.40 to 0.96; p=0.03 for cardiovascular mortality). The propensity score matched analysis showed similar results (5.6% vs 8.6%; HR 0.64; 95% CI 0.50 to 0.81; p<0.001 for all-cause mortality and 1.3% vs 2.2%; HR 0.58; 95% CI 0.35 to 0.95; p=0.03 for cardiovascular mortality). CONCLUSIONS: Cardiologist evaluation was associated with lower mortality in patients diagnosed with MINS. Further studies are needed to identify effective treatment strategies for MINS. TRIAL REGISTRATION NUMBER: KCT0004244.

Prediction model for postoperative atrial fibrillation in non-cardiac surgery using machine learning.

Some patients with postoperative atrial fibrillation (POAF) after non-cardiac surgery need treatment, and a predictive model for these patients is clinically useful. Here, we developed a predictive model for POAF in non-cardiac surgery based on machine learning techniques. In a total of 201,864 patients who underwent non-cardiac surgery between January 2011 and June 2019 at our institution, 5,725 (2.8%) were treated for POAF. We used machine learning with an extreme gradient boosting algorithm to evaluate the effects of variables on POAF. Using the top five variables from this algorithm, we generated a predictive model for POAF and conducted an external validation. The top five variables selected for the POAF model were age, lung operation, operation duration, history of coronary artery disease, and hypertension. The optimal threshold of probability in this model was estimated to be 0.1, and the area under the receiver operating characteristic (AUROC) curve was 0.80 with a 95% confidence interval of 0.78-0.81. Accuracy of the model using the estimated threshold was 0.95, with sensitivity and specificity values of 0.28 and 0.97, respectively. In an external validation, the AUROC was 0.80 (0.78-0.81). The working predictive model for POAF requiring treatment in non-cardiac surgery based on machine learning techniques is provided online (https://sjshin.shinyapps.io/afib_predictor_0913/). The model needs further verification among other populations.