Predicting In-Hospital Mortality in Patients Undergoing Percutaneous Coronary Intervention

Castro‐Dominguez, Yulanka; Wang, Yongfei; Minges, Karl E.; McNamara, Robert L.; Spertus, John A.; Dehmer, Gregory J.; Messenger, John C.; Lavin, Kimberly; Anderson, Cornelia; Blankinship, Kristina; Mercado, Nestor; Clary, Julie M.; Osborne, Anwar; Curtis, Jeptha P.; Cavender, Matthew A.

doi:10.1016/j.jacc.2021.04.067

Cited by 46 publications

(28 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For example, pre/post-implementation studies have shown that integrating a stratification by the NCDR-CathPCI bleeding model and using a bleeding avoidance strategy can reduce periprocedural bleeding 32 . Further, NCDR-risk scores have been updated when concerns are raised 5 , 33 , 34 . Considering the above, it would be difficult to justify using ML-based models instead of NCDR-CathPCI risk scores within our cohort.…”

Section: Discussionmentioning

confidence: 99%

Machine learning models for prediction of adverse events after percutaneous coronary intervention

Niimi

Sawano

Ikemura

et al. 2022

Sci Rep

View full text Add to dashboard Cite

An accurate prediction of major adverse events after percutaneous coronary intervention (PCI) improves clinical decisions and specific interventions. To determine whether machine learning (ML) techniques predict peri-PCI adverse events [acute kidney injury (AKI), bleeding, and in-hospital mortality] with better discrimination or calibration than the National Cardiovascular Data Registry (NCDR-CathPCI) risk scores, we developed logistic regression and gradient descent boosting (XGBoost) models for each outcome using data from a prospective, all-comer, multicenter registry that enrolled consecutive coronary artery disease patients undergoing PCI in Japan between 2008 and 2020. The NCDR-CathPCI risk scores demonstrated good discrimination for each outcome (C-statistics of 0.82, 0.76, and 0.95 for AKI, bleeding, and in-hospital mortality) with considerable calibration. Compared with the NCDR-CathPCI risk scores, the XGBoost models modestly improved discrimination for AKI and bleeding (C-statistics of 0.84 in AKI, and 0.79 in bleeding) but not for in-hospital mortality (C-statistics of 0.96). The calibration plot demonstrated that the XGBoost model overestimated the risk for in-hospital mortality in low-risk patients. All of the original NCDR-CathPCI risk scores for adverse periprocedural events showed adequate discrimination and calibration within our cohort. When using the ML-based technique, however, the improvement in the overall risk prediction was minimal.

show abstract

Section: Discussionmentioning

confidence: 99%

Machine learning models for prediction of adverse events after percutaneous coronary intervention

Niimi

Sawano

Ikemura

et al. 2022

Sci Rep

View full text Add to dashboard Cite

show abstract

“…A study by Brennan et al found that cardiogenic shock and procedure urgency were the most powerful predictors of inpatient mortality, whereas the presence of CTO was among the most significant angiographic predictors in the high-risk PCI subset [ 12 ]. The latest iteration of the risk score based on the NCDR, namely, the CathPCI Registry, reported that procedural urgency, cardiovascular instability, and the level of consciousness after cardiac arrest were the most predictive of in-hospital mortality in the pre-catheterisation model, de-emphasising the importance of coronary anatomy or procedural factors [ 13 ]. Two risk stratification models for in-hospital death in patients undergoing PCI based on the Japanese-PCI (J-PCI) registry were developed, namely, a full and a pre-catheterisation one [ 14 ].…”

Section: Discussionmentioning

confidence: 99%

“…Other angiographic features such as the culprit lesion type, bifurcation disease or vessel tortuosity, all indisputably increase the difficulty of a procedure and, thus, the risk of dissection; however, they are not harvestable in our databases, and clinical factors such as myocardial infarction or cardiogenic shock can be considered surrogates for the procedure’s difficulty. Moreover, many of the models described above were based exclusively on preprocedural factors or developed two types of risk scores, namely, a pre-catheterisation and a full one [ 7 , 13 , 15 ].…”

Section: Discussionmentioning

confidence: 99%

Forewarned Is Forearmed: Machine Learning Algorithms for the Prediction of Catheter-Induced Coronary and Aortic Injuries

Klaudel¹,

Klaudel

Glaza³

et al. 2022

IJERPH

View full text Add to dashboard Cite

Catheter-induced dissections (CID) of coronary arteries and/or the aorta are among the most dangerous complications of percutaneous coronary procedures, yet the data on their risk factors are anecdotal. Logistic regression and five more advanced machine learning techniques were applied to determine the most significant predictors of dissection. Model performance comparison and feature importance ranking were evaluated. We identified 124 cases of CID in electronic databases containing 84,223 records of diagnostic and interventional coronary procedures from the years 2000–2022. Based on the f1-score, Extreme Gradient Boosting (XGBoost) was found to have the optimal balance between positive predictive value (precision) and sensitivity (recall). As by the XGBoost, the strongest predictors were the use of a guiding catheter (angioplasty), small/stenotic ostium, radial access, hypertension, acute myocardial infarction, prior angioplasty, female gender, chronic renal failure, atypical coronary origin, and chronic obstructive pulmonary disease. Risk prediction can be bolstered with machine learning algorithms and provide valuable clinical decision support. Based on the proposed model, a profile of ‘a perfect dissection candidate’ can be defined. In patients with ‘a clustering’ of dissection predictors, a less aggressive catheter and/or modification of the access site should be considered.

show abstract

“…Risk-adjusted comparisons of in-hospital outcomes (death, AKI, transfusion, and major bleeding) were performed between the pre-pandemic and pandemic periods, using logistic regression models adjusting for baseline patient predicted risk, estimated from a recently updated version of our random forest model [ 27 ]. In addition, we performed a sensitivity analysis using patient risk estimates based on the recently published NCDR CathPCI Registry mortality risk model [ 28 ]. The results of these analyses are presented as odds ratios (OR) and 95% confidence intervals (CI).…”

Section: Methodsmentioning

confidence: 99%

Trends and outcomes of percutaneous coronary intervention during the COVID-19 pandemic in Michigan

et al. 2022

View full text Add to dashboard Cite

Background The COVID-19 pandemic has severely impacted healthcare delivery and patient outcomes globally. Aims We aimed to evaluate the influence of the COVID-19 pandemic on the temporal trends and outcomes of patients undergoing percutaneous coronary intervention (PCI) in Michigan. Methods We compared all patients undergoing PCI in the BMC2 Registry between March and December 2020 (“pandemic cohort”) with those undergoing PCI between March and December 2019 (“pre-pandemic cohort”). A risk-adjusted analysis of in-hospital outcomes was performed between the pre-pandemic and pandemic cohort. A subgroup analysis was performed comparing COVID-19 positive vs. negative patients during the pandemic. Results There was a 15.2% reduction in overall PCI volume from the pre-pandemic (n = 25,737) to the pandemic cohort (n = 21,822), which was more pronounced for stable angina and non-ST-elevation acute coronary syndromes (ACS) presentations, and between February and May 2020. Patients in the two cohorts had similar clinical and procedural characteristics. Monthly mortality rates for primary PCI were generally higher in the pandemic period. There were no significant system delays in care between the cohorts. Risk-adjusted mortality was higher in the pandemic cohort (aOR 1.26, 95% CI 1.07–1.47, p = 0.005), a finding that was only partially explained by worse outcomes in COVID-19 patients and was more pronounced in subjects with ACS. During the pandemic, COVID-19 positive patients suffered higher risk-adjusted mortality (aOR 5.69, 95% CI 2.54–12.74, p<0.001) compared with COVID negative patients. Conclusions During the COVID-19 pandemic, we observed a reduction in PCI volumes and higher risk-adjusted mortality. COVID-19 positive patients experienced significantly worse outcomes.

show abstract

Predicting In-Hospital Mortality in Patients Undergoing Percutaneous Coronary Intervention

Cited by 46 publications

References 27 publications

Machine learning models for prediction of adverse events after percutaneous coronary intervention

Machine learning models for prediction of adverse events after percutaneous coronary intervention

Forewarned Is Forearmed: Machine Learning Algorithms for the Prediction of Catheter-Induced Coronary and Aortic Injuries

Trends and outcomes of percutaneous coronary intervention during the COVID-19 pandemic in Michigan

Contact Info

Product

Resources

About