Machine learning prediction of hospitalization costs for coronary artery bypass grafting operations.

BACKGROUND: With the steady rise in health care expenditures, the examination of factors that may influence the costs of care has garnered much attention. Although machine learning models have previously been applied in health economics, their application within cardiac surgery remains limited. We evaluated several machine learning algorithms to model hospitalization costs for coronary artery bypass grafting. METHODS: All adult hospitalizations for isolated coronary artery bypass grafting were identified in the 2016 to 2020 Nationwide Readmissions Database. Machine learning models were trained to predict expenditures and compared with traditional linear regression. Given the significance of postoperative length of stay, we additionally developed models excluding postoperative length of stay to uncover other drivers of costs. To facilitate comparison, machine learning classification models were also trained to predict patients in the highest decile of costs. Significant factors associated with high cost were identified using SHapley Additive exPlanations beeswarm plots. RESULTS: Among 444,740 hospitalizations included for analysis, the median cost of hospitalization in coronary artery bypass grafting patients was $43,103. eXtreme Gradient Boosting most accurately predicted hospitalization costs, with R2 = 0.519 over the validation set. The top predictive features in the eXtreme Gradient Boosting model included elective procedure status, prolonged mechanical ventilation, new-onset respiratory failure or myocardial infarction, and postoperative length of stay. After removing postoperative length of stay, eXtreme Gradient Boosting remained the most accurate model (R2 = 0.38). Prolonged ventilation, respiratory failure, and elective status remained important predictive parameters. CONCLUSION: Machine learning models appear to accurately model total hospitalization costs for coronary artery bypass grafting. Future work is warranted to uncover other drivers of costs and improve the value of care in cardiac surgery.

Machine learning based predictive modeling of readmissions following extracorporeal membrane oxygenation hospitalizations.

Background: Despite increasing utilization and survival benefit over the last decade, extracorporeal membrane oxygenation (ECMO) remains resource-intensive with significant complications and rehospitalization risk. We thus utilized machine learning (ML) to develop prediction models for 90-day nonelective readmission following ECMO. Methods: All adult patients receiving ECMO who survived index hospitalization were tabulated from the 2016-2020 Nationwide Readmissions Database. Extreme Gradient Boosting (XGBoost) models were developed to identify features associated with readmission following ECMO. Area under the receiver operating characteristic (AUROC), mean Average Precision (mAP), and the Brier score were calculated to estimate model performance relative to logistic regression (LR). Shapley Additive Explanation summary (SHAP) plots evaluated the relative impact of each factor on the model. An additional sensitivity analysis solely included patient comorbidities and indication for ECMO as potential model covariates. Results: Of ∼22,947 patients, 4495 (19.6 %) were readmitted nonelectively within 90 days. The XGBoost model exhibited superior discrimination (AUROC 0.64 vs 0.49), classification accuracy (mAP 0.30 vs 0.20) and calibration (Brier score 0.154 vs 0.165, all P < 0.001) in predicting readmission compared to LR. SHAP plots identified duration of index hospitalization, undergoing heart/lung transplantation, and Medicare insurance to be associated with increased odds of readmission. Upon sub-analysis, XGBoost demonstrated superior disclination compared to LR (AUROC 0.61 vs 0.60, P < 0.05). Chronic liver disease and frailty were linked with increased odds of nonelective readmission. Conclusions: ML outperformed LR in predicting readmission following ECMO. Future work is needed to identify other factors linked with readmission and further optimize post-ECMO care among this cohort.

Highly Excretable Gold Supraclusters for Translatable In Vivo Raman Imaging of Tumors.

Raman spectroscopy provides excellent specificity for in vivo preclinical imaging through a readout of fingerprint-like spectra. To achieve sufficient sensitivity for in vivo Raman imaging, metallic gold nanoparticles larger than 10 nm were employed to amplify Raman signals via surface-enhanced Raman scattering (SERS). However, the inability to excrete such large gold nanoparticles has restricted the translation of Raman imaging. Here we present Raman-active metallic gold supraclusters that are biodegradable and excretable as nanoclusters. Although the small size of the gold nanocluster building blocks compromises the electromagnetic field enhancement effect, the supraclusters exhibit bright and prominent Raman scattering comparable to that of large gold nanoparticle-based SERS nanotags due to high loading of NIR-resonant Raman dyes and much suppressed fluorescence background by metallic supraclusters. The bright Raman scattering of the supraclusters was pH-responsive, and we successfully performed in vivo Raman imaging of acidic tumors in mice. Furthermore, in contrast to large gold nanoparticles that remain in the liver and spleen over 4 months, the supraclusters dissociated into small nanoclusters, and 73% of the administered dose to mice was excreted during the same period. The highly excretable Raman supraclusters demonstrated here offer great potential for clinical applications of in vivo Raman imaging.

Renally Excretable and Size-Tunable Silver Sulfide Nanoparticles for Dual-Energy Mammography or Computed Tomography.

Significant effort has been focused on developing renally-clearable nanoparticle agents since efficient renal clearance is important for eventual clinical translation. Silver sulfide nanoparticles (Ag2S-NP) have recently been identified as contrast agents for dual energy mammography, computed tomography (CT) and fluorescence imaging and probes for drug delivery and photothermal therapy with good biocompatibility. However, most Ag2S-NP reported to date are not renally excretable and are observed in vivo to accumulate and remain in the reticuloendothelial system (RES) organs, i.e. liver and spleen, for a long time, which could negatively impact their likelihood for translation. Herein, we present renally-clearable, 3.1 nm Ag2S-NP with 85% of the injected dose (ID) being excreted within 24 hours of intravenous injection, which is amongst the best clearance of similarly sized nanoparticles reported thus far (mostly between 20-75% of ID). The urinary excretion and low RES accumulation of these nanoparticles in mice were indicated by in vivo CT imaging and biodistribution analysis. In summary, these ultrasmall Ag2S-NP can be effectively eliminated via urine and have high translational potential for various biomedical applications.

Estudio comparativo sobre la prevalencia de hipoplasia e hipocalcificación en la dentición temporal completa en niños prematuros y nacidos a término / Enamel defects in primary dentition of premature infants and infants born on term

La formación del esmalte en el humano comienza en el útero. Este proceso, que ocurre en diferentes etapas, puede verse afectado por cambios fisiológicos, sistémicos externos, o disturbios mayores, antes o durante el período de formación dental, manifestándose como hipoplasia e hipocalcificación. El propósito de este estudio fue comparar los defectos del esmalte en los dientes temporales de niños prematuros con las anomalías dentales presentes en niños nacidos a término. Se tomaron 33 niños para cada grupo, los cuales tenían dentición temporal completa. Se analizaron los dientes afectados, la superficie y el tercio en el cual se presentaba el defecto. Estos datos fueron registrados en historias clínicas individuales y luego en tablas comparativas entre los dos grupos en estudio. Se examinaron una sola vez por dos observadores clínicos. Para el análisis de los resultados se elaboraron tablas de distribución de frecuencias para cada variable de estudio, para determinar la existencia de diferencias estadísticamente significativas entre el grupo estudio y el grupo control, se aplicó una prueba de chi cuadrado con un nivel de significancia de P<0.05. La frecuencia de displasia del esmalte fue mayor en el grupo prematuro 76 por ciento comparado con el 24 por ciento en el grupo control con un Chi cuadrado de 17,5151 y P<0,05 que fue significativo estadísticamente. Sin embargo la diferencia entre hipoplasia no fue significativa entre los dos grupos mientras que la prevalencia de hipocalcificación fue significativamente mayor en el grupo prematuro.