Bhattacharyya, Anirban

Background: Impella 5.5® with Smart Assist is a minimally invasive Left Ventricular Assist Devices (LVAD) approved by the Food and Drug Administration (FDA) for treating ongoing cardiogenic shock for up to 14 days. The Impella® intends to reduce ventricular workload and provide the circulatory support necessary for myocardial recovery. Research Question: Compared to standard practice, does adding an extension piece to the purge tube side arm of the Impella® Device decrease the incidence of device failure and positively impact the health outcome of adult patients receiving Impella® support? Study Design and Methods: A retrospective chart review of ICU patients was done at a tertiary care center from August 2018 to August 2022 to assess the differences in patient outcomes related to Impella® Device utilization before and after the implementation of the extension piece to the purge tube sidearm. Among patients reviewed, a total of 20 were included in our review, with seven not having the purge tube side arm extension added, while 13 patients had the extension. Results: The two study groups had no significant difference in patient health outcomes. Additionally, there were no instances of device failure requiring explanation without the extension tubing. However, there were no cases of the purge cassette cracking with the addition of the extension tubing. Conclusion: The addition of extension tubing to the purge cassette of the Impella® Device did not impact patient health outcomes or the incidence of device failure. There was a complete reduction in the incidence of the purge cassette cracking, which could reduce the potential for infection or device failure over a long period of mechanical support. There is a need for long-term prospective studies to confirm the results.

Abstract Rationale Current guidelines recognize the utility of provocative maneuvers during right heart catheterization to aid the diagnosis of pulmonary hypertension. Few studies have compared the performance of different provocation maneuvers. Objectives To assess the hemodynamic correlation among three provocative maneuvers, including their effect on pulmonary hypertension classification. Methods This prospective trial was conducted between October 2016 and May 2018. Adult patients underwent three provocative maneuvers during right heart catheterization: passive leg raise (PLR), load-targeted supine bicycle exercise, and rapid crystalloid fluid infusion. Patients were classified as follows: no pulmonary hypertension, precapillary pulmonary hypertension, isolated postcapillary pulmonary hypertension, combined pre- and postcapillary pulmonary hypertension, and uncategorized pulmonary hypertension. We assessed the hemodynamic changes associated with each maneuver. We also assessed whether provocative maneuvers led to hemodynamic reclassification of the patient to either postcapillary pulmonary hypertension with provocation or exercise pulmonary hypertension. Results Eighty-five patients (mean age 62 ± 12 years, 53% women) were included. Correlation between exercise and fluid challenge was moderate to strong (0.49–0.82; P < 0.001) for changes in right atrial pressure, mean pulmonary arterial pressure, pulmonary arterial wedge pressure, and cardiac index from baseline. Correlation between PLR and exercise (0.4–0.65; P < 0.001) and between PLR and fluid challenge (0.45–0.6; P < 0.001) was moderate for changes in right atrial pressure, mean pulmonary arterial pressure, pulmonary arterial wedge pressure, pulmonary vascular resistance, and cardiac index. Hemodynamic correlation between other provocative maneuvers was poor. Depending on provocative maneuver and classification criteria, there was significant variation in the number of patients reclassified as having exercise pulmonary hypertension (3–50%) or postcapillary pulmonary hypertension with provocation (11–48%). Conclusions Hemodynamic determinations during exercise and fluid challenge showed moderate to strong hemodynamic correlation. Moderate hemodynamic correlation was seen between PLR and exercise or fluid challenge. Although some provocative maneuvers demonstrate good hemodynamic correlation, there is inconsistency when using these maneuvers to identify patients with postcapillary or exercise pulmonary hypertension.

Abstract Introduction Delirium occurrence is common and preventive strategies are resource intensive. Screening tools can prioritize patients at risk. Using machine learning, we can capture time and treatment effects that pose a challenge to delirium prediction. We aim to develop a delirium prediction model that can be used as a screening tool. Methods From the eICU Collaborative Research Database (eICU-CRD) and the Medical Information Mart for Intensive Care version III (MIMIC-III) database, patients with one or more Confusion Assessment Method-Intensive Care Unit (CAM-ICU) values and intensive care unit (ICU) length of stay greater than 24 h were included in our study. We validated our model using 21 quantitative clinical parameters and assessed performance across a range of observation and prediction windows, using different thresholds and applied interpretation techniques. We evaluate our models based on stratified repeated cross-validation using 3 algorithms, namely Logistic Regression, Random Forest, and Bidirectional Long Short-Term Memory (BiLSTM). BiLSTM represents an evolution from recurrent neural network-based Long Short-Term Memory, and with a backward input, preserves information from both past and future. Model performance is measured using Area Under Receiver Operating Characteristic, Area Under Precision Recall Curve, Recall, Precision (Positive Predictive Value), and Negative Predictive Value metrics. Results We evaluated our results on 16 546 patients (47% female) and 6294 patients (44% female) from eICU-CRD and MIMIC-III databases, respectively. Performance was best in BiLSTM models where, precision and recall changed from 37.52% (95% confidence interval [CI], 36.00%–39.05%) to 17.45 (95% CI, 15.83%–19.08%) and 86.1% (95% CI, 82.49%–89.71%) to 75.58% (95% CI, 68.33%–82.83%), respectively as prediction window increased from 12 to 96 h. After optimizing for higher recall, precision and recall changed from 26.96% (95% CI, 24.99%–28.94%) to 11.34% (95% CI, 10.71%–11.98%) and 93.73% (95% CI, 93.1%–94.37%) to 92.57% (95% CI, 88.19%–96.95%), respectively. Comparable results were obtained in the MIMIC-III cohort. Conclusions Our model performed comparably to contemporary models using fewer variables. Using techniques like sliding windows, modification of threshold to augment recall and feature ranking for interpretability, we addressed shortcomings of current models.