Effects of early inhaled epoprostenol therapy on pulmonary artery pressure and blood loss during LVAD placement.

OBJECTIVE: Several strategies have been used to reduce the incidence of right ventricular failure after left ventricular assist device (LVAD) placement, including pulmonary vasodilation. The inhaled prostacyclin, epoprostenol, selectively dilates the pulmonary vasculature of ventilated areas of the lung, but also has been shown to inhibit platelet aggregation.(1) The authors evaluated the impact of early initiation of epoprostenol administration during LVAD placement on pulmonary artery pressures, use of vasoactive drugs, and blood loss. DESIGN: Retrospective data review. SETTING: Single center, university hospital. PARTICIPANTS: A total of 37 consecutive patients undergoing LVAD (HeartMate II) placement were included. INTERVENTIONS: In the first group of 23 patients (group 1), inhaled epoprostenol was not initiated until weaning from cardiopulmonary bypass (CPB). In a subsequent group of 14 patients (group 2), inhaled epoprostenol was started shortly after induction of anesthesia and continued throughout and post-CPB. MEASUREMENTS: Mean and systolic pulmonary artery pressures (mPAP, sPAP), vasoactive drugs, as well as hemodynamic parameters, blood loss, and use of blood products were recorded at the following time points: Baseline (BL), pre-CPB, post-CPB, and during postoperative days (POD) 0, 1, and 2. Data are presented as mean±SD or median [25%, 75%]. RESULTS: Groups did not differ in demographic characteristics and comorbidities. BL sPAP (41±13 v 46±15 mmHg; p = 0.051) and mPAP (32±8 v 34±8 mmHg; p = 0.483) values were not different between the groups. Systolic and mPAP in group 1 were significantly lower in the postoperative period compared with BL (sPAP on POD 0: 34±6 mmHg; p<0.001; mPAP on POD 0, 1, and 2: 24±4 mmHg, 25±4 mmHg, 27±6 mmHg; p<0.001-0.003)). In contrast, in group 2, sPAP as well as mPAP were significantly lower during weaning from CPB (sPAP: 37±8; p = 0.002; mPAP: 28±5 mmHg; p = 0.016) as well as in the postoperative period (sPAP on POD 0, 1 and 2: 34±7, 35±7, and 37±10 mmHg; p<0.001-0.004; mPAP on POD 0, 1, and 2: 24±4 mmHg, 25±5 mmHg, 27±6 mmHg; p<0.001-0.006). Blood loss on postoperative day 0 was significantly lower in group 1 (1646 mL [1137, 2300] v 2915 mL [2335, 6155]; p = 0.006). Epoprostenol was a significant predictor of blood loss in the regression model (p<0.001) but did not predict a change in sPAP. CONCLUSIONS: Inhaled prostacyclin reduces sPAP and mPAP in the postoperative period after LVAD placement regardless of the timing of initiation. Early initiation seems to reduce sPAP as well as mPAP more effectively during the weaning process from CPB. However, early initiation is associated with an increased blood loss in the immediate postoperative period. The concept of preventively "bathing" the lung in prostacyclin should be evaluated critically in a prospective fashion to adequately examine this question.

The association between brain injury, perioperative anesthetic exposure, and 12-month neurodevelopmental outcomes after neonatal cardiac surgery: a retrospective cohort study.

BACKGROUND: Adverse neurodevelopmental outcomes are observed in up to 50% of infants after complex cardiac surgery. We sought to determine the association of perioperative anesthetic exposure with neurodevelopmental outcomes at age 12 months in neonates undergoing complex cardiac surgery and to determine the effect of brain injury determined by magnetic resonance imaging (MRI). METHODS: Retrospective cohort study of neonates undergoing complex cardiac surgery who had preoperative and 7-day postoperative brain MRI and 12-month neurodevelopmental testing with Bayley Scales of Infant and Toddler Development, Third Edition (Bayley-III). Doses of volatile anesthetics (VAA), benzodiazepines, and opioids were determined during the first 12 months of life. RESULTS: From a database of 97 infants, 59 met inclusion criteria. Mean ± sd composite standard scores were as follows: cognitive = 102.1 ± 13.3, language = 87.8 ± 12.5, and motor = 89.6 ± 14.1. After forward stepwise multivariable analysis, new postoperative MRI injury (P = 0.039) and higher VAA exposure (P = 0.028) were associated with lower cognitive scores. ICU length of stay (independent of brain injury) was associated with lower performance on all categories of the Bayley-III (P < 0.02). CONCLUSIONS: After adjustment for multiple relevant covariates, we demonstrated an association between VAA exposure, brain injury, ICU length of stay, and lower neurodevelopmental outcome scores at 12 months of age. These findings support the need for further studies to identify potential modifiable factors in the perioperative care of neonates with CHD to improve neurodevelopmental outcomes.

Why do users override alerts? Utilizing large language model to summarize comments and optimize clinical decision support.

OBJECTIVES: To evaluate the capability of using generative artificial intelligence (AI) in summarizing alert comments and to determine if the AI-generated summary could be used to improve clinical decision support (CDS) alerts. MATERIALS AND METHODS: We extracted user comments to alerts generated from September 1, 2022 to September 1, 2023 at Vanderbilt University Medical Center. For a subset of 8 alerts, comment summaries were generated independently by 2 physicians and then separately by GPT-4. We surveyed 5 CDS experts to rate the human-generated and AI-generated summaries on a scale from 1 (strongly disagree) to 5 (strongly agree) for the 4 metrics: clarity, completeness, accuracy, and usefulness. RESULTS: Five CDS experts participated in the survey. A total of 16 human-generated summaries and 8 AI-generated summaries were assessed. Among the top 8 rated summaries, five were generated by GPT-4. AI-generated summaries demonstrated high levels of clarity, accuracy, and usefulness, similar to the human-generated summaries. Moreover, AI-generated summaries exhibited significantly higher completeness and usefulness compared to the human-generated summaries (AI: 3.4 ± 1.2, human: 2.7 ± 1.2, P = .001). CONCLUSION: End-user comments provide clinicians' immediate feedback to CDS alerts and can serve as a direct and valuable data resource for improving CDS delivery. Traditionally, these comments may not be considered in the CDS review process due to their unstructured nature, large volume, and the presence of redundant or irrelevant content. Our study demonstrates that GPT-4 is capable of distilling these comments into summaries characterized by high clarity, accuracy, and completeness. AI-generated summaries are equivalent and potentially better than human-generated summaries. These AI-generated summaries could provide CDS experts with a novel means of reviewing user comments to rapidly optimize CDS alerts both online and offline.