Red blood cell transfusion and outcomes in acute pulmonary embolism

Prediction of in-hospital adverse clinical outcomes in patients with pulmonary thromboembolism, machine learning based models

Background

Pulmonary thromboembolism (PE) is the third leading cause of cardiovascular events. The conventional modeling methods and severity risk scores lack multiple laboratories, paraclinical and imaging data. Data science and machine learning (ML) based prediction models may help better predict outcomes.

Materials and methods

In this retrospective registry-based design, all consecutive hospitalized patients diagnosed with pulmonary thromboembolism (based on pulmonary CT angiography) from 2011 to 2019 were recruited. ML based algorithms [Gradient Boosting (GB) and Deep Learning (DL)] were applied and compared with logistic regression (LR) to predict hemodynamic instability and/or all-cause mortality.

Results

A total number of 1,017 patients were finally enrolled in the study, including 465 women and 552 men. Overall incidence of study main endpoint was 9.6%, (7.2% in men and 12.4% in women; p-value = 0.05). The overall performance of the GB model is better than the other two models (AUC: 0.94 for GB vs. 0.88 and 0.90 for DL and LR models respectively). Based on GB model, lower O2 saturation and right ventricle dilation and dysfunction were among the strongest adverse event predictors.

Conclusion

ML-based models have notable prediction ability in PE patients. These algorithms may help physicians to detect high-risk patients earlier and take appropriate preventive measures.

Platelet transfusion in adults: An update

Many patients worldwide receive platelet components (PCs) through the transfusion of diverse types of blood components. PC transfusions are essential for the treatment of central thrombocytopenia of diverse causes, and such treatment is beneficial in patients at risk of severe bleeding. PC transfusions account for almost 10% of all the blood components supplied by blood services, but they are associated with about 3.25 times as many severe reactions (attributable to transfusion) than red blood cell transfusions after stringent in-process leukoreduction to less than 10⁶ residual cells per blood component. PCs are not homogeneous, due to the considerable differences between donors. Furthermore, the modes of PC collection and preparation, the safety precautions taken to limit either the most common (allergic-type reactions and febrile non-hemolytic reactions) or the most severe (bacterial contamination, pulmonary lesions) adverse reactions, and storage and conservation methods can all result in so-called PC "storage lesions". Some storage lesions affect PC quality, with implications for patient outcome. Good transfusion practices should result in higher levels of platelet recovery and efficacy, and lower complication rates. These practices include a matching of tissue ABH antigens whenever possible, and of platelet HLA (and, to a lesser extent, HPA) antigens in immunization situations. This review provides an overview of all the available information relating to platelet transfusion, from donor and donation to bedside transfusion, and considers the impact of the measures applied to increase transfusion efficacy while improving safety and preventing transfusion inefficacy and refractoriness. It also considers alternatives to platelet component (PC) transfusion.

Outcome Prediction in Critically-Ill Patients with Venous Thromboembolism and/or Cancer Using Machine Learning Algorithms: External Validation and Comparison with Scoring Systems

Intensive care unit (ICU) patients with venous thromboembolism (VTE) and/or cancer suffer from high mortality rates. Mortality prediction in the ICU has been a major medical challenge for which several scoring systems exist but lack in specificity. This study focuses on two target groups, namely patients with thrombosis or cancer. The main goal is to develop and validate interpretable machine learning (ML) models to predict early and late mortality, while exploiting all available data stored in the medical record. To this end, retrospective data from two freely accessible databases, MIMIC-III and eICU, were used. Well-established ML algorithms were implemented utilizing automated and purposely built ML frameworks for addressing class imbalance. Prediction of early mortality showed excellent performance in both disease categories, in terms of the area under the receiver operating characteristic curve (AUC–ROC): VTE-MIMIC-III 0.93, eICU 0.87, cancer-MIMIC-III 0.94. On the other hand, late mortality prediction showed lower performance, i.e., AUC–ROC: VTE 0.82, cancer 0.74–0.88. The predictive model of early mortality developed from 1651 VTE patients (MIMIC-III) ended up with a signature of 35 features and was externally validated in 2659 patients from the eICU dataset. Our model outperformed traditional scoring systems in predicting early as well as late mortality. Novel biomarkers, such as red cell distribution width, were identified.

Impact of persistent anaemia on mortality in patients hospitalised with acute pulmonary embolism: an Australian retrospective observational study

OBJECTIVES

Anaemia is associated with increased mortality in acute pulmonary embolism (PE) patients. However, prior studies have not examined the prognostic impact of trends in plasma haemoglobin during admission. This study investigates the impact of changes in haemoglobin level on mortality during hospital stay in acute PE.

STUDY DESIGN

A retrospective observational study.

SETTING

Tertiary-referral centre in Australia.

PARTICIPANTS

Consecutive patients from 2000 to 2012 admitted with confirmed acute PE were identified from a dedicated PE database. Haemoglobin levels on days 1, 3-4, 5-6 and 7 of admission were retrieved. Patients without both baseline haemoglobin and subsequent haemoglobin levels were excluded (n=327), leaving 1099 patients as the study cohort. Anaemia was defined as haemoglobin <130 g/L for men and <120 g/L for women. There were 576 patients without anaemia throughout admission, 65 with transient anaemia (anaemic on day 1, but subsequently normalised during admission), 122 with acquired anaemia (normal on day 1 but developed anaemia during admission) and 336 with persistent anaemia. A total of 71 patients received blood transfusion during admission.

MAIN OUTCOME MEASURE

6-month mortality was tracked from a state-wide death database and analysed using multivariable modelling.

RESULTS

After adjusting for transfusion, patietns with persistent anaemia had a significantly increased 6-month mortality risk (adjusted HR 1.97, 95% CI 1.26 to 3.09, p=0.003) compared with patients without anaemia. There was no difference in mortality between patients with transient or acquired anaemia and patients without anaemia.

CONCLUSION

Among patients who had anaemia during their admission for acute PE, only the subgroup with persistent anaemia demonstrated worse outcomes.