Establishment of a risk prediction model for prolonged mechanical ventilation after lung transplantation: a retrospective cohort study

Abstract

Background Prolonged mechanical ventilation (PMV), mostly defined as mechanical ventilation > 72 h after lung transplantation with or without tracheostomy, is associated with increased mortality. Nevertheless, the predictive factors of PMV after lung transplant remain unclear. The present study aimed to develop a novel scoring system to identify PMV after lung transplantation. Methods A total of 141 patients who underwent lung transplantation were investigated in this study. The patients were divided into PMV and non-prolonged ventilation (NPMV) groups. Univariate and multivariate logistic regression analyses were performed to assess factors associated with PMV. A risk nomogram was then established based on the multivariate analysis, and model performance was further examined regarding its calibration, discrimination, and clinical usefulness. Results Eight factors were finally identified to be significantly associated with PMV by the multivariate analysis and therefore were included as risk factors in the nomogram as follows: the body mass index (BMI, P = 0.036); primary diagnosis as idiopathic pulmonary fibrosis (IPF, P = 0.038); pulmonary hypertension (PAH, P = 0.034); primary graft dysfunction grading (PGD, P = 0.011) at T0; cold ischemia time (CIT P = 0.012); and three ventilation parameters (peak inspiratory pressure [PIP, P < 0.001], dynamic compliance [Cdyn, P = 0.001], and P/F ratio [P = 0.015]) at T0. The nomogram exhibited superior discrimination ability with an area under the curve of 0.895. Furthermore, both calibration curve and decision-curve analysis indicated satisfactory performance. Conclusion A novel nomogram to predict individual risk of receiving PMV for patients after lung transplantation was established, which may guide preventative measures for tackling this adverse event. Graphic Abstract

Pancreas Rejection in the Artificial Intelligence Era: New Tool for Signal Patients at Risk

“…As the Naive Bayes model provides a good AUC and CA, we then constructed a nomogram plot to predict pancreas rejection in SPK transplantation, as shown in Figure 7, based on clinical and laboratory features. This nomogram-based strategy has been used in other types of transplants, e.g., kidney [22], lung [23], and heart [24], as a means to predict transplant-related events. In our model the 10 top features identified using information gain were incorporated, and our endpoint was the ability to identify or generate risk stratification for pancreas rejection in SPK transplantation, and to our knowledge this is one of the first attempts of model-based risk stratification for pancreas rejection.…”

A novel nomogram for predicting prolonged mechanical ventilation in lung transplantation patients using extracorporeal membrane oxygenation

“…The occurrence of PMV in in-patients using ECMO as a bridge to LTx was 21.8% in the present study. Gao et al reported that the incidence of PMV (defined as > 72 h) was 31.9%, excluding ECMO-supportive patients 10 . Since PMV is not clearly defined in the literature, the PMV in this study was defined according to the previous large-scale retrospective study, where the mean ventilation time of LTx patients was 5 days 11 .…”

Long-term survival and development of chronic graft dysfunction after donor lung perfusion repair in isolated lungs-a meta-analysis

“…EVLP lung patients require a ventilator for a long time, spending more time in the ICU. In addition to lengthy hospital and ICU stays, long-term mechanical ventilation (PMV) was linked to worse outcomes after LTx [47]. It was reported that 14% of 690 LTx patients who participated in a single-centre study between 2005 and 2012 suffered long-term mechanical ventilation for more than 21 days [48], where the patient outcomes who lived to be discharged were unaffected.…”