Session Time: 2:30pm-4:00pm
Presentation Time: 2:30pm-2:42pm
Location: Room 208
*Purpose: Donation after circulatory death (DCD) is most often a controlled procedure after planned withdrawal of life-sustaining treatment (WLST) at an intensive care unit. In The Netherlands, in more than 20% of the DCD patients the agonal phase persists too long after WLST and the donation procedure is cancelled. Not-effectuated donations have deep impact on grieving families and are labor-intensive for transplantation teams. Therefore, identification of patients who will die within 60 min after WLST is necessary. The aim of this study was to externally validate two previously published prediction models of Rabinstein et al.,
*Methods: This multicenter retrospective study analyzed all patients who underwent WLST from 2010-2015 in six intensive care units in The Netherlands. The first prediction model used a simple scoring system (DCD N-score), in which points are attributed to: absence of cough reflex, absence of cornea reflex, absent extensor or motor reflex to pain and an oxygenation index (OI) above 3.0. The second model, the linear prediction model (LPDCD) yielded a percentage chance of the patient dying within 60 min after WLST and used OI as categorical variable. We determined discrimination (area under the receiver-operator curve, ROC) and calibration (Hosmer-Lemeshow (HL) test) of these models to predict the chance of dying within 60 min after WLST.
*Results: This study included 394 patients, of whom 284 (72%) died within 60 min after WLST. ROC analyses of the DCD-N score showed good discrimination for death within 60 min (Area under the curve (AUC) 0.81; 95% CI 0.74-0.88). The LPDCD model showed good discrimination for death within 60 min (AUC 0.78; 95% CI 0.70-0.85). Calibration of the LPDCD model showed that the model under and over predicted the probability of death (HL-test p=0.02).
*Conclusions: In conclusion, validation of the DCD-N score and the LPDCD model showed both good discrimination, but poor calibration for the prediction of the probability of death within 60 min. Construction of a new prediction model on a large data set is needed to obtain better calibration.
To cite this abstract in AMA style:Nijhoff MF, Pol RA, Volbeda M, Kotsopoulos AM, Sonneveld JP, Dongen Jvan, Abdo WF, Silderhuis V, Moers C. External Validation of the DCD-N Score and a Linear Prediction Model to Identify Potential Candidates for Donation after Circulatory Death: A Nation-Wide Cohort Study [abstract]. Am J Transplant. 2019; 19 (suppl 3). https://atcmeetingabstracts.com/abstract/external-validation-of-the-dcd-n-score-and-a-linear-prediction-model-to-identify-potential-candidates-for-donation-after-circulatory-death-a-nation-wide-cohort-study/. Accessed April 18, 2021.
« Back to 2019 American Transplant Congress