Improving the predictability of time to death in controlled donation after circulatory death lung donors

Shuji Okahara, Gregory I. Snell, Mark McDonald, Rohit D’Costa, Helen Opdam, David V. Pilcher, Bronwyn Levvey

Research output: Contribution to journalArticlepeer-review

Abstract

Although the use of donation after circulatory death (DCD) donors has increased lung transplant activity, 25–40% of intended DCD donors do not convert to actual donation because of no progression to asystole in the required time frame after withdrawal of cardiorespiratory support (WCRS). No studies have specifically focussed on DCD lung donor progression. This retrospective study reviewed intended DCD lung donors to make a prediction model of the likelihood of progression to death using logistic regression and classification and regression tree (CART). Between 2014 and 2018, 159 of 334 referred DCD donors were accepted, with 100 progressing to transplant, while 59 (37%) did not progress. In logistic regression, a length of ICU stay ≤ 5 days, severe infra-tentorial brain damage on imaging and use of vasopressin were related with the progression to actual donation. CART modelling of the likelihood of death within 90-minute post-WCRS provided prediction with a sensitivity of 1.00 and positive predictive value of 0.56 in the validation data set. In the nonprogressed DCD group, 26 died within 6 h post-WCRS. Referral received early after ICU admission, with nonspontaneous ventilatory mode, deep coma and severe infra-tentorial damage were relevant predictors. The CART model is useful to exclude DCD donor candidates with low probability of progression.

Original languageEnglish
Pages (from-to)906-915
Number of pages10
JournalTransplant International
Volume34
Issue number5
DOIs
Publication statusPublished - May 2021
Externally publishedYes

Keywords

  • donation after circulatory death
  • lung transplant
  • organ donation
  • progression within time frame

ASJC Scopus subject areas

  • Transplantation

Fingerprint

Dive into the research topics of 'Improving the predictability of time to death in controlled donation after circulatory death lung donors'. Together they form a unique fingerprint.

Cite this