Derivation and performance of an end-of-life practice score aimed at interpreting worldwide treatment-limiting decisions in the critically ill

Abstract

Background: Limitations of life-sustaining interventions in intensive care units (ICUs) exhibit substantial changes over time, and large, contemporary variation across world regions. We sought to determine whether a weighted end-of-life practice score can explain a large, contemporary, worldwide variation in limitation decisions.

Methods: The 2015-2016 (Ethicus-2) vs. 1999-2000 (Ethicus-1) comparison study was a two-period, prospective observational study assessing the frequency of limitation decisions in 4952 patients from 22 European ICUs. The worldwide Ethicus-2 study was a single-period prospective observational study assessing the frequency of limitation decisions in 12,200 patients from 199 ICUs situated in 8 world regions. Binary end-of-life practice variable data (1 = presence; 0 = absence) were collected post hoc (comparison study, 22/22 ICUs, n = 4592; worldwide study, 186/199 ICUs, n = 11,574) for family meetings, daily deliberation for appropriate level of care, end-of-life discussions during weekly meetings, written triggers for limitations, written ICU end-of-life guidelines and protocols, palliative care and ethics consultations, ICU-staff taking communication or bioethics courses, and national end-of-life guidelines and legislation. Regarding the comparison study, generalized estimating equations (GEE) analysis was used to determine associations between the 12 end-of-life practice variables and treatment limitations. The weighted end-of-life practice score was then calculated using GEE-derived coefficients of the end-of-life practice variables. Subsequently, the weighted end-of-life practice score was validated in GEE analysis using the worldwide study dataset.

Results: In comparison study GEE analyses, end-of-life discussions during weekly meetings [odds ratio (OR) 0.55, 95% confidence interval (CI) 0.30-0.99], end-of-life guidelines [OR 0.52, (0.31-0.87)] and protocols [OR 15.08, (3.88-58.59)], palliative care consultations [OR 2.63, (1.23-5.60)] and end-of-life legislation [OR 3.24, 1.60-6.55)] were significantly associated with limitation decisions (all P < 0.05). In worldwide GEE analyses, the weighted end-of-life practice score was significantly associated with limitation decisions [OR 1.12 (1.03-1.22); P = 0.008].

Conclusions: Comparison study-derived, weighted end-of-life practice score partly explained the worldwide study's variation in treatment limitations. The most important components of the weighted end-of-life practice score were ICU end-of-life protocols, palliative care consultations, and country end-of-life legislation.

Keywords: End-of-life care; End-of-life practice score; Intensive care unit; Life-sustaining therapy; Medical ethics; Palliative care; ROC analysis.

Fig. 1

Flowchart of the employed analytic methodology. ICU, intensive care unit; GEE, generalized estimating equations; EPV, end-of-life practice variable; ROC receiver operating characteristic, EPS end-of-life practice score, CPR cardiopulmonary resuscitation. *The weighted EPS was determined by first multiplying the comparison study’s [4] GEE-derived EPV coefficients by the 0 or 1 response grades of the 12 EPVs from the worldwide dataset [5], and then by summing up the aforementioned products. ^†The EPS rescaling formula is presented in Table 1. ^‡The original, unweighted EPS was calculated as the sum of the 0 or 1 response grades of the 12 EPVs from the worldwide dataset [5]; author consensus definitions of the EPVs are provided in Table 1

Fig. 2

Receiver operating characteristic curve based on the comparison study’s [4] generalized estimating equations model

Fig. 3

ROC curves of the 4 generalized estimating equations models of the worldwide study [5]. ROC receiver operating characteristic, EPS end-of-life practice score, EPV end-of-life practice variable, AUC area under the curve, CI confidence interval. A: Model with weighted and rescaled EPS (worldwide model 1); B: Model with EPVs (worldwide model 2); C: Reference model without EPVs or EPS (worldwide model 3); D: Model with original, unweighted EPS (worldwide model 4)