This prospective study revealed a NDE incidence of 15% in patients discharged from the ICU after a prolonged stay, no matter what critical illness they survived. This incidence is consistent with the 10–23% incidence observed in cardiac arrest survivors [8,9,10,11,12], but higher than the studies reporting a 4–8% prevalence estimation among the general population [13, 14].

In univariate analyses, we observed that seven variables were associated with the emergence of NDE; however, only a higher frequency of dissociative symptoms and a greater spiritual and personal well-being were the strongest predictors for the recall of NDE using multivariate analysis (Fig. 1). It is then reasonable to hypothesize that a propensity to dissociative states and to spiritual beliefs and practices make people more likely to report NDEs when exposed to certain physiological conditions. This corroborates Greyson’s retrospective study [6] showing more (non-pathological) dissociative symptoms in NDE population. As suggested by Noyes and Slymen [15], dissociation would offer a less distressing “reality” to people facing with a potential danger. This is also consistent with the hypothesis suggesting that NDEs may have a specific biological benefit when facing life-threatening situations [16]. Spiritual well-being is an important coping resource as well [17].

Summary of the results of this 1-year prospective study. *p < 0.05; NDE = near-death experience; ICU = intensive care unit; DES = Dissociative Experience Scale; WHOQOL-SRPB = World Health Organization Quality of Life-Spirituality, Religiousness and Personal Beliefs; EQ-5D-3L = EuroQol five-dimensional questionnaire; PaO2 = partial pressure of arterial oxygen; PaCO2 = partial pressure of arterial carbon dioxide; SpO2 = peripheral oxygen saturation; OR = odds ratios; CI = confidence interval

Yet, it is not thought possible to explain NDEs only in terms of psychological processes. Absence of proof is not proof of absence. The current literature converges to say that several neurophysiological mechanisms may provoke the occurrence of NDEs [1, 5]. Obviously, it was impossible to objectively determine when exactly during the ICU stay were NDEs experienced and if NDEs were related to the initial event leading to ICU admission or to subsequent events during the ICU stay. This may explain that none of the recorded physiological parameters have been found as a risk factor for NDE.

One year later, the ICU-related NDE was not significantly associated with HRQoL measured. Similarly, the perception of a life-threatening situation was not impacted a posteriori by the previous occurrence of NDE. Finally, for all our patients, the Greyson total score did not change 1 year later.

NDEs are typically reported as transforming [18] and may be associated with negative emotions [19]. This is why we consider clinically meaningful to interview patients about any potential memory upon awakening.

Some limitations need to be acknowledged. First, the number of patients who experienced a NDE were limited. This could have limited the analysis of NDE risk factors, as well as making false negatives more likely. Yet, patients whose experience did not reach the validated cut-off score of 7/32 for a typical NDE were not categorized as NDE experiencers, and some patients may have denied or forgotten to have lived a NDE. Second, the large number of comparisons increases the probability of type 1 errors; however, the level of significance of individual comparisons for the questionnaires was well below the 0.05 threshold, thereby reducing the likelihood of false positives. Finally, delirium during the ICU stay was not routinely assessed. It is still unknown if delirium could be a risk factor for NDE.

In conclusion, we observed a NDE incidence of 15% and that cognitive and spiritual outweighed medical parameters as predictors of the emergence of NDE. Further studies are needed to confirm these findings in larger cohorts or in survivors of a shorter ICI stay.