The effects of intensive care environment on postoperative nightmare

Shota Sonobe, Satoki Inoue, Masahiko Kawaguchi

We retrospectively investigated the incidence of postoperative nightmares and evaluated the impact of postoperative intensive care on the incidence of during subsequent hospital stay. To reduce the effect of selection bias, we compared the incidence of nightmares in propensity-matched pairs with postoperative management in ICUs or in surgical wards. This is a retrospective review of an institutional registry containing 21,606 anesthesia cases and was conducted with ethics board approval. Outcomes of surgical patients treated in ICUs and in postsurgical wards (ICU admission vs non ICU admission) were compared first for nightmares using the initial 12,508 patients. To avoid channeling bias, propensity score analysis was used to generate a set of matched cases (ICU admission) and controls (non ICU admission), yielding 642 matched patient pairs. The incidence rate of nightmares was compared as the primary outcome. Before adjusting patients' characteristics, ICU environment exposure increased the incidence of nightmares compared with non-ICU environment during subsequent hospital stay [ICU vs non-ICU: 101/718 (12.3 %) vs 1147/10,542 (9.81 %)]. The odds ratio (95 % CIs) for ICU was 1.29 (1.03-1.61) for nightmares (p = 0.022). After propensity score matching, however, an equal rate of nightmares occurred in the ICU environment exposure compared to the non-ICU environment [ICU vs non-ICU: 81/561 (12.6 %) vs 73/569 (11.4 %)]. The odds ratio and 95 % CIs for ICU were 1.13 (0.80-1.58) for nightmares (p = 0.54). The incidence of nightmares did not become more evident during subsequent hospital stay after ICU environment exposure.