Factual memories of ICU: recall at two years post-discharge and comparison with delirium status during ICU admission – a multicentre cohort study

Abstract

Aims and objective. To examine the relationship between observed delirium in ICU and patients’ recall of factual events up to two years after discharge.

Background. People, the environment, and procedures are frequently cited memories of actual events encountered in ICU. These are often perceived as stressors to the patients and the presence of several such stressors has been associated with the development of reduced health‐related quality of life or post‐traumatic stress syndrome.

Design. Prospective cohort study using interview technique.

Method. The cohort was assembled from 152 patients who participated in a previously conducted multi‐centre study of delirium incidence in Australian ICUs. The interviews involved a mixture of closed‐ and open‐ended questions. Qualitative responses regarding factual memories were analysed using thematic analysis. A five‐point Likert scale with answers from ‘always’ to ‘never’ was used to ask about current experiences of dream, anxiety, sleep problems, fears, irritability and/or mood swings. Scoring ranged from 6 to 30 with a mid‐point value of 18 indicating a threshold value for the diagnosis of post‐traumatic stress syndrome. A P‐value of <0·05 was considered significant for all analyses.

Results. Forty‐one (40%) out of 103 potential participants consented to take part in the follow‐up interview; 18 patients (44%) had been delirious and 23 patients (56%) non‐delirious during the ICU admission. The non‐participants (n = 62) formed a control group to ensure a representative sample; 83% (n = 34) reported factual memories either with or without recall of dreaming. Factual memories were significantly less common (66% cf. 96%) in delirious patients (OR 0·09, 95%CI 0·01–0·85, p = 0·035). Five topics emerged from the thematic analysis: ‘procedures’, ‘staff’, ‘comfort’, ‘visitors’, and ‘events’. Based on the current experiences, five patients (12%, four non‐delirious and one delirious) scored ≥18 indicative of symptoms of post‐traumatic stress syndrome; this did not reach statistical significance. Memory of transfer out of ICU was less frequent among the delirious patients (56%, n = 10) than among the non‐delirious patients (87%, n = 20) (p = 0·036).

Conclusion. Most patients have factual memories of their ICU stay. However, delirious patients had significantly less factual recall than non‐delirious patients. Adverse psychological sequelae expressed as post‐traumatic stress syndrome was uncommon in our study. Every attempt must be made to ensure that the ICU environment is as hospitable as possible to decrease the stress of critical illness. Post‐ICU follow‐up should include filling in the ‘missing gaps’, particularly for delirious patients. Ongoing explanations and a caring environment may assist the patient in making a complete recovery both physically and mentally.