Summary of findings for the main comparison. Information or education intervention versus no information or education intervention.

Information or education intervention versus no information or education intervention
Patient or population: adult ICU patients and their carers Settings: ICUs in Turkey, Germany, Taiwan, and the USA Intervention: information or education intervention Comparison: no information or education intervention
Outcomes	Effects of information or education interventions for adult ICU patients and their carers	No. of analysed participants (studies)	Certainty of the evidence (GRADE)	Comments
Severity of anxiety in patients HADS‐A (1 week after hospital discharge); scale from 0 to 20 CINT (admission to regular ward); scale from 0 to 100 BSRS (in ICU, time point not specified); scale from 0 to 20 Lower scores in all scales indicate less anxiety.	In 1 study, mean anxiety scores in the intervention group were 3.20 lower (3.38 to 3.02 lower). 2 studies reported mean anxiety scores with little or no difference between groups (0.40 lower in the intervention group, 4.75 lower to 3.95 higher; and 1.00 lower in the intervention group, 2.94 lower to 0.94 higher).	332 patient participants (3 studies)	⊕⊝⊝⊝ very lowa	We did not pool data: statistical heterogeneity was high (I² = 99%); we noted possible clinical differences in illness severity of participants (e.g. whether patient participants were intubated), and methodological differences in types of information provision (e.g. whether information was tailored, and what type, and how often, it was presented).
Severity of depression in patients HADS‐D (1 week after hospital discharge); scale from 0 to 20 BSRS (in ICU, time point not specified); scale from 0 to 20 Lower scores in both scales indicate less depression.	In 2 studies, mean depression scores in the intervention group were 2.90 lower (4.00 to 1.80 lower); and 1.27 lower (1.47 to 1.07 lower).	160 patient participants (2 studies)	⊕⊝⊝⊝ very lowb	We did not pool data: statistical heterogeneity was high (I² = 99%); we noted possible clinical differences in illness severity of participants (e.g. whether patient participants were intubated), and methodological differences in types of information provision (e.g. what type, and how often the information was presented).
Knowledge acquisition (patients and carers)	Not measured	‐	‐
Severity of PTSD in patients treated in ICUs	Not measured	‐	‐
Patient or carer satisfaction with information provided (e.g. self reported)	Not measured	‐	‐
Health‐related quality of life (HRQoL) SF‐12 MCS (3 months after ICU discharge); scale from 0 to 100 Lower scores indicate reduced HRQoL.	In 1 study, mean HRQoL score in the intervention group was 1.30 lower (4.99 lower to 2.39 higher).	143 patient participants (1 study)	⊕⊝⊝⊝ very lowc
Adverse effects	Not measured	‐	‐
GRADE Working Group grades of evidence High certainty: Further research is very unlikely to change our confidence in the estimate of effect. Moderate certainty: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low certainty: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low certainty: We are very uncertain about the estimate.
BSRS: Brief Symptom Rating Scale; HADS‐A: Hospital Anxiety and Depression Scale ‐ anxiety subscale; HADS‐D: Hospital Anxiety and Depression Scale ‐ depression subscale; ICU: intensive care unit; MCS: mental health component summary; PTSD: post‐traumatic stress disorder; SF‐12: 12‐item Short Form Health Survey

^aTwo studies reported insufficient information on randomisation methods and allocation concealment; we could not judge risk of selective reporting bias due to insufficient reporting; and we were unclear if lack of blinding would have influenced outcome assessment; we downgraded by one level for study limitations. Few studies with a small sample size reported outcome data, and we could not combine data; we downgraded one level for imprecision. We noted statistical heterogeneity in outcome data between studies; we downgraded one level for inconsistency.
 ^bBoth studies reported insufficient information on randomisation methods and allocation concealment; we could not judge risk of selective reporting bias due to insufficient reporting; and we were unclear if lack of blinding would have influenced outcome assessment; we downgraded by one level for study limitations. Few studies with a small sample size reported outcome data; we downgraded one level for imprecision. We noted statistical heterogeneity in outcome data between studies; we downgraded one level for inconsistency.
 ^cWe were unclear if lack of blinding would have influenced outcome assessment, and we noted high attrition; we downgraded by one level for study limitations. Data were from a single study with a small sample size, and wide confidence interval; we downgraded two levels for imprecision.