Summary of findings 1. Non‐vocal communication aid compared to usual care without an aid for adults requiring an artificial airway with or without mechanical ventilator support.

Non‐vocal communication aid compared to usual care without an aid for adults requiring an artificial airway with or without mechanical ventilator support
Patient or population: adults requiring an artificial airway with or without mechanical ventilator support Setting: – Intervention: non‐vocal communication aid Comparison: usual care without an aid
Outcomes	№ of participants (studies)	Impact	Certainty of the evidence (GRADE)	Comments
Proportion able to communicate or phonate	—	—	—	No study measured this outcome.
Health‐related quality of life	—	—	—	No study measured this outcome.
Emotional and psychological distress	90 (2 RCTs)	1 study of 60 participants reported 9 (30%) intervention participants using a communication board were quite to very distressed compared to 24 (80%) control participants. 1 study of 30 participants demonstrated a 15‐point reduction in HADS‐A (anxiety) (mean score 18.1 (SD 1.8) to 3.0 (SD 1.8) measured at baseline and after provision of a communication board for 48 hours compared to a 5‐point reduction in the control group (mean 16.9 (SD 2.4) to 12.0 (SD 4.3) measured at baseline and 48 hours. The HADS‐A is the anxiety subscale of the HADS. Scores range from 0 to 21 with higher scores indicating greater anxiety.	⊕⊝⊝⊝ Very lowa,b	We downgraded 2 levels due to very serious risk of bias and 1 level due to imprecision.
ICU length of stay (days)	1500 (2 RCTs)	1 study of 1440 participants reported no difference in the unadjusted (0.20, 95% CI –1.18 to 1.59) or adjusted (–0.08, 95% CI –1.28 to 1.13) median ICU length of stay measured in days. 1 study of 60 participants reported a reduction in ICU length of stay (MD –0.21, 95% CI –0.29 to –0.13).	⊕⊕⊝⊝ Lowc,d,e	Downgraded 1 level due to serious risk of bias and 1 level due to inconsistency of results.
Costs	1440 (1 RCT)	1 study reported cost‐adjusted charges from hospital administrative claims were slightly higher in the intervention group, but this difference was not statistically significant (unadjusted intervention effect USD 6380, 95% CI USD 579 to USD 13,339; P = 0.07; adjusted intervention effect USD 5797 (‐USD 936 to USD 12,529) (adjusting for participant age, sex, race, admission APACHE III, and neurological disorder as admitting diagnosis).	⊕⊕⊝⊝  Lowc,f	Downgraded 1 level due to  serious risk of bias, and 1 level due to imprecision.
Adverse events	1440 (1 RCT)	1 study of 1440 participants measured days of upper extremity restraint use and found no difference between intervention and control groups in number of ICU days physical restraint was used (50.1 (36.5%) ICU days with intervention vs 47.9 (36%) ICU days with control.	⊕⊕⊝⊝ Lowg	Downgraded 1 level due to serious risk of bias and 1 level due to imprecision.
APACHE III: Acute Physiology, Age, Chronic Health Evaluation III; CI: confidence interval; HADS‐A: Hospital Anxiety and Depression Scale – Anxiety; ICU: intensive care unit; MD: mean difference; RCT: randomised controlled trial; SD: standard deviation.
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

^aThe two studies contributing data to this outcome were assessed at high risk of selection, performance, and detection bias.
^bThe two studies contributing data to this outcome uses different measures of different emotions with different time‐points and were based on a small number of participants.
^cOne study was at high risk of performance bias; however, length of stay and costs are objective outcomes and outcome assessors were blinded.
^dOne study was at high risk of bias across multiple domains.
^eIncluded studies indicated either no effect or a small non‐clinically (i.e. 0.2 days) important reduction in ICU length of stay.
^fOne study reported this outcome with wide CIs.
^gOne study was at high risk of performance bias; however, duration of physical restraint use is an objective outcome and outcome assessors were blinded.