Rodriguez 2016.

Study characteristics
Methods	2‐centre, controlled before‐after study
Participants	115 participants Inclusion criteria: intubated airway, surgery, or other event causing sudden speechlessness lasting for ≥ 8 hours; aged ≥ 21 years; able to read English or Spanish; ability to see and have use of ≥ 1 arm; no permanent speech disability and already using an adaptive speech device; RASS scores within acceptable range of +1 to –1; absence of delirium measured by CAM‐ICU. Exclusion criteria: people who had participated in a previous study cohort or had an admitting diagnosis of a DSM‐IV major mental illness documented in the medical record.
Interventions	Intervention Received a speech‐generating app incorporated in a tablet device with 3 communication functions that included: touch selection of pictorial hot‐buttons (graphic pictures/symbols associated with a message) with prerecorded spoken messages representing symptoms or basic needs commonly experienced by patients with sudden speechlessness; handwriting on a separate screen with finger or stylus; typewriting on a separate screen using an onscreen keyboard. Also provided with a freestanding urgent button (a push button that announced "I need help" when activated). Control Usual practice that included access to a call light and provision of pen and paper to write messages.
Outcomes	Ease of communication. Frustration with communication. Satisfaction with communication method. Unit clerk's understanding of messages.
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	Sequential sampling used.
Allocation concealment (selection bias)	High risk	No allocation concealment.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Not possible to blind participants or personnel due to nature of intervention.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Blinding of outcome assessors not described.
Incomplete outcome data (attrition bias) All outcomes	High risk	Missing data were not addressed.
Selective reporting (reporting bias)	Low risk	No evidence of selective reporting.
Other bias	Low risk	No evidence of other bias. Funding: NINR SBIR grants 1R43NR01084201 and 9R44DC01227502A1. This work was supported in part by the NIH/NCATS Clinical and Translational Science Awards to the University of Florida KL2 TR000065 and UL1 TR000064 (not considered source of bias). Author conflict of interest: none stated.

CAM‐ICU: Confusion Assessment Method – ICU; COPD: chronic obstructive pulmonary disease; DSM‐IV: Diagnostic and Statistical Manual of Mental Disorders, fourth edition; EMR: electronic medical record; FiO₂: fraction of inspired oxygen; GCS: Glasgow Coma Scale; HADS: Hospital Anxiety Depression Scale; ICU: intensive care unit; PEEP: positive end expiratory pressure; RASS: Richmond Agitation Sedation Scale; SLP: speech language pathologist.