Koszalinski 2020.

Study characteristics
Methods	Quasi‐randomised controlled trial
Participants	36 participants Inclusion criteria: hospitalisation on any of the participating units for any length of time; RASS score between −1 and +1 (awake, aware, and not agitated); able to use Speak For Myself‐Voice for 48 hours; able to manipulate a computer tablet; ability to read and write English. Exclusion criteria: hospitalised on units other than those participating in the study; RASS score less than −1 or exceeding +1; unwilling to use Speak For Myself‐Voice for 48 hours; unable to manipulate the computer tablet; unable to read and write English.
Interventions	Intervention Participants provided the Speak for Myself‐Voice communication tablet app provided on an i‐Pad. This included an advanced care planning menu, pain indication, basic needs requests, a free‐text section app was used on i‐Pads. Participants were provided the Speak for Myself‐Voice app for 48 hours or until they no longer needed communication assistance. Control Alphabet and picture communication board as usual care.
Outcomes	Depression. Anxiety.
Notes	Secondary analyses paper reported each of the HADS items in detail (Koszalinki 2020).
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	Assignment to groups was not entirely random, the randomisation schedule was described as beginning with first participant assigned to the intervention group, second to control, third to intervention, etc.).
Allocation concealment (selection bias)	High risk	Allocation was known to the research team but concealed from the clinical team and participants.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Unable to blind due to nature of intervention.
Blinding of outcome assessment (detection bias) All outcomes	High risk	Blinding of outcome assessors not performed (information provided by corresponding author).
Incomplete outcome data (attrition bias) All outcomes	Low risk	No indication of incomplete data.
Selective reporting (reporting bias)	Low risk	No evidence of selective reporting.
Other bias	Low risk	No evidence of other bias. Partially funded by the University of Tennessee, Center for Health Sciences Research – not considered a source of bias. Author conflict of interest: none stated.