Allocation sequence was random (computer generated random numbers). Allocation occurred after enrolment by an investigator not involved in the participants' clinical management. Higher number of participants admitted with a stroke in the intervention group (39 vs. 16), in addition, the 'Others' subgroup the Barthel index is higher in the control group than the intervention group (29.1 vs. 23.1). This is thought to be compatible with chance.

Its assumed participants and the Occupational Therapist (OT) delivering the intervention were aware of the assigned intervention allocation. All participants received their allocated intervention. Therefore, even though intention to treat analysis not specified, analysis deemed appropriate.

Given the nature of this population (older adults during and after acute hospitalisation) we considered a threshold of 90% of participants with data as sufficient. We felt this would be consistent with measurements outside of a trial context. When accounting for mortality, 95% and 94% assessed at discharge in intervention and control arms respectively.

The use of the Barthel Index to measure independence with ADLs is considered appropriate, and there were no differences in measurement or ascertainment of the outcomes between groups. The OT assessor was blinded.

The data analyst was blinded, and followed a pre‐specified statistical plan. It is not thought that results were from multiple outcome measurements or multiple analyses.

The study is judged to be at low risk of bias in all domains for this outcome.

Allocation sequence was random (computer generated random numbers) and sequence concealed (opaque sealed envelope). Participant characteristics were balanced.

Participants and those delivering the interventions were aware of intervention assignments. Seventy‐nine participants were not admitted to the unit to which they were assigned. Deviations may have affected the effect estimate, however they were well‐balanced between groups. Intention to treat analysis was used.

Functional data was obtained in 1476 of 1483 of surviving patients (99.5%) at discharge.

The Katz ADL score is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were not blinded, and it was therefore considered likely that knowledge of the intervention could influence the outcome, given the likely strong belief in the benefits of the intervention ward.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at high risk of bias due to measurement of the outcome.

Randomisation was based on the availability of beds, and there was no allocation concealment as participants were allocated a ward prior to enrolment. Participant characteristics appeared balanced.

Participants and clinicians delivering care were aware of treatment assignments. There was no evidence of deviations from intended interventions and intention to treat analysis was used.

No missing data at hospital discharge.

The ADL Staircase is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were not blinded, and it was therefore considered likely that knowledge of the intervention could influence the outcome, given the likely strong belief in the benefits of the intervention ward.

A detailed pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at high risk of bias due to bias in the measurement of the outcome and bias arising from the randomisation process.

Allocation sequence was random (computer generated random numbers) and balanced participant characteristics, but no information was provided regarding allocation concealment.

It is assumed participants and clinicians were aware of the treatment assignments. There was no evidence of deviations from the assigned interventions and evidence of an intention to treat approach being used for analysis.

Accounting for mortality (24 died in each arm) all participants had outcome data at discharge.

The Katz ADL score is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were not blinded, and it was therefore considered likely that knowledge of the intervention could influence the outcome, given the likely strong belief in the benefits of the intervention ward.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at high risk of bias in measurement of the outcome for this outcome.

No information provided on randomisation methods or sequence concealment other than to say participants were randomised. Differences in participant characteristics appear compatible with chance.

It is assumed that participants and those delivering the interventions were aware of intervention assignments. However, all participants received their intended intervention and intention to treat was specified in the analysis.

When accounting for mortality, data was available for only 72% and 81% of intervention and control group respectively at T1. Some missing data is likely to be dependent on its true value, as the main reason for missing data was adverse events, and we consider participants who experience adverse events to be more likely to have a lower ADL score.

The Katz ADL score is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were not blinded, and it was therefore considered likely that knowledge of the intervention could influence the outcome, given the likely strong belief in the benefits of the intervention ward.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at high risk of bias due to missing outcome data and measurement of the outcome.

No description of randomisation sequence generation, but the paper describes a block randomisation strategy and sequence concealed (sealed envelopes). The patient characteristics appear well‐balanced between groups.

We assume both participants and those delivering the interventions were aware of intervention assignments. 6 participants did not receive allocated intervention, this is considered to be consistent with what could occur outside the trial context, and intention to treat analysis was used.

No missing data.

The Katz ADL score is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were blinded.

The study protocol details statistical analysis plan which is in accordance with results presented, and it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at low risk of bias in all domains for this outcome.

Allocation sequence was random (generated using an online system) and the allocation sequence was concealed until participants were enrolled (opaque concealed and sealed envelopes). Participant characteristics were well‐balanced.

Participants and those delivering the interventions were aware of intervention allocations. There was no evidence of deviations from the intended interventions and intention to treat analysis used.

No missing data after accounting for mortality.

The Barthel Index is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were not blinded, and it was therefore considered likely that knowledge of the intervention could influence the outcome, given the likely strong belief in the benefits of the intervention ward.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at high risk of bias due to measurement of the outcome.

Allocation sequence was random (computer generated random numbers) and sequence concealed (blinded project coordinator allocated participants). Participant characteristics were balanced.

Participants and those delivering the interventions were aware of intervention allocations. There was no evidence of deviations from the intended interventions and as such it is presumed intention to treat analysis was used.

At discharge data was available for 76% of reablement group, (78% of reminder group) and 80% of the control group. Only participants who had complete data (i.e. data collected at baseline, discharge and follow‐up) had outcomes reported. Although missing data is well‐balanced between all three groups, it is considered likely that reason for missing data is related to its true value.

The Katz ADL score is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were blinded.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at high risk of bias due to missing outcome data.

Allocation not random and sequence predictable as based on alternation. There is limited data on participant characteristics, but available data suggests balanced characteristics between groups.

Participants and those delivering the interventions were aware of intervention allocations. There was no evidence of deviations from the intended interventions. Per‐protocol analysis appears to have been used. 2 participants in the intervention group were excluded from analysis as they completed less than 70% of their walks. The 2 participants that were excluded accounted for 7% of the remaining sample, it is therefore considered that there was potential for a significant impact on the results.

Only 71% of intervention and 74% of control group had outcome data. A significant proportion of missing data was related to early discharge from hospital. These participants could be expected to have a higher functional level than those who remained in hospital. However, the missing data was well‐balanced between the two groups.

The Barthel Index is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were not blinded, and it was therefore considered likely that knowledge of the intervention could influence the outcome, given the likely strong belief in the benefits of the intervention ward.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at high risk of bias in three domains for this outcome.

Allocation of wards was random (coin toss) and the allocating officer unaware of study. Baseline differences between groups were thought to be compatible with chance.

Participants and clinicians delivering care were believed to be aware of treatment assignments. There was no evidence of deviations from intended interventions and intention to treat analysis was used.

After accounting for mortality, discharge Barthel Index (BI) scores were available for 75% in intervention group and 70% in control group. Although it is feasible that following discharge patients with lower levels of functional ability are more likely to be missing, we do not think this applies in this situation as assessment prior to discharge. We do not see a situation where it would be more likely to miss an assessment due to high/low level of disability at discharge.

The use of the BI to measure independence with ADLs is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were not blinded, and it is considered likely that knowledge of the intervention could influence the outcome, given the likely strong belief in the benefits of the intervention.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at high risk of bias in measurement of the outcome.

Sequence concealment was achieved via a member of the research team not involved in recruitment and using sealed envelopes. There was no information regarding method of generating random sequence, other than to say that there was randomisation, given details of allocation concealment, assumption made that appropriate method used. Patient characteristics between groups appear well‐balanced.

Participants and those delivering the interventions were likely to be aware of intervention allocations. One participant was not allocated to a group due to an administrative error and 35 participants did not receive the intervention as planned (17 in the intervention group). This is thought to be consistent with what could occur outside of the trial context. Intention to treat analysis was used.

No missing data.

The Barthel Index is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were blinded.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to raise some bias concerns in selection of the reported result.

Allocation sequence was random (computer generated random numbers). Sequence allocation was not concealed, but performed by a member of staff independent of the enrolment procedures. ). Participant characteristics were balanced between groups and differences compatible with chance.

It is assumed that both participants and clinicians delivering care were aware of assigned interventions. There is no evidence of deviations from intended interventions and intention to treat analysis used.

Admission and discharge modified Barthel Index (mBI) scores were available for 78.8% of patients (126/160). Although we felt that patients with lower levels of functional ability were more likely to be missing at follow‐up, we do not think this applies in this situation as assessment was prior to discharge. We do not see a situation where it would be more likely to miss an assessment due to high/low level of disability at discharge

The use of the mBI to measure independence with ADLs is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were blinded.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to raise some concerns due to missing outcome data and in selection of the reported results.

Allocation sequence was random (www.randomizer.org) and personal communication with author confirms allocation concealment. Baseline differences between groups are thought to be compatible with chance.

Both participants and clinicians delivering care were aware of assigned interventions. There is no evidence of deviations from intended interventions and intention to treat analysis used.

Assuming that those that discontinued the study did not provide Barthel Index (BI) outcome data, and accounting for mortality, 80% of intervention group and 78% of control group had outcome data. Although there were participants who discontinued the study due to medical deterioration, which could have and is considered likely to have biased the results (i.e. medical deterioration being associated with poor functional outcomes), the numbers were relatively low (only 6% of the sample) and were well‐balanced between groups. We therefore feel that this was unlikely to bias the results.

The use of the BI to measure independence with ADLs is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were blinded.

The trial protocol reflects the analyses as that were conducted, and it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to raise some concerns due to missing outcome data.

Pseudo‐randomisation, allocation was based on admitting unit and bed availability, and admitting unit was determined by a rotating roster. No evidence of baseline differences in patient characteristics other than those thought to be compatible with chance.

Both participants and clinicians delivering care were aware of assigned interventions. There is no evidence of deviations from intended interventions and intention to treat analysis used.

No missing data.

The use of the modified Barthel Index to measure independence with ADLs is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were blinded.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at high risk of bias due to the randomisation process.

The allocation sequence was based on recruitment blocks of 4‐8 weeks. Activity of daily living (ADL) scores at baseline and admission were significantly lower in the intervention group than the control group, this is thought to be an important prognostic factor and therefore a concern regarding the randomisation process.

Methods were designed to blind participants from group assignments however due to the nature of the intervention and lack of sham intervention the success of this blinding is felt unlikely. Ward staff and research staff were not blinded to the participant's assigned intervention. There were no deviations from the intended interventions due to trial context described and although intention to treat analysis not specified given the lack of deviations it is assumed.

After adjusting for mortality 97% in intervention group and 97% in control group had outcome data.

The Katz ADL score is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were not blinded, and it was therefore considered likely that knowledge of the intervention could influence the outcome, given the likely strong belief in the benefits of the intervention ward.

Analysis reflects plan in trial registration and study protocol, and results are not thought to be from multiple outcome measures or multiple analyses.

The study is judged to be at high risk of bias due to the randomisation process and measurement of the outcome.

Allocation sequence was random (computer‐generated block randomisation), and allocation sequence concealed from the investigators. Participant characteristics were well‐balanced between groups.

Participants and clinicians were aware of the treatment assignments. There was no evidence of deviations from the assigned interventions and an intention to treat approach was used for analysis.

When accounting for mortality only 93% of participants in intervention group and 67% in control group completed measures at discharge. The difference in the proportion of missing data, and that the assessments were carried out home after discharge, it was judged likely that the reasons for missing data may have depended on its true value.

The Barthel Index is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were blinded.

The analysis is in accordance with a pre‐specified analysis in a published protocol, and results are not thought to be from multiple outcome measures or multiple analyses.

The study is judged to be at high risk of bias due to missing outcome data.

Allocation sequence was random (computer generated random numbers) and sequence concealed (opaque sealed envelope). Participant characteristics were balanced.

Participants and those delivering the interventions were aware of intervention assignments. Seventy‐nine participants were not admitted to the unit to which they were assigned. Deviations may have affected the effect estimate, however the were well‐balanced between groups. Intention to treat analysis was used.

Functional data was obtained in 1476 of 1483 of surviving patients (99.5%) at discharge.

The Physical Performance and Mobility Examination (PPME) is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were not blinded, and it was therefore considered likely that knowledge of the intervention could influence the outcome, given the likely strong belief in the benefits of the intervention ward.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at high risk of bias due to measurement of the outcome.

Allocation sequence was random (generated using an online system) and the allocation sequence was concealed until participants were enrolled (opaque concealed and sealed envelopes). Participant characteristics were well‐balanced.

Participants and those delivering the interventions were aware of intervention allocations. There was no evidence of deviations from the intended interventions and intention to treat analysis used.

No missing data after accounting for mortality.

The Barden Activity Subscale to measure functional mobility is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were not blinded, and it was therefore considered likely that knowledge of the intervention could influence the outcome, given the likely strong belief in the benefits of the intervention ward.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at high risk of bias due to measurement of the outcome.

Allocation sequence was random (www.randomization.com) and the allocation sequence was only known by the chief investigator who was not involved with screening patients. Baseline differences between groups are thought to be compatible with chance.

Both participants and clinicians delivering care were aware of assigned interventions. There is no evidence of deviations from intended interventions and intention to treat analysis not explicitly stated, but all participants appear to have been analysed in the group to which they were randomised.

96% of participants had complete data sets.

The use of the Elderly Mobility Scale (EMS) to measure functional mobility is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were not blinded, and it is considered likely that knowledge of the intervention could influence the outcome, given the likely strong belief in the benefits of the intervention.

The trial protocol and registration reflects the analyses as that were conducted, and it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to raise some concerns due to the measurement of the outcome.

Allocation of wards was random (coin toss) and the allocating officer unaware of study. Baseline differences between groups were thought to be compatible with chance.

Participants and clinicians delivering care were believed to be aware of treatment assignments. There was no evidence of deviations from intended interventions and intention to treat analysis was used.

After accounting for mortality, discharge Functional Ambulation Classification (FAC) scores were available for 75% in intervention group and 70% in control group. Although it is feasible that following discharge patients with lower levels of functional ability are more likely to be missing, we do not think this applies in this situation as assessment prior to discharge. We do not see a situation where it would be more likely to miss an assessment due to high/low level of disability at discharge.

The use of the FAC to measure functional mobility is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were not blinded, and it is considered likely that knowledge of the intervention could influence the outcome, given the likely strong belief in the benefits of the intervention.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at high risk of bias in measurement of the outcome.

Allocation sequence was random (www.randomizer.org) and personal communication with author confirms allocation concealment. Baseline differences between groups are thought to be compatible with chance.

Both participants and clinicians delivering care were aware of assigned interventions. There is no evidence of deviations from intended interventions and intention to treat analysis used.

Assuming that those that discontinued the study did not provide Short Physical Performance Battery (SPPB) outcome data, and accounting for mortality, 83% of intervention group and 86% of control group had outcome data. Although there were participants who discontinued the study due to medical deterioration, which could have and is considered likely to have biased the results (i.e. medical deterioration being associated with poor functional outcomes), the numbers were relatively low (only 6% of the sample) and were well‐balanced between groups. We therefore feel that this was unlikely to bias the results.

The use of the SPPB to measure functional mobility is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were blinded.

The trial protocol reflects the analyses as that were conducted, and it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to raise some concerns due to missing outcome data.

Allocation sequence was random (computer‐generated) and allocation sequence was concealed. Baseline differences are thought to be compatible with chance.

Both participants and clinicians delivering care were aware of assigned interventions. There is no evidence of deviations from intended interventions and intention to treat analysis used.

90% of intervention group and 94% of the control group had Short Physical Performance Battery (SPPB) data. Given the nature of this population (older adults during and after acute hospitalisation) we considered a threshold of 90% of participants with data as sufficient.

The use of SPPB to measure functional mobility is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were blinded.

The trial protocol reflects the analyses as that were conducted, and it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at low risk of bias for all domains for this outcome.

The allocation sequence was based on recruitment blocks of 4‐8 weeks. Activity of daily living (ADL) scores at baseline and admission were significantly lower in the intervention group than the control group, this is thought to be an important prognostic factor and therefore a concern regarding the randomisation process.

Methods were designed to blind participants from group assignments however due to the nature of the intervention and lack of sham intervention the success of this blinding is felt unlikely. Ward staff and research staff were not blinded to the participant's assigned intervention. There were no deviations from the intended interventions due to trial context described and although intention to treat analysis not specified given the lack of deviations it is assumed.

After adjusting for mortality 97% in intervention group and 97% in control group had outcome data.

The use of the Short Physical Performance Battery (SPPB) for measuring functional mobility is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were not blinded, and it was therefore considered likely that knowledge of the intervention could influence the outcome, given the likely strong belief in the benefits of the intervention ward.

Analysis reflects plan in trial registration and study protocol, and results are not thought to be from multiple outcome measures or multiple analyses.

The study is judged to be at high risk of bias due to the randomisation process and measurement of the outcome.

Allocation sequence was random (computer‐generated block randomisation), and allocation sequence concealed from the investigators. Participant characteristics were well‐balanced between groups.

Participants and clinicians were aware of the treatment assignments. There was no evidence of deviations from the assigned interventions and an intention to treat approach was used for analysis.

When accounting for mortality only 93% of participants in intervention group and 67% in control group completed measures at discharge. The difference in the proportion of missing data, and that the assessments were carried out home after discharge, it was judged likely that the reasons for missing data may have depended on its true value.

The use of the de Morton Mobility Index (DEMMI) is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were blinded.

The analysis is in accordance with a pre‐specified analysis in a published protocol, and results are not thought to be from multiple outcome measures or multiple analyses.

As per RoB2 algorithm, the study is judged to be at high risk of bias due to missing outcome data.

Allocation sequence was random (computer generated random numbers). Allocation occurred after enrolment by an investigator not involved in the participants' clinical management. Higher number of participants admitted with a stroke in the intervention group (39 vs. 16), in addition, the 'Others' subgroup the Barthel index is higher in the control group than the intervention group (29.1 vs. 23.1). This is thought to be compatible with chance.

Its assumed participants and the Occupational Therapist (OT) delivering the intervention were aware of the assigned intervention allocation. All participants received their allocated intervention. Therefore, even though intention to treat analysis not specified, analysis deemed appropriate.

Given the nature of this population (older adults during and after acute hospitalisation) we considered a threshold of 90% of participants with data as sufficient. We felt this would be consistent with measurements outside of a trial context. When accounting for mortality, 93% and 91% assessed at discharge in intervention and control arms respectively.

The use of the Confusion Assessment Method (CAM) to assess for delirium is considered appropriate, and there were no differences in measurement or ascertainment of the outcomes between groups. The OT assessor was blinded.

The data analyst was blinded, and followed a pre‐specified statistical plan. It is not thought that results were from multiple outcome measurements or multiple analyses.

The study is judged to be at low risk of bias in all domains for this outcome.

The method of allocation sequence is not described, but the use of sealed envelopes suggests a random component was used. Participant characteristics were balanced.

Participants and those delivering the interventions were aware of intervention assignments. Twenty‐five patients were excluded due to them not meeting the set eligibility criteria, however, these: protocol violations are not expected to influence the effect estimate of the outcome as per protocol analyses used. The per protocol analyses was not thought to have a substantial impact on the result given the main reason for exclusion was inappropriate recruitment. Excluding this 25, the other 6 exclusions represent approximately 1% of the total sample size.

Data at discharge from hospital was available for 98% of participants.

The confusion assessment method (CAM) instrument was considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were not blinded, and it was therefore considered likely that knowledge of the intervention could influence the outcome, given the likely strong belief in the benefits of the intervention ward.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at high risk of bias due to measurement of the outcome.

No description of randomisation sequence generation, but the paper describes a block randomisation strategy and sequence concealed (sealed envelopes). The patient characteristics appear well‐balanced between groups.

We assume both participants and those delivering the interventions were aware of intervention assignments. 6 participants did not receive allocated intervention, this is considered to be consistent with what could occur outside the trial context, and intention to treat analysis was used.

No missing data.

The Confusion Assessment Method (CAM) score is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were blinded at admission, discharge and follow up, but not the research assistants as they were delivering the intervention. “The research assistants also used the CAM at each patient visit throughout the hospital stay to ensure that patients in either group did not develop incident delirium.” Therefore, assessors not considered to be blinded, and it is likely given that there is judgement involved in scoring of the CAM, that knowledge of the intervention could influence the scoring of the CAM given the likely strong beliefs in the benefits of the intervention ward.

The study protocol details statistical analysis plan which did not refer to assessment of delirium after baseline assessments, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at high risk of bias in measurement of the outcome.

Sequence concealment was achieved via a member of the research team not involved in recruitment and using sealed envelopes. There was no information regarding method of generating random sequence, other than to say that there was randomisation, given details of allocation concealment, assumption made that appropriate method used. Patient characteristics between groups appear well‐balanced.

Participants and those delivering the interventions were likely to be aware of intervention allocations. One participant was not allocated to a group due to an administrative error and 35 participants did not receive the intervention as planned (17 in the intervention group). This is thought to be consistent with what could occur outside of the trial context. Intention to treat analysis was used.

No missing data.

The Confusion Assessment Method (CAM) is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were blinded.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to raise some concerns in the selection of the reported results.

Allocation sequence was random (www.randomizer.org) and personal communication with author confirms allocation concealment. Baseline differences between groups are thought to be compatible with chance.

Both participants and clinicians delivering care were aware of assigned interventions. There is no evidence of deviations from intended interventions and intention to treat analysis used.

Assuming that those that discontinued the study did not provide Barthel Index (BI) outcome data, and accounting for mortality, 84% of intervention group and 86% of control group had outcome data. Although there were participants who discontinued the study due to medical deterioration, which could have and is considered likely to have biased the results (i.e. medical deterioration being associated with poor functional outcomes), the numbers were relatively low (only 6% of the sample) and were well‐balanced between groups. We therefore feel that this was unlikely to bias the results.

The use of the Confusion Assessment Method (CAM) to measure delirium is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were blinded.

The trial protocol reflects the analyses as that were conducted, and it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to raise some concerns due to missing outcome data.

Allocation sequence was random (computer‐generated) and allocation sequence was concealed. Baseline differences are thought to be compatible with chance.

Both participants and clinicians delivering care were aware of assigned interventions. There is no evidence of deviations from intended interventions and intention to treat analysis used.

90% of intervention group and 94% of the control group had outcome data. Given the nature of this population (older adults during and after acute hospitalisation) we considered a threshold of 90% of participants with data as sufficient.

The use of the Six‐item Cognitive Impairment Test (6CIT) to assess for delirium is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were blinded.

The trial protocol reflects the analyses as that were conducted, and it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at low risk of bias in all domains for this outcome.

Pseudo‐randomisation, allocation was based on admitting unit and bed availability, and admitting unit was determined by a rotating roster. No evidence of baseline differences in patient characteristics other than those thought to be compatible with chance.

Both participants and clinicians delivering care were aware of assigned interventions. There is no evidence of deviations from intended interventions and intention to treat analysis used.

No missing data.

Identifying delirium according to chart review using validated methodology is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were blinded.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at high risk of bias due to the randomisation process.

Randomisation was based on the availability of beds, and there was no allocation concealment as participants were allocated a ward prior to enrolment. Participant characteristics appeared balanced.

Participants and clinicians delivering care were aware of treatment assignments. There was no evidence of deviations from intended interventions and intention to treat analysis was used.

No missing data.

The EQ‐VAS is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were not blinded, and it was therefore considered likely that knowledge of the intervention could influence the outcome, given the likely strong belief in the benefits of the intervention ward.

A detailed pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at high risk of bias due to bias in the measurement of the outcome and bias arising from the randomisation process.

Allocation sequence was random (computer generated random numbers) and sequence concealed (blinded project coordinator allocated participants). Participant characteristics were balanced.

Participants and those delivering the interventions were aware of intervention allocations. There was no evidence of deviations from the intended interventions and as such it is presumed intention to treat analysis was used.

At discharge data was available for 76% of reablement group, (78% of reminder group) and 80% of the control group. Only participants who had complete data (i.e. data collected at baseline, discharge and follow‐up) had outcomes reported. Although missing data is well‐balanced between all three groups, it is considered likely that reason for missing data is related to its true value.

The EQ5D is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were blinded.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at high risk of bias due to missing outcome data.

Allocation sequence was random (www.randomizer.org) and personal communication with author confirms allocation concealment. Baseline differences between groups are thought to be compatible with chance.

Both participants and clinicians delivering care were aware of assigned interventions. There is no evidence of deviations from intended interventions and intention to treat analysis used.

Assuming that those that discontinued the study did not provide e EQ‐5D outcome data, and accounting for mortality, 83% of intervention group and 86% of control group had outcome data. Although there were participants who discontinued the study due to medical deterioration, which could have and is considered likely to have biased the results (i.e. medical deterioration being associated with poor functional outcomes), the numbers were relatively low (only 6% of the sample) and were well‐balanced between groups. We therefore feel that this was unlikely to bias the results.

The EQ‐5D is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were blinded.

The trial protocol reflects the analyses as that were conducted, and it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to raise some concerns due to missing outcome data.

Allocation sequence was random (computer‐generated) and allocation sequence was concealed. Baseline differences are thought to be compatible with chance.

Both participants and clinicians delivering care were aware of assigned interventions. There is no evidence of deviations from intended interventions and intention to treat analysis used.

90% of intervention group and 94% of the control group had outcome data. Given the nature of this population (older adults during and after acute hospitalisation) we considered a threshold of 90% of participants with data as sufficient.

The use of EQ‐5D5L to measure quality of life is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were blinded.

The trial protocol reflects the analyses as that were conducted, and it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at low risk of bias for all domains for this outcome.

No information on randomisation methods other than to say participants were randomised, however, patient characteristics appear well‐balanced between groups and differences compatible with chance.

It is assumed that participants and those delivering the interventions were aware of intervention assignments. There were 15 protocol deviations in the CIRACT group and 8 in the THB‐Rehab group. The deviations were thought to be in keeping with what would be expected outside of a trial context. Intention to treat not explicitly stated though it appears that the 1 participant who did not receive the intended intervention was analysed within the group to which they were assigned.

No missing data

Methods of measuring the number of falls was considered appropriate and there were no differences in measurement or ascertainment of the outcomes between groups. The assessor was not blinded, but it is thought that due to the lack of judgement in the measurement, that the knowledge of the intervention did not influence the outcomes.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to raise some concerns in the selection of the reported results and the randomisation process.

No description of randomisation sequence generation, but the paper describes a block randomisation strategy and sequence concealed (sealed envelopes). The patient characteristics appear well‐balanced between groups.

We assume both participants and those delivering the interventions were aware of intervention assignments. 6 participants did not receive allocated intervention, this is considered to be consistent with what could occur outside the trial context, and intention to treat analysis was used.

No missing data.

Methods of measuring the number of falls was considered appropriate and there were no differences in measurement or ascertainment of the outcomes between groups. The assessor was blinded.

The study protocol details statistical analysis plan which is in accordance with results presented, and it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at low risk of bias in all domains for this outcome.

Allocation sequence was random (generated using an online system) and the allocation sequence was concealed until participants were enrolled (opaque concealed and sealed envelopes). Participant characteristics were well‐balanced.

Participants and those delivering the interventions were aware of intervention allocations. There was no evidence of deviations from the intended interventions and intention to treat analysis used.

No missing data after accounting for mortality.

Methods of measuring the number of falls was considered appropriate and there were no differences in measurement or ascertainment of the outcomes between groups. The assessor was not blinded, but it is thought that due to the lack of judgement in the measurement, that the knowledge of the intervention did not influence the outcomes.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to raise some concerns in the selection of the reported results .

Allocation not random and sequence predictable as based on alternation. There is limited data on participant characteristics, but available data suggests balanced characteristics between groups.

Participants and those delivering the interventions were aware of intervention allocations. There was no evidence of deviations from the intended interventions. Per‐protocol analysis appears to have been used. 2 participants in the intervention group were excluded from analysis as they completed less than 70% of their walks. The 2 participants that were excluded accounted for 7% of the remaining sample, it is therefore considered that there was potential for a significant impact on the results.

Only 71% of intervention and 74% of control group had outcome data. A significant proportion of missing data was related to early discharge from hospital. These participants could be expected to have a higher functional level than those who remained in hospital. However, the missing data was well‐balanced between the two groups.

Methods of measuring the number of falls was considered appropriate and there were no differences in measurement or ascertainment of the outcomes between groups. The assessor was not blinded, but it is thought that due to the lack of judgement in the measurement, that the knowledge of the intervention did not influence the outcomes.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at high risk of bias arising from the randomisation process and due to deviations from the intended interventions.

Allocation of wards was random (coin toss) and the allocating officer unaware of study. Baseline differences between groups were thought to be compatible with chance.

Participants and clinicians delivering care were believed to be aware of treatment assignments. There was no evidence of deviations from intended interventions and intention to treat analysis was used.

No missing data.

Methods of measuring the number of falls was considered appropriate and there were no differences in measurement or ascertainment of the outcomes between groups. The assessor was not blinded, but it is thought that due to the lack of judgement in the measurement, that the knowledge of the intervention did not influence the outcomes.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to raise some concerns in the selection of the reported result.

Allocation sequence was random (computer generated random numbers). Sequence allocation was not concealed, but performed by a member of staff independent of the enrolment procedures. ). Participant characteristics were balanced between groups and differences compatible with chance.

It is assumed that both participants and clinicians delivering care were aware of assigned interventions. There is no evidence of deviations from intended interventions and intention to treat analysis used.

After accounting for mortality, falls data available for 99% of control group and 93% of intervention group. Given the nature of this population (older adults during and after acute hospitalisation) we considered a threshold of 90% of participants with data as sufficient.

Methods of measuring the number of falls was considered appropriate and there were no differences in measurement or ascertainment of the outcomes between groups. The assessor was blinded.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to raise some concerns due to the selection of the reported results.

Allocation sequence was random (www.randomizer.org) and personal communication with author confirms allocation concealment. Baseline differences between groups are thought to be compatible with chance.

Both participants and clinicians delivering care were aware of assigned interventions. There is no evidence of deviations from intended interventions and intention to treat analysis used.

No missing data

Methods of measuring the number of falls was considered appropriate and there were no differences in measurement or ascertainment of the outcomes between groups. The assessor was blinded.

The trial protocol reflects the analyses as that were conducted, and it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at low risk of bias in all domains for this outcome.

Allocation sequence was random (computer‐generated) and allocation sequence was concealed. Baseline differences are thought to be compatible with chance.

Both participants and clinicians delivering care were aware of assigned interventions. There is no evidence of deviations from intended interventions and intention to treat analysis used.

No missing data.

Methods of measuring the number of falls was considered appropriate and there were no differences in measurement or ascertainment of the outcomes between groups. The assessor was blinded.

The trial protocol reflects the analyses as that were conducted, and it is not thought that the results were from multiple outcome measures or multiple analyses.

As per RoB2 algorithm, the study is judged to be at low risk of bias for all domains for this outcome.

Pseudo‐randomisation, allocation was based on admitting unit and bed availability, and admitting unit was determined by a rotating roster. No evidence of baseline differences in patient characteristics other than those thought to be compatible with chance.

Both participants and clinicians delivering care were aware of assigned interventions. There is no evidence of deviations from intended interventions and intention to treat analysis used.

No missing data.

Methods of measuring the number of falls was considered appropriate and there were no differences in measurement or ascertainment of the outcomes between groups. The assessor was blinded.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at high risk of bias due to the randomisation process.

Allocation sequence was random (computer generated random numbers). Allocation occurred after enrolment by an investigator not involved in the participants' clinical management. Higher number of participants admitted with a stroke in the intervention group (39 vs. 16), in addition, the 'Others' subgroup the Barthel index is higher in the control group than the intervention group (29.1 vs. 23.1). This is thought to be compatible with chance.

Its assumed participants and the Occupational Therapist (OT) delivering the intervention were aware of the assigned intervention allocation. All participants received their allocated intervention. Therefore, even though intention to treat analysis not specified, analysis deemed appropriate.

Given the nature of this population (older adults during and after acute hospitalisation) we considered a threshold of 90% of participants with data as sufficient. We felt this would be consistent with measurements outside of a trial context. When accounting for mortality, 93% and 91% assessed at discharge for confusion in intervention and control arms respectively.

The use of the Confusion Assessment Method (CAM) to assess for delirium is considered appropriate, and there were no differences in measurement or ascertainment of the outcomes between groups. The OT assessor was blinded.

The data analyst was blinded, and followed a pre‐specified statistical plan. It is not thought that results were from multiple outcome measurements or multiple analyses.

The study is judged to be at low risk of bias in all domains for this outcome.

The method of allocation sequence is not described, but the use of sealed envelopes suggests a random component was used. Participant characteristics were balanced.

Participants and those delivering the interventions were aware of intervention assignments. Twenty‐five patients were excluded due to them not meeting the set eligibility criteria, however, these: protocol violations are not expected to influence the effect estimate of the outcome as per protocol analyses used. The per protocol analyses was not thought to have a substantial impact on the result given the main reason for exclusion was inappropriate recruitment. Excluding this 25, the other 6 exclusions represent approximately 1% of the total sample size.

Data at discharge from hospital was available for 98% of participants.

The confusion assessment method (CAM) instrument was considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were not blinded, and it was therefore considered likely that knowledge of the intervention could influence the outcome, given the likely strong belief in the benefits of the intervention ward.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at high risk of bias due to measurement of the outcome.

No description of randomisation sequence generation, but the paper describes a block randomisation strategy and sequence concealed (sealed envelopes). The patient characteristics appear well‐balanced between groups.

We assume both participants and those delivering the interventions were aware of intervention assignments. Six participants did not receive allocated intervention, this is considered to be consistent with what could occur outside the trial context, and intention to treat analysis was used.

No missing data.

Measures used to measure medical deterioration are considered appropriate and there and there were no differences in the measurement or ascertainment between groups. The assessors were blinded.

The study protocol details statistical analysis plan which is in line with the analyses conducted, and it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at low risk of bias in all domains for this outcome.

Allocation sequence was random (computer generated random numbers) and sequence concealed (blinded project coordinator allocated participants). Participant characteristics were balanced.

Participants and those delivering the interventions were aware of intervention allocations. There was no evidence of deviations from the intended interventions and as such it is presumed intention to treat analysis was used.

No missing data for medical deterioration during hospitalisation.

The measures are considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were blinded.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to have some concerns in the selection of the reported result.

Allocation of wards was random (coin toss) and the allocating officer unaware of study. Baseline differences between groups were thought to be compatible with chance.

Participants and clinicians delivering care were believed to be aware of treatment assignments. There was no evidence of deviations from intended interventions and intention to treat analysis was used.

No missing data.

The measures are considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were not blinded, but it is thought that due to the lack of judgement in the outcome, that the knowledge of the intervention did not influence the outcomes.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to raise some concerns in the selection of the reported result.

Sequence concealment was achieved via a member of the research team not involved in recruitment and using sealed envelopes. There was no information regarding method of generating random sequence, other than to say that there was randomisation, given details of allocation concealment, assumption made that appropriate method used. Patient characteristics between groups appear well‐balanced.

Participants and those delivering the interventions were likely to be aware of intervention allocations. One participant was not allocated to a group due to an administrative error and 35 participants did not receive the intervention as planned (17 in the intervention group). This is thought to be consistent with what could occur outside of the trial context. Intention to treat analysis was used.

No missing data.

The Confusion Assessment Method (CAM) is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were blinded.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to raise some concerns in the selection of the reported results.

Allocation sequence was random (computer generated random numbers). Sequence allocation was not concealed, but performed by a member of staff independent of the enrolment procedures. ). Participant characteristics were balanced between groups and differences compatible with chance.

It is assumed that both participants and clinicians delivering care were aware of assigned interventions. There is no evidence of deviations from intended interventions and intention to treat analysis used.

No missing data.

Measurement methods are considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were blinded.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to raise some concerns in the selection of the reported results.

Allocation sequence was random (www.randomizer.org) and personal communication with author confirms allocation concealment. Baseline differences between groups are thought to be compatible with chance.

Both participants and clinicians delivering care were aware of assigned interventions. There is no evidence of deviations from intended interventions and intention to treat analysis used.

The methods of measurement are considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were blinded.

The trial protocol reflects the analyses as that were conducted, and it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at low risk of bias in all domains for this outcome.

The study is judged to be at low risk of bias in all domains for this outcome.

Allocation sequence was random (computer‐generated) and allocation sequence was concealed. Baseline differences are thought to be compatible with chance.

Both participants and clinicians delivering care were aware of assigned interventions. There is no evidence of deviations from intended interventions and intention to treat analysis used.

90% of intervention group and 94% of the control group had outcome data. Given the nature of this population (older adults during and after acute hospitalisation) we considered a threshold of 90% of participants with data as sufficient.

The use of the Six‐item Cognitive Impairment Test (6CIT) to assess for delirium is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were blinded.

The trial protocol reflects the analyses as that were conducted, and it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at low risk of bias in all domains for this outcome.

Pseudo‐randomisation, allocation was based on admitting unit and bed availability, and admitting unit was determined by a rotating roster. No evidence of baseline differences in patient characteristics other than those thought to be compatible with chance.

Both participants and clinicians delivering care were aware of assigned interventions. There is no evidence of deviations from intended interventions and intention to treat analysis used.

No missing data.

Identifying delirium according to chart review using validated methodology is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were blinded.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at high risk of bias due to the randomisation process.

Allocation sequence was random (computer‐generated block randomisation), and allocation sequence concealed from the investigators. Participant characteristics were well‐balanced between groups.

Participants and clinicians were aware of the treatment assignments. There was no evidence of deviations from the assigned interventions and an intention to treat approach was used for analysis.

No missing data.

The measure is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were blinded.

The analysis is in accordance with a pre‐specified analysis in a published protocol, and results are not thought to be from multiple outcome measures or multiple analyses.

The study is judged to be at low risk of bias in all domains for this outcome.

Allocation sequence was random (computer generated random numbers). Allocation occurred after enrolment by an investigator not involved in the participants' clinical management. Higher number of participants admitted with a stroke in the intervention group (39 vs. 16), in addition, the 'Others' subgroup the Barthel index is higher in the control group than the intervention group (29.1 vs. 23.1). This is thought to be compatible with chance.

Its assumed participants and the Occupational Therapist (OT) delivering the intervention were aware of the assigned intervention allocation. All participants received their allocated intervention. Therefore, even though intention to treat analysis not specified, analysis deemed appropriate.

No missing data.

The method is considered appropriates, and there were no differences in measurement or ascertainment of the outcomes between groups. The OT assessor was blinded.

The data analyst was blinded, and followed a pre‐specified statistical plan. It is not thought that results were from multiple outcome measurements or multiple analyses.

The study is judged to be at low risk of bias in all domains for this outcome.

The method of allocation sequence is not described, but the use of sealed envelopes suggests a random component was used. Participant characteristics were balanced.

Participants and those delivering the interventions were aware of intervention assignments. Twenty‐five patients were excluded due to them not meeting the set eligibility criteria, however, these: protocol violations are not expected to influence the effect estimate of the outcome as per protocol analyses used. The per protocol analyses was not thought to have a substantial impact on the result given the main reason for exclusion was inappropriate recruitment. Excluding this 25, the other 6 exclusions represent approximately 1% of the total sample size.

Data at discharge from hospital complete for 98% of participants.

The measurement of length of stay is considered appropriate, and there were no differences in measurement or ascertainment between groups. Assessors were not blinded, but is thought that due to the lack of judgement in scoring a 'hard outcome' like length of stay, that it is unlikely that the knowledge of the intervention could influence the outcome.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to raise some concerns due to deviations from intended interventions and the selection of the reported results.

Allocation sequence was random (computer generated random numbers) and sequence concealed (opaque sealed envelope). Participant characteristics were balanced.

Participants and those delivering the interventions were aware of intervention assignments. Seventy‐nine participants were not admitted to the unit to which they were assigned. Deviations may have affected the effect estimate, however they were well‐balanced between groups. Intention to treat analysis was used.

No missing data.

The methods of measuring length of stay are considered appropriate, and there were no differences in measurement or ascertainment between groups. Assessors were not blinded, but is thought that due to the lack of judgement in scoring a 'hard outcome' like length of hospital stay, that it is unlikely that the knowledge of the intervention could influence the outcome.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to raise some concerns in the selection of the reported results and deviations from the intended interventions.

Randomisation was based on the availability of beds, and there was no allocation concealment as participants were allocated a ward prior to enrolment. Participant characteristics appeared balanced.

Participants and clinicians delivering care were aware of treatment assignments. There was no evidence of deviations from intended interventions and intention to treat analysis was used.

No missing data.

The methods of measuring length of hospital stay are considered appropriate, and there were no differences in measurement or ascertainment between groups. Assessors were not blinded, but is thought that due to the lack of judgement in scoring a 'hard outcome' like length of hospital stay, that it is unlikely that the knowledge of the intervention could influence the outcome.

A detailed pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at high risk of bias arising from the randomisation process.

“Patients were randomised only when both a treatment and control bed were available”. Although not stated, we believe the necessity for available beds on both wards indicate that a random component was likely used. Sequence allocation concealment is not discussed, but no significant differences in patient characteristics other than those compatible with chance were observed.

Participants and those delivering the interventions were aware of intervention assignments. There were 30 randomisation errors which has been interpreted to mean 'allocated to the ward that they were not randomised to' however there is no explanation of this term. These deviations were thought likely to have affected the outcome, though were relative well‐balanced. It appears a per protocol analyses was used, though given the low number (7%) and relative balance between groups, the omission of the participants randomised incorrectly probably did not have substantial impact on the results.

All participants accounted for at discharge.

The methods of measuring length of stay are considered appropriate, and there were no differences in measurement or ascertainment between groups. Assessors were not blinded, but is thought that due to the lack of judgement in scoring a 'hard outcome' like length of stay, that it is unlikely that the knowledge of the intervention could influence the outcome.

A detailed pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to raise some concerns in deviations from the intended interventions, methods of randomisation and selection of the reported result.

Allocation sequence was random (computer generated random numbers) and balanced participant characteristics, but no information was provided regarding allocation concealment.

It is assumed participants and clinicians were aware of the treatment assignments. There was no evidence of deviations from the assigned interventions and evidence of an intention to treat approach being used for analysis.

Accounting for mortality (24 died in each arm) all participants had outcome data at discharge.

The methods of measuring length of hospital stay are considered appropriate, and there were no differences in measurement or ascertainment between groups. Assessors were not blinded, but is thought that due to the lack of judgement in scoring a 'hard outcome' like length of hospital stay, that it is unlikely that the knowledge of the intervention could influence the outcome.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to have some concerns due to the methods of randomisation and in the selection of the reported result.

No information on randomisation methods other than to say participants were randomised, however, patient characteristics appear well‐balanced between groups and differences compatible with chance.

It is assumed that participants and those delivering the interventions were aware of intervention assignments. There were 15 protocol deviations in the CIRACT group and 8 in the THB‐Rehab group. The deviations were thought to be in keeping with what would be expected outside of a trial context. Intention to treat not explicitly stated though it appears that the 1 participant who did not receive the intended intervention was analysed within the group to which they were assigned.

No missing data

The methods of measuring length of hospital stay are considered appropriate, and there were no differences in measurement or ascertainment between groups. Assessors were not blinded, but is thought that due to the lack of judgement in scoring a 'hard outcome' like length of hospital stay, that it is unlikely that the knowledge of the intervention could influence the outcome.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to raise some concerns in the selection of the reported results and the randomisation process.

Randomisation methods are described only as: “an alternating randomisation procedure”. However, there was a centrally administered method to allocate interventions, and baseline differences did not appear to show a significant problem with randomisation process other than those thought to be compatible with chance.

Participants and clinicians delivering care were aware of allocated interventions. There was no evidence of deviations from intended intervention, and although intention to treat analysis was not specified, there was no information to suggest participants were not assessed in the group to which they were randomised.

After accounting for mortality (6 patients died), data was available for 91% of participants.

The methods of measuring length of hospital stay in hospital are considered appropriate, and there were no differences in measurement or ascertainment between groups. Assessors were not blinded, but is thought that due to the lack of judgement in scoring a 'hard outcome' like length of stay, that it is unlikely that the knowledge of the intervention could influence the outcome.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to raise some concerns in the randomisation process and in the selection of the reported results.

The allocation sequence was based on bed availability, and therefore not random, and sequence is believed to be predictable. However, baseline differences did not appear to show a significant problem with randomisation process other than those thought to be compatible with chance.

Participants and clinicians delivering care were aware of treatment allocations. There is no evidence of deviations from intended intervention, or that participants moved between the intervention and usual care units during the study period. Intention to treat analysis was not specified, however there was no information to suggest participants were not assessed in the group to which they were randomised.

No missing data.

The methods of measuring length of stay are considered appropriate, and there were no differences in measurement or ascertainment between groups. Assessors were not blinded, but is thought that due to the lack of judgement in scoring a 'hard outcome' like length of stay, that it is unlikely that the knowledge of the intervention could influence the outcome.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at high risk of bias due to bias arising from the randomisation process.

No description of randomisation sequence generation, but the paper describes a block randomisation strategy and sequence concealed (sealed envelopes). The patient characteristics appear well‐balanced between groups.

We assume both participants and those delivering the interventions were aware of intervention assignments. 6 participants did not receive allocated intervention, this is considered to be consistent with what could occur outside the trial context, and intention to treat analysis was used.

No missing data.

The methods of measuring length of hospital stay are considered appropriate, and there were no differences in measurement or ascertainment between groups. The assessor was blinded.

The study protocol details statistical analysis plan which is in accordance with results presented, and it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at low risk of bias in all domains for this outcome.

Allocation sequence was random (generated using an online system) and the allocation sequence was concealed until participants were enrolled (opaque concealed and sealed envelopes). Participant characteristics were well‐balanced.

Participants and those delivering the interventions were aware of intervention allocations. There was no evidence of deviations from the intended interventions and intention to treat analysis used.

No missing data after accounting for mortality.

Methods of measuring length of hospital stay are considered appropriate and there were no differences in measurement or ascertainment of the outcomes between groups. The assessor was not blinded, but it is thought that due to the lack of judgement in the measurement, that the knowledge of the intervention did not influence the outcomes.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to raise some concerns in the selection of the reported results .

Allocation sequence was random (computer generated random numbers) and sequence concealed (blinded project coordinator allocated participants). Participant characteristics were balanced.

Participants and those delivering the interventions were aware of intervention allocations. There was no evidence of deviations from the intended interventions and as such it is presumed intention to treat analysis was used.

No missing data for length of hospital stay.

The methods of measuring legnth of hospital stay are considered appropriate, and there were no differences in measurement or ascertainment between groups. The assessors were blinded.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to have some concerns in the selection of the reported result.

Allocation sequence was random (www.randomization.com) and the allocation sequence was only known by the chief investigator who was not involved with screening patients. Baseline differences between groups are thought to be compatible with chance.

Both participants and clinicians delivering care were aware of assigned interventions. There is no evidence of deviations from intended interventions and intention to treat analysis not explicitly stated, but all participants appear to have been analysed in the group to which they were randomised.

96% of participants had complete data sets.

The measurement of length of hospital stay is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were not blinded, but is thought that due to the lack of judgement in scoring a 'hard outcome' like length of hospital stay, that it is unlikely that the knowledge of the intervention could influence the outcome.

The trial protocol and registration reflects the analyses as that were conducted, and it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to raise some concerns due to the measurement of the outcome.

Allocation sequence was random (computer generated random numbers) and the allocation sequence concealed until participants were enrolled (via project coordinator blinded to baseline data). Participant characteristics appear well‐balanced.

Participants and clinicians delivering care were believed to be aware of treatment assignments. 6/64 participants in the intervention group did not receive allocated treatment, the reasons are considered consistent with what could occur outside the trial context, and intention to treat analysis was used.

No missing length of stay data.

The measurement of hospital length of stay is considered appropriate, and there were no differences in measurement or ascertainment between groups. The assessors are believed to be blinded to treatment allocation.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to raise some concerns in the selection of the reported result.

Allocation of wards was random (coin toss) and the allocating officer unaware of study. Baseline differences between groups were thought to be compatible with chance.

Participants and clinicians delivering care were believed to be aware of treatment assignments. There was no evidence of deviations from intended interventions and intention to treat analysis was used.

No missing data.

The measure of hospital length of stay considered appropriate, and there were no differences in measurement or ascertainment between groups. Assessors were not blinded, but is thought that due to the lack of judgement in scoring a 'hard outcome' like length of stay, that it is unlikely that the knowledge of the intervention could influence the outcome.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to raise some concerns in the selection of the reported result.

Sequence concealment was achieved via a member of the research team not involved in recruitment and using sealed envelopes. There was no information regarding method of generating random sequence, other than to say that there was randomisation, given details of allocation concealment, assumption made that appropriate method used. Patient characteristics between groups appear well‐balanced.

Participants and those delivering the interventions were likely to be aware of intervention allocations. One participant was not allocated to a group due to an administrative error and 35 participants did not receive the intervention as planned (17 in the intervention group). This is thought to be consistent with what could occur outside the trial context. Intention to treat analysis was used.

No missing data.

Measurement of length of hospital stay is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were blinded.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to raise some concerns in the selection of the reported results.

Allocation sequence was random (computer generated random numbers). Sequence allocation was not concealed, but performed by a member of staff independent of the enrolment procedures. ). Participant characteristics were balanced between groups and differences compatible with chance.

It is assumed that both participants and clinicians delivering care were aware of assigned interventions. There is no evidence of deviations from intended interventions and intention to treat analysis used.

After accounting for mortality, length of stay data available for 99% of control group and 93% of intervention group. Given the nature of this population (older adults during and after acute hospitalisation) we considered a threshold of 90% of participants with data as sufficient.

Measurement methods are considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were blinded.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to raise some concerns in the selection of the reported results.

Allocation sequence was random (www.randomizer.org) and personal communication with author confirms allocation concealment. Baseline differences between groups are thought to be compatible with chance.

Both participants and clinicians delivering care were aware of assigned interventions. There is no evidence of deviations from intended interventions and intention to treat analysis used.

No missing data.

The methods of measurement are considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were blinded.

The trial protocol reflects the analyses as that were conducted, and it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at low risk of bias in all domains for this outcome.

Allocation sequence was random (computer‐generated) and allocation sequence was concealed. Baseline differences are thought to be compatible with chance.

Both participants and clinicians delivering care were aware of assigned interventions. There is no evidence of deviations from intended interventions and intention to treat analysis used.

No missing data.

The measure is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were blinded.

The trial protocol reflects the analyses as that were conducted, and it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at low risk of bias in all domains for this outcome.

Pseudo‐randomisation, allocation was based on admitting unit and bed availability, and admitting unit was determined by a rotating roster. No evidence of baseline differences in patient characteristics other than those thought to be compatible with chance.

Both participants and clinicians delivering care were aware of assigned interventions. There is no evidence of deviations from intended interventions and intention to treat analysis used.

No missing data.

The method is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were blinded.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at high risk of bias due to the randomisation process.

The allocation sequence was based on recruitment blocks of 4‐8 weeks. Activity of daily living (ADL) scores at baseline and admission were significantly lower in the intervention group than the control group, this is thought to be an important prognostic factor and therefore a concern regarding the randomisation process.

Methods were designed to blind participants from group assignments however due to the nature of the intervention and lack of sham intervention the success of this blinding is felt unlikely. Ward staff and research staff were not blinded to the participant's assigned intervention. There were no deviations from the intended interventions due to trial context described and although intention to treat analysis not specified given the lack of deviations it is assumed.

No missing outcome data.

The measure is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were not blinded, but it is thought that due to the lack of judgement in the outcome, that the knowledge of the intervention did not influence the outcome.

Analysis reflects plan in trial registration and study protocol, and results are not thought to be from multiple outcome measures or multiple analyses.

The study is judged to be at high risk of bias due to the randomisation process.

Allocation sequence was random (computer‐generated block randomisation), and allocation sequence concealed from the investigators. Participant characteristics were well‐balanced between groups.

Participants and clinicians were aware of the treatment assignments. There was no evidence of deviations from the assigned interventions and an intention to treat approach was used for analysis.

No missing data

The measure is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were blinded.

The analysis is in accordance with a pre‐specified analysis in a published protocol, and results are not thought to be from multiple outcome measures or multiple analyses.

The study is judged to be at low risk of bias for all domains for this outcome.

The method of allocation sequence is not described, but the use of sealed envelopes suggests a random component was used. Participant characteristics were balanced.

Participants and those delivering the interventions were aware of intervention assignments. Twenty‐five patients were excluded due to them not meeting the set eligibility criteria, however, these: protocol violations are not expected to influence the effect estimate of the outcome as per protocol analyses used. The per protocol analyses was not thought to have a substantial impact on the result given the main reason for exclusion was inappropriate recruitment. Excluding this 25, the other 6 exclusions represent approximately 1% of the total sample size.

Data at discharge from hospital complete for 98% of participants.

The measurement of new institutionalisation is considered appropriate, and there were no differences in measurement or ascertainment between groups. Assessors were not blinded, but is thought that due to the lack of judgement in scoring a 'hard outcome' like new institutionalisation, that it is unlikely that the knowledge of the intervention could influence the outcome.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to raise some concerns due to deviations from intended interventions and the selection of the reported results.

Allocation sequence was random (computer generated random numbers) and sequence concealed (opaque sealed envelope). Participant characteristics were balanced.

Participants and those delivering the interventions were aware of intervention assignments. Seventy‐nine participants were not admitted to the unit to which they were assigned. Deviations may have affected the effect estimate, however they were well‐balanced between groups. Intention to treat analysis was used.

No missing data.

The methods of measuring length of stay are considered appropriate, and there were no differences in measurement or ascertainment between groups. Assessors were not blinded, but is thought that due to the lack of judgement in scoring a 'hard outcome' like length of hospital stay, that it is unlikely that the knowledge of the intervention could influence the outcome.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to raise some concerns in the selection of the reported results and deviations from the intended interventions.

“Patients were randomised only when both a treatment and control bed were available”. Although not stated, we believe the necessity for available beds on both wards indicate that a random component was likely used. Sequence allocation concealment is not discussed, but no significant differences in patient characteristics other than those compatible with chance were observed.

Participants and those delivering the interventions were aware of intervention assignments. There were 30 randomisation errors which has been interpreted to mean 'allocated to the ward that they were not randomised to' however there is no explanation of this term. These deviations were thought likely to have affected the outcome, though were relative well‐balanced. It appears a per protocol analyses was used, though given the low number (7%) and relative balance between groups, the omission of the participants randomised incorrectly probably did not have substantial impact on the results.

All participants accounted for at discharge.

The methods of measuring new institutionalisation are considered appropriate, and there were no differences in measurement or ascertainment between groups. Assessors were not blinded, but is thought that due to the lack of judgement in scoring a 'hard outcome' like new institutionalisation, that it is unlikely that the knowledge of the intervention could influence the outcome.

A detailed pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to raise some concerns in deviations from the intended interventions, methods of randomisation and selection of the reported result.

Allocation of wards was random (coin toss) and the allocating officer unaware of study. Baseline differences between groups were thought to be compatible with chance.

Participants and clinicians delivering care were believed to be aware of treatment assignments. There was no evidence of deviations from intended interventions and intention to treat analysis was used.

No missing data.

The measure of new institutionalisation considered appropriate, and there were no differences in measurement or ascertainment between groups. Assessors were not blinded, but is thought that due to the lack of judgement in scoring a 'hard outcome' like new institutionalisation, that it is unlikely that the knowledge of the intervention could influence the outcome.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to raise some concerns in the selection of the reported result.

Pseudo‐randomisation, allocation was based on admitting unit and bed availability, and admitting unit was determined by a rotating roster. No evidence of baseline differences in patient characteristics other than those thought to be compatible with chance.

Both participants and clinicians delivering care were aware of assigned interventions. There is no evidence of deviations from intended interventions and intention to treat analysis used.

No missing data.

The method is considered appropriate, and there were no differences in the measurement or ascertainment between groups. The assessors were blinded.

A pre‐specified statistical plan was not found, but it is not thought that the results were from multiple outcome measures or multiple analyses.

The study is judged to be at high risk of bias due to the randomisation process.