The safety and scientific validity of this study is the responsibility of the study sponsor and investigators. Listing a study does not mean it has been endorsed by the ANZCTR. Before participating in a study, talk to your health care provider and refer to this information for consumers
Trial registered on ANZCTR

Registration number
Ethics application status
Date submitted
Date registered
Date last updated
Type of registration
Prospectively registered

Titles & IDs
Public title
CHERISH Collaborative for Hospitalised Elders: Reducing the Impact of Stays in Hospital
Scientific title
Do older hospital inpatients admitted to acute care wards implementing the "Eat Walk Engage" quality improvement programme have reduced geriatric syndromes, shorter length of stay and greater likelihood of discharge home than older inpatients admitted to control wards?
Secondary ID [1] 286584 0
Universal Trial Number (UTN)
Trial acronym
Linked study record

Health condition
Health condition(s) or problem(s) studied:
delirium 294847 0
functional decline 294848 0
malnutrition 294849 0
Condition category
Condition code
Public Health 295114 295114 0 0
Health service research

Study type
Description of intervention(s) / exposure
"Eat Walk Engage" is a quality improvement programme designed to enhance uptake of evidence-based processes of care for older inpatients. The target processes are early mobility; adequate oral nutritional intake; and meaningful, cognitively stimulating activities. The programme uses enabling facilitation based on the i-PARIHS implementation framework, and engages members of the multidisciplinary team on intervention wards to identify barriers to these processes, trial solutions, and embed successful solutions into practice. Solutions may entail both individual behaviour change and system redesign, and are individualised to the context. The facilitator supports change using methods such as marketing, team building, active reflection, process mapping, audit and feedback and education.
The facilitator is recruited within the study hospital, but resourced, trained and supported by the project team which includes expert facilitators. This includes a 3 day formal training course; fortnightly in-person or telephone support by the expert facilitator team; 6 monthly forum for sharing skills and experiences between sites; and ad hoc telephone support as required. Facilitator contacts will be monitored by the project manager to measure fidelity. The intervention will take place over 12 months. Facilitator roles include monthly 1-hour group meetings between the site facilitator and the ward multidisciplinary team; identification of local barriers and enablers; facilitation of the planning and implementation of small scale improvement cycles as agreed by the team to support the programme aims; identification of delegation tasks suitable for an additional part time health care assistant resourced by the project; audit and feedback of structured process measures related to the improvement aims (patient interviews, activity audits and mealtime audits); and escalation of barriers and solutions if necessary to higher level decision makers. Meeting minutes and field notes by the novice and expert facilitators will be used to monitor intervention fidelity, and 6 monthly structured process audits will measure implementation effectiveness.
Intervention code [1] 291700 0
Comparator / control treatment
Usual ward care, including any facility based improvement programmes
Control group

Primary outcome [1] 294880 0
length of hospital stay: number of days between inpatient admission and discharge from hospital based on hospital record
Timepoint [1] 294880 0
at discharge from hospital
Primary outcome [2] 294881 0
any geriatric syndrome (composite of delirium, functional decline, in-hospital falls, in-hospital pressure injury and new incontinence) based on participant assessment at admission, day 5 and discharge (basic activities of daily living scale, short portable mental status questionnaire, 3D-CAM delirium assessment tool, and recall of incontinence, falls and skin injuries); as well as review of clinical documentation in hospital record
Timepoint [2] 294881 0
at admission, day 5 and at discharge
Secondary outcome [1] 314273 0
death or discharge to higher level of care based on hospital record
Timepoint [1] 314273 0
at hospital discharge
Secondary outcome [2] 314274 0
delirium (3D-CAM and chart review)
Timepoint [2] 314274 0
by day 5 of admission and by discharge
Secondary outcome [3] 314275 0
functional decline (any increase in basic activities of daily living assistance compared to baseline) by patient (or proxy) self-report
Timepoint [3] 314275 0
by day 5, discharge, and 30 days after discharge
Secondary outcome [4] 314276 0
hospital readmission based on state-wide hospital admissions information system
Timepoint [4] 314276 0
within 30 days of discharge
Secondary outcome [5] 314277 0
mortality based on hospital records and telephone follow-up
Timepoint [5] 314277 0
in hospital or within 30 days of discharge
Secondary outcome [6] 314278 0
cost-utility analysis based on EQ5D and hospital costing database
Timepoint [6] 314278 0
30 days after discharge
Secondary outcome [7] 341998 0
Mortality based on statewide death registry data
Timepoint [7] 341998 0
6 months after hospital admission
Secondary outcome [8] 341999 0
Hospital readmissions from state-wide hospital admissions registry
Timepoint [8] 341999 0
6 months from discharge

Key inclusion criteria
Admitted to hospital for 3 or more days, with admission to nominated intervention or control ward
Minimum age
65 Years
Maximum age
No limit
Both males and females
Can healthy volunteers participate?
Key exclusion criteria
Discharged from hospital within 2 days; palliative intent of care

Study design
Purpose of the study
Allocation to intervention
Randomised controlled trial
Procedure for enrolling a subject and allocating the treatment (allocation concealment procedures)
As this is a system-level improvement intervention, the study is a cluster randomised controlled trial, comparing changes measured before and after intervention on 4 randomly selected intervention wards with the changes occurring over the same time period in 4 hospital-matched control wards. Randomisation occurs at the level of ward (each hospital providing a control and intervention ward) and will be undertaken by the off-site study statistician using a random number generator i.e. allocation is concealed. Participant allocation to each ward (and therefore to treatment or control) will occur based on usual hospital processes (diagnosis, treating team on call schedules, and bed availability).
Methods used to generate the sequence in which subjects will be randomised (sequence generation)
Random number generation by off-site study statistician
Masking / blinding
Blinded (masking used)
Who is / are masked / blinded?

The people assessing the outcomes
The people analysing the results/data
Intervention assignment
Other design features
Detailed process of care measure will supplement the outcome measures to help determine the effectiveness of the implementation process at each site
Not Applicable
Type of endpoint/s
Statistical methods / analysis
Detailed statistical analysis plan

Primary comparison: The primary comparison will use post-implementation period data on the 4 wards (general medicine, respiratory medicine, orthopaedic and general surgery) implementing Eat Walk Engage compared to 4 wards (matched for hospital: two general medicine, specialty medicine and general surgery) not implementing the program, controlling for age, gender, Charlson comorbidity score, admission ADL status and admission cognitive status (SPMSQ score) and adjusting for clustering by ward.
Secondary comparison 1: will include data from the pre-implementation cohorts as additional controls, controlling for time period. This analysis will provide greater precision to the estimates from the post-intervention comparison. This approach may induce a bias because of unrelated temporal trends in length of stay over time due to other organisational factors, which could be wrongly attributed to the program.
Sensitivity analysis will use the primary analysis in the post-implementation cohort but include a time since intervention variable to identify whether there is an increasing effect on outcomes over the 6 month post-implementation period as the “dose” of intervention may have been increasing as the model matured. The change over time may be non-linear and therefore we will involve a range of non-linear shapes using the fractional polynomial approach . The best fitting change over time will be estimated using the deviance information criterion (DIC).

Pre-specified subgroup comparisons: will be examined using interaction terms within the primary outcome models and are:
• age under 75 years versus age 75 and older;
• frailty subgroups (less than 0.25 non-frail, 0.25-0.40 mildly frail, 0.40 and above moderately-severely frail) based on a deficit accumulation frailty index;
• the four hospitals.

Primary outcomes:
1. Length of stay (treating unit): described as median time to discharge and median differences between groups; analysed using Bayesian parametric survival analysis
2. Composite outcome of any “hospital associated complication of older people” (HAC-OP) which will consist of:
a. Hospital-associated delirium (delirium documented either by assessment or chart review, first recorded more than 1 day after admission)
b. Hospital-associated functional decline (increase in count of ADL requiring human assistance at discharge compared to 2 weeks prior to admission, by patient self-report; or in-hospital death or new residential care)
c. Hospital-associated incontinence (urinary or faecal incontinence present at discharge which was not present 2 weeks prior to admission, by patient self-report)
d. Hospital-associated pressure ulcer (identified by patient report or chart documentation, not present at admission assessment)
e. Hospital-associated fall (identified by patient report or chart documentation after admission)
This outcome will be modelled as a dichotomous outcome using logistic regression. Each of the five syndromes will be modelled in the same logistic regression model using a mixed model with a random intercept per participant to adjust for correlated data from the same participant. This regression model will estimate the effect of the intervention on the overall syndrome. In a sensitivity analysis we will add an interaction between the intervention and each syndrome to examine whether the intervention had a stronger effect on some syndromes. The models will control for age, gender, comorbidity score, admission ADL status and admission cognitive status (SPMSQ score) and adjust for clustering by ward.

Secondary outcomes:
1. Individual HAC-OP as defined above
2. Death or discharge to institutional care (new residential care, continuing acute, rehabilitation or convalescent care) versus discharge home
3. 30 day functional recovery (return to baseline ADL and IADL status)
4. 30 day all-cause hospital readmission
5. 30 day all-cause mortality
3. Quality of life (EQOL5D) at 30 days
4. 6 month all-cause hospital readmission
5. 6 month all-cause mortality

A “scrambled” analysis (based on simulated intervention groups) will be undertaken and shared with the investigator group for final refinement of methods before commencing full analysis. This aims to reduce the bias of making changes after the full results are available.

Missing data:
The small amount of missing data for ADL and SPMSQ at baseline will be imputed using Multivariate Imputation by Chained Equations (MICE). This is to ensure that the maximum amount of available data are used and to help avoid selection biases caused by participants with partially missing data (e.g., sicker patients being excluded). The variables used by MICE to impute ADL and SPMSQ will be age, IADL at baseline and Charlson comorbidity index.

We will use logistic regression to examine the missing outcome data and see what variables predict missing using treatment group, ward, age, gender, Charlson comorbidity score, admission ADL status and admission cognitive status (SPMSQ score). If strong associations exist we will use inverse-probability weighting to adjust the primary and secondary outcomes to compensate for the non-random missingness. This will be an additional sensitivity analysis.

Recruitment status
Date of first participant enrolment
Date of last participant enrolment
Date of last data collection
Sample size
Accrual to date
Recruitment in Australia
Recruitment state(s)
Recruitment hospital [1] 4079 0
Royal Brisbane & Womens Hospital - Herston
Recruitment hospital [2] 4080 0
Nambour General Hospital - Nambour
Recruitment hospital [3] 4081 0
The Prince Charles Hospital - Chermside
Recruitment hospital [4] 4082 0
Caboolture Hospital - Caboolture

Funding & Sponsors
Funding source category [1] 291702 0
Government body
Name [1] 291702 0
Queensland Department of Science, Information Technology and Innovation (DSITIA)
Country [1] 291702 0
Funding source category [2] 291703 0
Name [2] 291703 0
Queensland University of Technology
Country [2] 291703 0
Funding source category [3] 291704 0
Name [3] 291704 0
Metro North Hospitals and Health Services
Country [3] 291704 0
Funding source category [4] 291705 0
Name [4] 291705 0
Sunshine Coast Hospital and Health Service
Country [4] 291705 0
Funding source category [5] 291706 0
Name [5] 291706 0
Hornsby Ku-Ring-Gai Health Service
Country [5] 291706 0
Primary sponsor type
Professor Alison Mudge
Building C28 Level 1
Roal Brisbane and Women's Hospitals
Herston Queensland 4029
Secondary sponsor category [1] 290381 0
Name [1] 290381 0
Queensland University of Technology
Office of Research
Address [1] 290381 0
Level 4, 88 Musk Ave
Kelvin Grove Qld 4059
Country [1] 290381 0
Other collaborator category [1] 278535 0
Name [1] 278535 0
University of Queensland
Centre for Research in Geriatric Medicine
Address [1] 278535 0
Level 2, Building 33 Princess Alexandra Hospital
Ipswich Rd
Woolloongabba Qld 4102
Country [1] 278535 0
Other collaborator category [2] 278536 0
Name [2] 278536 0
University of Adelaide
School of Nursing
Address [2] 278536 0
Level 3, Eleanor Harrald Bldg
Royal Adelaide Hospital
North Terrace, Adelaide 5000
Country [2] 278536 0
Other collaborator category [3] 278537 0
Name [3] 278537 0
University of Melbourne
Department of Medicine
Address [3] 278537 0
Clinical Sciences Building, Royal Melbourne Hospital
Royal Parade
Parkville Vic 3050
Country [3] 278537 0
Other collaborator category [4] 278538 0
Name [4] 278538 0
University of Sydney
Sydney Medical School
Address [4] 278538 0
Northern Clinical School
Kolling Building, Royal North Shore Hospital
St Leonards NSW 2065
Country [4] 278538 0
Other collaborator category [5] 278539 0
Other Collaborative groups
Name [5] 278539 0
Hebrew Senior Life Centre
Institute for Ageing Research
Address [5] 278539 0
1200 Centre St
Massachussets 02131
Country [5] 278539 0
United States of America

Ethics approval
Ethics application status
Ethics committee name [1] 293225 0
RBWH Human Research Ethics Committee
Ethics committee address [1] 293225 0
Level 7, Block 7
Royal Brisbane and Women's Hospitals
Herston Qld 4029
Ethics committee country [1] 293225 0
Date submitted for ethics approval [1] 293225 0
Approval date [1] 293225 0
Ethics approval number [1] 293225 0
Ethics committee name [2] 293226 0
QUT University Human Research Ethics Committee
Ethics committee address [2] 293226 0
Office of Research Ethics and Integrity
Level 44, 88 Must Avenue
Kelvin Grove Qld 4059
Ethics committee country [2] 293226 0
Date submitted for ethics approval [2] 293226 0
Approval date [2] 293226 0
Ethics approval number [2] 293226 0
Ethics committee name [3] 293227 0
University of Queensland Ethics Committee
Ethics committee address [3] 293227 0
Ethics committee country [3] 293227 0
Date submitted for ethics approval [3] 293227 0
Approval date [3] 293227 0
Ethics approval number [3] 293227 0

Brief summary
Older people (age 65 and older) account for more than half of hospital bed days, and have longer stays and more hospital adverse events that younger people. A hospital stay is often a decisive point in an older person’s health, with hospitalisation accounting for half of newly acquired disability in elders. Geriatric syndromes (including delirium, functional decline, falls, incontinence and pressure injury) result in longer hospitalisations and greater risk of death and institutionalisation. Research clearly shows that “simple” strategies (early mobilisation, adequate oral nutrition, and meaningful cognitive activities) are effective to reduce geriatric syndromes, improve outcomes and reduce costs. While such strategies have been effectively incorporated in specialist “acute care for elders” wards, only a limited number of patients have access to these specialist services. In order to optimise care we need to embed these principles in all acute wards caring for older people. However, this requires systematic changes in acute care staff attitudes, practices and systems of care.

We have piloted a programme of enabling facilitation, based on the i-PARIHS implementation framework, to embed this evidence into practice. The “Eat Walk Engage” programme supports a ward-based multidisciplinary team to identify barriers, trial solutions and embed successful strategies into practice using evidence-based quality improvement methods. In two pilot wards at the Royal Brisbane and Women’s Hospital we have shown promising reductions in length of stay, geriatric syndromes and adverse events accompanying process improvements.

The CHERISH (Collaborative for Hospitalised Elders: Reducing the Impact of Stays in Hospital) study is a cluster randomised controlled trial of the “Eat Walk Engage” programme across 4 sites, and will provide robust evidence of the transferability, scalability, effectiveness and cost-effectiveness of the programme to inform further implementation. Comparing 4 intervention wards with control wards in the same hospitals to account for other sources of variation, we aim to demonstrate a reduction in hospital stay, geriatric syndromes, and discharge to a higher level of care within 12 months of implementing the “Eat Walk Engage” programme. The project is supported by a Queensland Accelerate Partnership Grant from the Department of Science, Information Technology, Innovation and the Arts, administered by Queensland University of Technology.
Trial website
Trial related presentations / publications
Mudge A, McRae P, Cruickshank M. Eat Walk Engage: an interdisciplinary collaborative model to improve care of hospitalised elders. Am J Med Qual 2015; 30 (1):5-13
Mudge AM, Banks MD, Barnett AG, Blackberry I, Graves N, Green T et al. CHERISH (Collaboration for Hospitalised Elders Reducing the Impact of Stays in Hospital): protocol for a multi-site improvement program to reduce geriatric syndromes in older inpatients. BMC Geriatrics 2017; 17:11
Public notes

Principal investigator
Name 56746 0
Prof Alison Mudge
Address 56746 0
Building C28 Level 1
Royal Brisbane and Women's Hospitals
Herston Queensland 4029
Country 56746 0
Phone 56746 0
Fax 56746 0
Email 56746 0
Contact person for public queries
Name 56747 0
Ms Prue McRae
Address 56747 0
Internal Medicine Research Unit, level 6 block 7
Royal Brisbane and Women's Hospital Herston Queensland 4029
Country 56747 0
Phone 56747 0
Fax 56747 0
Email 56747 0
Contact person for scientific queries
Name 56748 0
Prof Alison Mudge
Address 56748 0
Internal Medicine Research Unit, level 6 block 7
Royal Brisbane and Women's Hospital Herston Queensland 4029
Country 56748 0
Phone 56748 0
Fax 56748 0
Email 56748 0

No information has been provided regarding IPD availability

What supporting documents are/will be available?

No Supporting Document Provided

Results publications and other study-related documents

Documents added manually
No documents have been uploaded by study researchers.

Documents added automatically
SourceTitleYear of PublicationDOI
EmbaseEffect of a Ward-Based Program on Hospital-Associated Complications and Length of Stay for Older Inpatients: The Cluster Randomized CHERISH Trial.2022
EmbaseImplementing a ward-based programme to improve care for older inpatients: process evaluation of the cluster randomised CHERISH trial.2023
N.B. These documents automatically identified may not have been verified by the study sponsor.