Introduction

Older Australians who require admission to hospital are at risk of functional decline [1]. Post a hospital discharge, many older adults require engagement in a community-based rehabilitation program to restore or optimize their function within their home and community environments [2]. Established in 2005, the Transition Care Programmed (TCP) is one such Australian multi-disciplinary post–hospital service for older adults who would benefit from short-term slow-stream rehabilitation [3]. The TCP is a national initiative jointly funded by the Australian Federal and State and Territory governments, with the Australian Federal government providing program policy oversight while States and Territories operationalizing the program [3]. Australia’s TCP model of care is similar to those delivered under other international health care authorities such as the United Kingdom’s National Health Service Intermediate Care and Rabblement programs, Canada’s regional health authorities restorative care units, and early supported discharge under the governance of New Zealand’s District Health Boards [4]. As of 2019, there was 4060 TCP packages operational in Australia [5].

The TCP aims to provide a goal-centered community-based service that assists to optimize the daily functioning of older people to reduce premature entry into residential aged care facilities [3]. The TCP provides both a home-based slow-stream rehabilitation to clients and/or a residential-based program for clients who cannot directly transition home from hospital. To access a TCP package, an older person must be assessed and deemed eligible for the program by a health professional from the Aged Care Assessment Team (ACAT). This assessment must occur whilst a person is admitted to a public or private hospital/virtual ward and medically stable. As per the National TCP Guidelines, the program is limited to 12 weeks, however an extension of a further 6 weeks may be granted by ACAT under extenuating circumstances [3]. A case management framework is utilized to deliver this flexible care and rehabilitation program; clients may access nursing and personal care, and low intensity therapy to optimize their restorative process [3]. As per the TCP Guidelines and dependent on their individualized care plan, clients have access to health professions such as occupational therapy, physiotherapy, speech pathology, dietetics, podiatry, nursing, and social work. They may also be able to access equipment hire, continence aids, and domestic and personal care assistance which may include community access assistance to complete shopping or attend appointments. As the TCP is a flexible care package, each client’s input is individualized to their needs and their goals for the program [3]. Mandatory information and consumer outcome measures reported to government include the Modified Barthel Index (MBI) at entry to and exit dates from TCP, hospital length of stay (LOS), TCP LOS, TCP discharge destination [3]. Goal-type and achievement information is not reported to government funding bodies but may be collected as part of local processes.

The types of goals that older adults set post hospital admission can be categorized according to the International Classification of Functioning, Disability and Health Framework [6]. Goals related to meaningful daily activities and engagement in these activities has been demonstrated to have a significant impact on health outcomes for older adults [7].  Goal achievement in activities of daily living prevents hospital re-admission, defers residential care, increases functional capacity, and improves heath related quality of life [2, 6, 8-11]. The addition of goal setting to a community-based exercise group has been shown to decrease frailty and increase quality of life for older adults, potentially reducing the risk of needing facility-based care [10].  Additionally, Comas, Peel [9] found that older people continue to recover with ongoing improvement in quality of life for at least six months post completion of the program. There is evidence to support the efficacy of the TCP in enabling clients to achieve or partially achieve a range of occupational goals [6].

To facilitate improved functional capacity, individualized and collaborative goal setting underpins the multidisciplinary team intervention provided the TCP clients.  Clients may choose to work towards achieving a myriad of goals which can be classified as either in-home or community-based goals. Anecdotally, it is thought that clients who develop community-based goals improve their overall functioning when compared to those who do not, however there is little research on this in the literature. It is not currently known whether the TCP clients who do and do not identify community-based goals are a different population with regards to their baseline characteristics. Despite goal setting being a critical component of the clinical rehabilitation process, little is known about differences in cohorts who create in-home goals only and those who create in-home and community-based goals. This information is critical because it may dictate how clients with different goal types may be better managed by TCP. The aim of this study was to understand the differences between consumer cohorts who develop home-based goals only and those who develop at least one community-based goal in community TCP in an outer metropolitan area of an Australian capital city.

Materials and Methods

Study design

This retrospective cohort study utilized information from a local database of the Transition Care Program at an outer metropolitan hospital site (‘Transdata’) and patient integrated electronic medical records (ieMR).

Setting

This single-site study was conducted at a medium sized metropolitan Australian Hospital and involved clients admitted to the local TCP from 1st July 2014 to 31st December 2019.

The health service is a part of a larger capital city metropolitan area and services the local suburban and coastal communities. As of 2020, the local population numbered 160,331 with approximately 19% of people aged 65 years or older with an annualized growth rate of 3.6% for people aged 70 years and above, representing the largest growth across all ages groups [12].

Between 2014 and 2019, the TCP team servicing this community comprised of nurses, personal care workers, occupational therapists, physiotherapists, and allied health assistants. Staff provided both clinical and case management services to program participants. A brokerage model was utilized to support TCP to ensure required personal care, nursing and allied health services were provided to participants based on individual needs. This TCP site services 35 community TCP packages and 15 residential TCP packages located within a local residential aged care facility. During the study period, this TCP site reported to have provided services to an average of 325 participants per year.

Participants

Participants were eligible to be included in this study if they were admitted to either residential or community TCP and discharged from the community TCP in the study period (e.g. participants must not have transferred to another TCP, sent back to hospital and discharged from there, or died during TCP admission).

Dataset

‘Transdata’ is a local database that is used to record information about clients and their goals when they are admitted to TCP. This database was developed in 2014 and is maintained by TCP staff including nursing and allied health. It is part of well ingrained internal processes to utilize this database as part of the admission and discharge process from TCP. Two researchers (NB and EMc) reviewed the database together and cleaned data where appropriate. NB and EMc reviewed and reclassified goals as either community or home goals, where there was uncertainty related to how a goal should be classified a third researcher (SS) was brought in to review and assist with decision making. 

Measurements

Variables captured included:

• Demographic and health data (age, gender, living arrangements, primary diagnosis, number of comorbidities on admission to TCP)
• Details on hospital admission (length of stay in hospital prior to TCP, readmission to hospital while admitted to TCP, number of admissions to hospital at 3- and 6-months post TCP discharge),
• Details on TCP admission (reason for admission, length of stay in TCP)
• Details on goals (types of goals – home or community, number of goals, goals attained),
• Details on independence (from Modified Barthel Index (MBI)) at admission and discharge. MBI is a 100-point rating scale of a patient’s ability to perform ten kinds of activities of daily living. A lower score indicates less independence, while higher scores indicate greater independence [13].

For the logistic regression, primary diagnosis were grouped into three categories; orthopedic (admissions where related to orthopedic surgery), general surgery (admissions related to any other surgery), and general medical (admissions related to medical issues).   

Goals

Goals were categorized in the ‘Transdata’ database according to the International Classification of Functioning, Disability and Health Framework. The research team classified a home goal as anything that happened within the confines of the home boundary, whereas a community goal was anything that happened outside the home boundary. Using this definition, similar goal types could be classified as either a home or community goal. For example, ‘to walk around the house’ would be a home goal, while ‘to walk around the neighborhood’ would be a community goal.

Participants were divided into two groups, those with ‘home goals’ only and those with ‘at least one community goal’. Participants who had at least one community goal, usually also made home goals.

On discharge, goals were classified as ‘did not attain’, ‘partially attained’, or ‘fully attained’

Sample size

Sample size was based on the current admissions to TCP over the study period that were available in the Transdata database.

Statistical analysis

Data were collected in Excel (Microsoft) and analyzed using SPSS (version 28). A p-value was considered significant if it was less than 0.05. Data was descriptively summarized depending on the type of data. For continuous variables median (IQR) was used to describe non-normally distributed data while mean (sd) was used for normally distributed data.  

 For comparing across groups (community-based goals vs. home-goals only), categorical data was compared using Chisquare tests of independence (χ2). For comparing continuous data across groups Man-Whitney U tests or Independent T-tests will be used depending on the data ability to meet test assumptions.

Logistic regression models were used to determine predictors of nominating a community goal. The dependent variables were dichotomized to at least one community goal or home goals only. Independent variables were age, gender, hospital length of stay, number of goals, MBI on admission to TCP, whether the patient had a carer, number of comorbidities, and reason for admission (surgical, orthopedic, or medical). The area under receiver operating characteristics (AUC) was calculated for the final models to examine the predictive ability of each model, which ranged from 0.0 to 1.0, where 1 indicates perfect predictability of a model [14].

Results

Of the 1057 TCP episodes, 1051 (99.3%) had goals while 6 (0.7%) had no goals. Of the 1051 who had goals, 694 (66%) had at least one community goal while the remaining 363 (34%) made home goals only. The characteristics of individuals with home goals only or at least one community goal are compared in (Table 1).

Table 1: Characteristics of those individuals with home goals only and those with at least one community goals.

The functional outcomes between the two groups were significantly different (Table 2).

Table 2: Functional information of individuals who had home goals only and those who had at least one community goal

There was a total of 4117 valid goals in the database, 1092 (26%) of these were community-based goals, while 3025 (74%) were home-based goals. (Table 3) outlines the types of goals that were made and their classifications. Supplementary information 1 provides examples of home and community goals across the goal categories.

Table 3: Number of home and community goals that were attained

The differences of the types of goals and goals achieved across those who have home goals only and those with at least once community goal are described in (Table 4).

Table 4: Types of goals and goal attainment compared between Individuals with home goals only and those with at least one community goal.

The multivariable logistic regression model for predicting having at least one community goal (Table 5) found that significant independent factors were age, number of goals nominated, and MBI on admission to TCP. The predictive ability of the model, calculated by AUC was 0.8, indicating good predictability.

Table 5: Predictors for those who make at least one community goal

Discussion

This study found there were differences between those who make at least one community goal and those who make only home goals as part of their TCP rehabilitation. Between these two groups differences exist between the older adults, their functionality, and the categories of goals that they make. The regression model found that number of goals, age on admission, and MBI on admission were significant predictors of whether a patient would make community goals. Despite the age, number of comorbidities, and gender of participants being equally represented in both groups, there were differences of length of stay in hospital, functionality, and discharge location.

There were some differences in the types of home-goals compared to previous research into this topic. The three most common home goals in our study were: 1) mobility inside the home environment, 2) self-care (e.g. showering, dressing), and 3) domestic life (e.g. housework and meal preparation).  This differed to Rietkerk, Uittenbroek [15] findings where community dwelling older adults were less likely to develop self-care goals compared with physical health, mobility, or support goals. Waldersen, Wolff [16] also found most common goals developed by low-income community dwelling older adults were related to physical function (transferring, changing, or maintaining body position, and star climbing). This difference in goal types may be due to the acuity of the TCP client’s functional decline and the local aged-care services context where formal services for self-care support may be difficult to obtain, meaning that returning to independence with self-cares can be imperative to remain residing in the community. The three most common community goals were related to: 1) community life (e.g., shopping or returning to work), 2) mobility outside the home environment, and 3) recreation and leisure (e.g. activities at the local leisure centre). Mobility, either inside or outside the home environment was a common goal-type for both cohorts.

There were also interesting differences between the types of goals the two groups of older adults made. Those with community goals were more likely to have a mobility goal and a domestic life goal and more likely to achieve these goals. This is likely a reflection of the better mobility needed to navigate the community or take on additional tasks at home, and potential higher baseline function of those with community goals. Individuals with home goals only were more likely to have a ‘return or remain home’ goal but were less likely to achieve this goal than their counterparts who had community goals. People with community goals were also more likely to have recreation and leisure goals, and more likely to achieve these goals. This raises some concerns about the recreation and leisure activities that those with home goals only can or do undertake if they have less interaction with their community. In a small sample study, Athan, Bissett [17] found TCP clients may be influenced by and reliant on family members, friends, or community services to assist to engage the person in communitybased activities. Perhaps the provision of education regarding their potential influence on older adults’ leisure goals to family members, friends or community-based services by TCP clinicians may assist to improve community interaction for this cohort of older adults. 

Overall, goal attainment was higher for people who made at least one community goal.  Smith & Fields [18] found that those who develop community/recreation/leisure/play related goals, impairment outcome measures were more likely to remain unchanged, however this study was small in comparison with only 43 participants developing community goals.  Waldersen, Wolff [16] study found that goal attainment was not associated with age, sex, education, depressive symptoms, or health related quality of life measures, but was negatively associated with severe pain. The proportion of goals attained in this study was higher (83%, 3423/4117) than the study by Waldersen, Wolff [16] which was 73%.

There were several goal categories that there was no difference across. These included emotional functions, wound and skin management, self-care, pain management, weight management, and communication. Despite low numbers of these types of goals it was surprising that they were not more frequently attributed to individuals with home only goals. However, a larger sample size may be required to detect changes between these types of goals. Rietkerk, Uittenbroek [15] study investigated the impact of pain on goal attainment and found that the reported presence of severe pain significantly impacted a participant’s goal attainment. This finding was reflected in this study as overall goal achievement was 74% of goals (n= 836) with those who identified pain as a health concern (n=68) achieved 68% of their goals. 

Interestingly, those with home goals had a larger percentage increase in their MBI score on discharge from TCP compared to those with community and home goals. While this was a statistically significant difference, the clinical significance of this is unclear. The functional status on discharge of patients who had home goals was similar to that of patients who had community goals (MBIs of 90 and 95 respectively). Despite also being statistically significant, it is unclear if this difference is clinically significant, as this would indicate that regardless of the goals that patients make, they may discharge from TCP with similar functional ability. However, given that MBI measures activities of daily living this is likely somewhat expected that a ceiling effect regarding higher levels of function would occur within this data. Smith & Fields [18] studied the impact of TCP on functional outcomes and found that although the MBI is the outcome measure utilized for government reporting, the MBI does not suitably measure activity, participation, or psychosocial domains. A key recommendation of a government commissioned report by KPMG in 2019, was that the use of the MBI as a sole functional outcome measure does not measure the true value of the TCPs impact on clients [4]. According to Yang et al [19]. When using the MBI, a percentage change in MBI is more likely to reflect clinically meaningful change than a change in raw numbers.  Yang, Wang [19] proposed that a change of at least 19% in MBI score was required to reflect actual change in functionality. Using this cut-off for our data, it could be deduced that those with home goals had a change in functionality while those who had community goals did not have a clinically significant change in their functionality. 

Unsurprisingly, a higher proportion of people with at least one community goal were discharged back to the community with support, while a higher proportion of people with home goals only were discharged to an aged care facility. This could be due to differences in baseline function seen in this study with MBI score on admission to TCP being higher for those patients with community goals. This incongruence in findings likely provides further justification regarding the limitations of the MBI in measuring clinically significant changes in functionality.

Length of stay in hospital was longer for those with both at least one community goal. This could be due to patients with community goals having higher baseline levels of function prior to admission but experiencing a medical issue (whether acute or chronic) requiring a longer period of rehabilitation to return to their usual baseline (which would see them achieving their home goals) before moving on to rehab and starting to incorporate community goals. This is somewhat reinforced by those with community goals having more goals than those with just home goals. This likely indicates their need to have some goals but wanting to return to pre-hospital admission community activities. This contrasts with an Australian study by Cations, Lang [20] who found improvement in MBI scores in TCP participants was more common in younger age, less frail, women, and those with shorter hospital stay prior to TCP admission. Unfortunately baseline function prior to TCP admission was not collected as part of the maintenance of this database. Additionally given the codes collected for TCP admission, it was impossible to tell between those admitted with an acute condition (e.g., new cerebrovascular event) or a chronic condition (exacerbation of heart failure and deconditioning). It is possible those with an acute condition that precipitated their admission to TCP may be more likely to create community goals, if that is what they were accustomed to prior to admission.

Limitations

As with all research, this project has several limitations. Firstly, the coding of goals is reliant on the discretion of the clinician entering this information with some goals being hard to separate resulting in overlap between some groups. For example, ‘attending lawn bowls with friends’ would be coded as both ‘community life’ and ‘recreation and leisure’. Depending on the factor that was inhibiting lawn bowl attendance (not recorded in database) there may also be several underlying goals associated with each goal, for lawn bowls attending lawn bowls may also be reliant on mobility, or on financial ability to pay for lawn bowls.

A further limitation is that functional assessment tools used in this research (MBI) are solely focused on home life. Therefore, trying to use the MBI to reflect improvements in community life goals is not possible. Additionally, TCP programs have been rolled out in different ways across hospital and health services in Australia, therefore extrapolation of results may require consideration of the local context. However this study’s results reflect the Queensland experience as per the Australian Government Productivity Commission Queensland Aged Care Services data for from 2014 to 2019 which demonstrated average length of stay in TCP as 55 days, similar MBI on entry (73 for Queensland data), but lower MBI on discharge (84 for Queensland data) [5].

Finally, this project measured goal attainment and functionality after a set period of time (e.g. at the end of the program) and does not take into account goals which may have been achieved after the end of the rehabilitation program. Comans, Peel [9] found that older person receiving a post-discharge program continued to experience improvements in quality of life at least 6 months after the program ended. While goals in this study were not measured, it is likely that goals would also be achieved if quality of life improved. Yuri, Takabatake [10] found a similar post discharge effect, although their study found the positive effects ceased at 9 months post program.

Conclusion

Older adults who develop community-based goals are more likely have a higher MBI score upon admission and discharge from TCP and are also more likely to remain living in the community. However those who develop home-based goals only have a larger change in this functional outcome measure during TCP despite being more likely to discharge to a residential aged care facility compared to those who develop at least one communitybased goal. Encouraging individuals to make goals is an important aspect of practice and further research is warranted to investigate if setting community-based goals with all older adults will further improve outcomes for those who are more likely to set home-based goals only. Further research is also required to better understand the impact that the change in baseline functioning prior to TCP admission has on the predictors of positive functional outcomes particularly as the reliance on measures such as the MBI for individual and service outcomes is unlikely to truly reflect the clinical significance of this positive impact.

Acknowledgements

The authors would like to thank Wendy Marshall for assisting with provision of the data set utilized for this paper.

Ethical Guidelines: This project was approved by the local research ethics board, approval reference number HREC/2020/ QMD/63442.

Data Availability Statement: The authors are unable to share the data set utilized for this study as the researcher has not completed planned or expected analyses for future publications.

References

1. Geyskens L, Jeuris A, Deschodt M, Van Grootven B, Gielen E, et al. (2022) Patient-related risk factors for in-hospital functional decline in older adults: A systematic review and meta-analysis. Age Ageing. 51.
2. Gray LC, Peel NM, Crotty M, Kurrle SE, Giles LC, et al. (2012) How effective are programs at managing transition from hospital to home? A case study of the Australian transition care program. BMC Geriatrics. 12:6.
3. Department of Health and Aged Care. (2022) Transition Care Programme Guidelines. In: Department of Health and Aged Care, editor.
4. KPMG. (2019) Review of the Transition Care Programme.
5. Productivity Commission. Report on Government Services 2023: PART F, SECTION 14. In: Productivity Commission, editor.: Australian Government; 2023.
6. Salih SA, Peel NM, Marshall W. (2015) Using the International Classification of Functioning, Disability and Health framework to categorise goals and assess goal attainment for transition care clients. Australasian Journal on Ageing. 34:E13-E6.
7. Matuska K, Giles-Heinz A, Flinn N, Neighbor M, Bass-Haugen J. (2003) Outcomes of a pilot occupational therapy wellness program for older adults. The American Journal of Occupational Therapy. 57:2204.
8. Comans TA, Peel NM, Cameron ID, Gray L, Scuffham PA. (2015) Healthcare resource use in patients of the Australian Transition Care Program. Aust Health Rev. 39:411-6.
9. Comans TA, Peel NM, Gray LC, Scuffham PA. (2013) Quality of life of older frail persons receiving a post-discharge program. Health Qual Life Outcomes. 11:58.
10. Yuri Y, Takabatake S, Nishikawa T, Oka M, Fujiwara T. (2016) The effects of a life goal-setting technique in a preventive care program for frail community-dwelling older people: a cluster nonrandomized controlled trial. BMC Geriatrics. 16:101.
11. Parsons J, Rouse P, Robinson EM, Sheridan N, Connolly MJ. (2012) Goal setting as a feature of homecare services for older people: does it make a difference? Age and ageing. 41:24-9.
12. Region summary: Redland [Internet]. Commonwealth of Australia. 2022.
13. Lee SY, Kim DY, Sohn MK, Lee J, Lee SG, et al. (2020) Determining the cut-off score for the Modified Barthel Index and the Modified Rankin Scale for assessment of functional independence and residual disability after stroke. PLoS One. 15:e0226324.
14. Hanley JA, McNeil BJ. (1983) A method of comparing the areas under receiver operating characteristic curves derived from the same cases. Radiology. 148:839-43.
15. Rietkerk W, Uittenbroek RJ, Gerritsen DL, Slaets JPJ, Zuidema SU, et al. (2021) Goal planning in person-centred care supports older adults receiving case management to attain their health-related goals. Disability and rehabilitation. 43:1682-91.
16. Waldersen BW, Wolff JL, Roberts L, Bridges AE, Gitlin LN, et al. (2017) Functional goals and predictors of their attainment in low-income community-dwelling older adults. Archives of physical medicine and rehabilitation. 98:896-903.
17. Athan V, Bissett M, Boland N, Owen K, Collins J. (2022) Engagement of Older Adults in out-of-Home Occupations: Transitioning from Hospital to Home. Physical & Occupational Therapy In Geriatrics. 40:94-111.
18. Smith HN, Fields SM. (2020) Changes in older adults’ impairment, activity, participation and wellbeing as measured by the AusTOMs following participation in a Transition Care Program. Australian Occupational Therapy Journal. 67:517-27.
19. Yang CM, Wang YC, Lee CH, Chen MH, Hsieh CL. (2022) A comparison of test-retest reliability and random measurement error of the Barthel Index and modified Barthel Index in patients with chronic stroke. Disability and rehabilitation. 44:2099-103.
20. Cations M, Lang C, Crotty M, Wesselingh S, Whitehead C, et al. (2020) Factors associated with success in transition care services among older people in Australia. BMC geriatrics. 20:1-10.