Study selection
Our search revealed 3271 publications (Figure 1), of which 23 publications met our inclusion criteria. These publications referred to eight RCTs and 11 qualitative studies.
Study characteristics
Eight RCTs with 1072 participants were included [2, 4, 8, 23–30]. The size of the studies ranged from ten to 400 participants. The cohort study alongside one RCT was not considered in the analysis [8].
Eleven qualitative studies with a total of 346 participants (range: n=4 [36] to n=122 [40]) were included. All the studies used interview techniques, one used participant observation additionally [37], and another one used a semistructured survey [40]. Three studies explored perceptions of patients, therapists and older adults in community dwellings with regard to the use of virtual reality (VR) applications in PDHA [10, 33, 34]. Three studies explored factors considered by OTs when deciding about stroke patients’ need for a predischarge home assessment visit, as well as clinical reasoning and practice of PDHA [34, 38–40]. One study focused on older adults’ and carers’ perception of and involvement in PDHA decision-making processes [32]. Another study also highlighted the patients’ perspective on PDHA [36]. A summary of characteristics of the included quantitative and qualitative studies is displayed in Table 1.
Table 1: Study characteristics of included studies
Reference
|
Study
design,
country,
setting
|
Participants
Number,
age in years, percent female (♀%)
|
Intervention description
|
Outcomes/aim of qualitative research
|
Outcome measures
|
FU
in mo
nths
|
Clemson et al., 2016 [2]
|
RCT
Australia
Acute care, unspecified
|
n=400
Intervention 80.2 (±6.4, range n.r.)
♀ 59.6% Control
80.7 (± 5.7, range n.r.) ♀ 63.9%
|
Intervention
In-hospital rapport building, interview, ADL-assessment, predischarge home visit, post-discharge home-visit, telephone calls
Control
Usual care, in-hospital interview, ADL assessment, access visit if required
|
Functional independence, participation in ADL, unplanned readmissions, emergency department visits, recommendations
|
NEADL [41], Late Life Disability Index (LLDI) - sub scores: frequency and limitation [60], number of: recommendations, unplanned readmissions, emergency department visits, falls, process outcomes (e.g., number of prescribed and tried equipment; effects not estimated)
|
3
|
Drummond et al., 2013 [8]
|
RCT
UK
Stroke rehabilitation unit
|
n=126 Intervention
70.64 (± 14.29, range 34-88)
♀ 54.7%
Control
73.65 (± 16.06, range 41-99) ♀ 47.8%
|
Intervention
One or two predischarge home visits
Control
Structured home assessment interview
|
ADL/IADL, mobility, unplanned readmissions, falls, emotional distress in medical settings, depressed mood of clients with stroke and significant aphasia, caregiver strain
|
NEADL [41], Barthel Index [61], RMI [46], Number of unplanned readmissions, GHQ-28 [49], SADQ-10 [51], Caregiver Strain Index [52]
|
1
|
Hagsten et al., 2004 [23] Hagsten et al., 2006 [24]
|
RCT
Sweden
Acute care, hip fractures
|
n=100 Intervention
81 (± 23, range 68-91)
♀ 84% Control
79 (± 30, range 65-95) ♀ 76%
|
Intervention Individual daily training, including use of technical aids, single predischarge home visit
Control
One session of walking instruction when in hospital
|
ADL/IADL, health-related quality of life
|
ADL [62], EQ5D [63], IADL single scales: moving around indoors; performance of light housework; getting in and out of a car, SWED-QUAL [43] subscales
|
2
|
Lannin et al., 2007 [25]
|
RCT
Australia
Rehabilitation unit, mixed (cardiac, orthopedic trauma, neurological, orthopedic joint surgery, spinal or deconditioned)
|
n=10
Intervention
80.0 (± 7, range n.r.)
♀ 100%
Control
82.4 (± 7, range n.r.) ♀ 60%
|
Intervention
Single predischarge home visit Control
In-hospital consultation prior to discharge
|
ADL/IADL, mobility, unplanned readmissions, falls, fear of falling, community support, health-related quality of life
|
NEADL [41], FIM [64], RNLI - Reintegration to Normal Living Index [65], Tinnetti [45], number of unplanned readmissions, number of falls, FES-I [47], EQ5D [63], EQ-5D VAS
[44]
|
7
|
Lockwood et al., 2019
[26]
|
RCT
Australia
Acute and rehabilitation (cardiac, orthopaedic trauma, neurological,
orthopaedic joint surgery, spinal or deconditioned)
|
n=77
Intervention
83.4 (± 7.1, range n.r.)
♀ 76%
Control
80.9 (± 7.3, range n.r.) ♀ 68%
|
Intervention
Single predischarge home visit and usual care
Control
usual care
|
ADL/IADL, health-related quality of life, unplanned readmissions, number of days in hospital after index discharge, falls, fear of falling, adverse events
|
NEADL [41], SMAF [42], FIM [66], FES-I [47], EQ5D[63], EQ-5D VAS [44]
|
6
|
Lockwood et al., 2020
[27]
£
|
RCT, process evaluation
Australia
Acute and rehabilitation (cardiac, orthopaedic trauma, neurological,
orthopaedic joint surgery, spinal or deconditioned)
|
n=77
Intervention
83.4 (± 7.1, range n.r.)
♀ 76%
Control
80.9 (± 7.3, range n.r.)
♀ 68%
|
Intervention
Single predischarge home visit and usual care
Control
usual care
|
Recommendations, adherence to recommendations
|
Number of recommendations, Number of adherence to recommendations
|
1
|
Nikolaus et al., 2003 [9]
|
RCT
Germany
Geriatric acute care, unspecified
|
n=360 Intervention
81.2
(±6.2, range 84.9-87.5)
♀ 72.4%
Control 81.9
(±6.5, range 74.4- 88.4) ♀ 74.3%
|
Intervention Predischarge home visit and post-discharge follow-up visit(s), comprehensive in-hospital geriatric assessment
Control
Comprehensive geriatric assessment and usual care
|
Falls, recommendations
|
Number of falls, compliance with recommendations after 12 months
|
12
|
Pardessus et al., 2002 [28]
|
RCT
France
Geriatric acute care, unspecified
|
n=60 Intervention
83.51 (± 9.08, range n.r.) ♀ 76%
Control
82.9 (± 6.33, range n.r.) ♀ 80%
|
Intervention
Predischarge home visit
Control
Usual care
|
ADL/IADL,
Falls, re-hospitalization, institutionalization
|
IADL [67], SMAF subscales [42] ADL subscales [68] number of recurring falls, mean number of fall recurrence in former fallers, number of re-hospitalizations, number of institutionalizations
|
12
|
Provencher et al., 2020 [29] §
|
RCT, Post-hoc analysis
Australia
Acute care, unspecified
|
n=400
Intervention 80.2 (±6.4, range n.r.)
♀ 59.6%
Control
80.7 (± 5.7, range n.r.)
♀ 63.9%
|
Intervention
In-hospital rapport building, interview, ADL-assessment, predischarge home visit, post-discharge home visit, telephone calls
Control
Usual care, in-hospital interview, ADL assessment, access visit if required
|
Functional independence,
participation in ADL, unplanned readmissions, emergency department visits, recommendations
|
NEADL [34], Late Life Disability Index (LLDI) - sub scores: frequency and limitation [52],
number of: recommendations, unplanned readmissions, emergency department visits, falls, process outcomes (e.g., number of prescribed and tried equipment; effects not estimated)
|
3
|
Threapleton et al., 2018 [4]
|
RCT
UK
Stroke ward, acute care
|
n=16 Intervention 72 (±21.08, range 38-90)
♀ 75% Control
70 (± 12.6, range 46-86) ♀ 37%
|
Intervention Single predischarge virtual home assessment
Control
Usual care
|
ADL/IADL, overall independence,
mobility,
fear of falling,
health-related quality of life
|
NEADL [41], Barthel- Index [61], MRS [69], Rivermead Mobility Index [46], FES-I [47], EQ5D [63]
|
6
|
Wales et al., 2018 [30] §
|
RCT, economic evaluation
Australia
Acute care, unspecified
|
n=400
Intervention 80.2 (±6.4, range n.r.)
♀ 59.6% Control
80.7 (± 5.7, range n.r.) ♀ 63.9%
|
Intervention
In-hospital rapport building, interview, ADL-assessment, predischarge home visit, post-discharge home visit, telephone calls
Control
Usual care, in-hospital interview, ADL assessment, access visit if required
|
Costs for predischarge home visits
|
costs for occupational therapy
time, travel, community follow up, hospital readmission
|
3
|
Atwal et al., 2008 [32]
|
Semi-structured interview
UK
Geriatric acute care
|
Patients, main carers n = 15 86,46 years
(range 73-97) ♀ 60%
|
Intervention
Single predischarge home visit
|
- To explore older adults’ and carers’ involvement in decisions that were made during the home visit;
- To explore older adults´ and carers' perceptions of the home visit process
|
n.a.
|
n.a.
|
Atwal et al., 2014a [33]
|
Semi-structured interview; think aloud technique
UK
Acute care and community care
|
OTs
n=7 ♀ 71%
social services, older persons, mental health, acute care, pediatrics
|
Intervention
Virtual reality predischarge home assessment with interior design application
|
- To explore occupational therapists’ perceptions of a virtual reality interior design application (VRIDA);
- to gain insights into the feasibility of using VRIDA as a tool to aid the predischarge home visit (perceived usefulness, perceived ease of use, actual use)
|
n.a.
|
n.a.
|
Atwal et al., 2014b [34]
|
Semi-structured interview
UK
Acute care, intermediate care, rehabilitation,
older patients, mental health (older people)
|
OTs n = 21
|
Intervention
Predischarge home visit / access visit
|
- To explore occupational therapists' perceptions of home visits;
- To ascertain occupational therapists’ clinical reasoning with respect to conducting home visits
|
n.a.
|
n.a.
|
Cameron et al., 2014 [35]
|
In-depth interview, semi-structured, focus groups
Canada
Rehabilitation facility, stroke
|
Patients
n = 16 62 years
(range 25 - 87) ♀ 75%
Family caregivers n=15 41 years
(range 23 - 75) ♀ 86,7%
Multiple health professionals
n=20
|
Intervention
Single predischarge home visit or preparation in hospital and single/multiple predischarge weekend passes
|
- To explore stroke survivors’, caregivers’, and healthcare professionals’ perceptions of weekend passes offered during inpatient rehabilitation and its role in facilitating the transition home
|
n.a.
|
n.a.
|
Davis et al., 2019 [40]
|
Semi-structured survey
Republic of Ireland
acute settings, rehabilitation settings and convalescence settings, adult patients (over 18 years)
|
OTs
n=122
|
Intervention
Pre-discharge home visit
|
- To investigate clinical practice during DPHV and the clinical reasoning guiding occupational therapists within an Irish context
|
n.a. for semistructured part
quantitative part of survey: use of standardized tool, contents of home visit bag, numbers of recommendations, consensus on clinical practice
|
n.a.
|
Godfrey et al., 2019 [39]
|
Focus-group interviews
Australia
Acute or sub-acute settings from three facilities
|
OTs
n=19
Multidisciplinary stakeholders
n=8
|
Intervention
Pre-discharge home visit
|
- To understand both occupational therapists’ and multidisciplinary stakeholders’ perceptions and contemporary practice regarding decision-making and pre-discharge home visits through exploration of experience and current practice in the Australian context. Investigation of factors associated with when, how and to whom pre-discharge home visits are provided
|
n.a.
|
|
Hibberd, 2008 [36]
|
Semi-structured interview
UK
Intermediate care unit
|
Patients
n=4 65 years and older ♀ 50%
|
Intervention
Predischarge home visit / access visit
|
Part of an evaluation study;
- To gain patient perspectives on home visiting process -to ensure service meets needs
|
n.a.
|
n.a.
|
Money et al., 2015 [10]
|
Semi-structured interviews, thinking aloud
UK
Community dwelling
|
Community dwelling older people
n=10 56-80 years
♀ 50%
|
Intervention
Virtual reality predischarge home assessment with interior design application
|
- To explore community-dwelling, older adults’ perceptions of using a computerized 3D interior design application (perceived usefulness, ease of use, and actual use)
- To consider the potential barriers and opportunities of using CIDA as an assistive tool within the predischarge home visits process
|
n.a.
|
n.a.
|
Nygard et al., 2004 [37]
|
Interviews, focus groups, participant observation
Sweden
Geriatric acute care, mixed diagnoses
|
Patients
n=23 78 years
(range 68-86) ♀ 50%
Living alone
n=12
OTs
n=9
|
Intervention
Single predischarge home visit
|
- To describe and illustrate, from both clients’ and therapists’ perspectives, the occupational therapy interventions and recommendations that were undertaken and followed-up in common practice during predischarge home visits;
- To gain insight in the accuracy of expectations of therapists and in perceived usefulness of predischarge home visits to clients
|
n.a.
|
n.a.
|
Threapleton et al., 2017 [3]
|
Semi-structured interview
UK
Acute care, rehabilitation, community,
stroke
|
Patients
n=8
68 years
(range 44-92)
♀ 75%
Stroke survivors n=4 70 years
(range 61-79) ♀ 75% OTs n=13
|
Intervention
Virtual predischarge home visit
|
- To explore perceptions concerning the acceptability, potential utility and limitations of the use of a virtual reality interior design application from the perspectives of therapists and patients
|
n.a.
|
n.a.
|
Whitehead et al., 2014 [38]
|
Semi-structured interview
UK
Acute, rehabilitation, mixed, hyper acute, stroke
|
OTs
n=20
|
Intervention
Predischarge home assessment visits
|
- To explore what factors occupational therapists consider when deciding which patients with a stroke need a predischarge home assessment visit
|
n.a.
|
n.a.
|
FU: latest time point of follow-up; n.a.: not applicable; § refers to the original RCT of Clemson et al., 2016 [2]; £ refers to the original RCT of Lockwood et al., 2019 [26]
|
Setting and participants
The studies were published between 2002 and 2020, and the majority were conducted in the UK [3, 4, 8, 10, 32–34, 36, 38, 40] and Australia [2, 25–27, 29–31, 39]. One study each was carried out in Germany [9], France [28] and Canada [35], and two studies were conducted in Sweden [23, 24, 37].
Participants in RCTs were recruited in acute care settings [2, 4, 9, 23, 24, 26–30] and rehabilitation units [8, 25], and for qualitative studies in the community [10], in rehabilitation [35], in acute care [3, 32] and in intermediate care [36].
Diagnoses were mixed, not specified or not sufficiently reported in five RCTs [2, 9, 25–28, 30] and in seven qualitative studies [32–34, 36, 37, 39, 40]. In two RCTs and three qualitative studies, participants had suffered from a stroke [3, 4, 8, 35, 38]. The diagnosis was hip fracture in another RCT [23, 24].
The qualitative studies reported on participants’ views [3, 31, 35–37] and on views of OTs [3, 32, 33, 35, 37, 38], families [31, 35], and older community-dwelling people [10].
Types of interventions
Interventions comprised a single predischarge home visit only [4, 8, 25, 26] as well as additional supportive interventions through in-hospital activities [9, 23, 24, 28], including extended assessment [2, 9, 25] and / or extended training [9, 24]. Further intervention components were patient education [4, 8, 25, 28] and post-discharge follow-ups [2, 9]. All the PDHAs were conducted by OTs alone, or with additional professionals allied to health care (physiotherapists, nurses, social workers) [9, 28]. The patients were present during the home assessment in seven out of eight RCTs [2, 4, 8, 23–26, 28]. All but one of the interventions were conducted in the patient's home, and included functional assessment [9]. Virtual home visits, conducted at the hospital, were investigated in one study [4]. The intervention details are available from the corresponding author.
Types of comparators
Usual care in Australia was described as an in-hospital access to multidisciplinary care [26], as well as a structured interview with the OT, including two structured assessments and an access visit if more information was required, such as measurements for rails [2] or additional patient education and information about equipment use and community services [25]. Usual care in the UK was described as structured interviews and general discussions about potential problems, and referring to agencies [8]. One study [4] reported additional home / access visits as a control, if required. Usual care in Sweden [23, 24] comprised nursing care and instruction from a physiotherapist for walking aids. Usual care in Germany [9] comprised comprehensive geriatric assessment and recommendations. Usual care in France was not described [28].
Risk of bias within studies
The results of the risk of bias assessment are summarized in Figure 2 and are presented in more detail in Additional File 3. Risk of selection bias was low in all but one study, where it was unclear [28]. For the outcome IADL/ADL, the risk of performance bias was unclear in five studies [2, 4, 8, 26, 28], and high in two of the seven studies addressing this outcome. For quality of life, the risk of performance bias was high in two studies [23–25] and unclear in three of five studies addressing this outcome [4, 8]. Risk of readmission and risk of falling were not biased in all six studies addressing this outcome [2, 8, 9, 23–26, 28]. Mobility was detected in two studies with a low or unclear performance bias, respectively. Three studies assessed fear of falling with a high or unclear risk of bias, respectively [4, 25, 26, 28]. Risk of detection bias was unclear in two studies [23, 24, 28]. Risk of attrition bias was high in one study
[23, 24]. Risk of other bias was unclear in one study [2].
The quality appraisal of the qualitative studies is shown in Additional File 4. The quality of the studies did not influence the analysis since all the studies were considered as being valuable for our research question.
Effectiveness of PDHA versus usual care
Eight RCTs including 1149 participants compared PDHA with usual care [2, 4, 8, 9, 23–26, 28]. Forest plots for comparisons are displayed in Additional File 6. Meta-analysis was performed for Instrumental Activities of Daily Living (IADL) and Activities of Daily Living (ADL), quality of life (Qol), mobility, fear of falling, risk of falling and risk of readmission. Details on the GRADE judgment are reported in Additional File 5.
The summary of findings for the main outcomes is presented in Table 2.
Assessment of reporting bias through funnel plot analysis was not appropriate due to the small number of studies.
IADL/ADL (Instrumental) Activities of Daily Living (IADL/ADL) were measured in seven of eight studies on patients with stroke, hip fractures, or mixed or unspecified diagnoses respectively [2, 4, 8, 23–26, 28]. Five studies used the Extended Activities of Daily Living scale (NEADL) [41], another used the Functional Autonomy Measurement System (SMAF) [42], each as a full questionnaire. One study used the subscale Physical Function from the Swedish Health-Related Quality of Life Survey (SWED-QUAL), which assesses a patient’s ADL performance (e.g. dressing, climbing stairs) and is therefore comparable to the content of included ADL-measures [43]. There was no overall effect in (instrumental) functions of daily living for participants at the latest follow-up after receiving PDHA when measured with various scales (655 participants, SMD -0.17, 95% CI [-0.87 to 0.53], p=0.64, I2=91%). The quality of evidence was judged to be very low due to concerns about risk of bias (blinding of outcome assessment), inconsistency and imprecision with considerable heterogeneity. A sensitivity analysis of five studies using the same scale (NEADL) confirmed the results (MD -0.32 [-1.26 to 0.61], p=0.50, I2=0%) with very low heterogeneity [2, 4, 8]. GRADE assessment indicated low quality due to high risk of bias (blinding of outcome assessment) and imprecision.
Quality of life (QoL) Three studies used the EQ-5D overall score [44] and another three the subscales of the EQ-5D measure of health status from the EuroQol Group (EQ-5D) or SWED-QUAL [43], respectively. Pooling all studies with any Qol measure [4, 23–25] showed no statistically significant group differences of PDHA compared to usual care for patients with stroke, hip fractures, or mixed diagnoses respectively, with moderate heterogeneity (263 participants, SMD 0.06, 95% CI [-0.30 to 0.42], p=0.74, I2=42%). Applying the GRADE approach, we assessed the quality of the evidence to be very low due to a risk of bias (unblinded participants and personnel) and imprecision of results. A sensitivity analysis of three studies using the same scale (EQ-5D overall score) did not significantly affect the Qol outcome (186 participants MD 0.03, 95% CI [-0.08 to 0.15], p=0.56, I2=0%). The quality of the evidence for these results was low due to inconsistency and imprecision.
Mobility. Two studies assessed mobility through Performance-Oriented Assessment of Mobility Problems (Tinetti) or The Rivermead Mobility Index (RMI) rating scale for patients with mixed diagnoses or stroke, respectively [4, 25] Pooling these studies showed no improvement at the latest time points of follow-up at one and three months (26 participants, SMD 1.24, 95% CI [-0.69 to 3.17], p=0.21, I2=78%). However, the quality of the evidence was rated very low due to inconsistency and high imprecision based on a very small number of participants with high heterogeneity.
Three studies measured fear of falling in participants with a stroke or mixed diagnoses, respectively, using the Falls Efficacy Scale - International (FES-I) [47]. There might be a slight trend towards an increase in fear of falling in participants who received the PDHA intervention. Applying the fixed effect Model (FEM) resulted in a statistically significant effect in favor of the control group (85 participants, MD -4.74 95% CI [-8.30 to -1.18], p=0.002) with moderate heterogeneity (I2=51%). When a pre-specified random effects model (REM) was used, there was no difference between groups in pooled effects for fear of falling (85 participants, MD -4.01, 95% CI [-10.4, 2.05], p=0.51) with moderate heterogeneity (I2=51%). Using the GRADE approach, we assessed the quality of the evidence for this outcome to be very low due to a risk of performance bias (unblinded participants and personnel) and imprecision of results resulting from the very small number of participants.
Risk of falling The overall effect of PDHA on reducing risk of falling was not statistically significant (523 participants, RR 0.88, 95% CI [0.69 to 1.13], p=0.32), I2=0%). Included were patients with mixed or unspecified diagnoses, respectively, hip fractures and stroke [8, 9, 25, 26, 28]. The quality of evidence was assessed as moderate because considerable harm and benefit were included in the confidence intervals of all the studies. We were therefore concerned with regard to imprecision.
Risk of readmission Pooling five studies showed no statistically significant effect on the reduction of readmissions throughout an average of 5 months after receiving PDHA (590 participants, RR 1.09, 95% CI [0.64 to 1.87], p=0.70, I2=43%) in patients with unspecified or mixed diagnoses or stroke, respectively [2, 8, 25, 26, 28] . Applying the GRADE approach, the quality of evidence was assessed as moderate because significant harm and benefit were included in the confidence intervals of all the studies. For this reason, we were concerned with regard to imprecision.
Outcomes from single studies:
Overall independence was assessed with the Modified Ranking Scale [48] in one study with a missing significant difference between the groups at one month after discharge (16 participants, MD -0.20 95% CI [-0.65 to 0.25], p=0.38) [4].
Psycho-social outcomes
One study reported on three different psycho-social outcomes, although all had missing significant differences at one month after discharge: Emotional distress in medical settings was measured through the GHQ-28 [49, 50] in 85 participants (in the intervention group with median 19; IQR 12.25–23.75 vs. median 23; IQR 15.5–31.5 in the control group; p=0.10). Depression was measured through The Stroke Aphasic Questionnaire [51] in 85 participants (in the intervention group with median 6; IQR 3.25–9.75 vs. median 7; IQR 4–11 in the control group; p=0.37). Caregiver strain was measured though the Caregiver Strain Index [52] in 85 participants (in the intervention group with median 5.5; IQR 1.75–7 vs. median 6; IQR 5–8 in the control group; p=0.11).
Process outcomes
The number of recommendations was reported in two studies with significant increases in the number of modifications in the intervention group compared to the control group at 30 or 90 days after discharge, respectively (average number of modifications 2.8 (1.6 to 3.9), p <.001in one study and range 0-13 in intervention vs. 0-7 in controls, p=0.001 in another study) [2].
Admissions to hospitals and care facilities
The number of emergency department visits was reported in one study with missing significant differences between the groups at 90 days after discharge (337 participants; RR=1.06, 95% CI [0.73 to 1.55], p=0.73 [2].
One study (86 participants) reported missing significant differences in the number of institutionalizations after 12 months (60 participants, RR= 0.58; 95% CI 0.26 to 1.27; p=0.17) [28].
The number of patients receiving community support was reported in one study, which stated that, three months after discharge, a total of three patients across groups received community support (seven patients across groups received support at baseline) [25].
Table 2: Summary of findings
PDHA compared with usual care for adults with any diagnosis at all
|
|
Patients or population: adults with any diagnosis at all (except mental disorders only) Setting: acute / sub-acute hospital care or rehabilitation unit Intervention: predischarge home assessment Comparison: usual care
|
|
Outcomes
|
SMD* or MD* or RR, [95% CI], I2, p
|
Number of participants (number of studies)
|
GRADE
|
Comments
|
|
|
IADL/ADL. Various scales. Including studies with NEADL, NEADL (60), SMAF, SWED-QUAL Subscale Physical function. Higher score indicates better function. Mean duration of follow-up: 8 months (range 1-12 months)
|
SMD -0.17 [-0.76, 0.42], I2=90% p=0.58
|
655 (7)
|
⨁OOO very lowa
|
|
|
IADL/ADL. NEADL Score 0-22 points. Higher score indicates better results. Mean duration of follow-up: 2.8 months (range 1-6 months)
|
MD -0.35 [-1.31, 0.61], I2=79% p=0.34
|
510 (5)
|
⨁OOO very lowc
|
|
|
Quality of life. Various scales: EQ-5D overall score, EQ-5D subscale VAS, SWED-QUAL subscale general health perception. Higher score indicates better health status. Mean duration of follow-up: 2.6 months (range 1-6 months)
|
SMD 0.06 [-030, 0.42] I2=42% p=0.74
|
263 (5)
|
⨁OOO lowc
|
|
|
Quality of life. EQ-5D overall score 0-1. Higher score indicates better health status. Mean duration of follow-up: 2.6 months (range 1-6 months)
|
MD 0.03 [-0.08, 0.15], I2=0% p=0.56
|
186 (3)
|
⨁⨁OO lowb
|
|
|
Mobility. Various scales: Tinetti (scale 4-24) and RMI (0-15). Higher scores indicate better mobility. Mean duration of follow-up: 2 months (range 1-3 months)
|
SMD 1.24 ['-0.69, 3.17], I2=78% p=0.21
|
26 (2)
|
⨁OOO very lowb
|
|
|
Fear of falling. FES-I Score 10-100. Higher scores indicate more confidence. Mean duration of follow-up: 3.3 months (range 1-6 months)
|
MD -4.01 [-10.4, 2.05], I2=51% p=0.19
|
85 (3)
|
⨁OOO very lowc
|
Fixed effect model: (MD -4.74 [-8.30, -1.18] I2=51%, p=0.009
|
|
Risk of falling Mean duration of follow-up: 9.2 months (range 1-12 months)
|
RR 0.88 [0.70, 1.09], I2=0% p=0.25
|
501 (5)
|
⨁⨁⨁Od moderate
|
|
|
Risk of readmission: Mean duration of follow-up: 5 months (range 1-12 months)
|
RR 1.09 [0.64, 1.87], I2=43% p=0.74
|
590 (5)
|
⨁⨁⨁Od moderate
|
|
|
Adverse effects of intervention:
|
Zero adverse events in both groups were reported in one study.
|
59 (1)
|
|
|
|
a downgraded due to unblinded personnel and participants, inconsistency and imprecision of results b downgraded due to inconsistency and high imprecision of results c downgraded due to downgrade because of unblinded participants and personnel, and imprecision of results d downgraded due to imprecision of results; FE: fixed effect model, RE: random effects model
|
|
Qualitative synthesis
Based on four comprehensive descriptive themes, seven analytical themes were identified regarding the barriers and facilitators of the PDHA process. Details are reported in Additional File 7 (Summary of the descriptive themes) and Additional File 8 (Overview on analytical themes).
Barriers and facilitators in PDHA process, analytical themes
The safety assessment of the home environment. Participants highlighted the importance of safety after hospital discharge [40]. The aim was to identify any required provisions and adaptations before going home and to identify and eliminate risks within the home [3, 32, 34, 40], as well as to assess whether the home environment was suitable for the required equipment [34]. The facilitators were the structured identification of risk factors and patient/family’s awareness of these factors following education, therefore enabling practical recommendations [40]. VR was identified as a useful tool to educate patients in order to identify and discuss risk factors, thus increasing patients’ awareness [3]. Therefore, we inferred the implication: 1 “Use environmental assessments together with patients to provide education about hazards.”
Functional assessment of the patient at home as a reality check. The aim of a functional assessment in general was to assess whether the patient is able to manage within his / her home [34]. On the whole, the predischarge home visit was a chance for therapists to gain a realistic view of the patients’ functions [34, 35, 37–40]. But this also applied vice versa: "It’s making them [the patients] aware of that impact and how they might be able to overcome the problems they will encounter. [...] We do get patients who say ´Oh, once I´m home I´ll be fine...`, but I don`t think they`ll always appreciate the limitations they`re going to encounter." [34]. PDHA gave information on future therapy sessions and helped to tailor individual rehabilitation goals [35, 37, 39]. Visiting their home motivated patients to do the therapy so that they could return home [32, 34, 35, 37, 40]. Performance tests at home can cause the patients to become anxious about failure, so the social skills of OTs are definitely needed [32, 34]. However, the preparations for the functional requirements for carrying out activities of daily living at home in the context of PDHA offered the chance to reduce anxiety [40]. We inferred the implications: 2 Conduct a functional assessment that includes the living reality of the patient and helps the patient to find individual participation goals for therapy, and 2.1 Consider potential patient anxieties regarding the assessment situation.
Intervention planning and evaluation. Novice therapists in particular struggled with the aim and content of PDHA [40]. The actual timing of PDHA was highly dependent on organizational factors and resource availability [34, 35, 39]. There were often pragmatic aspects, like the availability of supportive network, patients’ preferences or “gut feeling” to consider when deciding about whether or not to conduct a PDHA [37–40]. Working with community players often led to dissatisfaction with devices [37], without having the chance to follow up with the patient [39]. The facilitators were identified as: clear aims and assessment tasks, early patient identification and planning and a decision support tool [39], further use of standardized protocols during PDHA and collaboration with community services [40], as well as a formal evaluation after the PDHA [35]. The use of a digital interface to transmit environmental information could encourage the communication between the various stakeholders [10]. Therefore, we derived the implication 3: Use standardized procedures and materials to guide the PDHA process. Digital solutions might support the collaboration between hospital and community service providers.
Patient information about the home assessment procedure. Older people felt insufficiently informed prior to and after the home visit. Lack of information about the aims, the outcomes and the next steps of the process of PDHA made them feel insecure and anxious and excluded from the process [32, 34, 36]. Even during the home visit, there were situations in which the carer, but not the patient, was included in the process [32]. Written information about PDHA was seen as a facilitator by patients [35, 36]. Some patients and therapists felt a lack of real informed choice about the assessment [32]. Therefore we derived the implication 4. Provide adequate (verbal and written) patient information about aim, process, assessment, results and consequences of the predischarge home assessment.
Patients’ and family carers’ acceptance of home modifications and aids. The concerns of the patients that the OT’s modifications might hinder them in performing ADL in the usual and preferred way was identified as a barrier [32]. The use of a patient’s know-how on where to use an aid most effectively in their home environment was a facilitator for acceptance [37]. From the OT’s view it was seen as challenging to propose and communicate potential adaptations to the patients [10]. The lack of imagination regarding home modifications and adaptations [3] was seen as a barrier for acceptance. OTs and older people estimated that the use of visualization with a 3-D interior design software application would enable patients to better understand assistive technologies and adaptations [10, 33]. OTs considered a virtual reality tool as superior to drawings and photographs [33]. Consequently, a more clear visualization was seen as a facilitator for OTs to communicate better about modifications or even explain better decisions against a patient bbeeing discharged to his/her own home [10]. In addition, a clear visualization as a joint basis for discussion was seen as a facilitator to include patients in the decisions about home modifications and aids, giving them a chance to give immediate feedback on proposed changes, thus leading to shared decision-making [3, 10, 33] . Therefore we inferred implication 5: Provide tailored adaptations based on shared decision-making and involve explicitly patients’ ideas, solutions and expectations in planning home modifications, and 5.1 Provide appropriate visualization for discussing recommended aids and home modifications.
Matching PDHA and clinical patient conditions. Different patient conditions in terms of diagnosis and related kinds of impairments as well as the levels of impairment may be factors that facilitate or inhibit the performance of certain PDHA approaches. For example, sensory and visual limitations might be an indication for a home visit. However, the same limitations may have an adverse effect on the use of a virtual home assessment. Too low or too high levels of functional limitations spoke rather against home visits and in favor of ward-based assessments or access visits. Summing up, different patient conditions required different approaches for assessment. We inferred implication 6: Tailor the intervention components and mode of delivery to patients’ level and kind of impairments.
Context factors in daily routine of PDHA. Many of the qualitative studies identified factors that may have a beneficial or impeding effect on the decision of whether and how to conduct PDHA. Lack of resources (staff, time, secretarial backup, technical resources for virtual assessment) hampers the process of organization and execution [3, 32, 34, 36, 38, 39]. A virtual approach to PDHA could partially overcome some of the obstacles (e.g. out of hospital catchment zone, car availability, safety requirements for allowing a home visit with patient) [33]. Factors such as risks while making home visits and the organization of appropriate PDHA attendants have an impact on the process of PDHA [34, 37, 39, 40]. Therefore, we derived the implication 7: Consider specific context factors in PDHA-design.
Integrative synthesis
An overview of the results of the analysis at the individual study level with regard to the respective qualitative results (whether the PDHA intervention had considered implications 1-7) and the outcome effects in the patient outcomes is shown in table 3.
Table 3. Synthesis of practice implications and RCT interventions
|
Integrative synthesis of qualitative and quantitative results in studies in PDHA
|
|
|
|
Implications for interventions
|
Outcomes
|
Reference
|
Setting, diagnosis
|
Intervention
|
1
|
2
|
2.1
|
3
|
4
|
5
|
5.1
|
6
|
7
|
IADL/ADL, scale: NEADL
|
IADL/ADL, Various scales
|
Quality of Life, scale: EQ-5D
|
Quality of Life various scales
|
Mobility
|
Fear of Falling
|
Risk of falling
|
Risk of readmission
|
AE
|
Effect (MD or RR), 95% CI
|
Studies
|
|
[2]
|
Acute care, unspecified
|
Pre- and post- discharge visit + follow up phone calls
|
Nr
|
Un
|
Nr
|
Ad
|
Nr
|
Nr
|
Nr
|
Nr
|
Ad
|
MD-0.30 [-1.32 to 0.72]
|
MD-0.06 [-0.27 to 0.15]
|
-
|
-
|
-
|
-
|
-
|
RR 1.12 [0.76 to 1.66]
|
-
|
|
Stroke rehabilitation unit, stroke
|
Predischarge home visit
|
Un
|
Ad
|
Nr
|
Nr
|
Nr
|
Un
|
Nr
|
Nr
|
Ad
|
MD -5.50 [-15.58 to 4.58]
|
MD -0.23 [-0.66 to 0.20]
|
MD 0.03 [-9.40 to 9.46]
|
MD 0.00 [-0.41 to 0.41]
|
-
|
-
|
RR 1.41 [0.67 to 2.98]
|
RR 3.91 [0.88 to 17.46]
|
-
|
[23, 24]
|
Acute care, hip fracture
|
Individual daily training including use of technical aids + predischarge home visit + instruction from physiotherapist how to walk with technical aids
|
Nr
|
Nr
|
Nr
|
N
|
Nr
|
Un und
Ad
|
Nr
|
Nr
|
Nr
|
-
|
MD 0.23 [-0.19 to 0.66]
|
-
|
MD -0.27 [-0.70 to 0.16]
|
-
|
-
|
-
|
-
|
-
|
[25]
|
Rehabilitation unit, mixed
|
Predischarge home visit
|
Ad
|
Ad
|
Nr
|
Ad
|
Nr
|
Un
|
Nr
|
Nr
|
Nr
|
MD 13.40 [6.88 to 19.92]
|
MD 2.30 [0.51 to 4.10]
|
-
|
MD 1.50 [0.00 to 3.00]
|
MD 0.27 [-0.98 to 1.51]
|
MD 7.00 [-17.36 to 31.36]
|
RR 0.50 [0.06 to 3.91]
|
RR 0.33 [0.02 to 6.65]
|
-
|
[26]
|
Hospital wards and community, hip fracture
|
Home assessment by occupational therapist prior to discharge
|
Un
|
Ad
|
Nr
|
Ad
|
Nr
|
Un
|
Nr
|
Ad
|
Nr
|
MD -0.50 [-3.47 to 2.47]
|
MD -0.08 [-0.59 to 0.43]
|
MD 0.00 [-0.13 to 0.13]
|
MD 0.00 [-0.51 to 0.51]
|
-
|
MD -1.20 [-6.72 to 4.32]
|
RR 0.83 [0.47 to 1.45]
|
RR 0.55 [0.25 to 1.18]
|
0
|
[9]
|
Geriatric acute care, unspecified
|
Predischarge home visit + post-discharge follow up visit(s) + comprehensive in-hospital geriatric assessment
|
Ad
|
Nr
|
Nr
|
Un
|
Nr
|
Un
|
Nr
|
Ad
|
Nr
|
-
|
-
|
-
|
-
|
-
|
-
|
RR 0.84 [0.63 to 1.12]
|
-
|
-
|
[28]
|
Geriatric acute care, unspecified
|
Predischarge home visit + contacting potential social support + physical therapy + therapeutic modifications
|
Ad
|
Nr
|
Nr
|
Nr
|
Nr
|
Un
|
Nr
|
Nr
|
Ad
|
-
|
MD -2.48 [-3.16 to -1.80]
|
-
|
-
|
-
|
-
|
RR 0.87 [0.50 to 1.49]
|
RR 1.50 [0.61 to 3.69]
|
-
|
[4]
|
Stroke ward / acute care, stroke
|
Predischarge Virtual Home Assessment
|
Ad
|
Nr
|
Nr
|
Un
|
Nr
|
Ad
|
Ad
|
Un
|
Ad
|
MD 12.00 [-10.57 to 34.57]
|
MD 0.49 [-0.51 to 1.49]
|
MD 0.19 [-0.08 to 0.46]
|
MD 0.64 [-0.37 to 1.65]
|
MD 2.24 [0.91 to 3.56]
|
MD -7.80 [-12.54 to -3.06]
|
-
|
-
|
-
|
Implications: 1) Use environmental assessments together with patients to provide education about hazards. 2) Conduct a functional assessment that includes living reality of the patients and helps the patient to find individual participant goals for therapy. 2.1) Consider potential patient anxieties regarding the assessment situation. 3) Use standardized procedures and materials to guide the PDHA process. Digital solutions could support the collaboration between hospital and community service providers. 4) Provide adequate (verbal and written) patient information about aim, process, assessment, results and consequences of the predischarge home assessment. 5) Provide tailored adaptations based on shared decision-making and involve explicitly patient’s ideas, solutions and expectations in planning home modifications. 5.1) Provide appropriate visualization for discussing recommended aids and home modifications. 6) Tailor the intervention components and mode of delivery to patients’ level and kind of impairments. 7) Consider specific context factors in PDHA-design.
|
AE-Adverse effects of intervention, Ad – Addressed, Un – Unsure, Nr – Not reported
|