Improved functions and reduced length of stay after inpatient rehabilitation programs in older adults with stroke: A systematic review and meta-analysis of randomized controlled trials

Abstract

BACKGROUND: to examine the effects of inpatient rehabilitation programs on function and length of stay in older adults with stroke

METHODS: A total of five electronic databases were searched for relevant randomized controlled trials that examined the effects of inpatient rehabilitation programs on functional recovery, as measured by the functional independence measure and length of stay, which was measured in days. We included full-text articles written in English, and no time limit. The methodological quality and risk of bias were assessed using the Physiotherapy Evidence Database Scale and the Cochrane collaboration tools respectively. The effect sizes and confidence intervals were estimated using fixed-effect models

RESULTS: Eight randomized controlled trials involving 1,910 patients with stroke were included in the meta-analysis showed that patients who participated in the inpatient rehabilitation programs had significantly (p less than 0.05) higher functional independence measure scores (effect size = 0.10; 95 percent confidence interval = 0.01, 0.22) and shorter length of stay (effect size = 0.14; 95 percent confidence interval = 0.03, 0.22).

CONCLUSIONS: This systematic review provided evidence that inpatient rehabilitation programs have beneficial effects, improving functionality and reducing length of stay for older adults with stroke.

Keywords

Stroke length of stay rehabilitation

1 Introduction

According to Easton et al. (2009) stroke characterized as an “any central cerebral ischemic occasion with manifestations enduring less than 24 hours”. The absolute number of strokes, measured in disability-adjusted life years (DALYs), has continued to increase from approximately 38 million DALYs worldwide in 1990 to 61 million DALYs in 2020; this increase is the result of the aging population (Mackay & Mensah, 2004). Worldwide, 15million individuals endure a stroke annually. Of these, 3 million women and 2.5 million men die and another 5 million are permanently disabled, which places a burden on their families and the affected individuals every year (Mackay & Mensah, 2004).

Stroke is additionally driving reason for functional impairment. Twenty percent of survivors require inpatient rehabilitation programs (IRPs) after 3 months (Go et al., 2014). A systemic review (SR) also indicated the need for more health care services directed at patients with stroke, including rehabilitation services (Tummers, Schrijvers, & Visser-Meily, 2012). Different rehabilitation programs may be used to maximize an individual’s functions after stroke (Dobkin, Plummer-D’Amato, Elashoff, & Lee, 2010; Nadeau et al., 2013). Considerable controversy and debate encompass the adequacy of related programs after stroke (Pollock et al., 2014). According to the World Health Organization (WHO), rehabilitation interventions are used to maximize function and minimize the limitations in activity using neurofacilitation, functional and compensatory training strategies (Stucki, Cieza, & Melvin, 2007).

The stroke patient if medically stable, rehabilitation can begin in the acute care facility within approximately 72 hours. In a stroke unit, these interventions can prevent or minimize the effects of deconditioning and the risk of secondary impairment (Duncan et al., 2005). According to the final report of the National Study of Subarachnoid Haemorrhage on an audit completed in 34 neurosurgical units in the United Kingdom (UK) and Ireland between 2001-2002, more than 50% of older adults with stroke require additional IRPs after a successful intervention in acute care (Lindsay et al., 2006). Therefore, acute care consideration is all around acknowledged as a specific medical control for the treatment of more young patients with stroke compared to older adult patients who have reduced functional outcomes, which are associated with aging, cognitive problems, the length of stay (LoS), and multiple comorbidities (Chung, Niewczyk, DiVita, Markello, &Granger, 2012).

Despite, the effects of IPRs on maximizing functional independence and reducing the LoS in older adults with stroke are controversial or unclear. To our knowledge, the literature in this area has not been systematically collected and analyzed before. As a result, we performed a SR and meta-analysis (MA) to examine the effects of IRPs on the functional status and LoS in older adults with stroke based on the evidence from randomized controlled trials (RCTs).

2 Methods

This SR was conducted and reported as per the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Knobloch, Yoon, & Vogt, 2011) and the Consolidated Standards of Reporting Trials (CONSORT) (Schulz, Altman, & Moher, 2011) statements.

A strategic and systematic search was conducted using the Medline, CINAHL plus (EBSCO), PubMed, Physiotherapy Evidence Database (PEDro), and Cochrane Central Register of Controlled Trials (CENTRAL) databases (Guyatt, Rennie, Evidence-Based Medicine Working Group, & American Medical Association, 2002). The keywords used to perform the search were the following: stroke or cerebrovascular accident (CVA); post (–) acute, functional, function, or functional status; elderly, old, older, very old, or aged; inpatient, programs, facilities, multidisciplinary therapy, rehabilitation, physiotherapy, or occupational therapy; hospitalized; discharge; functional independence measure (FIM); and outcomes. The bibliographic search was restricted to RCTs published in English with available full-text and no restriction on the time. An illustration of the search technique utilized for one of the databases appears in Appendix 1.

Finally, the bibliographies of all eligible articles were evaluated by three independent reviewers (S.B., V.V., and E.F.).

Three evaluators (S.B., V.V., and E.F.) independently reviewed the articles identified in the search. The full contents of the conceivably relevant articles were obtained and assessed using the following inclusion criteria: subjects (Hatano, 1976), setting (Safer & Koseoglu, 2015), outcome (Yavuzer, Kucukdeveci, Arasil, & Elhan, 2001), methodology or design (Horwitz, 1995), type of publication (Treadwell, Singh, Talati, McPheeters, & Reston, 2011), and methodological quality (Chumney et al., 2010). Differences of opinion regarding the articles were resolved by discussion and consensus between two of the reviewers (S.B. and V.V.). The detailed inclusion and exclusion criteria were shown in Table 1. Excluded articles (n = 10) and major reason to exclusion appear in Appendix 2.

Table 1
Inclusion and exclusion criteria (Treadwell et al., 2011)

Criteria Inclusion Exclusion

Subjects A clinical diagnosis of stroke (World Health Organization; WHO) (Hatano, 1976) patients aged 65 years or older. All patients not diagnosed with stroke, Stroke patients aged younger than 65 years.

Setting (Inpatient rehabilitation care, also known as hospital-level or acute, and rehabilitation care (Safer & Koseoglu, 2015). Inpatient rehabilitation programs-defined as impatient multidisciplinary programs with active physiotherapy, occupational therapy, or both, according the WHO ICF framework (Stucki et al., 2007). Non-inpatient rehabilitation programs.

Results measures (Yavuzer et al., 2001). Primary outcomes (Admission, discharge, and follow-up). Concentrates on that did not cover the predefined result information, (for example, functional status or FIM).

Functional status or total functional independent measure (FIM).

FIM-motor.

FIM-cognitive.

Secondary outcome (optional).

Length of stay.

Methodology or design Randomized controlled trial. Non-randomized controlled trials.

(Horwitz, 1995).

Type of publication (Treadwell et al., 2011). Journal publications. Published and unpublished dissertations, theses. Books, or conference proceedings, and clinical trial protocols.

English language Studies published in languages other than English.

Full-text accessibility. The full-text was not available.

Methodological quality (Chumney et al., 2010). A score of 5 or higher on the Physiotherapy Evidence Database (PEDro) scale. Less than 5 on the PEDro scale.

Quantitative analysis (Meta-analysis). Outcome data presented as the means and standard deviations or effect sizes. Outcome data presented in a format other than means and standard deviations or effect sizes.

Criteria	Inclusion	Exclusion
Subjects	A clinical diagnosis of stroke (World Health Organization; WHO) (Hatano, 1976) patients aged 65 years or older.	All patients not diagnosed with stroke, Stroke patients aged younger than 65 years.
Setting (Inpatient rehabilitation care, also known as hospital-level or acute, and rehabilitation care (Safer & Koseoglu, 2015).	Inpatient rehabilitation programs-defined as impatient multidisciplinary programs with active physiotherapy, occupational therapy, or both, according the WHO ICF framework (Stucki et al., 2007).	Non-inpatient rehabilitation programs.
Results measures (Yavuzer et al., 2001).	Primary outcomes (Admission, discharge, and follow-up).	Concentrates on that did not cover the predefined result information, (for example, functional status or FIM).
	Functional status or total functional independent measure (FIM).
	FIM-motor.
	FIM-cognitive.
	Secondary outcome (optional).
	Length of stay.
Methodology or design	Randomized controlled trial.	Non-randomized controlled trials.
(Horwitz, 1995).
Type of publication (Treadwell et al., 2011).	Journal publications.	Published and unpublished dissertations, theses. Books, or conference proceedings, and clinical trial protocols.
	English language	Studies published in languages other than English.
	Full-text accessibility.	The full-text was not available.
Methodological quality (Chumney et al., 2010).	A score of 5 or higher on the Physiotherapy Evidence Database (PEDro) scale.	Less than 5 on the PEDro scale.
Quantitative analysis (Meta-analysis).	Outcome data presented as the means and standard deviations or effect sizes.	Outcome data presented in a format other than means and standard deviations or effect sizes.

The three reviewing authors independently extracted the data (S.B., V.V., and E.M.). Information on the following characteristics was extracted: study design (e.g., RCTs), quality (e.g., PEDro scale), participants (e.g., age and diseases diagnosed), setting (e.g., inpatient rehabilitation), and outcomes (FIM and LoS). The authors resolved disagreements by consensus and contacted the study authors for clarification when necessary.

The methodological nature of each included article was then autonomously assessed by two analysts utilizing the PEDro scale (Table 2). The PEDro scale was picked because it is a significant evaluation of the methodological nature of clinical trials and has been applied to treatments and contemplations for restorative interventions and outcomes. The PEDro scale ranges from 0 to 10 and is used to survey study quality. For this review, the PEDro score of 5 was picked as an optional cutoff in light of the fact that the experts felt this score would demonstrate a strong level of confirmation (Foley, Teasell, Bhogal, & Speechley, 2003) and high quality, as reported by Moseley et al. (2002) The reliability and validity of the PEDro scale have been reported elsewhere (Maher, Sherrington, Herbert, Moseley, & Elkins, 2003).

Table 2

Quality appraisal of the included studies (listed alphabetically by author) of older adults with stroke who partook in inpatient rehabilitation programs

Study	PEDro Criterion Score*											Score^†	Quality^‡
	1	2	3	4	5	6	7	8	9	10	11
Fong et al., 2007	Y	Y	N	Y	N	N	Y	Y	N	Y	Y	6	High
Liu et al., 2014	Y	Y	Y	Y	N	N	Y	Y	N	Y	N	6	High
Masiero et al., 2014	Y	Y	N	Y	N	N	Y	N	Y	Y	Y	6	High
Peiris et al., 2013	Y	Y	Y	Y	N	N	Y	Y	Y	Y	Y	8	High
Rabadi et al., 2008	Y	Y	Y	N	N	N	Y	Y	N	Y	Y	6	High
Rabadi et al., 2008	Y	Y	Y	N	N	N	Y	Y	N	Y	Y	6	High
Rawl et al., 1998	N	Y	N	Y	N	N	N	Y	N	Y	Y	5	High
Strasser et al., 2008	Y	Y	N	Y	N	N	N	Y	N	Y	Y	5	High
Sze et al., 2002	Y	Y	Y	Y	N	N	Y	Y	Y	Y	Y	8	High
Wikander et al., 1998	Y	Y	N	Y	N	N	N	Y	N	Y	Y	5	High

^*PEDro = physiotherapy evidence database assessment; 1 = eligibility, 2 = randomized allocation, 3 = concealed allocation, 4 = similar prognosis, 5 = blinded subjects, 6 = blinded therapist, 7 = blinded raters, 8 = key outcome, 9 = intention to treat, 10 = comparison between groups, 11 = precision and variability. Note: the eligibility item does not contribute to the total score. ^†Maximum score of 10. ^‡The included studies with a PEDro score ≥5 were considered high quality. Y = yes; N = no.

It is important to include a risk of bias when assessing the internal validity of each article in this SR (Higgins, Altman, & Sterne, 2011). Two of the review authors (S.B. and V.V.) independently documented the risk of bias of the included articles (Table 3). The risk of bias was assessed using the Cochrane collaboration apparatus as part of the domain-based assessment, with a low, high, or unclear risk of bias judgment for each item (Higgins et al., 2011). The evaluated areas were random sequence generation, allocation concealment, blinding of subjects, blinding of the therapist, blinding of raters, incomplete outcome data, and specific reporting. The results of this risk of bias tool did not influence the inclusion in the final output. The authors determined the differences by examination and contacted the articles’ authors for clarification when necessary.

Table 3

Risk of bias of all included studies (listed alphabetically by author)

	Fong et al., 2007	Liu et al., 2014	Masiero et al., 2014	Peiris et al., 2013	M. Rabadi et al., 2008	M. H. Rabadi et al., 2008	Rawl et al., 1998	Strasser et al., 2008	Sze et al., 2002	Wikander et al., 1998
Random sequence generation	+	+	+	+	+	+	+	+	+	+
Allocation concealment	–	+	–	+	+	+	–	–	+	–
Blinding of participants	–	–	–	–	–	–	–	–	–	–
Blinding of therapist	–	–	–	–	–	–	–	–	–	–
Blinding of raters	+	+	+	+	+	+	–	–	+	–
Incomplete outcome data	+	+	–	+	+	+	+	+	+	+
Selective reporting	+	+	+	+	+	+	+	+	+	+
Conclusion	–	–	–	–	–	–	–	–	–	–

Abbreviations: – = high risk of bias; + = low risk of bias.

Two reviewers (S.B. and V.V.) independently screened the selected articles. Crude understanding and the Cohen kappa measurement were utilized to quantify the agreement between the two analysts.

The primary outcome measure of interest was the functional status, as measured by FIM (FIM-motor, FIM-cognitive, and FIM-total at discharge). The secondary outcome measure of interest was LoS, measured in days. We extracted the means and standard deviations (SDs) of the admission and discharge scores for the primary and secondary outcomes from the included articles. We displayed all of the examined result measures as constant data. Standard mean differences (SMDs) were calculated for all of the outcomes. The Hedges g measurement was chosen because it remedies slight overestimations that may emerge from a small number of examples and the estimation variability of the selected articles was calculated using a balanced effect size (ES). The fixed effect was selected for all of the outcomes because the heterogeneity was low (I² <25%) (Higgins, Thompson, Deeks, & Altman, 2003). Then, the 95% confidence intervals (CIs) were calculated for the ESs, based on a variance outcome, to identify the overlapping CIs between studies, which confirmed that the heterogeneity was low. The centrality of the overall impact was tested using the z statistic (see Appendix 3 for the formulas used to calculate Hedges g using a Microsoft Excel spreadsheet) (Borenstein, 2009). The magnitude of the ES was categorized using the Cohen categories, with g < 0.5 as a small ES, g≥0.5 and ≤0.8 as a medium ES, and g > 0.8 as a large ES.

The heterogeneity among comparable trials was analyzed utilizing both the standard Cochran Q and Higgins I² statistics (Higgins & Green, 2008; Higgins et al., 2003). A noteworthy Q test just shows the region of heterogeneity among the data included, while the I² file measures its size (Higgins & Thompson, 2002). A genuine categorization of the qualities for I² would not be suitable for all circumstances, although it would likely relegate descriptive words of low, moderate, and high to I² values of 25, 50, and 75%, respectively (Higgins et al., 2003). A low p value (≤0.05) for the Q statistic was considered proof of heterogeneity of the treatment impacts. A fixed-effects MA was directed to decide the general ES since it is more preservationist, especially for little specimens (Higgins & Thompson, 2002). (see Appendix 4 for the formulas used to calculate the fixed models using a Microsoft Excel spreadsheet) (Borenstein, 2009; Hardy & Thompson, 1998). All of the statistics were computed using Microsoft Excel 2007, version 12.0 [Microsoft, Redmond, WA, USA] (Neyeloff, Fuchs, & Moreira, 2012).

3 Results

Figure 1 shows the procedure used to choose the 10 articles incorporated into the SR (Fong et al., 2007; Liu & Chan, 2014; Masiero, Armani, Ferlini, Rosati, & Rossi, 2014; Peiris, Shields, Brusco, Watts, & Taylor, 2013; Rabadi et al., 2008; Rabadi, Coar, Lukin, Lesser, & Blass, 2008; Rawl, Easton, Kwiatkowski, Zemen, & Burczyk, 1998; Strasser et al., 2008; Sze, Wong, Yi, & Woo, 2002; Wikander, Ekelund, & Milsom, 1998) and the 8 articles that were deemed suitable for the MA (Fong et al., 2007; Liu & Chan, 2014; Peiris et al., 2013; M. Rabadi et al., 2008; M. H. Rabadi et al., 2008; Rawl et al., 1998; Strasser et al., 2008; Wikander et al., 1998). The two articles (Masiero et al., 2014; Sze et al., 2002) that were excluded from the MA reported median (interquartile range) values for the outcomes. The agreement between reviewers on the specified inclusion and exclusion criteria (see Table 1) for all stages of article selection was nearly perfect (100% crude agreement, kappa = 1) (Viera & Garrett, 2005).

Fig.1

Flowchart of papers selected for the study.

The after effects of the methodological appraisal are compressed into Table 2. The agreement of the methodological quality between the two independent reviewers was perfect (100% crude agreement, kappa = 1) (Viera & Garrett, 2005). The PEDro scores of the included articles ranged from 5 to 8, with a median score of 5 classified as a high quality article. The most common limitations were the following: lack of blinding of the patient and therapist (Fong et al., 2007; Liu & Chan, 2014; Masiero et al., 2014; Peiris et al., 2013; M. Rabadi et al., 2008; M. H. Rabadi et al., 2008; Rawl et al., 1998; Strasser et al., 2008; Sze et al., 2002; Wikander et al., 1998), lack of blinding of the raters (Rawl et al., 1998; Strasser et al., 2008; Wikander et al., 1998), intention to treat (Fong et al., 2007; Liu & Chan, 2014; M. Rabadi et al., 2008; M. H. Rabadi et al., 2008; Rawl et al., 1998; Strasser et al., 2008; Wikander et al., 1998), concealed allocation (Fong et al., 2007; Masiero et al., 2014; Rawl et al., 1998; Strasser et al., 2008; Wikander et al., 1998), and a similar prognosis (M. Rabadi et al., 2008; M. H. Rabadi et al., 2008). The items that best adhered to the PEDro Scale were random allocation, key outcome, comparison between groups, precision, and variability, which were evident in almost all of the trials. The randomization methods, eligibility, statistical analysis criteria were clearly defined in all of the included articles (Fong et al., 2007; Liu & Chan, 2014; Masiero et al., 2014; Peiris et al., 2013; M. Rabadi et al., 2008; M. H. Rabadi et al., 2008; Rawl et al., 1998; Strasser et al., 2008; Sze et al., 2002; Wikander et al., 1998).

The evaluators agreed on the risk of bias of each included article. The risk of bias assessment of the included articles is illustrated in Table 3. The overall risk of bias assessment indicated that the risk of bias was high in all of the included articles. The most common risks of bias in all of the included articles were blinding (patients, therapists, or raters) (Fong et al., 2007; Liu & Chan, 2014; Masiero et al., 2014; Peiris et al., 2013; M. Rabadi et al., 2008; M. H. Rabadi et al., 2008; Rawl et al., 1998; Strasser et al., 2008; Sze et al., 2002; Wikander et al., 1998) and concealed allocation (Fong et al., 2007; Masiero et al., 2014; Rawl et al., 1998; Strasser et al., 2008; Wikander et al., 1998). Not surprisingly, the clinicians and subjects were not blinded in any of the studies. This finding contrasts with the conclusion of Boutron et al (Boutron, Tubach, Giraudeau, & Ravaud, 2004) who demonstrated that patient blinding is difficult to accomplish in nonpharmacological studies. However, whereas all of the articles were RCTs, only 5 reported concealment.

Table 4 summarizes the characteristics of the articles included in this review. All the 10 included articles with 2,309 participants were allocated to the intervention (n = 1,124) or control (n = 1,175) groups in either rehabilitation hospital or unit setting. The mean age of older adults with stroke ranged from 66.0 ± 11 to 74.3 ± 11.8 years, and the LoS ranged from 9 to 76 days. Despite the fact that men and women were incorporated into the studies, men represented the unmistakable sex. In all of the articles included in this review, the primary diagnosis of the participants was stroke, and they were admitted to an IRP for functional recovery, as measured with the FIM scale, and had an adequate follow-up (≥74%). The overall loss to follow-up from all of the included articles was 0.45% (n = 10). Most of the included studies were conducted in the US.

Table 4

Characteristics of the included randomized controlled trials (listed alphabetically by author) of an inpatient rehabilitation programs

Author, year country	Study design	Setting	Primary diagnosis	No. of patients (% female)	Intervention/control	Average age (yr) Mean ± SD	Length of stay (days)	Length of Intervention	Follow-up (%)	Intervention
Fong et al., 2007, China	Pre-post,60 day follow-up, single-blinded randomized controlled trial	Single-center inpatient rehabilitation hospital	Hemisphere stroke	60 (37)	39/15	71.2 ± 10.2	43	1 hr 5 min a week for30 days	76%	Intervention group: Nineteen patients got every day exploratory preparing in intentional trunk rotation (TR) for 1 hour five times each week for 30 days. Twenty patients got the same kind of treatment, together with half-field eye-fixing (TR + EP). Control group: Fifteen patients received conventional training with the same contact time.
Liu et al., 2014, Australia	Pilot randomized controlled trial	Rehabilitation unit	Acute post stroke	44 (50)	24/20	71.0 ± 3.4	13	5 hr for one week	94%	Patients were haphazardly appointed to the SR (n = 24) or practical restoration (control; n = 20) intervention. The SR intervention consisted of 1 week of therapist-supervised practices of daily tasks using the SR of one’s own performance (five 1-hr sessions). Patients in the control intervention practiced the same daily tasks with a therapist’s demonstration and guidance.
Masiero et al., 2014, Italy	Dose-matched, randomized controlled clinical trial	Stroke unit in a rehabilitation hospital	Hemorrhagic or ischemic cerebrovascular accident	34 (32)	14/16	66.2 ± 8.6	9	120 min, 5 days per week for5 weeks	74%	All of the participants received a total daily rehabilitation treatment of 120 min, 5 days per week for 5 weeks at baseline, 3 months, and 7 months. Control group: Received standard therapy for the upper limb. Intervention group: Received standard therapy (65% of exercise time) associated with robotic training (35% of exercise time).
Peiris et al., 2013, Australia	Multicenter, single-blinded, randomized controlled trial	Inpatient rehabilitation facilities	Orthopedic and neurological (stroke)	996 (64)	496/500	74.0 ± 13.0	23	4 days for the control group and 5 days for the intervention group	94%	The participants were randomly allocated to usual care, Monday to Friday rehabilitation (control), or to Monday to Saturday rehabilitation (intervention). The extra Saturday rehabilitation involved physiotherapy and occupational treatment.
M. Rabadi et al., 2008, USA	Prospective, randomized controlled study	Stroke unit in a rehabilitation hospital	Acute stroke	30 (37)	20/10	72.2 ± 9.7	28	3 hr 40 min a day	100%	The majority of the patients got standard, inpatient, post-stroke recovery preparing for 3 hours a day, in addition to 12 extra 40-minute sessions of the movement based treatment.
M. H. Rabadi et al., 2008, USA	Randomized, prospective, double-blinded, single center study	Stroke unit in a rehabilitation hospital	Hemorrhagic or ischemic stroke	116 (41)	58/58	74.3 ± 11.8	25	NA	88%	All of the patients received 2-minute and 6-minute timed walk tests measured at admission, upon discharge, and upon discharge disposition (home/not home).
Rawl et al., 1998, USA	Randomized controlled clinical trial	Stroke unit in a rehabilitation hospital	Cerebrovascular accident and hip fracture	100 (70)	49/51	69.2 ± 12.8	NA	NA	100%	The nurse-managed follow-up intervention program consisted of four contacts (three in-person contacts and one telephone contact) with the advanced practice nurse (APN) at 2 days, 30 days, and 4 months after discharge.
Strasser et al., 2008, USA	A cluster randomized trial	Veterans affairs medical centers	Stroke	789 (3)	350/439	66.0 ± 11.0	22	NA	92%	A multiphase, staff preparing program conveyed more than 6 months, including an off-site workshop underscoring group elements, critical thinking, and the utilization of execution input information; activity gets ready for procedure change; phone and video gathering; and particular group execution profiles, with proposals to utilize this data to alter the group process.
Sze et al., 2002, China	Prospective, randomized controlled study	Stroke rehabilitation unit	Acute stroke	106 (55)	53/53	70.8 ± 8.7	28	Intervention: 5 sessions per week for inpatients. 3 sessions per week for 8 weeks and 2 sessions per week for the other 2 weeks for day hospital patients. Control: PT 5 sessions per week (60 min per session)	87%	Intervention arm: The patients got the same standard treatment as depicted for the control arm, together with conventional Chinese needle therapy for 5 sessions for each week for inpatients, 3 sessions for every week to the end of eighth week, and 2 sessions for each week for the remaining 2 weeks for day clinic patients. The needle therapy session was haphazardly booked either before or after the OT session in the same half-day session and went on for 30 min. Control arm: For inpatients, the standard treatment included PT, 5 sessions for every week and 60 min for every session; OT, 5 sessions for each week and 45 min for each session; language instruction and mental guiding, as demonstrated; and gifted nursing care and a day by day medicinal round. For day healing center patients, the standard treatment included participation at a day clinic for 3 days a week for up to 8 weeks from incorporation and after those 2 days a week for the remaining 2 weeks. There were 60 min of PT and 45 min of OT per participation day. Treatment was not prespecified, and antiplatelet operators and anticoagulants were permitted at the carefulness of the going to doctor.
Wikander et al., 1998, Sweden	Randomized controlled trial	Stroke unit in a rehabilitation hospital	Hemispheric stroke	34 (56)	21/13	74.0 ± 1.5	76	NA	100%	The greater part of the patients were surveyed on affirmation and on release utilizing the Katz activities of daily living (ADL) file, the mental general prosperity file thing G of the FIM file (FIM-G), and a versatility score.

Abbreviations: FIM = functional independent measure; NA = not available; SD = standard deviation; yr = years; hr = hours; min = minutes.

The outcomes are summarized in Table 5. Of the 10 articles, 5 articles (Liu & Chan, 2014; Peiris et al., 2013; M. H. Rabadi et al., 2008; Strasser et al., 2008; Wikander et al., 1998) reported a significant functional improvement in the intervention group that was admitted to an IPR. Of these five articles, the results of two articles (Peiris et al., 2013; M. H. Rabadi et al., 2008) revealed that the participants in the intervention group had a significantly higher FIM-total score upon discharge. The discharge results from five articles (Liu & Chan, 2014; Peiris et al., 2013; M. H. Rabadi et al., 2008; Strasser et al., 2008; Wikander et al., 1998) showed that the participants in the intervention group exhibited a significantly improved FIM-motor subscale score. Three articles (Liu & Chan, 2014; Peiris et al., 2013; Wikander et al., 1998) showed that the participants in the intervention group exhibited a significantly improved FIM-cognitive subscale score at discharge. All of the included articles reported on the LoS, which varied between 9 and 76 days. Only one article reported that the intervention group may have had a shorter LoS upon discharge compared to the control group (Peiris et al., 2013).

Table 5

Summary of outcomes

Author	Sample	Outcome measure	Outcome (FIM)
Fong et al., 2007	n = 60 with right hemisphere stroke	Patients were assessed on days 0, 30, and 60 using the Behavioral Inattention Test, the Clock Drawing Test, and the FIM.	No significant difference was observed in the sub-scores and total scores of the FIM-motor measure among all groups.
	n = 39 intervention group
	n = 15 control group
Liu et al., 2014	n = 44 with acute unilateral stroke	Performance of tasks, including household and monetary transaction tasks, FIM, Fugl-Meyer Assessment (FMA), and Color Trails Test (CTT).	Significant improvement in the FIM in the intervention group (p < 0.001, r = 0.87 in the motor subscale; p < 0.001, r = 0.49 in the cognitive subscale) compared to the control group.
	n = 24 intervention group
	n = 20 control group
Masiero et al., 2014	n = 34 with first, single, ischemic, or hemorrhagic cerebrovascular accident	Muscle tone (Modified Ashworth Scale), strength (Medical Research Council), and synergism (Fugl-Meyer motor scores) at the impairment level. Dexterity (Box and Block Test and Frenchay Arm Test) and activity of daily living (FIM) at the activity level.	No significant between-group differences in the FIM at baseline, post-intervention, and follow-up were observed.
	n = 14 intervention group
	n = 16 control group
Peiris et al., 2013	n = 996 with orthopedic and neurology disability conditions	FIM, health-related quality of life (EQ-5D), and length of stay.	-Intervention group participants had a higher FIM (mean refinement 2.3, 95% certainty interim (CI) 0.5, 4.1; p = .01) on release and might have had a shorter length of stay by 2 days (95% CI 0, 4; p = 0.1) upon release contrasted with the control group
	n = 160 with stroke		-he intercession gathering was 17% more inclined to have fulfilled a clinically huge change in the FIM of 22 or more FIM centers (danger extent (RR) 1.17, 95% CI 1.03, 1.34) upon release contrasted with the control group.
	n = 81 intervention group
	n = 79 control group
M. Rabadi et al., 2008	n = 30 with ischemic or hemorrhagic stroke	Discharge scores in the Fugl-Meyer Assessment Scale for upper appendage disability. Engine Status Scale. FIM-all out, FIM-engine and FIM-discernment sub-scores.	No significant contrasts in the FIM-downright sub-score (p < .32), FIM-engine sub-score (p < 0.24), FIM-cognizance sub-score (p < .90), and length of stay (p < 0.37) were watched.
	n = 20 intervention group
	n = 10 control group
M. H. Rabadi et al., 2008	n = 116 with ischemic or hemorrhagic stroke	FIM-total, FIM-motor, and FIM-cognition sub-score.	-Significant differences in the FIM-total (p = 0.001) and FIM-motor subscale (p = 0.001) were observed compared to the control group.
	n = 58 intervention group		-No significant difference in the FIM-cognition subscale (p = 0.80) and length of stay (p = 0.77) was observed compared to the control group.
	n = 58 control group
Rawl et al., 1998	n = 100 half had a cerebrovascular accident and one-third had a hip fracture	FIM score, rehospitalization, and anxiety.	There was no significant difference between groups, but FIM scores at discharge and at 4 months showed a comparable improvement in the functional abilities of the subjects over time.
	n = 49 intervention group
	n = 51 control group
Strasser et al., 2008	n = 789 with stroke	FIM and length of stay.	-A significant difference was observed in the improvement of the functional outcome between the 2 groups, with a significantly increased rate of stroke patients increasing more than a middle FIM addition of 23 focuses contrasted with the mediation bunch (distinction in expansion, 13.6%; p = 0.032).
	n = 350 intervention group		- There was no huge distinction in the length of of stay.
	n = 439 control group
Sze et al., 2002	n = 106 with acute stroke	Fugl-Meyer Assessment of Physical Performance (FMA), Barthel Index, Functional Independence Measure (FIM), Abbreviated Mental Test (AMT), and NIH Stroke Scale (NIHSS).	- No factually noteworthy contrasts in the 2 arms for any of the result measures at week 10 or result changes after some time were watched
	n = 53 intervention group
	n = 53 control group
Wikander et al., 1998	n = 34 with acute unilateral hemisphere stroke	Katz ADL index, urinary incontinence, Psychological General Well-Being index (PGWB), mobility, and FIM.	-There were no differences between groups in the ADL (FIM) upon admission.
	n = 21 intervention group		- Twenty patients in the intercession bunch recaptured self control before release contrasted with 3 (p < 0.01) in the control bunch.
	n = 13 control group		- There was likewise a more noteworthy change in the prosperity of the intercession bunch contrasted with the control bunch (p < 0.01).

Abbreviations: FIM = functional independent measure; NA = not available; SD = standard deviation; yr = years; NIH, National Institutes of Health.

Eight RCTs with 1,910 patients were included in MA. The data from all of the trials were combined to analyze four distinctive outcome measures: FIM-motor, FIM-cognitive, and FIM-total at discharge as well as LoS. The settled ESs obtained from the fixed-effects model is introduced in the forest plots. The effects are summarized in the combined forest plot for all of the included articles in Fig. 2. A single plot for each outcome and the full details are available in Appendix 5.

Fig.2

Consolidated forest plot of all outcomes. Abbreviations: FIM, functional independent measure; ES, effect size; CI, confidence interval.

The small ES for FIM-motor (6 RCTs, 1,776 participants; Hedges g ES = 0.13; 95% CI = 0.05, 0.22; p = 0.001), FIM-cognitive (3 RCTs, 166 participants; Hedges g ES = 0.26; 95% CI = 0.09, 0.22; p = 0.002), and FIM-total (5 RCTs, 1,797 participants; Hedges g ES = 0.10; 95% CI = 0.01, 0.22; p = 0.016) at discharge were significant under the fixed-effects theory. The small ES for LoS was also significant (5 RCTs, 1,187 participants; Hedges g ES = 0.14; 95% CI = 0.03, 0.22; p = 0.006). Therefore, favored the IRPs compared to the control and were effective in maximizing the functions and reducing the LoS. The sample sizes for the intervention and control groups were 834 and 947 participants, respectively. No evidence of heterogeneity was observed, yet the discoveries were critical for FIM-motor (Q value = 14.8; p = 0.01; I² = 0%), FIM-cognitive (Q value = 4.08; p = 0.13; I² = 0%), and FIM-total (Q value = 17.09; p = 0.00; I² = 0%) as well as LoS (Q value = 9.67; p = 0.05; I² = 0%).

Generally, there was no distribution predisposition, and the studies did not perform quantitative tests (Classic Fail-safe N tests, Begg and Mazumdar rank correlation test, Egger test of the intercept, and Trim and Fill test) (Sterne, Gavaghan, & Egger, 2000). Because the tests for funnel plot asymmetry are not prescribed when there are fewer than 10 concentrations, a sample funnel plot is displayed in Appendix 6 (Sterne et al., 2011).

4 Discussion

This review aimed to examine the effects of IRPs on the functional status and LoS of older stroke patients. The IRP setting of each included article was a hospital or unit. Our evaluation showed that all of the articles performed adequate random sequence generation. The present review showed that IRPs might be beneficial for maximizing the functions in older adults with stroke. Regarding the overall outcomes, the IRPs showed a short-term effect at discharge.

We used the PEDro scale, a scoring system to evaluate the methodology of the included articles. The advantages of this scoring framework are its convenience and capacity to easily compare the scores from different studies. However, there is no best quality level to confirm the authenticity of these scores, making it hard to figure out if they are legitimately measuring the methodological quality. Also, these scores don’t consider the relative importance of the differing things fused into the scales, and they frequently join things that are not related to the study quality or inclination assessment. The Cochrane review now considers risk of bias apparatuses with regard to the methodological quality evaluation scales (Higgins & Green, 2008).

The judgment of the risk of predisposition was extremely difficult to determine because of the poor or brief reporting of the data. More than half of the area construct assessment using the Cochrane collaboration apparatus was judged to be at low risk of bias for determination, location, and steady loss bias; however, for most of the studies, the risk of bias was judged to be high in blinding participants, therapists, raters and the concealed allocation domain of the Cochrane collaboration tool. In this review, none of the patients from the included articles were blinded to the study and aims, which could confound the results of the studies. In the event that the participants were blinded, the nature of the rehabilitation interventions and the moral prerequisite to obtain informed consent often make it difficult to maintain blinding. For the most part, it is difficult to blind the treating physicians or specialists because they must be acquainted with the intervention that they are controlling. Three of the included articles (Rawl et al., 1998; Strasser et al., 2008; Wikander et al., 1998) did not use a blinded assessor, which can introduce considerable bias into the results of the study. In particular, specialists frequently have solid convictions with regard to their support of a specific methodology. The overall conclusion assessment indicated that the risk of bias was high in all of the included articles. The inclusion of these articles did not affect the results, and the methodological features of the RCTs of rehabilitation interventions were improved.

The results from this review provided strong evidence that the IRPs may maximize functional independence and shorten the LoS in older adults with stroke. A previous SR of the literature on stroke patients aged 18 years or older indicated that the patient’s age, cognition, and functional level were associated with prognosis and rehabilitation admission. This SR included studies in which the methodological quality was poor (Hakkennes, Brock, & Hill, 2011). This is reflected in the International Classification of Function, Disability, and Health from the WHO, which indicates that social and environmental factors also affect a person’s functional status (Stucki et al., 2007). A thorough, basic survey of the therapeutic writing from 1950 to 1998 uncovers considers that were connected with practical result after stroke. The early start of rehabilitation administrations and rehabilitation in an interdisciplinary versus a multidisciplinary setting after stroke seems to have a solid association with enhanced practical result at discharge (Cifu & Stewart, 1999). Thus, age, functional level, and rehabilitation approaches were the main factors that influenced the patient’s maximum functional independence after stroke.

Despite the differences among IRPs across all of the articles included in the MA, the results demonstrate significant effects on the primary and secondary outcomes. To our knowledge, this is the first MA to provide strong evidence of the effects of IRPs on the functional recovery and reduced LoS in older adults with stroke, and the results support its effectiveness. Other MAs have either been exceptionally broad, acquired a solid positive conclusion (Collaboration, 2013; Ferrarello et al., 2011; Luker, Lynch, Bernhardsson, Bennett, & Bernhardt, 2015; Pereira et al., 2012) or have had diverse participants, settings, and results measures; however, none confirmed the effects of IRPs for older adults with stroke. Recently, Pollock et al. (2014) reported that a physical rehabilitation program significantly maximized the functions in adults with stroke aged 18 years or older. The authors measured the functional outcomes and presented them as SMDs and 95% CIs for 10,401 participants using different scales, including the FIM for Activity of Daily Living (ADL). Bachmann et al. (2010) reported that inpatient rehabilitation that is specifically designed for geriatric patients with hip fracture aged 55 years or older showed beneficial effects on the functional improvement. The authors measured the functional outcomes using the Berthel Index for ADL and presented the results as odds ratios and 95% CIs for 4,780 participants. In contrast with the present MA, there was strong evidence that IRPs can maximize functions in older stroke patients. Katrak et al. (2009) presented a review of a prospectively maintained database of all stroke patients admitted to a rehabilitation unit in an Australian university hospital during a 9.5-year period. This review showed that patients achieved significantly greater gains in function after inpatient rehabilitation, which was measured with the FIM scale. These differences in results may be a result of the inclusion criteria that were adopted in this review. In the present review, the mean change score (Appendix 3) was measured for each primary outcome using the FIM and compared using Hedges g and 95% CIs; however, the differences in the data can affect the results (Garg, Hackam, & Tonelli, 2008).

Results like our discoveries have been distributed for older adults with stroke (Liu & Chan, 2014). Schnitzler et al. (2014) conducted a study in France of 28,201 patients aged >65 years who were admitted to a multidisciplinary inpatient rehabilitation (neurological or geriatric rehabilitation) service in 2009. This study affirms, on a national level, the practical advantage of specific rehabilitation in a neurological rehabilitation center. Lee et al. (2012) demonstrated that a prolonged IRP (6 months) achieved functional improvements in stroke patients older than 60. Foley et al. (2012) reported a study of 123 stroke patients admitted to an IRP over a 6-month period and revealed that there was a large increase in FIM focus in the IRP. Granger et al. (2009) provided benchmark data for a vast national specimen of patients who received inpatient rehabilitation following stroke from 893 medical rehabilitation facilities located in the US. These results were obtained from the Uniform Data System for Medical Rehabilitation (UDSMR) from 2000– 2007. The report on stroke rehabilitation showed an improvement in the patients’ functional independence during their rehabilitation stay. Thus, older adults with stroke recover their functionality during inpatient rehabilitation stay.

The other point of this review was to look at the effect of IRPs on LoS. Compared to the control, the IRPs were able to significantly reduce the LoS. Tan et al. (2009) determined that a patient’s socioeconomic status and family structure are more likely to influence LoS. Somerford et al. (2004) published that the LoS is more likely to be longer for female patients, whereas patients who reside in rural areas had a shorter LoS. Brosseau et al. (1996) found that IRPs were indirect predictors of the LoS. Overall, the available literature revealed that IRPs were strongly influenced by the socioeconomic and environmental factors of stroke patients (Ottenbacher & Jannell, 1993; Reistetter et al., 2014).

As discussed above, the identification of all relevant trials was confounded by a number of factors and the search strategy. Therefore, this review may have publication bias towards specific types of studies and distributions. For example, we may be missing the following types of articles: those distributed without the full-text, those distributed in a language other than English, or those distributed in journals that are excluded from the electronic databases that we searched.

The clinical implications of the discoveries in our review warrant further consideration. Our findings suggest that IRPs are effective at maximizing functional independence in older adults with stroke. Thus, these findings indicate a need for IRP development, particularly in older adults with stroke. Because few studies indicate that age is the strongest predictor of functional outcome in an inpatient rehabilitation setting (Abdul-sattar & Godab, 2013; Bagg, Pombo, & Hopman, 2002), additional analyses should be performed to determine whether IPRs truly improve the functional outcome and reduce the LoS in older adults with stroke. If confirmed, this may be a new intervention that improves the outcomes in older adults with stroke, and it should be a high priority for implementation in practice.

A few limitations of this review should to be noted. The total number of included articles in this review was small and limited to older adult patients with stroke. This may decrease its generalizability. Two studies were excluded from the quantitative analysis because the outcomes of these studies were reported as medians and interquartile ranges. If we had included these studies, they may have influenced the results of the MA because the data were not normally distributed. The lack of publication bias underpins the production inclination evaluation. We did not examine the patients’ demographic criteria (e.g., age, sex, race, and socioeconomic status), which can influence the functional outcomes in IRPs. There were differences in the interventions and settings between studies. However, we attempted to address these disparities. The articles that were included in this review were limited to full-text articles written in English; it is possible that some studies may have been missed. Another limitation is that we did not examine the types of inpatient rehabilitations, which can affect functional outcome, in this study. For example, in their critical review of rehabilitation interventions comparing different types of inpatient rehabilitations of stroke care, Cifu and Stewart (1999) demonstrated differences in the functional outcomes and LoS that depended on the setting in which the inpatient rehabilitation is provided. However, we did not include studies that aimed to determine the effect of neglect on functional outcomes in stroke patients. This may have contributed to the variation in the results.

In summary, this SR provided evidence that IRPs have beneficial effects on improving functionality and reducing the LoS for older adults with stroke. Regarding the overall outcomes, IRPs showed a short-term effect at discharge. However, a consensus on the IRP definition and its components is lacking. This methodology provides a significant amount of information with which to guide future practice.

Conflict of interest

The authors declare that there is no conflict of interest.

Footnotes

Appendix 1

An example of the search strategy used for the PubMed database

Limited: English language, human, full-text, and 65+ years.

Appendix 2

Excluded studies (listed alphabetically by author)

Appendix 3

Example of the formulas used to calculate Hedges g using a Microsoft Excel spreadsheet

Appendix 4

Example of computing the combined effect and related statistics for a fixed effect model using a Microsoft Excel spreadsheet

Abbreviations: ES, effect size; SE, standard error; CI, confidence interval; df, degree of freedom; w, weight; k, no. of studies.

Column (Cell)	Label	Excel Formula	Formula
(Section 1) Statistical data
A	Study	Entered
B	Effect size	Entered	(0.1)
C	Standard error	Entered
D	Lower 95% CI	=B – 1.96*C
E	Upper 95% CI	=B + 1.96*C
F	Z-value	=B/C
G	p-Value	=(1-(NORMDIST(ABS(F),0,1,TRUE)))*2
(Section 2) Each study Fixed Effect of WT and WT*ES
H	Variance	=SUMSQ(C)	SE²
I	Variance within	=$H
J	WT	=1/I3	W_i = 1/V_i or 1/SE²
K	ES*WT	=B*J
J9	WT	=SUM(J3:J8)
K9	ES*WT	=SUM(K:K8)
(Section 3) Calculating the joint impact and related statistics for the fixed effect model
I13	Effect size	=K9*J9	T. =Σ^k_i = 1W_i T_i /Σ^k_i = 1W_i
I14	Variance	=1/J9	V = 1/Σ^k_i = 1W_i
I15	Standard error	=SQRT(K14)	SE (T_.) = $\sqrt{V}$ _.
I16	Lower 95% CI	=K13 - 1.96*K15	T_.— 1.96 * SE (T_.)
I17	Upper 95% CI	=K1 + 1.96*K15	T_.+1.96 * SE (T_.)
I18	Z-value	=K13/K15	Z = T_./SE (T_.)
I19	p-value (1-tailed)	=(1-(NORMDIST((K18),0,1,TRUE)))	P=1- φ (Z)
I20	p-value (2-tailed)	=(1-(NORMDIST(ABS(K18),0,1,TRUE)))*2	P = 2[1- φ (_I Z _I)]
(Section 4) Computing the values required for the tau-squared test
L	ES²*WT	=SUMSQ(B)*J
M	WT²	=SUMSQ(J)
L9	SUM	=SUM(L3:L9)
M9	SUM	=SUM(M3:M9)
(Section 5) Computing heterogeneity Q and I² for the fixed model
L13	Q	=L9-SUMSQ(K9)/J9	Q =Σ^k_i = 1W_i T_i – ²/Σ^k_i = 1W_i
L14	Df	=B10-1	Df = (Number of studies) – 1
L15	Numerator	=MAX(M13-M14,0)
L16	C	=J9-M9/J9	C =ΣW_i – ΣW_i²/ΣW_i
L17	Tau-sq	=M15/M16	T² = Q – df/C
L18	I²	=(M1-M14)/M1*100	I² = (Q-df)/Q*100

Abbreviations: FIM, functional independent measure; CI, confidence interval; SUMSQ, sum square root; MAX, maximum; SQRT, square root; Df, degree of freedom; ES, effect size, WT, weight.

Appendix 5

All related statistics and forest plots for the outcomes.

Appendix 6

Example of a funnel plot (FIM-total at discharge outcome)

Acknowledgments

The authors extend their appreciation to the Research Centre, College of Applied Medical Sciences and the Deanship of Scientific Research at King Saud University.

References

Abdul-sattar

A. B.

, & Godab

(2013). Predictors of functional outcome in Saudi Arabian patients with stroke after inpatient rehabilitation. NeuroRehabilitation, 33(2), 209–216. doi: 10.3233/NRE-130947

Bachmann

, Finger

, Huss

, Egger

, Stuck

A. E.

, & Clough-Gorr

K. M.

(2010). Inpatient rehabilitation specifically designed for geriatric patients: Systematic review and meta-analysis of randomised controlled trials. BMJ, 340, c1718. doi: 10.1136/bmj.c1718

Bagg

, Pombo

A. P.

, & Hopman

(2002). Effect of age on functional outcomes after stroke rehabilitation. Stroke, 33(1), 179–185.

Borenstein

(2009). Introduction to meta-analysis. Chichester, U.K.: John Wiley & Sons.

Boutron

, Tubach

, Giraudeau

, & Ravaud

(2004). Blinding was judged more difficult to achieve and maintain in nonpharmacologic than pharmacologic trials. J Clin Epidemiol, 57(6), 543–550 10.1016/j.jclinepi.2003.12.010. doi: 10.1016/j.jclinepi.2003.12.010

Brosseau

, Philippe

, Potvin

, & Boulanger

Y. L.

(1996). Post-stroke inpatient rehabilitation. I. Predicting length of stay. Am J Phys Med Rehabil, 75(6), 422–430.

Chumney

, Nollinger

, Shesko

, Skop

, Spencer

, & Newton

R. A.

(2010). Ability of Functional Independence Measure to accurately predict functional outcome of stroke-specific population: Systematic review. J Rehabil Res Dev, 47(1), 17–29.

Chung

D. M.

, Niewczyk

, DiVita

, Markello

, & Granger

(2012). Predictors of discharge to acute care after inpatient rehabilitation in severely affected stroke patients. Am J Phys Med Rehabil, 91(5), 387–392. doi: 10.1097/PHM.0b013e3182aac27

Cifu

D. X.

, & Stewart

D. G.

(1999). Factors affecting functional outcome after stroke: A critical review of rehabilitation interventions. Arch Phys Med Rehabil, 80(5 Suppl 1), S35–S39.

10.

Collaboration S. U. T. (2013). Organised inpatient (stroke unit) care for stroke. Cochrane Database Syst Rev, 9, CD000197. doi: 10.1002/14651858.CD000197.pub3

11.

Dobkin

B. H.

, Plummer-D’Amato

, Elashoff

, & Lee

(2010). International randomized clinical trial, stroke inpatient rehabilitation with reinforcement of walking speed (SIRROWS), improves outcomes. Neurorehabil Neural Repair, 24(3), 235–242. doi: 10.1177/1545968309357558

12.

Duncan

P. W.

, Zorowitz

, Bates

, Choi

J. Y.

, Glasberg

J. J.

, Graham

G. D.

, … & Reker

(2005). Management of Adult Stroke Rehabilitation Care: A clinical practice guideline. Stroke, 36(9), e100–143. doi: 10.1161/01.STR.0000180861.54180.FF

13.

Easton

J. D.

, Saver

J. L.

, Albers

G. W.

, Alberts

M. J.

, Chaturvedi

, Feldmann

, … & Sacco

R. L.

(2009). Definition and evaluation of transient ischemic attack: A scientific statement for healthcare professionals from the American Heart Association/American Stroke Association Stroke Council; Council on Cardiovascular Surgery and Anesthesia; Council on Cardiovascular Radiology and Intervention; Council on Cardiovascular Nursing; and the Interdisciplinary Council on Peripheral Vascular Disease. The American Academy of Neurology affirms the value of this statement as an educational tool for neurologists. Stroke, 40(6), 2276–2293. doi: 10.1161/STROKEAHA.108.192218

14.

Ferrarello

, Baccini

, Rinaldi

L. A.

, Cavallini

M. C.

, Mossello

, Masotti

, & ...

Di Bari, M.

(2011). Efficacy of physiotherapy interventions late after stroke: A meta-analysis. J Neurol Neurosurg Psychiatry, 82(2), 136–143. doi: 10.1136/jnnp.2009.196428

15.

Foley

, McClure

J. A.

, Meyer

, Salter

, Bureau

, & Teasell

(2012). Inpatient rehabilitation following stroke: Amount of therapy received and associations with functional recovery. Disabil Rehabil, 34(25), 2132–2138. doi: 10.3109/09638288.2012.676145

16.

Foley

N. C.

, Teasell

R. W.

, Bhogal

S. K.

, & Speechley

M. R.

(2003). Stroke rehabilitation evidence-based review: Methodology. Top Stroke Rehabil, 10(1), 1–7.

17.

Fong

K. N.

, Chan

M. K.

, Ng

P. P.

, Tsang

M. H.

, Chow

K. K.

, Lau

C. W.

, … & Chan

(2007). The effect of voluntary trunk rotation and half-field eye-patching for patients with unilateral neglect in stroke: A randomized controlled trial. Clin Rehabil, 21(8), 729–741. doi: 10.1177/0269215507076391

18.

Garg

A. X.

, Hackam

, & Tonelli

(2008). Systematic review and meta-analysis: When one study is just not enough. Clin J Am Soc Nephrol, 3(1), 253–260. doi: 10.2215/CJN.01430307

19.

A. S.

, Mozaffarian

, Roger

V. L.

, Benjamin

E. J.

, Berry

J. D.

, Blaha

M. J.

, … & Turner

M. B.

(2014). Executive summary: Heart disease and stroke statistics–2014 update: A report from the American Heart Association. Circulation, 129(3), 399–410. doi: 10.1161/01.cir.0000442015.53336.12

20.

Granger

C. V.

, Markello

S. J.

, Graham

J. E.

, Deutsch

, & Ottenbacher

K. J.

(2009). The uniform data system for medical rehabilitation: Report of patients with stroke discharged from comprehensive medical programs in 2000-2007. Am J Phys Med Rehabil, 88(12), 961–972. doi: 10.1097/PHM.0b013e3181c1ec38

21.

Guyatt

, Rennie

, Evidence-Based Medicine Working Group, & American Medical Association (2002). Users’ guides to the medical literature: A manual for evidence-based clinical practice. Chicago: AMA Press.

22.

Hakkennes

S. J.

, Brock

, & Hill

K. D.

(2011). Selection for inpatient rehabilitation after acute stroke: A systematic review of the literature. Arch Phys Med Rehabil, 92(12), 2057–2070. doi: 10.1016/j.apmr.2011.07.189

23.

Hardy

R. J.

, & Thompson

S. G.

(1998). Detecting and describing heterogeneity in meta-analysis. Stat Med, 17(8), 841–856.

24.

Hatano

(1976). Experience from a multicentre stroke register: A preliminary report. Bull World Health Organ, 54(5), 541–553.

25.

Higgins

, & Green

(2008). Cochrane handbook for systematic reviews (5.0 ed.). West Sussex, U.K.: Wiley-Blackwell.

26.

Higgins

J. P.

, & Thompson

S. G.

(2002). Quantifying heterogeneity in a meta-analysis. Stat Med, 21(11), 1539–1558. doi: 10.1002/sim.1186

27.

Higgins

J. P.

, Thompson

S. G.

, Deeks

J. J.

, & Altman

D. G.

(2003). Measuring inconsistency in meta-analyses. BMJ, 327(7414), 557–560. doi: 10.1136/bmj.327.7414.557

28.

Higgins

J. P. T.

, Altman

D. G.

, & Sterne

J. A.

(2011). Assessing risk of bias in included studies. In Higgins

J.P.T.

, & G. S (Eds.), Cochrane handbook for systematic reviews of interventions (5.1.0 ed.). West Sussex, U.K.: Wiley-Blackwell.

29.

Horwitz

R. I.

(1995). Large-scale randomized evidence: Large, simple trials and overviews of trials”: Discussion. A clinician’s perspective on meta-analyses. J Clin Epidemiol, 48(1), 41–44.

30.

Katrak

P. H.

, Black

, & Peeva

(2009). Do stroke patients with intracerebral hemorrhage have a better functional outcome than patients with cerebral infarction? PM R, 1(5), 427–433. doi: 10.1016/j.pmrj.2009.03.002

31.

Knobloch

, Yoon

, & Vogt

P. M.

(2011). Preferred reporting items for systematic reviews and meta-analyses (PRISMA) statement and publication bias. J Craniomaxillofac Surg, 39(2), 91–92. doi: 10.1016/j.jcms.2010.11.001

32.

Lee

J. H.

, Kim

S. B.

, Lee

K. W.

, & Lee

J. Y.

(2012). The effect of prolonged inpatient rehabilitation therapy in subacute stroke patients. Ann Rehabil Med, 36(1), 16–21. doi: 10.5535/arm.2012.36.1.16

33.

Lindsay

K.W.

, Langham

, Kirkpatrick

P.J.

, Shaw

M.D.M.

, Gholkar

A.R.

, Molyneux

, … & Horrocks

, (2006). National study of subarachnoid hemorrhage: Final report of an audit carried out in 34 neurosurgical units in the UK and Ireland between 14 September 2001 to 13 September 2002. London: The Royal College of Surgeons of England.

34.

Liu

K. P.

, & Chan

C. C.

(2014). Pilot randomized controlled trial of self-regulation in promoting function in acute poststroke patients. Arch Phys Med Rehabil, 95(7), 1262–1267. doi: 10.1016/j.apmr.2014.03.018

35.

Luker

, Lynch

, Bernhardsson

, Bennett

, & Bernhardt

(2015). Stroke survivors’ experiences of physical rehabilitation: A systematic review of qualitative studies. e. Arch Phys Med Rehabil, 96(9), 1610–. doi: 10.1016/j.apmr.2015.03.017

36.

Mackay

, & Mensah

(2004). Atlas of Heart Disease and Stroke. Geneva, Switzerland: World Health Organization.

37.

Maher

C. G.

, Sherrington

, Herbert

R. D.

, Moseley

A. M.

, & Elkins

(2003). Reliability of the PEDro scale for rating quality of randomized controlled trials. Phys Ther, 83(8), 713–721.

38.

Masiero

, Armani

, Ferlini

, Rosati

, & Rossi

(2014). Randomized trial of a robotic assistive device for the upper extremity during early inpatient stroke rehabilitation. Neurorehabil Neural Repair, 28(4), 377–386. doi: 10.1177/1545968313513073

39.

Moseley

A. M.

, Herbert

R. D.

, Sherrington

, & Maher

C. G.

(2002). Evidence for physiotherapy practice: A survey of the Physiotherapy Evidence Database (PEDro. Aust J Physiother, 48(1), 43–49.

40.

Nadeau

S. E.

, Wu

S. S.

, Dobkin

B. H.

, Azen

S. P.

, Rose

D. K.

, Tilson

J. K.

, … & Duncan

P. W.

(2013). Effects of task-specific and impairment-based training compared with usual care on functional walking ability after inpatient stroke rehabilitation: LEAPS Trial. Neurorehabil Neural Repair, 27(4), 370–380. doi: 10.1177/1545968313481284

41.

Neyeloff

J. L.

, Fuchs

S. C.

, & Moreira

L. B.

(2012). Meta-analyses and forest plots using a Microsoft Excel spreadsheet: Step-by-step guide focusing on descriptive data analysis. BMC Res Notes, 5, 52. doi: 10.1186/1756-0500-5-52

42.

Ottenbacher

K. J.

, & Jannell

(1993). The results of clinical trials in stroke rehabilitation research. Arch Neurol, 50(1), 37–44.

43.

Peiris

C. L.

, Shields

, Brusco

N. K.

, Watts

J. J.

, & Taylor

N. F.

(2013). Additional Saturday rehabilitation improves functional independence and quality of life and reduces length of stay: A randomized controlled trial. BMC Med, 11, 198–. doi: 10.1186/1741-7015-11-198

44.

Pereira

, Graham

J. R.

, Shahabaz

, Salter

, Foley

, Meyer

, & Teasell

(2012). Rehabilitation of individuals with severe stroke: Synthesis of best evidence and challenges in implementation. Top Stroke Rehabil, 19(2), 122–131. doi: 10.1310/tsr1902-122

45.

Pollock

, Baer

, Campbell

, Choo

P. L.

, Forster

, Morris

, … & Langhorne

(2014). Physical rehabilitation approaches for the recovery of function and mobility following stroke. Cochrane Database Syst Rev, 4, CD001920. doi: 10.1002/14651858.CD001920.pub3

46.

Rabadi

, Galgano

, Lynch

, Akerman

, Lesser

, & Volpe

(2008). A pilot study of activity-based therapy in the arm motor recovery post stroke: A randomized controlled trial. Clin Rehabil, 22(12), 1071–1082. doi: 10.1177/0269215508095358

47.

Rabadi

M. H.

, Coar

P. L.

, Lukin

, Lesser

, & Blass

J. P.

(2008). Intensive nutritional supplements can improve outcomes in stroke rehabilitation. Neurology, 71(23), 1856–1861. doi: 10.1212/01.wnl.0000327092.39422.3c

48.

Rawl

S. M.

, Easton

K. L.

, Kwiatkowski

, Zemen

, & Burczyk

(1998). Effectiveness of a nurse-managed follow-up program for rehabilitation patients after discharge. Rehabil Nurs, 23(4), 204–209.

49.

Reistetter

T. A.

, Karmarkar

A. M.

, Graham

J. E.

, Eschbach

, Kuo

Y. F.

, Granger

C. V.

, … & Ottenbacher

K. J.

(2014). Regional variation in stroke rehabilitation outcomes. Arch Phys Med Rehabil, 95(1), 29–38. doi: 10.1016/j.apmr.2013.07.018

50.

Safer

V. B.

, & Koseoglu

B. F.

(2015). Timing of inpatient rehabilitation initiation in stroke patients: Factors influencing early admission. J Phys Ther Sci, 27(6), 1913–1917. doi: 10.1589/jpts.27.1913

51.

Schnitzler

, Woimant

, Nicolau

, Tuppin

, & de Peretti

(2014). Effect of rehabilitation setting on dependence following stroke: An analysis of the French inpatient database. Neurorehabil Neural Repair, 28(1), 36–44. doi: 10.1177/1545968313497828

52.

Schulz

K. F.

, Altman

D. G.

, & Moher

(2011). CONSORT 2010 statement: Updated guidelines for reporting parallel group randomised trials. Int J Surg, 9(8), 672–677. doi: 10.1016/j.ijsu.2011.09.004

53.

Somerford

P. J.

, Lee

A. H.

, & Yau

K. K.

(2004). Ischemic stroke hospital stay and discharge destination. Ann Epidemiol, 14(10), 773–777.

54.

Sterne

J. A.

, Gavaghan

, & Egger

(2000). Publication and related bias in meta-analysis: Power of statistical tests and pre-valence in the literature. J Clin Epidemiol, 53(11), 1119–1129.

55.

Sterne

J. A.

, Sutton

A. J.

, Ioannidis

J. P.

, Terrin

, Jones

D. R.

, Lau

, … & Higgins

J. P.

(2011). Recommendations for examining and interpreting funnel plot asymmetry in meta-analyses of randomised controlled trials. BMJ, 343, d4002. doi: 10.1136/bmj.d4002

56.

Strasser

D. C.

, Falconer

J. A.

, Stevens

A. B.

, Uomoto

J. M.

, Herrin

, Bowen

S. E.

, & Burridge

A. B.

(2008). Team training and stroke rehabilitation outcomes: A cluster randomized trial. Arch Phys Med Rehabil, 89(1), 10–15. doi: 10.1016/j.apmr.2007.08.127

57.

Stucki

, Cieza

, & Melvin

(2007). The International Classification of Functioning, Disability and Health (ICF): A unifying model for the conceptual description of the rehabilitation strategy. J Rehabil Med, 39(4), 279–285. doi: 10.2340/16501977-0041

58.

Sze

F. K.

, Wong

, Yi

, & Woo

(2002). Does acupuncture have additional value to standard poststroke motor rehabilitation? Stroke, 33(1), 186–194.

59.

Tan

W. S.

, Heng

B. H.

, Chua

K. S.

, & Chan

K. F.

(2009). Factors predicting inpatient rehabilitation length of stay of acute stroke patients in Singapore. Arch Phys Med Rehabil, 90(7), 1202–1207. doi: 10.1016/j.apmr.2009.01.027

60.

Treadwell

J. R.

, Singh

, Talati

, McPheeters

M. L.

, & Reston

J. T.

(2011). A Framework for “Best Evidence” Approaches in Systematic Reviews. Rockville (MD).

61.

Tummers

J. F.

, Schrijvers

A. J.

, & Visser-Meily

J. M.

(2012). Economic evidence on integrated care for stroke patients; a systematic review. Int J Integr Care, 12, e193.

62.

Viera

A. J.

, & Garrett

J. M.

(2005). Understanding interobserver agreement: The kappa statistic. Fam Med, 37(5), 360–363.

63.

Wikander

, Ekelund

, & Milsom

(1998). An evaluation of multidisciplinary intervention governed by functional independence measure (FIMSM) in incontinent stroke patients. Scand J Rehabil Med, 30(1), 15–21.

64.

Yavuzer

, Kucukdeveci

, Arasil

, & Elhan

(2001). Rehabilitation of stroke patients: Clinical profile and functional outcome. Am J Phys Med Rehabil, 80(4), 250–255.