The Clinical, Quality of Life and Economic Outcomes of Inpatient Rehabilitation: A Systematic Review

title not relevant.


Introduction
Over a billion people, about 15% of the world's population, have some form of disability either due to injury or acute and chronic diseases [1]. Between 110 million and 190 million adults have significant difficulties in functioning. Rates of disability are increasing due to population ageing and raises in the prevalence of chronic health conditions, among other causes. Disability has a negative impact on social development and economic development [1].
Rehabilitation is instrumental in enabling people with limitations in functioning, to remain in or return to their home or community, live independently, and participate in education, the labour market and civil life. Access to rehabilitation can decrease the consequences of disease or injury, improve health and quality of life and decrease the use of health services [2].
Physical rehabilitation is a medical specialty focused on prevention, diagnosis, and therapy for patients who experience functional limitations resulting from injury, disease, or malformation. The benefits of rehabilitation could be clinical-physical, neurological, and cognitive related improvements-, functional-motor related improvements and economic-including patient's work productivity [3,4]. Rehabilitation programs can be provided in alternative settings including an acute hospital, sub-acute hospital, specialist facilities (inpatient or outpatient), or the patient's home.
While many countries have started taking action to improve the lives of people with disabilities, much remains to be done [2]. Increased collaboration amongst rehabilitation professionals in developed and developing countries is essential to implement appropriate and sustainable rehabilitative services.
In Greece, rehabilitation services are provided mainly by private specialized institutions, even though, there are also some integrated services in general hospital care public facilities. However, the provided services are fragmented, underdeveloped, underfunded and in many cases inadequate and much more attention needs to be paid to this particular medical specialty.
In this light, the objective of the present study was to systematically review the literature reporting evaluations on the clinical, quality of life, and economic benefits of inpatient rehabilitation for patients suffering from stroke, spinal cord injury (SCI), and multiple sclerosis (MS). the amount of information reported in the full-text continued to be insufficient to make a decision about inclusion, the studies were excluded. The study selection process was documented through a flow chart showing the number of studies/papers remaining at each stage.

Data extraction
A standardized data extraction form for each health condition (stroke, SCI, MS), developed for the purpose of this review, was used by the two reviewers to extract the data independently. Any disagreement in the data extraction form between the two reviewers was resolved through discussion between these two or by involving a third independent researcher. The aforementioned extraction form was designed to include data on the background information of the study, its methodological characteristic, and the key results.

Data synthesis
In this systematic review, the results are summarized in a qualitative manner collating data from studies. We synthesized the relevant and available data in a systematic manner following the review question, the inclusion and exclusion criteria.

Study selection
After removing duplicate citations, 1,764 unique citations remained for screening. The manual screening of all titles and abstracts yielded 84 articles that contained information about the benefits of inpatient rehabilitation. Of the latter full articles retrieved and reviewed by the investigators, 40 met the inclusion criteria. The reference lists of all relevant papers originally selected for inclusion in the review and relevant reviews were also searched manually to identify potentially relevant articles which were not identified by the original electronic search. Consequently, six additional studies of interest were collected in full text with agreement for inclusion in the systematic review, taking the total to 46. Details of literature search strategy are shown in Figure 1.
Overall 41 out of the 46 studies examined the clinical benefits of inpatient rehabilitation (16 for stroke patients, 15 for SCI patients, seven for MS patients, and three for mixed population of stroke and SCI patients) and five studies assessed the economic benefits for post-stroke patients. Notably, there was significant heterogeneity in terms of study designs and in the way that functionality was measured.

Stroke
Overall, 24 studies examined the impact of inpatient rehabilitation on stroke patients (Table 3).

Clinical outcomes
Four studies [7][8][9][10] assessed functional disability in stroke patients using the Barthel Index (BI), which is a standardized and well validated method of measuring a patient's level of physical independence. In  methodological and clinical expertise. Τhe literature search was conducted using three different combinations of keywords for: rehabilitation; outcomes and health condition, as presented in Table 1. The terms in the three major categories were combined by the Boolean "AND", whilst the terms utilized within each of the search categories were combined by the Boolean "OR". The filters "English" and "Humans" were added as to restrict our search to the relevant studies. There was no search limitation in terms of time and geographical location of the original studies. The search was limited to studies published up to December 2015. The Appendix presents the full search strategy used for MEDLINE, which was adapted appropriately for the rest of the databases.

Study selection
Consequently, the identified studies were reviewed and assessed for inclusion in the review by two independent researchers, based on the predetermined inclusion/exclusion criteria presented in Table  2. Clinical trials were excluded since the review was focused on realworld evidence data. Therefore, as presented in Table 2, observational studies were taken into consideration. The study selection procedure encompassed two stages: initially, all the identified studies were evaluated on the basis of titles and/or abstracts against the eligibility criteria; in the second stage, when the information provided by titles/ abstracts was insufficient to decide on inclusion/exclusion, or when the titles/abstracts indicated that the specific studies met the inclusion criteria, the full-papers were retrieved to be screened. In cases where all of these studies, patients demonstrated a statistically significant improvement in the BI score at discharge score relative to the one at admission.
Two studies used the Motor Assessment Scale, Item 6 Upper Arm Function (MAS6) to measure the upper arm disability [17,22]. In the first study, 83% of stroke patients demonstrated a statistically significant improvement in arm function at rehabilitation discharge, while 68% achieved a shift from severe to mild/moderate upper arm disability on discharge. The second study reported that 45% of patients had a statistically significant change in arm function recovery at rehabilitation discharge.
In addition, the study by Ee et al. [23] indicated that the percentage of totally dependent post stroke patients was statistically significantly lower in terms of the Rehabilitation Profile System (RPS) at discharge. Moreover, Gialanella et al. [24] demonstrated that patients had statistically significant improved mobility, measured by the Lindmark and the Rivermead Mobility Index (RMI), as well as neurological status, measured in terms of the National Institute of Health Stroke Scale (NIH). It is worth noting that 80.5% of patients were ambulatory independent at discharge contrary to 1.4% on admission, an impressive outcome improvement.
Four studies evaluated the long-term benefits of inpatient rehabilitation [8,21,25,26]. Sim et al. [8] demonstrated that the gains in patients' functional status were generally maintained one year after discharge, with a further statistically significant improvement in toileting. Furthermore, Mutai et al. [25] reported that 51.9% were classified as independent in terms of their Activities of Daily Living (ADL), 1-5 years after discharge. According to Mahler et al. [26] the

1.764
Articles identified through database searching (duplicates excluded).

1.687
Articles excluded after screening from abstract/title -not relevant.

84
Articles for further assessment (if full text available, relevant full text).

40
Studies excluded: 24 did not fit inclusion criteria.
3 irrelevant aim of interest.

insufficient results.
3 irrelevant outcomes of interest.
6 with abstract only available.

44
Articles accepted for review.

2
Additional articles identified through other sources (e.g. bibliographies of screened studies or relevant systematic reviews).

46
Studies included.    percentage of independent post stroke patients who underwent inpatient rehabilitation reached 81% as compared to 51% of the patients without inpatient rehabilitation, one year after stroke. Moreover, Graham et al. [21] indicated that 79.7% of patients were successfully maintaining their health status by themselves, 3-6 months after discharge.
Finally, Mahler et al. [26] assessed the effectiveness of inpatient rehabilitation as compared to traditional treatment on stroke patients. The patients who underwent the rehabilitation program had their BI score increased by 42 ± 29 points, as compared to patients without (inpatient) rehabilitation, whose functional level rose by 23 ± 26 points only (p<0.05).

Quality of life outcomes
Quality of life was assessed in three studies [9,18,25]. The study by Mutai et al. [25] indicated that 21.6% of patients suffered from depression 1-3 years after stroke. On the other hand, Kuptniratsaikul et al. [9] highlighted that the number of patients with anxiety and depression was statistically significant lower at discharge. More specifically, 25.5% of patients had anxiety and 37.8% had depression on admission. At discharge, the percentages of patients with anxiety and depression decreased to 6.8% and 16.3%, respectively. In addition, the same study reported that the quality of life scores at discharge were significantly higher than those on admission. Finally, the study by Madden et al. [18] reported that the mean improvement between admission and discharge at the SF-36 (patient-reported survey of patient health) scores was statistically significant.

Economic outcomes
A significant proportion of patients with stroke returned home after discharge. More specifically, four studies [14][15][16]18] reported that the percentage of patients who returned to their home (without further institutionalization 74%, 84%, 43.4%, and 81% respectively. The percentage of the third study [16] is much lower than the others because it referred to patients with severe stroke who were totally ambulatory dependent. According to the cost analysis of Mahler et al. [26], inpatient rehabilitation is the most significant part of the total health insurance costs in the first year after stroke (37%). However, inpatient rehabilitation's crucial benefit related to the high percentage of independent patients after one year (81%), which in turn may be associated with reduced health care long-term costs. Moodie et al. [27] compared costs and outcomes of stroke patients who received either conventional care or mobile service or stroke unit care (as below). The study demonstrated that although acute Stroke Care Unit (SCU) was more expensive, it was found to be cost-effective compared to a mobile service or conventional care. Khiaocharoen et al. [28] who conducted a cost-utility analysis of rehabilitation for stroke patients in Thailand, concluded that inpatient rehabilitation services for stroke survivors were cost-effective as compared with conventional care. Patel et al. [29] highlighted that the percentages of patients who avoided death/ institutionalization were 87%, 69%, and 78% in the stroke unit, stroke team, and domiciliary care groups, respectively. Finally, Andersson et al. [30] compared the outcomes of two rehabilitation groups, hospitaland home-based respectively. Although the home-based group had significantly lower costs, the number of acute care ward days after a decision about rehabilitation was made was three days in the hospitalbased group and nine in the home-based group and the difference was significant. The hospital-based group thereafter had a mean duration of 28 in-hospital rehabilitation days and the home-based group had 36 days of home rehabilitation (Table 3).

SC
Overall 18 studies examined the impact of inpatient rehabilitation on SCI patients (Table 4).

Clinical outcomes
Five studies assessed physical and cognitive disability with the FIM scale [19][20][21]31,32]. Two of them reported a statistically significant improvement in patients' functional status from rehabilitation admission to rehabilitation discharge [31,32]. The remaining three studies indicated that the patients' total FIM score (physical and  cognitive) was significantly higher at rehabilitation discharge [19][20][21]. Additionally, Graham et al. [21] reported that follow-up (3-6 months) FIM total ratings remained from table to slightly increased over time in 75.4% of patients.
Two studies indicated that the improvement of patients' functional ability was statistically significant after admission to rehabilitation program, as measured by the BI scale [33,34]. Furthermore, the study by Scivoletto et al. [34] demonstrated that all functional and neurological scales showed statistically significant improvements in SCI patients, despite the delayed onset, of rehabilitation treatment.
Four studies reported that a great number of patients showed a significant improvement in ambulation and achieved independence or assisted dependence in walking at rehabilitation discharge [34][35][36][37]. More specifically, Scivoletto et al. [34] reported that at admission only 11% of patients were able to walk independently relative to 41% at discharge whereas the same percentages were reported to be 5.3% and 45.2% respectively in the study by Yen et al. [37] Regarding bladder status [34,35,37], self-care [35], and activities of daily living [37], inpatient rehabilitation had a significantly positive impact on patients' ability to perform independently the aforementioned activities.

Quality of life outcomes
Franceschini et al. [38] presented data indicating that SCI patients reported to be satisfied with their current quality of life (6,5 QoL score, 10 max score) and that 48.6% were satisfied with their partner relationships, 6 years after rehabilitation discharge. Additionally, 67% of patients were satisfied with their quality of life, 2-12 years after rehabilitation discharge, as reported by Schonherr et al. [39]. With regard to sports and hobbies, 86.3% of patients had at least one hobby 8-15 years after rehabilitation discharge whereas 41% of them were still active in sport, as reported by van Asbeck et al. [40].

Economic outcomes
As far as productivity loss is concerned, eight studies evaluated patients' ability to return to some form of vocation within a reasonable period of time after injury [37][38][39][40][41][42][43][44]. More specifically, Yen et al. [37] indicated that 21.6% of SCI patients returned to some form of vocation one year post-injury while Franceschini et al. [38] showed that 29.5% were employed six years post-injury. The study by Schonherr et al. [39] demonstrated that most people with SCI were able to resume work 2-12 years after injury. In particular, 60% of patients had a job at the time of follow-up. Franceschini et al. [41] reported that 42.1% of SCI patients were employed at the time of follow-up (3.8 years). Finally, 32.4% of patients were employed and 36.7% were housekeeping 8-15 years after rehabilitation as reported by van Asbeck et al. [40].
Three studies in where all the patients were employed at the time of injury demonstrated that the percentages of them who were able to return to paid work for at least 1 hour/week within 5 years after discharge from inpatient rehabilitation were 33%, 44.7%, and 50.9% respectively.
According to the study by Scivoletto et al. [34] 90% of patients who underwent rehabilitation returned to their home while Citterio et al. [45] and Yen et al. [37] reported those percentages to be 73% and 87.9% respectively.
Finally, there were no economic evaluation studies identified that assessed inpatient rehabilitation's outcomes for SCI patients (Table 4).

Multiple Sclerosis
Overall, seven studies examined the impact of inpatient rehabilitation on MS patients (Table 5).

Clinical outcomes
Two studies found that the patients' discharge neurological status was not significantly different from the admission's as evaluated by means of Expanded Disability Status Scale (EDSS) [46,47]. According to the study by Freeman et al., improvements were maintained in disability (Functional Independent Measure) and handicap (London Handicap Scale) for 6 months after discharge but neurological status (EDSS) demonstrated a gradual deterioration within 1 year after discharge [48]. The study by Kiddetal indicated that 17% of MS patients were improved on the EDSS [49], while Aisenetal reported that MS patients were achieved statistically significant improvement between admission and discharge EDSS mean scores [50].  Two studies showed that statistically significant improvements occurred to MS patients as evaluated by means of the ΒΙ [46,47]. Five studies used the FIM to assess MS patients in terms of functional ability [48][49][50][51][52]. In four of the aforementioned studies (the fifth is a cohort study [51] the FIM score at discharge was higher than the one at admission, while in two of them [50,52] the FIM gain is statistically significant. It is worth noting that significant improvements also occurred in FIM subgroupings: self-care (eating, dressing, grooming, bathing), sphincter control (bladder, bowel), and locomotion (ambulation, stair climbing, wheel chair management) for all patients [50].

Background information of study
Additionally, two studies highlighted the statistically significant improvement that MS patients demonstrated in functional independence and disability, as evaluated by means of the Functional Systems, the Rivermead Mobility Index, and the London Handicap Scale [47,50].
Finally, according to Freeman et al. [51], MS patients with the same neurological status were randomized to a treatment or a control group. In terms of disability and handicap level improvements, there was a statistically significant difference between the two groups, 6 weeks later. In terms of the percentage of patients who improved, deteriorated, or remained the same, overall, 53%of the treatment group had improved their total handicap score, 3% remained the same, and 44% deteriorated. In contrast, 23% of the control group improved, 12% stayed the same, and 65% deteriorated.

Quality of life outcomes
With regard to health related quality of life measurement, in terms of the SF-36, Freeman et al. [48] reported that 54% of patients achieved maximum scores at 3 months after discharge and 28.2% at 6 months (in the physical dimension). In contrast, in the mental dimension, 21% of patients peaked at 3 months, with most (61%) peaking at 6 months.

Economic outcomes
Finally, there were no economic evaluation studies identified that assessed inpatient rehabilitation's outcomes for MS patients (Table 5).

Discussion
We systematically reviewed the available literature containing studies that evaluated the clinical, functional, and economic benefits of inpatient rehabilitation for stroke, SCI, and MS patients. This study is important as it may provide insights into the evidence produced so far and what needs further research and future studies. We identified and included 46 articles in the review. Although the types of methodologies, measures and populations studied varied widely, we were able to identify clear health and economic benefits stemming from physical inpatient rehabilitation both for patients and healthcare systems.
In particular, there was strong evidence supporting the functional and neurological benefits of post-acute inpatient rehabilitation for all patient groups and situations. Also, there was moderate evidence to report that patients had a statistically significant gain in health-related quality of life outcomes. Finally, it was shown that the gains in patients' functional and disability status were generally maintained after discharge except for the MS patients in whom neurological status demonstrated a gradual deterioration after rehabilitation discharge over time. Moreover, the evidence indicates that the effectiveness of inpatient rehabilitation may be influenced by factors such the age of patients, their medical history, socio-economic status and onset of rehabilitation.
Our findings are in line with these presented in previously conducted systematic reviews which examined specifically the outcomes of inpatient rehabilitation on stroke [53,54], SCI [55], and MS [56] patients. More specifically, the study by Knecht et al. [54] reported that well-organized acute and intermediate rehabilitation after stroke can provide patients with the best functional results. Furthermore, the study by Lam et al. [55] showed that inpatient rehabilitation focused on gait training can offer the greatest benefits to functional ambulation in sub-acute or chronic spinal cord injury. Moreover, a study by Khan et al. [56] indicated that inpatient rehabilitation does not change the level of impairment, but can improve the experience of people with multiple sclerosis in terms of activity and participation.
In terms of the economics, there is very scarce evidence. Notwithstanding, the low number of studies, it appeared that in certain settings rehabilitation may be cost-effective in patients with stroke and spinal injury. There were no studies available for multiple sclerosis patients.
In terms of the studies available, it appears that the majority of studies assessing the effectiveness of inpatient rehabilitation were prospective, with most of them, up to a year. Moreover, there is no consistency in terms of how effectiveness was quantified as many different measures were utilized. Finally, it should be noted that programs were not standardized and were also delivered in different settings. In terms of the economic studies, from a methodological point of view, most of them are short term and they mainly focus on the health care system, based on cost-effectiveness or cost minimization modelling. Hence, they may underestimate the economic benefits of rehabilitation as it is associated with longer economic benefits for the health system due to resource utilization reductions and indirect benefits for the economy and society due to higher productivity, superior functioning and return to employment. Hence, long term cost-benefit analyses are more appropriate for evaluating it. Therefore, despite the availability of several studies in the field concerning the effectiveness of inpatient rehabilitation on the three health conditions of interest, it is evident that there is lack of economic evaluations and long term studies whereas there is increased variability in terms of the outcomes considered. Therefore, further research is required in order to establish more vividly the benefits of inpatient rehabilitation and influence decision making and patient management.
The results of this review must be interpreted in light of the methodological pitfalls of studies of this kind. We should acknowledge the possibility of publication bias due to the fact that only published studies, written in English language, were incorporated in our review. In addition, the search was limited to free databases. Moreover, the studies which were identified in this review covered a wide range of methodologies, outcome measures, and patient populations and consequently the heterogeneity of these studies prevented us from any quantitative estimates, of the overall benefits of inpatient rehabilitation and from performing a formal meta-analysis. Also, our review did not take into account information such as severity of disease, intensiveness of intervention, and length of stay. Furthermore, the information regarding the perspective of economic evaluation studies is not available. Finally, it should be also acknowledged that, unlike other treatments such as drugs, rehabilitation is not homogenous and standard therapy across different settings, and often data on the content and related information on rehabilitation programs evaluated is missing or differs across studies.

Conclusion
Despite the heterogeneity of outcomes and the limitations of this systematic review, there is abundant and clear evidence supporting the effectiveness and benefits of inpatient rehabilitation. In summary, inpatient rehabilitation improves clinical outcomes for patients with disability or impairment due to stroke, spinal cord injury, and multiple sclerosis. There is also scarce evidence that inpatient rehabilitation may be cost saving or highly cost-effective, especially for patients with stroke. Additional effectiveness and economic evaluation studies may contribute more to the evidence supporting the issue of rehabilitation for patients cost to inform policy and decision making and to improve patient access and outcomes of therapy.