Effectiveness and feasibility of early physical rehabilitation programs for geriatric hospitalized patients: a systematic review

Background Old adults admitted to the hospital are at severe risk of functional loss during hospitalization. Early in-hospital physical rehabilitation programs appear to prevent functional loss in geriatric patients. The first aim of this review was to investigate the effect of early physical rehabilitation programs on physical functioning among geriatric patients acutely admitted to the hospital. The second aim was to evaluate the feasibility of early physical rehabilitation programs. Methods Two searches, one for physical functioning and one for feasibility, were conducted in PubMed, CINAHL, and EMBASE. Additional studies were identified through reference and citation tracking. To be included articles had to report on in-hospital early physical rehabilitation of patients aged 65 years and older with an outcome measure of physical functioning. Studies were excluded when the treatment was performed on specialized units other than geriatric units. Randomized controlled trials were included to examine the effect of early physical rehabilitation on physical functioning, length of stay and discharge destination. To investigate feasibility also non randomized controlled trials were added. Results Fifteen articles, reporting on 13 studies, described the effect on physical functioning. The early physical rehabilitation programs were classified in multidisciplinary programs with an exercise component and usual care with an exercise component. Multidisciplinary programs focussed more on facilitating discharge home and independent ADL, whereas exercise programs aimed at improving functional outcomes. At time of discharge patients who had participated in a multidisciplinary program or exercise program improved more on physical functional tests and were less likely to be discharged to a nursing home compared to patients receiving only usual care. In addition, multidisciplinary programs reduced the length of hospital stay significantly. Follow-up interventions improved physical functioning after discharge. The feasibility search yielded four articles. The feasibility results showed that early physical rehabilitation for acutely hospitalized old adults was safe. Adherence rates differed between studies and the recruitment of patients was sometimes challenging. Conclusions Early physical rehabilitation care for acutely hospitalized old adults leads to functional benefits and can be safely executed. Further research is needed to specifically quantify the physical component in early physical rehabilitation programs.


Background
The rapidly growing population of old adults in Western countries has become a major concern for health care systems. Due to a poorer health status, old adults consume a disproportionate amount of medical care. In some European countries, more than 40% of patients admitted to the hospital for an overnight stay are aged 65 years and older, while their total share of the population is less than 20% [1].
Unfortunately, old adults admitted to the hospital are at severe risk of functional decline, both during hospitalization and after discharge [2,3]. A number of studies found that approximately 33% of the patients have severe functional deterioration at time of discharge compared to their status before hospital admission [3][4][5]. For patients 90 years or older this number even increases to 63% [4]. Functional decline during and after hospital stay has shown to be an important risk factor for nursing home placement [6,7].
The decline in functional capacity seems to be partly the result of the hospitalization itself, unrelated to diagnostic or therapeutic interventions. Older patients have decreased physiological and functional reserves that make them more vulnerable to the effects of bed rest and decrease in dietary intake, which both are highly prevalent during hospitalization. Due to immobilisation, muscle strength and aerobic capacity tend to decline rapidly. After only ten days of bed rest healthy old adults lose 12-14% of both their VO 2max and lower extremity muscle strength [8]. Without any voluntary muscle contractions muscle strength can even decrease by 5% per day [2].
Altogether, functional decline is a common problem that is significantly associated with negative outcomes such as institutionalization, re-hospitalization and subsequent mortality [3]. The primary focus of hospital care is treating acute and chronic illnesses. A physical rehabilitation intervention that may preserve physical function is often not part of the treatment. To preclude a rapid decline in physical function it is important that hospital programs are also directed explicitly towards activating the older patient early after hospital admission. Early physical rehabilitation might help to prevent decline in physical functioning arising from immobility and prolonged bed rest [9].
Over the years, several multidisciplinary and exercise types of early rehabilitation interventions have been studied. Previous studies showed that early rehabilitation programs improved both patient (e.g. physical functioning) and hospital outcomes (e.g. reducing costs) for acute ill geriatric patients [10,11]. However, an important issue not yet addressed in the current literature is the feasibility of in-hospital early exercise programs for acute geriatric patients. To start an early physical rehabilitation program, knowledge of which patient population benefit from the program is required. Furthermore, it is important to know if there will be adverse events during early physical rehabilitation programs in terms of falls or other injuries and what the adherence rate of the patients will be during the treatment sessions. Therefore, the first aim of this review is to evaluate the effects of early physical rehabilitation programs on physical functioning of geriatric patients acutely admitted to a hospital. In this review early physical rehabilitation in acutely ill patients refers to physical therapy, occupational therapy, and physical exercises initiated immediately upon achieving physiologic stability and continued throughout the hospital stay. Such activities start within 1 or 2 days after hospitalization. The second aim of the present review is to evaluate the feasibility of early physical rehabilitation programs in the hospitalized geriatric patients.

Search strategy
A systematic literature search was conducted in three electronic databases, PubMed, CINAHL, and EMBASE in August 2013. Keywords used to perform the search were: "aged" or "elderly" or "geriatric", "hospital" or "hospitalized", "exercise" or "rehabilitation", "ADL" or "physical functioning" or "mobility" or "physical performance", and "acute" or "acutely". The articles included were randomized controlled trials (RCT), written in English, including participants aged 65 years or older who were acutely admitted to the hospital. The interventions investigated in the articles needed to include a physical exercise part with a physical functioning measure as outcome. Studies were excluded if the included patients required treatment on a specialized unit other than an acute geriatric unit or when the evaluated intervention aimed at a specific disorder or surgical process. Articles about the feasibility of early physical rehabilitation of inpatients were retrieved by adding the keywords "feasibility" or "feasible" or "adherence rate" or "safety" to the keywords mentioned above. The articles about feasibility had the same in and exclusion criteria as the articles addressing physical functioning with the exception that also non randomized controlled trials were included. For the inclusion process title and abstract were examined and when necessary the full article was obtained and read. Additional studies were identified through reference and citation tracking. Two reviewers independently screened title, abstract and full text. Disagreement about inclusion of articles was resolved by discussion and consensus between the two reviewers.

Data extraction and analysis
Data were extracted against pre-defined categories by two researchers. The data compiled from the studies included information on: study design, characteristics of participants and setting, the intervention and control group treatment, time of assessment, ADL, physical performance, length of stay and discharge destination. Furthermore, the feasibility outcomes were the ability to enrol patients into the rehabilitation program, and the adherence rate and safety of the patients during the therapy sessions. The information extracted from the articles was organized into tables and systematically compared.

Methodological quality
The methodological quality of the included RCTs was assessed using the Delphi scale [12]. The Delphi scale is a quality assessment tool for RCTs and has shown to be valid and reliable [13]. It consists of 9 different criteria which can be scored positive, negative, or unclear ("yes", "no", and "don't know"). One point was given for each "yes" and zero points for each "no" or "don't know", the total quality sore ranged from 0 (low quality) to 9 (high quality).

Selected studies
The literature search for physical functioning yielded a total of 772 papers ( Figure 1). After removing 765 articles based on title and abstract, 9 articles were qualified for full text reading. Four articles were removed after full text reading and ten articles were added after reference checking. The remaining 15 articles, describing 13 studies, were included to this review. The search for feasibility studies yielded 50 papers. After removing 47 articles by title, abstract and full text reading and adding one article after reference checking, four studies were included for the feasibility assessment. The article of Laver et al. [14] was included in both the physical functioning and the feasibility section. PRISMA guidelines were followed in this systematic review [15].

Methodological quality
In Table 1 the quality scores on the Delphi Scale for the different RCT studies are reported. Total quality scores ranged from 3 to 7 with a median score of 5. The methodological quality was moderate for most studies. Randomization methods and eligibility criteria were clearly defined in all 13 studies. The studies scored particularly low on blinding the assessor, the care provider and the participant. The Delphi scores were good for the concealed treatment allocation, the similarity of the intervention and control groups at baseline and the clarity of the specified eligibility criteria. Table 2 summarizes the characteristics of patients, study settings, early physical rehabilitation programs and outcomes of the included RCT studies in this review. The mean age for patients admitted to acute care in the hospital for a general medical condition varied between 78 and 86 years old. The most common reasons for admission were cardiac problems, respiratory problems, gastrointestinal problems, neurological problems, infections and injuries caused by a fall. The living situation of the patients before they were admitted to the hospital varied, patients came from nursing homes and other types of institutionalised care or from the community where they lived alone or with family. However, the  studies did not include all patients in the intervention, reasons for exclusion were medical instability [18,22,23], need for specialized care [17,19,22,24], living in nursing homes [19,20,22,23,25,26,28], small survival chance or need for palliative care [18,22,23,25,28], and being diagnosed with an illness causing functional impairment [18,23,28]. Overall, there was a great heterogeneity among the participants between the different studies. Table 2 gives an overview of the settings and interventions of each included randomized controlled trial. The studies were performed on acute geriatric units, geriatric units and medical units of university hospitals, city hospitals and acute care hospitals. Early physical rehabilitation programs could be divided into two categories, (1) multidisciplinary programs with an exercise component and (2) usual care with an additional exercise program.

Setting and intervention
In the review we refer to these categories as multidisciplinary programs and exercise programs, respectively. The main aim of the multidisciplinary programs was to maintain or obtain independent ADL and encourage returning home. Multidisciplinary intervention teams usually consisted of a geriatrician, (geriatric) nurses, social workers, physical therapists and occupational therapists [17,19,24,26,29]. The specific exercise component of the multidisciplinary intervention studies is hardly described, and information on intensity, duration, and frequency of exercises is often lacking. Usual care with an additional exercise program was provided in eight studies [14,16,18,20,22,23,25,28]. The patients in the exercise programs were supervised by allied health assistants, a physiotherapist or an occupational therapist. Patients performed exercises five times a week up to twice daily.
The aim of the exercise programs was predominantly to improving functional outcomes by training strength, mobility, and balance. Strength exercises were progressed by increasing the number of sets and repetitions and walking exercises were progressed in intensity (from slow to moderate pace) or duration (from 5 to 30 minutes). In one study an interactive video gaming program was used to exercise. A Nintendo Wii fit group trained 25 min/day, 5 days/week under supervision of a physical therapist [14]. Some exercise programs supported participants to continue exercise after discharge [20,25,28]. This was achieved by educational materials, by (two or more regular) encouraging phone calls and home visits [20,28], or by a follow-up treatment, twice a week up to twice a day, including physical and occupational therapy [25]. The control groups of the studies generally received usual care according to the general routines of the hospital they were admitted to.

Physical functioning
All included studies provided at least one outcome measure related to physical functioning. Most studies used measures of activities of daily living (ADL) and physical performance to describe physical functioning. Additional information about length of stay and discharge destination were also documented.

Activities of daily living
The included studies gained information on patient's ADL by conducting interviews and (self-administered) questionnaires. Table 3 gives on overview of the used instruments to measure (I)ADL. Eleven of the thirteen studies reported the effects of their intervention on (I)ADL at time of hospital discharge [14,[16][17][18][19][21][22][23][24][25]29].        Two multidisciplinary programs found that patients in exercise group compared to the patients in usual care group significantly improved more and worsened less in the number of basic ADL activities they were able to perform [24,29]. The other studies did not find a significant group difference in ADL at time of discharge. However, one study with an exercise intervention found a significant interaction effect between group and admission scores on the modified Barthel index meaning that patients with a low admission score who received an exercise intervention showed greater improvement in ADL than those patients who received only usual care [23]. Two studies reported a ceiling effect for the Barthel index and a floor effect for the Katz index [22,23].
Remaining studies did not found a significant difference on ADL between early physical rehabilitation programs and usual care (see Table 2).
Of the eight studies that reported follow-up data on (I)ADL [17][18][19][20][24][25][26]28] only the three studies with an exercise intervention that provided patients with follow-up treatment after hospital discharge reported larger improvements in the exercise group than in the usual care group after 1, 6 and 12 months [20,25,28]. One study that did not provide patients with follow-up treatment also found a positive effect of the intervention on ADL at 3 months post-discharge. However, this study included patients who died by assigning them a score of zero, whereas the mortality rate was higher in the usual care group than the multidisciplinary group [26].

Physical performance
Seven studies reported measures of physical performance; Table 3 gives an overview of the used measurement instruments [14,19,20,22,23,28,29]. Three (2 multidisciplinary programs and 1 exercise programs) of the five studies describing physical performance at time of admission and discharge found that the intervention groups improved more or declined less than the usual care groups [19,23,29]. However, for one study the difference in the Timed Up and Go (TUG) was not significant after adjusting for confounders such as patient characteristics, admission modified Barthel Index, comorbidity, and mental scores [23].
For the TUG a floor effect was found. In one study, almost 40% of the patients were physically unable to perform the TUG at both admission and discharge [23], whereas in another study 23% of all patients were unable to perform the TUG at admission to the hospital [22].
Two studies reported follow-up results on physical performance, but the results are ambiguous [20,28]. One study [28] reported no significant difference between the patients in the exercise group and in the usual care group in the change in physical performance scores obtained 2 weeks before admission and one month after discharge. Conversely, the other study [20] reports that the exercise group had greater improvement over time (up to 24 weeks after discharge) in walking distance, walking speed and stair climbing. Similar contradicting results were found on more general outcome measures of physical functioning, provided by a health questionnaire that contained physical well-being and physical activity. No group differences were found for the National Health Survey Physical Activity Scale at one month follow-up [28], while the other study found that the exercise group scored higher than the usual care group on physical health related quality of life at 4, 12, and 24 weeks after discharge [21]. In short, follow-up results on physical performance are contradictory. Nevertheless, at time of discharge most studies show a greater improvement for patients in the multidisciplinary and exercise group than for the patients in the usual care group.

Length of stay and discharge destination
All included randomized controlled trials reported about the length of stay, which varied between the 4.7 days and 42.7 days. Four studies (three multidisciplinary programs) reported a significant difference shorter length of stay of the exercise group than for the usual care group [17,25,27,29].
From the seven studies [17,19,[22][23][24][25]29] that reported on discharge destination, five studies found that a higher proportion of patients in the early physical rehabilitation group than in the usual care group were discharged home, instead of being transferred to additional (subacute) hospital treatment or to institutionalized care, however, results were significant in three studies (Table 2). Two multidisciplinary programs found that respectively 14% and 18% of the patients in the multidisciplinary group were discharged to a long-term care institution opposed to 22% and 27% of usual care patients [24,29]. The third study, an exercise intervention, included only patients who had lived at home before admission and found that 4% of the patients in the exercise group and 8% of the patients in the usual care group were discharged to longterm care institutions [25].

Feasibility
Feasibility of early physical rehabilitation programs, was explicitly assessed in four studies (Table 4) comparing usual care with an exercise intervention program [14,[30][31][32]. In addition included randomized controlled trails with an exercise program reported on enrolment (n=3), adherence rate (n=2) and adverse events (n=5), while multidisciplinary studies reported on enrolment (n=3) only.
Subjects enrolled in the studies explicitly assessing feasibility were respectively above the age of 60 or 70. The most common exclusion criteria were severe impairments in physical performance and cognition, requiring palliative care, expected short length of stay, and medical instability. One study specifically targeted frail patients that were at risk of functional decline [32].
Feasibility was measured quantifying patients enrolment, patient adherence to the program and at patient's safety in the context of the exercise program. Nine of the 11 studies found that between the 14% and 48% of the admitted patients met the inclusion to be enrolled in the programs, and between 3 and 19% of the patients were not willing to participate [14,17,19,[22][23][24]28,31,32]. In general patients not willing to participate stated that they did not feel like exercising or that they did not believe they could exercise. They felt unwell and/or were scared that exercising would make them feel worse.
In one study [30] only 2% of all admitted patients were included. Of the 76 patients that met the inclusion criteria only 10 consented to participating while 87% of the eligible patients refused to take part in the program. In fact none of the included patients were able to start with the exercise program. Reasons for not participating in the exercise program were not of a physical nature.
Instead, patients were discharged before therapy could begin or were unavailable due to diagnostic tests or appointments with healthcare professionals.
A final measure of feasibility is patient safety during the exercise program. One feasibility study and six RCT studies included in this review reported on potential side effects such as injuries, accidents, and more specifically, fall incidents, related to participating in the early physical rehabilitation programs [14,16,18,20,23,28,31]. None of the studies found any differences in the number of incidents between the exercise groups and usual care groups.

Discussion
The first aim of this review was to evaluate the effect of early physical rehabilitation programs for geriatric hospitalized patients on physical functioning. A total of 15 articles, reporting on 13 studies, were included that provided early physical rehabilitation programs in a hospital setting. The present review shows that early physical rehabilitation programs might be beneficial to prevent rapid decline in physical functioning.
A classification could be made between multidisciplinary programs with an exercise component and usual care with an exercise component. Multidisciplinary programs reduced the length of stay for patients in the exercise groups compared to the patients receiving usual care. Additionally, the patients in the multidisciplinary programs were less likely to be discharged to a nursing home or other forms of institutionalized care than patients in the usual care group. The usual care programs with an exercise intervention had the main aim to improve functional outcomes, some of the studies showed indeed an improvement on ADL and physical performance.
The two types of exercise interventions, e.g. multidisciplinary programs with an exercise component and usual care with an exercise component, did not find different results in physical functioning at time of discharge. At time of discharge, results on physical performance were to some extent contradicting, but the majority of the included studies showed that patients in the exercise groups had better ADL and physical performance than patients in the usual care groups, although those results were not always significant. Follow-up results on ADL and physical performance showed that persistent positive effects were mainly found in studies that provided patients with continuous interventions after hospital discharge. The studies that provided patients with only in-hospital exercise interventions found little or no effect on ADL and physical performance at follow-up examinations. These results are in line with other studies that investigate the effects of care in geriatric units and of inpatient rehabilitation for geriatric patients on functional parameters [10,33]. The effects of only in-hospital interventions were clearly positive at time of discharge but were greatly reduced during follow-up. These results suggest that the recovery of patients could further benefit from a community based or in-home intervention programs which build on inhospital programs. Such programs could consist of physical or occupational therapy. In addition, the present review shows in home visits and follow-up telephone calls might be effective for adherence for home-based exercise [20]. Further research is needed to clarify the effects and feasibility of community based and in-home intervention programs in old adults after hospital discharge.
The studies presented in this review included very heterogeneous groups. Half of all studies excluded nursing home patients. Probably, due to the aim of those studies to facilitate discharge home. The studies in this review that targeted frail older patients, patients with increased risk for readmission or patients with a high risk for nursing home admission found positive effects of an exercise program on ADL, length of stay and discharge destination [20,25,26]. This result indicates that frail old adults and nursing home patients may benefit from early rehabilitation. Targeting the right population for early physical rehabilitation may be seen as crucial.
The outcomes with regard to physical functioning were measured by a variety of instruments and at different points in time, some of these instruments demonstrated floor and ceiling effects. Floor and ceiling effects could influence the outcomes and distort the results. There were also cases in which information could only be collected with the help of close relatives and caregivers, because relevant information could not be obtained from the patients themselves. The limitations of the current used measurement instruments implies that there might be a need for more sensitive instruments that measure aspects of physical functioning in hospitalized old adults.
Finally, the second aim of this review was to investigate the feasibility of early physical rehabilitation programs for acute ill older patients. The early physical rehabilitation programs must be safe to perform and may not cause high numbers of drop outs. There were four studies identified reporting on the feasibility of inhospital exercise programs [14,[30][31][32]. Additionally, a part of the randomized controlled trials, used to determine the effect on physical functioning, reported also on some feasibility points. Several studies included in this review affirmed the safety of early rehabilitation programs. Patients in the early physical rehabilitation groups were not more often injured, nor did they experience more adverse events or falls than usual care patients. Results on patient recruitment were contradictory and should be interpreted with care due to the limited number of available studies. Adherence rates were high for most studies. Patient as well as the intervention provider's satisfaction was higher when patients were treated with a early physical rehabilitation program instead of usual care [19]. There was one study that encountered difficulties in recruiting patients and was unable to have any patients participate in the exercise program at all [30]. Patients refused to participate or were unavailable at scheduled time of therapy. One of the greatest barriers in the implementation of intervention research, according to the opinion of nurses, was a lack of awareness and knowledge [34]. Equally important in this regard were difficulties in the cooperation of interdisciplinary team members [34]. Education of staff and patients about the safety and the benefits of early physical rehabilitation as well as regular team conferences that improve coordination might help to increase the success rate of intervention programs with regard to participate.

Limitations
The total number of included studies in this review is small and the methodological quality for most studies was moderate. Most articles scored particularly low on blinding the assessor, the care provider, and the participant. However, it may be regarded as almost impossible to assure blinding in the context of early physical rehabilitative interventions. One of the major limitations of the included studies lies with the poor descriptions of exercise in early physical rehabilitation programs. Since there is often no clear description on type, duration, frequency, and intensity of exercises that patients receive. Furthermore, because many multidisciplinary programs are focused on outcomes beyond functionality, e.g. early discharge planning, it is difficult to determine what the effect of the exercise component on physical functioning is in this setting. Future research should provide quantification of exercises that older patients perform during their hospital stay, so that there would be a clearer view on the dose-response relationship of physical activity and functional outcomes.

Conclusion
Early physical rehabilitation programs for acutely hospitalized old adults have the potential to improve physical functioning, and also to prevent patients from being discharged to nursing homes or other forms of institutionalized care. Interventions including a follow-up program after hospital discharge increase the chance of maintaining positive effects on functionality for longer periods of time. Early physical rehabilitation for acutely hospitalized older adults seem to be safe to execute in terms of adverse events such as falls or other injuries, but recruiting the most suitable patients and getting them to participate regularly in the program can be a challenge. Therefore, the commitment and collaboration of staff is of great importance. Further research is needed to quantify the physical activity of patients in early physical rehabilitation programs and to determine the effects and feasibility of community-based and inhome exercise programs.