Effect of physical interventions on physical performance and physical activity in older patients during hospitalization: a systematic review

Background To counteract decline in physical performance and physical activity in older patients during hospitalization, multiple physical interventions were developed. However, it is unknown whether these are effective in this particular population. This systematic review aimed to identify the effect of physical interventions on physical performance and physical activity in older patients during hospitalization. Methods The systematic search included PubMed, EMBASE, Cinahl, the Trials database of The Cochrane Library and SPORTdiscus from inception to 22 November 2017. Studies were included if the mean age of the patient cohort was 65 years and older and the effect of physical interventions on physical performance or physical activity was evaluated during hospitalization. Results Fifteen randomized controlled trials met the inclusion criteria. Overall, the effect of physical interventions on physical performance was inconsistent. Patient tailored interventions, i.e. continuously adapted to the capabilities of the patient were not found to be superior over interventions that were not. Physical activity as outcome measure was not addressed. Reporting of intensity of the interventions and adherence were frequently lacking. Conclusions Evidence for the effect of physical interventions on physical performance in older patients during hospitalization was found uncertain. Further research on the efficacy of the intervention is needed, comparing types of intervention with detailed reporting of frequency, intensity and duration. Electronic supplementary material The online version of this article (10.1186/s12877-018-0965-2) contains supplementary material, which is available to authorized users.


Background
Older age is associated with a high prevalence of age related diseases and is a major risk factor to become admitted to hospitals [1]. Hospitalization is associated with a decline in physical performance [2], nursing home admission [3] and short term mortality [4]. A decline in physical performance is likely to be aggravated by low physical activity during hospitalization [5], which in turn affects activities of daily living [6] and occurrence of falls [7]. Older patients have decreased physiological and functional reserves that renders them vulnerable to negative effects of low physical activity during hospitalization [8].
In recent years attention has been paid to improving hospital outcomes especially for the vulnerable group of older patients [9]. Multiple physical interventions have been developed to enhance physical performance and physical activity in hospitalized patients. Physical interventions include examples such as exercise prescribed by health care professionals, supervised exercise sessions and physician counseling during hospitalization [10]. A meta-analysis showed that physical therapy of higher intensity, i.e. longer duration and higher frequency of sessions, reduce length of hospital stay and improve physical performance in patients older than 18 years with sub-acute and acute conditions [11]. In older hospitalized patients positive effects of multidisciplinary programs with an exercise component have been reported, however, the effects of solely a physical intervention were inconclusive [10,12]. Principal elements of these programs included goal setting tailored to the individual patient and interventions tailored to the patients' needs [13]. To improve patient outcome it is important to identify what type of physical interventions positively affects physical performance and physical activity in older patients during hospitalization.
This systematic review aimed to identify the effect of physical interventions on physical performance and physical activity in older patients during hospitalization. Additionally, we aimed to compare the effect of patient tailored physical interventions e.g. continuously adapted to the capabilities of the patient to the effect of non-patient tailored interventions. We hypothesize that physical interventions improve physical performance and physical activity and that physical interventions continuously adapted to the capabilities of the patient are superior over interventions that are not.

Search strategy
For this systematic review a literature search was performed by RO and KR. To identify relevant publications about physical interventions of hospitalized patients with an age of 65 years or more we performed searches in the bibliographic databases PubMed, EMBASE.com, Cinahl (via Ebsco), the Trials database of The Cochrane Library (via Wiley) and SPORTdiscus (via Ebsco) from inception to 2017 November 22nd. Search terms included controlled terms (as MeSH in PubMed and Emtree in Embase etc.) as well as free text terms. We used free text terms only in The Cochrane Library. Search terms expressing the age of the patients were used in combination with search terms for physical interventions and search terms for hospitalization. The search was then limited to randomized controlled trials. The full search strategies for all databases can be found in Additional file 1.

Study selection
Studies were assessed for eligibility by screening title and abstract by two researchers (KR and KS) for the following inclusion criteria: the study population or sub group consisted of hospitalized older patients with a mean age of 65 years and older, and the studies contained physical interventions during hospitalization with physical performance or physical activity as outcome measures. Physical performance was defined as the ability to perform a physical task at a desired level. Physical activity was defined as any bodily movement produced by skeletal muscles requiring energy expenditure. Studies were excluded when: other study designs than randomized controlled trials (RCTs) were used, pre and post measurements were not performed during hospitalization, articles were written in other languages than English or Dutch and physical interventions were performed to improve disease related outcomes of patients affected by Chronic Obstructive Pulmonary Disease (COPD), chronic heart failure, stroke, hip fracture or knee replacement. These patient groups were excluded due to the disease specific interventions. In case of uncertainty for inclusion the articles were discussed with a third researcher (CM/AM).

Data extraction and analysis
Data extraction was completed by two researchers (KS and KR). Data extracted of each study included reason for hospitalization, setting, patient group characteristics (number and age of patients of intervention and control group), intervention characteristics (type of intervention, patient tailored, frequency, intensity and duration of the intervention), adherence, moment of pre and post measurement, outcomes measures (primary and secondary), change in physical performance and physical activity between pre and post measurement was extracted and PE-Dro score [14]. Interventions were considered patient tailored if the intervention was adapted to the capabilities of the patients prior to and during the intervention. Frequency was defined as the number of intended physical intervention sessions per week and intensity as the number of repetitions or level of exertion and duration of one session. Duration of a session was expressed as the number of minutes of one session and duration of the intervention as the number of days the intervention was performed. Adherence was defined as the percentage of the sessions the patient participated against the number of intended sessions during the intervention period.
The extracted data were structured in tables stratified by study characteristics, characteristics of the physical interventions and study outcomes. Conclusions were based on the significance level of the primary outcome and on the number of outcomes with a significant effect. P-values equal or lower than 0.05 were considered statistically significant. Studies were grouped by patient tailored and non-patient tailored interventions. Due to the heterogeneity of the interventions and outcomes grouping by type of intervention or outcomes measures was not appropriate. Level of evidence was based on the outcome of the quality assessment of studies and on population size.

Results
A total of 1645 studies including 581 duplicates were found. After screening titles, abstracts, and full text 1049 studies were excluded, resulting in 15 studies being included in this systematic review. The study selection process and reasons for exclusion are presented in Figure 1. Overall, type of intervention and outcome measures for physical performance varied widely. None of the studies measured physical activity as outcome. The number of the patients included in the RCTs was low, only one study included more than 200 patients in the intervention group. The study characteristics are presented in Table 1.

Characteristics of the physical interventions
The interventions consisted of horse riding simulation, physiotherapy in combination with whole body vibration training, physiotherapy based on elements of proprioceptive neuromuscular facilitation, a physiotherapy program with a backward or forward walking interval training cycle, exercise programs, a rehabilitation and nutritional intervention, interactive gaming program, progressive resistance strength training, electrical quadriceps stimulation or kinesiotherapy (or a combination of both) and exergames on balance, leg strength and flexibility. A description and the characteristics of the interventions are presented in Table 2. The median number of patients in the intervention groups was 30 . In 13 studies, control groups received usual care, one control group received ball exercise and one control group received self-regulated conventional exercises. All studies were assessed as sufficient or good quality RCTs as defined by PEDro score of 4 or higher (Additional file 2: Table S1).
All physical interventions were adapted to the capabilities of the patient prior to the intervention; in eight studies the interventions were adapted both prior to and during the intervention. Frequency of the interventions varied between three to 24 sessions per week. Duration of one session varied between one to 60 min and duration of the intervention between five to 56 days. Intensity was only specified in seven studies and adherence in five studies. Table 3 presents the effect of physical interventions on physical performance and physical activity in older patients during hospitalization. Four of the eight physical interventions that were continuously adapted to the patient's capabilities showed positive results on physical performance. Horse riding simulation showed a significantly improved gait ability, measured by Time Up and Go (TUG) and 10-Meter Walk Test (10MWT), and balance in the intervention group compared with the   Bürge [21] 2017 Physical exercise program: group training including strength, flexibility, walking and balance.

Effects of the physical interventions
Yes, intensity of the exercises increased gradually during the training and was adapted to individual patient abilities. Czyzewski [17] 2013 Physiotherapy based on elements of proprioceptive neuromuscular facilitation: respiratory exercises and change of position using manual resistance on sternum, upper and lower limbs, repeated initial stretch and bilateral symmetric moving standards of shoulder girdle and upper limbs.
Yes, exercises were individualized with an intensity in the range of 40-50% of maximal frequency of heart rate, and instructions for individual practice were provided. Haines [23] 2007 Additional exercise program: by applying therapeutic principles of tai chi with functional movements an activity visualization. Yes, exercises were tailored to the individual patient abilities. Progress was monitored during each session and the intervention was modified based on improvements in the patient's function.

5-10
NG Median (IQR) 15 (11.5, 20) 90 Tibaek [27] 2013 Yes, intensity of the interval training cycle was adjusted to individual patient abilities based on a stress test and a calculated training heart rate. The heart rate and blood pressure were continuously monitored. Workload increased gradually during the training.  - Step test-left, steps, median, IQR      control group [15]. Physiotherapy in combination with whole body vibration training showed positive results on Functional Independence Measures (FIM), however, no positive effect was found on muscle strength [16]. Physiotherapy based on elements of proprioceptive neuromuscular facilitation showed a positive result on return of functional independence of basic movement activities after surgery measured by the scale of independent postoperative patient's activity. No positive effect was found on gait ability, measured by TUG and 10MWT [17]. A physiotherapy program with a backward walking interval training cycle had a positive effect on walking distance after surgery measured with the 6-min Walk Test (6MWT), however, the group with a forward walking interval training cycle showed no positive effect on walking distance compared with the control group [18]. Exercise programs for upper limb, lower limb and trunk and continuously adapted to the patient's capabilities, had no positive effect on functional independence reflected by Barthel Index (BI) [19,20]. The exercise program had no effect on recovery from baseline in functional independence [19]. In addition, an exercise program which included group training on strength, flexibility, walking and balance did not show positive results on functional independence measured by BI en FIM [21]. A rehabilitation intervention in combination with nutritional supplementation had no effect on physical performance assessed by BI [22]. Five of the seven physical interventions that were not continuously adapted to the patient's capabilities showed positive results on physical performance measures. Additional physical activity by applying therapeutic principles of tai chi had a significant positive effect on fall rates and on balance measured by the Functional Reach Test (FRT), however, no effects were found on outcomes measures for balance, gait ability and muscle strength [23]. Interactive video gaming showed a significantly improved gait ability measured by TUG and balance in the intervention group compared with the control group, but no effect was found on physical performance reflected by Short Physical Performance Battery (SPPB) [24]. An exercise program including exercises for upper limb, lower limb and trunk showed a significant positive effect on functional performance reflected by TUG [25]. Progressive resistance strength training of the lower extremities was found to have a positive effect on balance measured by Berg Balance Scale (BBS) [26] and showed variable effects on functional independence as measured by BI while a significant positive effect on climbing stairs was found and no effect on transfer, walking and physical function [27]. No significant effect of kinesiotherapy or electrical stimulation of the quadriceps or a combination of both was found on balance, gait ability and muscle strength [28] and no effect was found of exergames on balance [29].

Discussion
Overall, the evidence found for the effect of physical interventions on physical performance in older patients during hospitalization was uncertain. Patient tailored physical interventions were not found to be superior over interventions that were not. Although, the studies were rated of sufficient to good quality, the number of studies and included patients was low and interventions were heterogeneous as well as outcome measures; reporting of intensity of the interventions and adherence were mostly lacking.
This systematic review, focusing on a generic and heterogeneous group of older patients, failed to identify the positive effects of patient tailored physical interventions which is in contrast to studies focusing on specific and more homogenous groups of hospitalized patients like COPD. Interventions like individualized resistance training sessions improved physical performance in hospitalized patients with exacerbation of COPD [30,31]. Individually tailored modifications to the prescribed intervention are found to contribute to a progressive challenge of the individual capability [32]. Although, next to adaption to the capabilities of the patient, other characteristics of physical interventions like frequency, intensity, duration of the intervention and adherence are critical determinants of its effect.
The frequency of the physical interventions varied widely and seems not to be related with the effect on physical performance in older patients during hospitalization. The physical interventions had a large variance in intensity due to the large variability in intervention types, i.e. progressive resistance training of the lower extremities vs. exercise using interactive video gaming, and in duration of intervention sessions. Previous studies showed that higher intensities result in greater functional improvement in older adults [33][34][35]. The reason that we were not able to find this may be due to the lack of reported intensities of interventions in the majority of the studies. In a systematic review on the effects of different exercise interventions on functional capacity in physically frail older adults it was concluded that resistance training exercises should include two to three sessions per week, with three sets of 8-12 repetitions at increasing intensity to 80% of one repetition maximum test to improve functional performance [36]. In most studies included in this systematic review the duration of the interventions was determined by the length of hospital stay. Other systematic reviews including physically frail older adults [36] and nursing home residents [33] reported interventions varying between ten weeks to one year and between two to four months respectively. In a systematic review on the effect of interventions with nutrition and exercise in different populations of older adults, a minimum duration of the intervention of three months was suggested to improve physical performance [37]. Considering the short length of hospital stay, the impact of in-hospital physical interventions might be limited by the intervention duration.
This systematic review included a number of studies including a low number of patients, however, the studies were rated of sufficient to good quality. Information on multiple characteristics of the physical interventions were lacking. Therefore, it is inconclusive if the effect of a physical intervention is depended on the intervention being adapted to the capabilities of the patient or that characteristics like frequency, intensity and duration of the physical intervention are more decisive. The question is whether the effects of physical interventions to avoid physical inactivity are sufficient or that more progressive and targeted physical interventions are required. Further research should focus on identifying the dose-effect relationship of both patient tailored and generic physical interventions. This is substantial for hospital policies considering the feasibility of the interventions in clinical practice and the cost efficiency. Physical interventions to avoid physical inactivity are likely to require less resources compared to more progressive and patient tailored interventions.
The outcomes measures used to express physical performance varied widely. As suggested by Cruz-Jentoft et al. (2014) [37], standardization and proper definition of outcome measures for physical performance is needed to compare the effects of physical interventions. None of the studies included an intervention measuring physical activity as an outcome, therefore no evidence was found of the effect of physical interventions on physical activity in older patients during hospitalization.
To the best of our knowledge this the first review focusing on the effect of type of physical interventions on physical performance and physical activity in older patients during hospitalization. It was not possible to perform sub analyses or a quantitative analysis of the data due to the heterogeneity of the interventions and outcomes measures.

Conclusions
Evidence for the effect of physical interventions on physical performance in older patients during hospitalization was found uncertain. Physical interventions continuously adapted to the capabilities of the patient were not found to be superior compared to interventions that were not.