Effects of comprehensive geriatric assessment on physical fitness in an acute medical setting for frail elderly patients.

INTRODUCTION: Frail elderly people often use emergency care. During hospitalization, physical decline is common, implying an increased risk of adverse health outcomes. Comprehensive Geriatric Assessment (CGA) has been shown to be beneficial for these patients in hospital care. However, there is very limited evidence about the effects on physical fitness. The aim was to compare effects on physical fitness in the acute care of frail elderly patients at a CGA unit versus conventional care, 3 months after discharge. PATIENTS AND METHODS: A clinical, prospective, controlled trial with two parallel groups was conducted. Patients aged ≥75 years, assessed as frail and in need of inpatient care, were assigned to a CGA unit or conventional care. Measurements of physical fitness, including handgrip strength (HS), timed up-and-go (TUG), and the 6-minute walk test (6-MWT) were made twice, at the hospital index care period and at the 3-month follow-up. Data were analyzed as the mean change from index to the 3-month follow-up, and dichotomized as decline versus stability/improvement in physical fitness.
RESULTS: In all, 408 participants, aged 85.7±5.4 years, were included. The intervention group improved significantly in all components of physical fitness. The controls improved in TUG and declined in HS and 6-MWT. When the changes were dichotomized the intervention group declined to a lesser extent; HS p<0.001, 6-MWT p<0.001, TUG p<0.003. The regression analysis showed the following odds ratios (ORs) for how these outcomes were influenced by the intervention; HS OR 4.4 (confidence interval [CI] 95% 2.2-9.1), 6-MWT OR 13.9 (CI 95% 4.2-46.2), and TUG OR 2.5 (CI 95% 1.1-5.4).
CONCLUSION: This study indicates that the acute care of frail elderly patients at a CGA unit is superior to conventional care in terms of preserving physical fitness at 3 months follow-up. CGA management may positively influence outcomes of great importance for these patients, such as mobility, strength, and endurance.

Introduction

Old people with multimorbidity and disability are frequent visitors to acute medical hospital departments.1 In this group of patients, admissions are often inevitable, they often require a longer length of stay compared with their younger counterparts, and the readmission rate is high.2–5 Frailty is a clinical syndrome reflecting the dependence and vulnerability of these people.6 There are different ways to define frailty. The phenotype model of frailty defined by Fried et al7 demonstrates predictive validity for the adverse outcomes that frail elderly people run the risk of experiencing, such as falls, hospitalizations, disability, institutionalization, and death. According to this definition, a person is considered frail when three or more of the following criteria are present: weakness (eg, low handgrip strength [HS]), poor endurance, slow walking speed, low physical activity, and shrinking. Another way to describe frailty is by the accumulation of deficits.8

Physical function describes a person’s capacity to carry out the physical activities of daily living (ADL).9 Physical fitness comprises a set of measurable health- and skill-related outcomes, such as cardiorespiratory endurance, muscle strength, muscle endurance, and balance.10 The relationship between frailty and physical fitness is evident and frailty is often said to herald physical decline.11 Age-related muscle loss, sarcopenia, is one key component of the frailty syndrome and it negatively affects physical fitness.12,13

In a recently published study, physical fitness was measured objectively and found to be severely impaired in acutely hospitalized frail elderly patients.14 It is well known that frail elderly patients risk further deterioration in connection with hospital care and the recovery rate appears to be low.5,15–19 A decrease in physical fitness is associated with an increased risk of disability and dependence20 and it negatively affects the ability to benefit from medical interventions, which worsens the prognosis still further.13,21–26 However, previous research indicates that it may be possible to reverse frailty.27 It is therefore of utmost interest to identify frailty in order to prevent, reduce, and postpone adverse health consequences.28

Today, the conventional acute medical ward is usually a specialized, organ-specific unit with the goal of providing care according to national and international guidelines adapted for specific diseases. Studies have shown that frail, multimorbid, and disabled patients are likely to benefit from a more holistic approach.29,30 In the context of hospital care, a health care model called Comprehensive Geriatric Assessment (CGA) has been shown to be consistently beneficial regarding mortality, disability, and cognitive functions.31 The CGA concept involves early identification of persons at the greatest risk of complications and adverse health outcomes.32 The initial purpose of CGA was to plan and/or deliver medical, psychosocial, and rehabilitative care and the model is defined as a It explicitly involves an early-rehabilitation perspective.34

multidimensional interdisciplinary diagnostic process intended to determine a frail elderly person’s medical, psychological and functional capabilities and limitations, in order to develop an overall plan for treatment and long-term follow-up.33

Previous studies have concluded that CGA implies functional benefits, in terms of ADL, for elderly patients with acute medical or orthopedic disorders, compared with conventional care.34–37 To our knowledge, there is no study which has objectively investigated how CGA in an acute medical setting specifically affects physical fitness, in severely frail elderly patients.

Aim

The aim was to compare the effects on physical fitness in the acute care of frail elderly patients at a CGA unit versus conventional care, 3 months after discharge.

Patients and methods

Design and setting

This is a clinical prospective controlled trial with two parallel groups carried out at the NÄL-Uddevalla (NU) Hospital Group, in western Sweden. The total population of the NU health care system is 280,000 inhabitants. The study was approved by the regional ethical review board in Gothenburg (Dnr: 8883-12, December 12, 2012) and registered at the Swedish National Database of Research and Development; identifier 113021 (http://www.researchweb.org/is/vgr/project/113021).

Participants

The study population was frail elderly patients included in the research project entitled “Is the treatment of frail elderly patients effective in an elderly care unit”. The inclusion criteria were patients ≥75 years, assessed to be in acute need of inhospital treatment and frail according to the FRail Elderly Support researcH group (FRESH) screening instrument.38,39 A patient clearly suited for care at an organ-specific medical unit, for example, patients with acute myocardial infarction, sepsis, or acute stroke, were excluded from the study, as were patients whose informed consent could not be obtained.

Data collection

The data collection has previously been described by Ekerstad et al40 and is therefore only briefly described. When the staff at the ambulance or the primary health care center identified a patient who met the inclusion criteria, they phoned a senior physician at the CGA unit or, if it was at night, the on-call physician. If the physician agreed that the patient fulfilled the inclusion criteria of the study protocol and there was a bed available at the CGA unit, the patient was admitted there directly and allocated to the intervention group. If no bed was available at the CGA unit, the patient was admitted to a conventional acute medical ward via the emergency room and allocated to the control group. As soon as possible after admission, written informed consent was obtained. A few patients were cognitively impaired and informed consent was then given by a next of kin, through proxy consent. When proxy consent via a next of kin could not be given, cognitively impaired patients were not included.

Procedure

The participants performed the physical performance tests twice – first, before discharge from the index hospital stay and then at the 3-month follow-up visit at hospital or in the patient’s home. In many previous studies of elderly patients, measurements were done at 3-month follow-up,29,35 which can be considered to be an appropriate point regarding follow-up of these patients. In addition, the FRESH screening instrument38,39 and Charlson’s comorbidity index (CCI)41 were administered at these time points. Clinical and demographic characteristics were documented and taken from the patient’s medical records and by questioning the patient or his/her next of kin.

The tests were carried out by experienced physical therapists or physicians who had all been trained and informed in a group about the test procedure before the data collection started and then repeatedly during the study. Because of the participants’ condition, it was not possible to standardize the day of testing. The intention was to perform the initial tests during the latter part of the hospital stay, before discharge.

In both study groups, standard diagnostic and therapeutic procedures were undertaken according to national and international guidelines. For both groups, this included bedside medical examinations, laboratory testing, X-ray examinations, electrocardiograms, medical referrals, nutritional therapy, blood transfusions, and oral or parenteral drug treatment.

Intervention

The intervention was acute medical care at a CGA unit (in Swedish Medicinsk ÄldreVårdsAvdelning [MÄVA]). Table 1 provides a comparison with conventional care. There are two CGA units in the NU Hospital Group which, in addition to care in accordance with guidelines, are characterized by a structured, systematic interdisciplinary CGA and care via validated instruments and evidence-based procedures. The CGA units at the hospital involve a person-centered approach and work in close collaboration with other health care providers in the municipalities and primary health care. Patients can be admitted directly without passing through the emergency room. At the CGA units, physical therapists and occupational therapists work as members of the team, close to other professions in the ward. These professionals implement a structured early-rehabilitation strategy and, soon after admission, every patient is assessed regarding ADL, walking ability, balance, and the need for assistive devices to find the best treatment for each patient. Information is usually given to patients and their relatives to reduce the risk of concern and uncertainty related to the risk of decline in physical activity. A team conference is held every day, enabling all the professions in the ward to share information and experiences, in order to use a more consistent, person-centered approach toward each patient with one of the goals of keeping the patient ambulatory and independent, to the greatest extent possible. Although the patients did not receive homework tasks, for example, via recording charts, there were indeed educational moments in the intervention, including advice regarding physical exercise after discharge.

Table 1

Comparison of management in the intervention group (CGA) and control group (conventional acute medical care)

	Comprehensive geriatric assessment and care	Conventional acute medical care
Department and facilities	Two MÄVA (acute elderly care CGA units) wards with a total of 48 beds; one, two, or four-bed rooms	Wards of internal and emergency medicine; one, two, or four-bed rooms
	Division of Internal Medicine and Emergency Care	Division of Internal Medicine and Emergency Care
Team members
Physicians	Yes. Specialists in internal medicine, family medicine and/or geriatrics	Yes. Specialists in internal medicine
Licensed practical nurses	Yes. Including specialized admission and discharge nurses	Yes
Occupational therapists	Yes	No. Only counseling
Physiotherapists	Yes	No. Only counseling
Nutritionists	No. Only counseling	No. Only counseling
Treatment	Systematic, structured interdisciplinary comprehensive geriatric assessment and care by validated instruments focusing on: somatic and mental health, medication review, functional and activity ability including early rehabilitation, social situation Early discharge planning	Following routines at departments of internal medicine and emergency care in accordance with guidelines
Admission route	Directly to the MÄVA ward via ambulance or primary care	Via the emergency ward

Notes: For both groups, standard management procedures in accordance with national and international guidelines were followed. Copyright © 2017. Dove Medical Press. Reproduced from Ekerstad N, Karlson BW, Dahlin Ivanoff S, et al. Is the acute care of frail elderly patients in a comprehensive geriatric assessment unit superior to conventional acute medical care? Clin Interv Aging. 2017;12:1–9.40

Abbreviations: CGA, Comprehensive Geriatric Assessment; MÄVA, Medicinsk ÄldreVårdsAvdelning.

Control group

The patients in the control group were treated at conventional acute medical care units, where standard procedures according to national and international guidelines were followed. All the patients were admitted to these medical care units via the emergency room. From these wards, care planning was performed prior to discharge, via the hospital’s central care planning unit. Physical therapists and occupational therapists are also linked to each of these medical care units, but, at these units, these professionals use a more consultative approach and only see patients after being actively contacted by physicians or nurses. The physical therapists or occupational therapists are not involved in any regular team meetings at the conventional acute medical care units.

Measurements

Frailty

Frailty was assessed using the FRESH screening instrument,38,39 which is a validated screening instrument that has been shown to be appropriate in an emergency hospital setting. It consists of five questions relating to tiredness, falls, endurance, needing support while shopping, and three or more visits to the emergency department in the past 12 months. If two or more of these questions were answered with a yes, the patient was considered to be frail. FRESH has been shown to have greater sensitivity (84%) and slightly lower specificity (75%) compared with the frailty phenotype indicators.38 FRESH screening was performed by a physician or a nurse by the time of inclusion. At the 3-month follow-up, it was performed by the assessing physician. Information was received by interviewing the patient or his/her next of kin and from the patient’s medical records.

Comorbidity

CCI41 is a frequently used measurement of the total burden of comorbidities. It consists of 19 comorbidities, each of which is given a severity weighting depending on the risk of dying associated with this condition. The CCI shows the sum of weighted items with a score between 0 and 10 and predicts the mortality for a patient. The CCI was completed from medical records by the assessing physician before discharge and at the 3-month follow-up.

Physical fitness

In relation to previously described frailty indicators, three different tests were performed to measure physical fitness: HS, functional mobility, and submaximal aerobic capacity.

HS

A hydraulic hand dynamometer (Saehan Corporation, Masan, Republic of Korea) with a grip position that could be adapted to suit each individual was used for measuring HS. The patient performed the test in a sitting position with the hand next to the body, the shoulder joint in a neutral position, the elbow flexed at 90°, and the wrist in a neutral position.42 The patient was instructed to squeeze his/her dominant hand as hard as possible and then relax. The outcome is the peak value (kg) of three attempts, with a short rest (about 1 minute) between each try. If the patient was not able to sit, he/she was allowed to lie in bed with the head end raised and the elbow supported. This test has been shown to have good validity when measuring muscle strength and good test-retest reliability in community-dwelling elderly and in older persons with dementia (ICC =0.97).42,43

Functional mobility

The timed up-and-go test (TUG) measures the time spent, in seconds, when a patient rises from a standard armchair (~46 cm), walks 3 m, turns around and walks back to the chair again. The 3-m distance should be clearly marked with tape that deviates from the floor color.44 In this study, the patients were instructed to walk as quickly and as safely as possible and walking aids were allowed. The TUG is an instrument with good test-retest reliability (ICC =0.99) in frail elderly adults. In community-dwelling elderly adults, it has been shown to be a valid measurement of functional mobility and correlates well with gait speed (Pearson’s r=0.75), ADL (Pearson’s r=−0.79), and balance (Pearson’s r=−0.72).44,45

Submaximal aerobic capacity

The 6-minute walk test (6-MWT) measures the total distance, in meters, during a 6-minute walk. The person walks back and forth along a 30-m corridor with markings every 5 m,46 after an instruction to walk as far as possible during 6 minutes. In the present study, the patients were allowed to stop and continue during the test, but the test was interrupted if the patient was unable to continue, mostly due to dyspnea and strong dizziness. Walking aids were allowed. The 6-MWT is a test with good test-retest reliability (ICC =0.95) for community-dwelling elderly45 and is commonly used to measure submaximal functional exercise capacity in patients with cardiorespiratory diseases.46

Analysis

The sample size calculation (Sample Size Calculator; ClinCalc LLC, Arlington Heights, IL, USA) was based on the primary outcome decline in ADL (Katz index) from baseline to 3 months after discharge. No previous study with an identical primary variable (different follow-up times) was found. However, one similar study was found,47 which focused on the ADL function in less frail patients during hospitalization. Using a two-sided test, 80% power, and a significance level of α=0.05, it was necessary to include 150 patients in each study group. To compensate for the uncertainty, due to longer follow-up times which may reduce the difference in ADL decline, that is, expected reduction of treatment effect size, it was estimated that 200 patients in each study group, 400 in total, had to be included.

The data were computerized and analyzed using the Statistical Package for Social Sciences (IBM SPSS Statistics for Windows, version 22.0; IBM Corporation, Armonk, NY, USA). Student’s t-test was used to compare continuous, parametric data between the groups and the chi-square test was used to compare categorical data. The 0- to 3-month changes were calculated first as the mean change and standard deviation for each group, after which they were dichotomized as decline versus stability/improvement. Adjustments for possible differences at baseline were made, using logistic regression models. Age, gender, CCI score, and the base-line value of measurement were counted as covariates in these models.

An intention-to-treat principle was followed.

Results

Baseline characteristics of the study population

From March 2013 to July 2015, a total of 419 evaluable patients were randomized of whom 408 patients were evaluable; 206 in the intervention group (CGA unit) and 202 in the control group (conventional care) (Figure 1).

Figure 1

Flowchart of data collection – physical fitness.

Notes: Flowchart completed here with information regarding participants’ performance of physical fitness instruments. Copyright © 2017. Dove Medical Press. Adapted from Ekerstad N, Karlson BW, Dahlin Ivanoff S, et al. Is the acute care of frail elderly patients in a comprehensive geriatric assessment unit superior to conventional acute medical care? Clin Interv Aging. 2017;12:1–9.40

Abbreviations: HS, handgrip strength; TUG, timed up-and-go; 6-MWT, 6-minute walk test.

Their mean age was 85.7±5.4 years and 56% were female. There were no significant differences between the groups regarding age, gender, degree of frailty, or percentage living alone. Both groups were heavily affected by diseases, particularly renal impairment and cardiovascular disease. The intervention group had a slightly higher comorbidity burden (CCI 7.4±2.1 vs 6.2±1.5, p<0.001). In unadjusted measurements of physical fitness, the groups did not differ at baseline in terms of HS and 6-MWT, but the control group performed the TUG more slowly (p<0.05) (Table 2).

Table 2

Baseline characteristics of the population

Variable	Intervention group (CGA, unit)N		Control group (conventional care)N		p-value
Age, years, mean (SD)	206	85.7 (5.3)	202	85.6 (5.4)	0.850
Gender, female, n (%)	206	122 (59)	202	108 (53)	0.241
Frailty screening score, mean (SD)	206	3.5 (0.9)	202	3.4 (0.9)	0.149
Charlson’s index score, mean (SD)	206	7.4 (2.1)	202	6.2 (1.5)	<0.001
Living alone, n (%)	206	139 (67)	202	132 (65)	0.649
Own living without home-help service, n (%)	206	60 (29)	202	77 (38)	0.055
Handgrip strength (kg), mean (SD)	184	18.8 (7.2)	153	18.0 (7.9)	0.330
6-MWT (m), mean (SD)	147	146 (103.4)	95	160 (100.0)	0.287
TUG (sec), mean (SD)	153	30.0 (23.2)	120	37.4 (28.6)	0.020
Reported reasons for admission, n (%)
Dyspnea	206	67 (32)	202	65 (32)
Worsened general condition/tiredness	206	48 (23)	202	43 (21)
Pain	206	29 (14)	202	24 (12)
Fever/infection	206	28 (14)	202	40 (20)
Vertigo/falling	206	27 (13)	202	30 (15)
Others	206	52 (25)	202	35 (17)

Notes: The baseline characteristics of the population divided by group, intervention group, and control group, and the main reasons that led to admission. Continuous data are presented as the mean ±1 SD. Nominal data are presented as number (%). Copyright © 2017. Dove Medical Press. Adapted from Ekerstad N, Karlson BW, Dahlin Ivanoff S, et al. Is the acute care of frail elderly patients in a comprehensive geriatric assessment unit superior to conventional acute medical care? Clin Interv Aging. 2017;12:1–9.40

Abbreviations: CGA, Comprehensive Geriatric Assessment; SD, standard deviation; 6-MWT, 6-minute walk test; TUG, timed up-and-go test.

The number of hospital days per patient during the index care episode was 11.2 (mean) in the intervention group; the number was 9.2 (mean) in the control group (p=0.002). At the 3-month follow-up, the total number (index +3 months after discharge) of hospital days were 16.2 in the intervention group, 16.9 in the control group (p=0.648).40

Physical fitness outcomes at the 3-month follow-up

Analyses of unadjusted continuous variables for the 0- to 3-month change showed a significant improvement in the intervention group in all components of physical fitness. Patients in the control group significantly improved in their ability to perform the TUG, but declined in HS and 6-MWT. When comparing groups, there were significant improvements in the intervention group compared with the control group in terms of HS (p<0.001) and the 6-MWT (p<0.001), but not for TUG (p=0.132) (Table 3).

Table 3

Change in physical fitness in 0–3 months

	Variable	Intervention group (CGA unit)					Control group					p-value (between groups)
		N	Unadjusted		Adjusted		N	Unadjusted		Adjusted		Unadjusted	Adjusted
			Mean (SD)	CI 95%	Mean	CI 95%		Mean (SD)	CI 95%	Mean	CI 95%
Change 0–3 months	Handgrip strength (kg)	133	+1.47 (5.0)	+0.72 to +2.22	+1.64	0.93–2.36	108	−0.69 (3.6)	−1.53 to +0.14	−0.90	−1.70 to −0.10	<0.001	<0.001
	6-MWT (m)	83	+19.9 (82.1)	+2.4 to +37.4	+21.4	5.8–37.0	52	−58.3 (77.8)	−80.4 to −36.3	−60.7	−80.6 to −40.9	<0.001	<0.001
	TUG (s)	105	+6.67 (19.2)	+3.07 to 10.28	+6.75	4.03–9.45	70	+2.30 (17.9)	−2.11 to +6.72	+2.19	−1.15 to 5.45	0.132	0.042

Notes: Continuous variables for change in physical fitness in 0–3 months. Adjustments were made for age, female gender, Charlson’s index, and the baseline value of measurement. Data are presented as the mean ±1 SD and 95% CI. +, improvement; −, decline.

Abbreviations: CI, confidence interval; CGA, Comprehensive Geriatric Assessment; SD, standard deviation; 6-MWT, 6-minute walk test; TUG, timed up-and-go test.

After adjustment for age, gender, CCI, and the baseline value of measurement, the intervention group had significantly improved in all components of physical fitness. The patients in the control group improved in the TUG but declined in HS and the 6-MWT. Comparing groups, there were significant advantages for the intervention group in all components of physical fitness, HS (p<0.001), 6-MWT (p<0.001), and TUG (p=0.042) (Table 3).

When dichotomizing the changes into the categories of decline versus stability/improvement, there were significant differences between groups for all three components of physical fitness. The intervention group declined to a lesser extent compared with the control group; HS p<0.001, 6-MWT p<0.001, TUG p=0.003. Figure 2 represents a visual picture of the results.

Figure 2

Changes in physical fitness in CGA unit and conventional care.

Notes: The number (%) of patients that declined, were stable, or improved in physical fitness in CGA unit (intervention group) and in conventional care (control group). We used the change of the study population, and if changed one quartile or more, it was assumed to be a relevant change. Handgrip strength (kg): decline: <−2 kg, stable: −1 to +2 kg, improvement: ≥+3 kg. 6-MWT (m): decline: <−50 m, stable: −49 to +23 m, improvement: >+24 m. TUG (s): decline: slower than +1.3 s, stable: +1.2 to −6.3 s, improvement: faster than −6.4 s.

Abbreviations: TUG, timed up-and-go; 6-MWT, 6-minute walk test; CGA, Comprehensive Geriatric Assessment.

In the regression analysis, the odds ratios (ORs) showed the extent to which the outcome was influenced by the intervention; the unadjusted ORs were: HS OR 3.2 (confidence interval [CI] 95% 1.7–6.1), 6-MWT OR 7.0 (CI 95% 2.8–17.7), and TUG OR 2.8 (CI 95% 1.3–5.9) (Table 4). After adjustment, the ORs were HS OR 4.4 (CI 95% 2.2–9.1), 6-MWT OR 13.9 (CI 95% 4.2–46.2), and TUG OR 2.5 (CI 95% 1.1–5.4) for the tests, respectively (Table 4).

Table 4

Decline in physical fitness in 0–3 months

	Variable	Intervention group		Control group		OR (CI 95%)
		N	N (%)	N	N (%)	Unadjusted	Adjusted
Decline	Handgrip strength (≥2.0 kg)	133	23 (17.3)	108	46 (42.6)	3.2 (1.7–6.1)	4.4 (2.2–9.1)
	6-MWT (≥50 m)	83	9 (10.8)	52	26 (50.0)	7.0 (2.8–17.7)	13.9 (4.2–46.2)
	TUG (≥1.3 s)	105	18 (17.1)	70	26 (37.1)	2.8 (1.3–5.9)	2.5 (1.1–5.4)

Notes: The 0–3 months change presented as dichotomized variables, decline versus non-decline, which denote preserved or improved physical fitness. Adjusted analyses were carried out with age, female gender, Charlson’s index, and baseline value of measurement as covariates. The data are presented as number (%), ORs, and 95% CI. We found no consensus definition of minimal clinical important change for frail elderly hospitalized patients. We stipulated a definition of the rationale for these terms from a statistical viewpoint. Thus, we used the change from index to follow-up of the study population, and if decreased one quartile or more, it was assumed to be a relevant decline. Handgrip strength (kg): decline >2.0 kg. 6-MWT (m): decline >50 m. TUG (s): decline >1.3 s.

Abbreviations: 6-MWT, 6-minute walk test; TUG, timed up-and-go test; OR, odds ratio; CI, confidence interval.

Discussion

This study indicates that the acute medical care of frail elderly patients at a CGA unit is superior to the care at a conventional acute medical care unit when it comes to preserving physical fitness, such as HS, submaximal aerobic capacity, and functional mobility, at the 3-month follow-up.

Moreover, the present study shows that it is possible to improve physical fitness in severely frail, hospitalized patients. The intervention group demonstrated improvements in submaximal aerobic capacity and HS, indicating that there is still rehabilitation potential. CGA management in hospital care may positively influence outcomes of great importance for the patients, such as walking ability, independence, and returning home.

To our knowledge, this is the first study which specifically and objectively evaluated CGA in terms of physical fitness. However, the results are consistent with previous research on the effects of CGA on ADL.34,35,37 ADL is usually used for the evaluation of medical care for frail elderly persons using questionnaires (eg, Barthel ADL index48 or Katz index49), which are filled in by asking the patient, or by proxy.29 Self-reports are often used in large studies because they are easy to administer, provide few missing data, and capture the patient’s own perspective. However, there is a risk of over- and underestimation, recall bias, and social desirability.50,51 One review52 compiled different instruments for measuring frailty. The physical domain of the frailty syndrome was included in all the studied instruments. It was shown that the degree of frailty can be indicated in individual physical fitness tests. Tests of strength, walking ability, and endurance are most commonly used.14,21

In one study, returning home, autonomy, and walking ability were factors of great value, when elderly patients ranked the outcomes they considered most important in post-acute geriatric hospital care.53 These outcomes are all related to frailty and highlight the importance of developing interventions that may positively affect physical fitness and prevent or delay the onset of progressive disability.20

Research has shown that endurance, strength, and muscle power training can prevent disability in frail elderly people.54 In a hospital setting, a Cochrane report55 found that multidisciplinary interventions involving exercise reduced hospital length of stay, cost of hospital stay, and increased proportion of patients discharged directly home compared with usual care. Further, a meta-analysis56 showed that extra physical therapy had beneficial effects, such as improved mobility, physical activity, and quality of life, compared with a standard physical therapy program in hospitalized patients with acute or subacute conditions.

The studied CGA units work by a structured early-rehabilitation strategy, which involves physical therapy and occupational therapy initiated immediately upon achieving physiologic stability, which continues throughout the hospital stay. Research on early rehabilitation has shown improved physical function and the intervention has been described as feasible and safe to execute.57 Many factors within the CGA concept probably influence, when it comes to preserving physical fitness. There may be several critical differences compared to conventional care, which may interact, and benefit frail elderly patients. However, the early-rehabilitation perspective including assessment, care, and educational efforts could be regarded as crucial for the prevention of functional decline. Consequently, more time was spent on physical training in these units. Good access to assistive devices and team meetings, which may enable the coordination of the efforts of the team members to understand the patient’s physical abilities and inspire security and encourage mobility and autonomy, are components that also may have an effect.16 However, it is still unknown exactly which components of CGA might have positive effects. It is likely that also components other than those related to physical fitness contribute to improved hospital outcomes.

Care at a CGA unit, like care at a stroke unit, is a complex multidisciplinary hospital intervention. Patients suffering from an acute stroke who are cared for at a stroke unit are more likely to survive, return home, and regain independence than those cared for in general wards.58 Langhorne et al59 studied the components that make stroke unit care effective. The results highlight the structured assessment procedures of all team members, an early-mobilization strategy, rehabilitation through a multidisciplinary team, and early planning for discharge.

The present study succeeded in including patients with severe comorbidity, disability, and cognitive impairments. It resembles the clinical reality of today’s emergency medical care and the generalizability is thought to be good. The outcomes are based on measurements of different components of physical fitness with well-validated instruments. The high proportion of missing data is, however, important to consider, although it is in line with previous research including physical performance tests in hospitalized, frail, elderly patients.55,60 The walking tests in particular had a lot of missing data. In this population, it might have been advantageous to have one additional test tolerated by nonambulatory patients. Despite this, a fairly large number of participants completed the tests. In order not to lose patients to follow-up, some visits were made in the patients’ homes. This made standardized performance difficult and resulted in some further dropouts.

The patients were randomized based on the availability of hospital beds, which was assumed to be the most clinically feasible method for including and evaluating patients, representative of today’s emergency hospital care. Randomizing through a lottery was considered utterly difficult to implement. It is known that a lottery-randomization method is challenging when evaluating clinical practice in complex hospital health care61 and frail elderly adults with comorbidities, disabilities, and cognitive impairments are unfortunately often excluded from scientific research.29 The allocation procedure we used seems to be confirmed as random to a satisfactory extent, as most of the baseline characteristics did not differ between the groups. However, the absence of a lottery procedure constituted a potential risk of bias. In the analysis, the data were adjusted for potential differences between groups at baseline.

Another weakness is that neither patients nor staff was blinded and this could potentially have influenced a few of the outcomes. At index, neither patients nor assessors could be blinded, since the wards were located in different hospital buildings. In practice, it would have also been very difficult to blind the assessors at the 3-month follow-up.

The CGA concept might be implemented in everyday hospital health care. Future research could also focus on how components of the CGA concept might be transferred to conventional care in order to preserve physical fitness and improve the prognosis for the benefit of more hospitalized patients.

Conclusion

Medical care for acutely ill frail elderly patients at a CGA unit appears to be superior to conventional acute care in terms of preserving physical fitness measured as HS, submaximal aerobic capacity, and functional mobility at 3 months follow-up.

In the acute care of frail elderly patients, more attention should focus on interdisciplinary teamwork with the emphasis on preserving physical fitness and encouraging ambulation and autonomy.

Clinical implications

This study shows that it is possible to improve physical fitness in severely frail, hospitalized patients. The CGA concept, with its focus on early rehabilitation, has been shown to benefit these patients in terms of submaximal aerobic capacity and HS. By showing there is still rehabilitation potential, targeted interventions suitable for these patients may be implemented in clinical health care, such as interdisciplinary teamwork including individualized physical therapist assessment and treatment. CGA management in hospital care may positively influence outcomes of great importance for the patients, such as walking ability and independence.