Extended physiotherapy after Intensive Care Unit (ICU) stay: A prospective pilot study with a before and after design

ABSTRACT

Objective

To examine whether extended compared to standard level of physiotherapy is feasible and has beneficial effects on physical function in ICU survivors.

Methods

This prospective pilot study with a before and after design included patients discharged from ICU to a surgical ward. The comparison group were recruited between January and April 2019 and received standard level of physiotherapy. The intervention group were recruited between May and December 2019 and received extended physiotherapy, corresponding to 50% additional physiotherapist, working 4 hours per weekday. The intervention participants received an individual rehabilitation plan developed in collaboration with a ward-based physiotherapist, and an extended number of sessions provided by the extra resource included practicing individualized exercises, for example walking and stair climbing. Physical function was measured with the Chelsea Critical Care Physical Assessment tool (CPAx) at ICU discharge, during hospital stay and discharge. Group differences were analyzed using the Mann–Whitney U-test and Chi² test.

Results

Out of 46 eligible patients, 39 (85%) fulfilled the study (comparison n = 12, intervention n = 27) and were included in the final analyses. No adverse events occurred, and the attendance rate was high (98.5%). There were no statistically significant differences between the groups regarding physical function, hospital stay, and readmissions, but there were tendencies to better outcomes in all these parameters in favor of the intervention group. Additionally, patients in the intervention group had statistically significantly higher scores in the CPAx items “transferring from bed to chair” (median 5 vs 4, p = .039) and “stepping” (median 5 vs 4, p = .005) at hospital discharge.

Conclusion

This pilot study indicates that extended physiotherapy after ICU discharge is feasible and does not entail patient safety risks. However, determining the potential beneficial effects for the patients remains to be evaluated in a larger trial.

KEYWORDS:

• Critical care
• ambulation
• rehabilitation
• physical therapy
• physical assessment

Introduction

Acute muscle wasting, muscle atrophy, and neuromuscular dysfunction represent key challenges that affect critically ill patients in intensive care units (ICU). A muscle mass loss up to 4% per day in the ICU is not uncommon (Griffiths and Hall, 2010). This is associated with increased risk of short-term complications, as well as delayed functional recovery and reduced health-related quality of life in a longer perspective (Hermans and den Berghe G, 2015). A previous study, including a follow-up of survivors of acute respiratory distress syndrome, showed that they performed only 66% of predicted walk distance after 1 year and 76% after 5 years (Herridge et al., 2011).

To prevent muscle mass deterioration and enhance recovery, rehabilitation and early mobilization is endorsed as soon as the patient’s vital status has been stabilized (Arias-Fernandez, Romero-Martin, Gomez-Salgado, and Fernandez-Garcia, 2018). Early mobilization in ICU has been shown to improve muscle strength, physical function, and reduce ICU and hospital stay (Doiron, Hoffmann, and Beller, 2018; Morris et al., 2008; Truong, Fan, Brower, and Needham, 2009). International guidelines endorse continued rehabilitation after ICU discharge, but attainments gained in the ICU are often lost shortly after discharge to the ward (Hopkins et al., 2012; Pandullo et al., 2015). This is a clinically important problem that most likely can be attributed to too little exercise after ICU discharge for the physically weakened ICU patients. However, to the best of our knowledge, we have only identified two interventional studies that have evaluated the effect of extended physiotherapy after ICU discharge. One rather small randomized controlled trial (RCT) found a non-significant tendency for shorter hospital stay related to extended physiotherapy (Gruther et al., 2017). Another trial showed improved patient satisfaction but not improved physical outcome after hospital discharge (Walsh et al., 2015). Despite these somewhat negative results, we believe that more physiotherapy can be beneficial for the newly discharged ICU patient, and we hypothesized that surgical patients with shorter ICU length of stay would probably benefit from extended physiotherapy at the hospital ward. In an overview of systematic reviews, it was concluded that the evidence to support the effect of physiotherapy delivered after ICU discharge was limited and more studies in this area of research were recommended (Connolly, O’Neill, Salisbury, and Blackwood, 2016). Therefore, we conducted a single-center pilot study including surgical ICU patients, aiming to examine whether extended level of physiotherapy delivered in the wards after ICU discharge would be feasible and have beneficial effects on physical function at hospital discharge.

Methods

Study design, setting, and participants

This was a prospective, single-center pilot study, with a before-and-after design with a non-randomized comparison group. The study was conducted in four general surgical wards at an urban emergency hospital in 2019, in Stockholm, Sweden. Critically ill patients treated in an ICU with planned discharge to the surgical ward were screened for inclusion. All ICU admissions were unplanned, even for the patients who had undergone elective surgery.

Inclusion criteria were adult patients with an ICU stay >48 hours, who understood the Swedish language. Patients who received palliative care with no intention of rehabilitation were excluded. Eligible patients received written and verbal information regarding the study and were asked to participate by a physiotherapist before the ICU discharge. All included patients provided written informed consent. Ethical approval was obtained from the Regional Ethical Review Board (No: 2018/2102-31/1) and the study was performed according to the Helsinki Declaration (World Medical Association, 2013). The study was registered in ClinicalTrial.gov (NCT05349994). The study received regional funding, covering one extra physiotherapist working 50% for 6 months.

Procedure

Comparison group

Patients in the comparison group were consecutively included between January 29^th and April 30^th, 2019. The participants received standard level of physiotherapy from ward-based physiotherapists, corresponding to a 1.75 employment (i.e. two physiotherapists working 40 and 30 hours/week Monday to Friday, respectively) distributed on 40–48 patients in four surgical wards approximately 10–12 patients per ward. The treatments included mobilization in or out of bed, breathing exercises with specialized devices and/or assisted coughing. The physiotherapists prescribed walking aids and evaluated the patients’ need of further rehabilitation. Those still in need of support in daily activities were referred to rehabilitation hospitals.

Intervention group

Inclusion of the intervention group started May 2, 2019, and ended December 16, 2019, with a pause, between June 10 and August 31 due to summer vacation. The intervention group received an extended level of physiotherapy – an extra physiotherapist resource corresponding to a 50% employment (4 h/weekday) which incorporated 1–2 sessions of approximately 30 minutes for 2–4 patients in the surgical wards during weekdays. In addition to the standard physiotherapist treatment, the intervention patients received an individual rehabilitation plan developed in collaboration between the patient and the ward-based physiotherapist at admission to the surgical ward or on the following day. A physiotherapist, with more than 20 years of experience of ICU as well as surgical ward physiotherapy, performed the interventions.

Data collection

Patient characteristics and clinical variables at baseline were obtained from medical records. The following variables were collected: age, sex, smoking habits, use of walking aids, and the Simplified Acute Physiology Score 3 (SAPS 3), a scoring system for severity of illness at ICU admission that can predict vital status at hospital discharge (Moreno et al., 2005).

Outcomes

Feasibility measures included the number of recruited patients out of eligible ones, the number of planned interventions not delivered, and the number of adverse events reported by the study physiotherapist after each session. Potential adverse events were defined as falls, unbearable pain, or experience of discomfort leading to interruption in mobilization.

Physical function was assessed with the Chelsea Critical Care Physical Assessment tool (CPAx), a tool validated for patients with critical illness (Corner, Soni, Handy, and Brett, 2014). The CPAx contains 10 items of functional ability, each graded on a scale from 0 to 5 where 0 is totally dependent and 5 is independent, generating a maximum score of 50. Thus, a higher score corresponds to better physical function. CPAx has been translated into Swedish and has a high inter-rater reliability (Holdar et al., 2021). In addition, the sub scores can be displayed on a radar chart that visualizes the patient’s physical progress and enables pedagogical discussions between the patient and the multidisciplinary team (Corner, Soni, Handy, and Brett, 2014). CPAx was measured at several time points: at ICU discharge, 3 days after admission to the ward, every fifth day during hospital ward stay, and at discharge from the hospital. The assessment at ICU discharge was performed by one of the three ICU physiotherapists (clinical experience ranging from 5 to 25 years). Other assessments were mostly done by the physiotherapist at the ward. Data regarding length of stay (LOS) at the ward (days), readmission to the ICU (yes or no), and discharge to independent living (yes or no) were also collected. ICU readmission was defined as readmission from the first ICU discharge until discharge from hospital.

Statistical analyses

Demographic and clinical data were described with medians and interquartile ranges, counts, and percentages. Differences were calculated with Mann–Whitney U-test and Chi² test or Fisher’s exact test as appropriate. Improvements in physical function (CPAx) between groups based on the delta value between CPAx at ICU discharge and CPAx at hospital discharge, and in absolute CPAx score (total and per item) at discharge from hospital were calculated with Mann–Whitney U-test. Physical functions within groups were depicted with radar charts. Differences between groups regarding LOS in ward, ICU readmission, and discharge to independent living were calculated with Mann–Whitney U-test, Chi² test, and Fisher's exact test, respectively. All analyses were performed per-protocol. A two-sided p-value of <0.05 was considered statistically significant. Statistical analyses were performed using SPSS statistics (version 28.0; IBM Corp. Armonk, NY).

Results

Study enrollment

Of the 443 patients who were discharged alive from the ICU during the study period, 264 were discharged to a surgical ward. Out of these, 46 were eligible for the study and all 46 consented to participate. One patient later withdrew participation, and one was excluded due to transfer to a medical ward. In total, 14 patients were included in the comparison group and 30 in the intervention group. During the study, two participants died in the hospital, two were discharged before final CPAx assessment, and one patient withdrew participation. In total, 12 participants in the comparison group and 27 in the intervention group (Figure 1) were included in the analyses. Most patients were older men undergoing acute or elective surgery for gastrointestinal or vascular reasons (Table 1). At ICU discharge, no patients obtained the minimal (0) or maximal (50) CPAx score. However, at hospital discharge, 1 patient in the comparison group and 11 in the intervention group received the maximal score (ceiling effect).

Table 1. Demographic and clinical characteristics of the comparison (n = 12) and intervention group (n = 27).^a

Variable	Comparison group (n = 12)	Intervention group (n = 27)	P value
Age, years, median (IQR)	73 (71–76)	71 (61–82)	.663
Sex, male/female, n (%)	7 (58.3)/5 (41.7)	19 (70.3)/8 (29.6)	.462
Smoking, yes/no, n (%)	3 (25.0)/9 (75.0)	7 (25.9)/20 (74.1)	.951
SAPS 3, median (IQR)	59 (52–66)	56 (45–71)	.620
ICU length of stay, median days (IQR)	8 (4–26)	5 (3–8)	.221
Walking aid (before hospitalization), yes/no, n (%)	1 (8.3)/11 (91.7)	8 (29.6)/19 (70.4)	.145
Diagnosis^b, n (%) Gastrointestinal Vascular Pancreatitis Multitrauma	7 (58.3) 4 (33.3) 1 (8.3) 0 (0.0)	14 (51.9) 9 (33.3) 3(11.1) 1 (3.7)	.903
Surgery, n (%) No surgery Elective surgery Emergency surgery	2 (16.7) 2 (16.7) 8 (66.7)	6 (22.2) 13 (48.1) 8 (29.6)	.080

^aIQR = interquartile range, SAPS 3 = Simplified Acute Physiology Score 3; ^b Description of diagnoses: intestinal ischemia (n = 2 vs. 0), perforated esophagus (n = 1 vs. 0) aortic aneurysm (n = 4 vs. 9), acute pancreatitis (n = 1 vs. 2), peritonitis (n = 1 vs. 4), septic shock (n = 1 vs. 0), vascular gastrointestinal disease (n = 1 vs. 2), aspiration (n = 1 vs. 1), multitrauma (n = 0 vs. 1), postoperative complications (n = 0 vs. 2), cholangitis (n = 0 vs. 1), and severe abdominal complications (n = 0 vs. 5).

Figure 1. Flow chart of study enrollment in the comparison group (a) and intervention group (b).

Intervention compliance

Patients in the intervention group received 1.3 sessions per patient and weekday compared to 0.7 sessions per patient and weekday in the comparison group. The extended number of sessions entailed practicing individualized exercises, for example, walking with or without walking aids and stair climbing. The number of attended physiotherapy sessions was 330 out of 335 planned within the intervention group (corresponding to 98.5% attendance rate). Reasons for not attending were because of intervening surgical procedures (n = 3), leave of absence (n = 1) and patient decline (n = 1). No adverse events occurred during any of the sessions. Characteristics of the ICU and the surgical wards did not change regarding patient mix, staffing, and nurse–patient ratio between the periods of data collection.

Effects of extended physiotherapy on physical function

Physical function (CPAx) at ICU discharge (p = .150) was similar between the two groups (Supplementary Table 1). At hospital discharge patients who received extended level of physiotherapy had a total median CPAx score of 48 (IQR 42–50) compared with 45 (IQR 36–46) for the comparison group (p = .056) (Table 2 and Supplementary Table 2). The scores for individual items “transferring from bed to chair” (p = .039) and “stepping” (p = .005) were statistically significantly higher for those with extended level of physiotherapy (Figure 2 and Supplementary Table 2). Improvements in CPAx score over time from ICU discharge to hospital discharge was 20 (IQR 15–23) for the intervention group versus 18 (IQR 9–26) for the comparison group (p = .663).

Table 2. Median CPAx score at ICU and hospital discharge, in total and per item, divided into comparison (n = 12) and intervention group (n = 27).^a

	Comparison group (n = 12)		Intervention group (n = 27)
Variable	Median score, ICU discharge (IQR)	Median score, hospital discharge (IQR)	Median score, ICU discharge (IQR)	Median score, hospital discharge (IQR)
Total CPAx score (0–50 p)	23 (17–25)	45 (36–46)	26 (19–32)	48 (42–50)
Respiratory function (0–5 p)	4 (4–5)	5 (5–5)	4 (4–5)	5 (5–5)
Cough	4 (4–5)	5 (5–5)	4 (3–5)	5 (5–5)
Moving within bed	3 (2–3)	5 (4–5)	3 (2–4)	5 (4–5)
Supine to sitting	2 (2–3)	4 (4–5)	3 (2–3)	5 (4–5)
Dynamic sitting	4 (3–4)	5 (5–5)	4 (3–4)	5 (5–5)
Standing balance	0 (0–3)	4 (3–5)	2 (0–3)	5 (4–5)
Sit to stand	0 (0–2)	4 (3–5)	2 (0–2)	5 (4–5)
Transferring bed to chair	0 (0–0)	4 (4–5)	0 (0–3)	5 (4–5)
Stepping	0 (0–1)	4 (3–4)	0 (0–2)	5 (4–5)
Grip strength	2 (1–5)	4 (3–5)	3 (2–5)	5 (3–5)

^aICU = Intensive care unit, IQR = interquartile range, CPAx = Chelsea Critical Care Physical Assessment Tool.

Figure 2. Radar chart displaying CPAx change from ICU discharge to hospital discharge, median per item, divided into comparison (a) and intervention group (b).

Length of ward stay and readmission to ICU

Among patients who received extended level of physiotherapy, the length of surgical ward stay was shorter (9 versus 16 days), ICU readmissions were fewer (18.5% versus 50.0%), and a higher number of patients were discharged to independent living (33.3% versus 16.7%) (Table 3). However, these improvements were not statistically significant.

Table 3. Length-of-stay in ward, readmissions to ICU, and discharge to independent living for the intervention and comparison group.^a

Variable	Comparison group (n = 12)	Intervention group (n = 27)	P value
LOS in ward, median (IQR)	16 (9–32)	9 (7–17)	.251
Readmission to ICU, n (%)	6 (50.0)	5 (18.5)	.221
Discharged to independent living, n (%)	2 (16.7)	9 (33.3)	.253

^aLOS = Length of stay, IQR = interquartile range, ICU = intensive care unit

Discussion

This pilot interventional study indicates that extended physiotherapy after ICU discharge is feasible and does not seem to have any negative effects for the patients. Out of 46 eligible patients 39 (85%) fulfilled the study and were included in the final analyses. No adverse events occurred, and the attendance rate was high (98,5%). A session attendance of at least 80% was defined as high attendance according to Kampshoff et al. (2016).

The indications regarding physical function are in line with a previous RCT with surgical ICU patients, where the intervention group received early goal-directed mobilization and improved their functional independence at hospital discharge compared to the standard care group (Schaller et al., 2016). However, the intervention was conducted already in the ICU, which limits comparability to our study were the intervention was ward-based and conducted after ICU stay. In our study, within the comparison group, some participants had a CPAx score of 3 in items involving transferring at discharge from hospital. That means they are dependent on support in daily activities, which would probably have an impact on quality of life for an individual. In a broader context, it has also been shown that extended physiotherapy with an extra 19 minutes per day in patients with acute or subacute conditions, not specifically post-ICU, appeared to reduce length of hospital stay and improve daily activity (Peiris et al., 2018). A rather small RCT with ICU survivors compared extended physiotherapy at a general ward in 23 patients with 27 patients who received standard care as suggested by a primary care team. In the intention-to-treat analysis, a tendency to shortened hospital stay was found, but the difference was not statistically significant (Gruther et al., 2017).

Somewhat in contrary, Walsh et al. (2015) found no difference in mobility assessed with the Rivermead Mobility Index at discharge from hospital, for an intervention group with increased physiotherapy and nutritional therapy during post-ICU hospital stay. This trial showed that patient satisfaction improved for those who received the intervention, but physical recovery and health-related quality of life were similar between the groups. However, the evaluation was conducted 3, 6, and 12 months after discharge and the multimodality of the intervention made evaluation of the sole effect of exercise more complicated (Walsh et al., 2015). Discrepancy in results could be due to different outcome measures and differences in interventions. In our study, both groups improved in overall physical recovery over time, but the comparison group spent more days at the hospital than the intervention group that might have affected the results in favor of the comparison group. Also, the comparison group received physiotherapy as standard care in the surgical ward. The proportion of surgery type differed somewhat between groups (i.e. 48% of the intervention group had elective surgery compared to only 17% in the comparison group). The difference was not statistically significant but should still be kept in mind when interpreting the results as undergoing elective compared to emergency surgery could indicate a more favorable postoperative recovery regarding LOS and discharge destination. Discharge to home is often of great importance to the individual and could lead to lower costs for health care and society compared to discharge to geriatric care or a rehabilitation unit. In a previous study by Corner, Soni, Handy, and Brett (2014), the CPAx score showed strong associations with hospital discharge location, with a median score of 39 for patients who returned home without rehabilitation needs, and a median score of 31 for patients who returned home with community support.

In our study, we could see a possible link in the intervention group between their median CPAx score and being able to return home. A greater number of participants in the intervention group could be discharged to independent living compared to the comparison group receiving standard care. We did, however, only examine discharge destination and not the need for support after hospital discharge. Comparability between studies is also limited due to international differences in hospital and rehabilitation organizations.

Despite the lack of statistically significant differences regarding LOS, the median was 7 days longer in the comparison group. The costs of extra physiotherapy related to fewer days spent in the ward should be taken into consideration when cost priorities are made. In a study with similar total sample size and a greater absolute difference in LOS between intervention and standard care groups, there was a tendency toward statistical significance (Gruther et al., 2017). Length of stay is a complex outcome to measure due to multiple influential factors such as traditions within hospitals, patients’ expectations, and number of beds in the rehabilitation units. Moreover, there are international differences in healthcare systems that hamper comparison between countries.

Evaluating patients with CPAx in surgical wards was possible. However, ceiling effects appeared when used at hospital discharge as 11 out of 27 (40.7%) reached the maximal CPAx score of 50 in the intervention group at discharge from hospital, compared to one in the comparison group. Thus, more than a third of the participants in the intervention group might have had an even greater physical improvement than measured with the CPAx. According to a previous article, the ceiling effect of the CPAx in the ICU is 0.8% and floor effect 3.2% (Corner, Soni, Handy, and Brett, 2014). CPAx is not yet validated for patients outside the ICU context, whereas assessment tools that capture change in physical function both while in the ICU and after transfer to a ward are desirable.

Strengths and limitations

The main strengths of this study were the clinically driven research question and that the study was conducted directly in clinical practice. The main limitation of this study was the non-randomized design, which increases the risk of confounding. With a randomized design, it would also have been possible to blind the assessing physiotherapist to group allocation. This will be taken into consideration when designing a larger trial.

In addition, the risk of selection bias was apparent as we only included patients transferred to surgical wards where postoperative mobilization is a routine among staff, and therefore all patients are encouraged and supported in mobilization. That could have had a positive effect in the comparison group, if they were frequently mobilized with staff, which we did not control for. In the study hospital, surgical patients are seldom treated for more than 48 hours in the ICU, which limited the number of eligible patients. A broader inclusion criterion adding not just discharge to surgical wards but other types of general wards would increase the number of patients eligible for inclusion in the study, which might increase generalizability and the recruitment rate for a larger trial in combination with a multicenter approach.

When patients were asked to participate, they received information about assessments of physical function, which could increase awareness among participants in both groups and possibly increase their physical activity at the ward. A more suitable measurement tool to evaluate physical function both at ICU and after discharge with less ceiling effect would have been desirable. Nevertheless, to our knowledge, no such measurement tool has been developed yet. There is always a risk of measurement error due to several assessors, and other factors such as medical procedures that might influence the patient’s condition and physical capacity during hospital stay. The physiotherapist who delivered the intervention also conducted most of the assessments. Perhaps, a more optimal approach would have been to let an independent person assess the outcomes.

Conclusion

This pilot study indicates that extended physiotherapy after ICU discharge is feasible and does not entail patient safety risks. However, determining the potential beneficial effects for the patients remains to be evaluated in a larger trial. The main modification when planning a larger trial would be adding a randomized design with an assessor blinded evaluation.

Acknowledgments

A special thanks to the physiotherapists Vidar Haugen, Matheo Johansson, Annelie Söderlund, and Marion Warnhoff. The authors would also like to thank Hans Järnbert-Pettersson and Mikael Andersson Franko for their expert advice regarding statistics. The work was funded as a quality improvement project by the Regional County Council.

Supplemental data

Supplemental data for this article can be accessed online at https://doi.org/10.1080/09593985.2022.2143251.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Regional County Council in Stockholm.