Early Mobilization Interventions in the Intensive Care Unit: Ongoing and Unpublished Randomized Trials

Background Critical care societies recommend early mobilization (EM) as standard practice in the intensive care unit (ICU) setting. However, there is limited randomized controlled trial (RCT) evidence supporting EM's effectiveness. Our objective was to identify ongoing or completed RCTs assessing EM's effectiveness in the ICU. Method We searched ClinicalTrials.gov and the Australian New Zealand Clinical Trials Registry for ongoing or completed but not published RCTs in an ICU setting with objective outcome measures. Results There were 14 RCTs included in the analysis. All studies were in the general or mixed ICU setting (N=14). Half of the studies (N=7) were small RCTs (<100 projected participants) and half (N=7) were medium-sized RCTs (100–999 participants). Inclusion criteria included mechanical ventilation use or expected use (N=13) and prehospital functional status (N=7). Primary EM interventions were standard physiotherapist-based activities (N=4), cycling (N=9), and electrical muscle stimulation (N=1). Only one study involved nurse-led EM. The most common assessment tool was the 6-minute walk test (N=6). Primary outcome measures were physiological (N=3), clinical (N=3), patient-centered (N=7), and healthcare resource use (N=1). Most studies (N=8) involved post-ICU follow-up measures up to 1-year posthospitalization. There were no studies targeting older adults or people with acute cardiac disease. Conclusion Identified studies will further the evidence base for EM's effectiveness. There is a need for studies looking at specific patient populations that may benefit from EM, such as older adults and cardiac patients, as well as for novel EM delivery strategies, such as nurse-led EM.


Background and Rationale
Early mobilization (EM) involves progressive mobilization activities that start upon hemodynamic and respiratory stabilization, typically within 24-48 hours of intensive care unit (ICU) admission [1]. e goal of EM is to prevent loss of muscle function, promote mobility recovery, and maintain or help patients regain prehospital functional capabilities.
ere is observational evidence that EM in the ICU is associated with improvements in muscle strength and physical function, decreased rates of delirium, and reduced healthcare resource utilization (shorter ICU and hospital stays and reduced hospital readmission rate) [2][3][4][5][6]. e safety and feasibility of EM in critically ill patients has also been established [7]. As a result, critical care professional societies recommend EM as part of routine ICU care [8,9].
Yet despite the evidence for safety and feasibility, there is limited randomized evidence supporting the effectiveness of EM in the ICU setting [10]. A recent Cochrane review of randomized controlled trials (RCTs) of EM interventions in the ICU found only 4 studies with 690 patients and concluded that there was insufficient randomized evidence for EM's effectiveness [10]. While there has been a considerable number of EM observational studies, there have been relatively few published RCTs ( Figure 1). As a result, there is limited good quality evidence for the ability of EM to maintain or improve physical function at discharge and patient-centred outcomes following hospital discharge. A lack of strong evidence for EM's effectiveness is a noted barrier to the implementation of EM [11].
Given the dissonance between EM's current evidence base and critical care societal recommendations for EM, there is a need to characterize upcoming studies that could provide stronger evidence for EM's benefits. us, the objective of this review was to identify ongoing or completed but not yet published EM RCTs in the ICU setting. We also identified important areas of EM research that currently lack data and that could be addressed in future research design.

Methods
We searched ClinicalTrials.gov and the Australian New Zealand Clinical Trials Registry (ANZCTR), two web-based databases that list ongoing and completed RCTs. ClinicalTrials.gov is operated by the United States National Library of Medicine and contains more than 130,000 trials from more than 170 countries worldwide. e ANZCTR is an Australian government-funded registry of clinical trials that includes studies primarily, but not exclusively, from Australia and New Zealand. e ANZCTR was included in the search since a preliminary literature search and prior systematic reviews identified a research group in Australia and New Zealand that was active in EM research. For both of these databases, we used the following search query: (early mobilization OR early mobilization OR early mobility OR early rehab OR early rehabilitation) AND (randomized). Studies were included if they were (1) randomized controlled trials, (2) ongoing or completed but not yet published, (3) in the ICU setting, and (4) had an objective outcome measure. Studies were excluded if they were observational studies, in a pediatric population, terminated prematurely, involved cohorts of patients post-orthopedic surgery, poststroke, and posttraumatic brain injury, as these patients possess specific deficits with specialized needs for mobilization and physical rehabilitation. Studies were also excluded if they were completed but unpublished for 5 or more years, as the likelihood of eventual publication was low. Two independent reviewers (J. M. and J. F.) reviewed all the studies that met inclusion criteria for suitability into the analysis; disagreements were resolved by a third reviewer (M. G.). To confirm unpublished status, PubMed and Web of Knowledge were queried for each eligible trial. If no published study was found, attempts to contact authors by e-mail were made. e search was performed on December 19, 2018. Trial characteristics were retrieved from the study listings and from trial methodology papers when available. Attempts to contact authors by e-mail were made for missing or unavailable data. Projected study size was defined as small (<100 participants), medium (100-999 participants), and large (≥1000 participants).

Results
ere were 392 studies identified by the search, of which 14 were included in the analysis ( Figure 2; Table 1). All studies (N � 14) involved the general ICU setting. e projected study size was small (N � 7) or medium (N � 7) RCTs, with study populations ranging in size from 28 to 772 participants. Nearly all studies (N � 13) involved mechanical ventilation (MV) as part of the inclusion criteria. Other inclusion criteria included prehospital functional status (N � 7) and expected prolonged MV (N � 7). Exclusion criteria involved moribund patients (N � 7), obesity (N � 7), neuromuscular and central nervous system disorders (N � 11), lower extremity and skin deficits (N � 8), and cognitive impairment (N � 5). Special populations included middle age/older adults (age > 42, N � 1), suspected sepsis (N � 2), and postsurgical (N � 3). e primary EM intervention was standard physiotherapy using active and passive activity (N � 4), cycling (N � 9), and electrical muscle stimulation (N � 1) ( Figure 3). e EM protocols and the comparator group activities are described in Table S1. Studies utilizing cycling also offered regular physiotherapy sessions to patients, and 3 studies offered both cycling and electrical stimulation. Other components of the intervention were electrical stimulation either in combination with cycling or alone (N � 3) and cognitive components (N � 3). EM interventions were led by physiotherapists (N � 2) and occupational therapists (N � 1) and co-led by nursing and physiotherapy (N � 1), with no specified leader in 10 studies. EM initiation was within 48 hours of ICU admission (N � 4), not specified (N � 4), or required the patient to be medically cleared prior to the initiation of EM (N � 6).

Discussion
We identified 14 RCTs from two web-based clinical trial registries that are evaluating the impact of EM in an ICU setting. e majority of studies focused on patients who are undergoing or have been liberated from mechanical ventilation. Studies are small to medium-sized. Patients are being excluded from studies due to obesity, neuromuscular issues, and poor prognosis. ere are no studies focused on older adults or acute cardiovascular disease patients. Cyclingbased and electrical stimulation therapies are being studied, in addition to standard active and passive physiotherapy regimens. Outcome measures collected are heterogeneous. Most studies included posthospitalization follow-up.    Critical Care Research and Practice 5

Critical Care Research and Practice
Prior EM RCTs have focused on critically ill patients receiving mechanical ventilation in a general, medical, or cardiac surgical ICU [1,10]. e patients included in these RCTs were typically adults with a mean age in the late 50 s. EM interventions included passive and active range of motion, mobility activities in and around the bed, upper limb exercises, and ambulation. ere were methodological issues with prior studies, such as poor description of intervention arms, high rates of participant dropout, and lack of blinding of outcome assessors. A variety of functional outcome measures were used in prior studies, which precluded meta-analysis to characterize the impact of EM on functional outcomes. A Cochrane review of EM studies that included four RCTs reported studies using the Barthel Index, a measure of independent functional status, and the Acute Care Index of Function, the Physical Function ICU test score, and the Short Physical Performance Battery, as measures of physical performance [10]. Our study similarly found that upcoming RCTs are using a number of different functional measurement scales.
ere is a need to standardize functional measurement tools for reporting the effect of mobility interventions on functional status in the ICU.
Previous RCTs focused on standard physiotherapy-based protocols [1,10]. Cycling-based activities and electrical stimulation, in addition to or as a replacement of standard physiotherapy, have emerged as therapeutic techniques for EM interventions. Cycling-based activities generally started with passive cycling with flexion/extension of lower extremities at a specific rate (i.e., # cycles/minute) for a predetermined duration with gradual increase in activity duration and intensity as tolerated. e advantages of cycling-based activities include the ability to deliver EM to patients who are mechanically ventilated and sedated as passive cycling. However, the patient is generally required to be hemodynamically stable prior to initiating cycling activities and patient participation is necessary when being advanced to active cycling. Electrical stimulation-based activities require electrodes to be attached to certain muscle groups such as hamstrings, quadriceps, and calves such that electrical impulses can be delivered to stimulate the muscle groups, which can be seen as visible contractions. e key advantages of electrical muscle stimulation are that it is an external stimulus, which does not require any participation from the patient and makes it feasible to be delivered in a critical care setting and can be delivered to patients without causing any hemodynamic instability. Prior studies investigating the impact of both cycling and EMS on muscle weakness and function in the ICU have had mixed findings. Both techniques are safe with low adverse event rates, but their impact on muscle wasting and strength have been mixed [12,13]. A recent randomized trial in over 300 critically ill patients combined early in-bed cycling and EMS and did not find an improvement in muscle strength [14].
ICU survivors have been shown to suffer from postintensive care syndrome, which includes physical, psychological, and cognitive impairments leading to functional decline [15]. Functional outcomes, a key patient-centered measure, were frequently included as primary and/or secondary outcomes in the included studies. ere is increasing recognition that many patients prioritize individual quality of life and functional independence over other more conventional societal measures [16]. EM may also improve mental health outcomes. An RCT looking at physiotherapyled EM in mechanically ventilated patients found a trend toward improved mental health outcomes [17]. Several upcoming studies included cognitive components in addition to physical therapy, such as providing positive feedback to patients, performing cognitive tasks, or following strict checklists to allow spontaneous awakening trials and to address pain and delirium [18,19]. A pilot RCT that combined cognitive therapy and early physical therapy with usual care in ICU patients with respiratory failure or shock did not show improvement in the intervention arm [18]. e effect of EM in other important critical care patient populations needs to be considered in future research design ( Table 2).
ere are knowledge gaps in understanding the impact of EM, such as the timing of initiation of EM and the type and method of EM delivery.
ere are also certain populations that may stand to benefit from EM but have not been adequately explored, such as older adults, people with acute cardiovascular disease, people with preexisting mobility impairment, and nonmechanically ventilated critically ill patients [20]. Older adults may be more likely to benefit from an intervention that impacts functional status due to the higher burden of preexisting functional impairments and increased risk of functional decline during hospitalization. Prior EM studies largely excluded older adults and focused on younger overall cohorts [1,21]. e absolute number and percentage of older adults in the ICU is growing as well [22]. ere are also few studies evaluating EM in patients with acute CV disease [23]. In addition, patients with underlying neuromuscular/cerebral pathologies were excluded due to anticipated difficulty in weaning. ese patients may still benefit from EM, as they are starting off at a disadvantage and therefore would be important to minimize further set-backs. To evaluate the impact of EM in the ICU on key outcome measures, there is a need for collaborative groups to obtain funding support to conduct large, multicenter trials to obtain high-quality data. However, achieving recruitment of large study population may be challenging in the ICU setting. We found two proposed studies with recruitment targets of over 700 patients. Previous RCTs in the ICU have shown considerable difficulty at meeting their target sample sizes for the primary outcome and many had to be stopped early due to insufficient accrual or logistical issues [24]. Strategies to boost study recruitment and to deal with potential logistical challenges, such as multicenter enrolment and the use of existing critical care trial research networks, should be considered in trial design.
Our study has a number of limitations. First, we used only two web-based clinical trial registries. ere are other clinical trial registries available. Clinicaltrials.org is an international registry and contains the largest number of trial listings. In addition, a preliminary web and literature search indicated that there were potential studies to include in the ANZCTR registry and there were few potential studies in other trial registries. Second, we excluded certain ICU settings, such as poststroke and orthopedics, where a specialized approach to rehabilitation is required.
ird, we attempted to contact the author to ensure that there was intent to publish the completed but unpublished studies. However, it is possible that some of these studies might not be published. One possible barrier to publish may have been difficulty establishing adequate follow-up. We also excluded studies that were completed over 5 years ago but were not yet published.

Conclusion
EM in the ICU has the potential to improve outcomes but current data on its' effectiveness are limited. e studies identified can yield more information regarding which EM modalities may be feasible, as well as the potential short-and long-term benefits. is review has identified areas that can be further explored, such as looking at specific study populations and looking at more innovative EM delivery methods.

Data Availability
e data used to support the findings of this study are included within the article, under the figure and table sections.

Conflicts of Interest
e authors declare that they have no conflicts of interest.

Supplementary Materials
Supplementary Table 1 (S1): a detailed description of the intervention and comparison for each clinical trial. Details of the nature of early mobilization are described, such as the type of activity and timing of initiation and the duration and frequency of the intervention, if details were provided. (Supplementary Materials)