Early mobilization after stroke: early adoption but limited evidence.

Abstract

HomeStrokeVol. 46, No. 4Early Mobilization After Stroke Free AccessResearch ArticlePDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissionsDownload Articles + Supplements ShareShare onFacebookTwitterLinked InMendeleyReddit Jump toSupplementary MaterialsFree AccessResearch ArticlePDF/EPUBEarly Mobilization After StrokeEarly Adoption but Limited Evidence Julie Bernhardt, PhD, Coralie English, PhD, Liam Johnson, PhD and Toby B. Cumming, PhD Julie BernhardtJulie Bernhardt From the Florey Institute of Neuroscience and Mental Health, University of Melbourne, Australia (J.B., C.E., L.J., T.B.C.); International Centre for Allied Health Evidence, University of South Australia, Australia (C.E.); Institute of Sport, Exercise and Active Living, Victoria University, Melbourne, Australia (L.J.). Search for more papers by this author , Coralie EnglishCoralie English From the Florey Institute of Neuroscience and Mental Health, University of Melbourne, Australia (J.B., C.E., L.J., T.B.C.); International Centre for Allied Health Evidence, University of South Australia, Australia (C.E.); Institute of Sport, Exercise and Active Living, Victoria University, Melbourne, Australia (L.J.). Search for more papers by this author , Liam JohnsonLiam Johnson From the Florey Institute of Neuroscience and Mental Health, University of Melbourne, Australia (J.B., C.E., L.J., T.B.C.); International Centre for Allied Health Evidence, University of South Australia, Australia (C.E.); Institute of Sport, Exercise and Active Living, Victoria University, Melbourne, Australia (L.J.). Search for more papers by this author and Toby B. CummingToby B. Cumming From the Florey Institute of Neuroscience and Mental Health, University of Melbourne, Australia (J.B., C.E., L.J., T.B.C.); International Centre for Allied Health Evidence, University of South Australia, Australia (C.E.); Institute of Sport, Exercise and Active Living, Victoria University, Melbourne, Australia (L.J.). Search for more papers by this author Originally published17 Feb 2015https://doi.org/10.1161/STROKEAHA.114.007434Stroke. 2015;46:1141–1146Other version(s) of this articleYou are viewing the most recent version of this article. Previous versions: January 1, 2015: Previous Version 1 In the last decade, increasing attention has been paid to understanding the components of care that might contribute to the stroke unit effect. Early mobilization, in its many guises, is one component of care proposed to contribute to the survival and recovery benefits of stroke unit care.1 This topical review provides an overview of the current evidence, research, and practice recommendations for early mobilization after stroke.We Need to Define Early MobilizationAs a term, early mobilization is problematic. There is no common understanding of the meaning of early (eg, hours, days, weeks, months) or mobilization (movement of, eg, cells, joints, limbs, people). A recurring theme in this review, inadequate definition currently limits our ability to synthesize information on the topic. For example, in some clinical trials of mobility interventions started soon after stroke, mobilization is used to describe a program of task-specific standing and walking retraining (rehabilitation) delivered by therapists or nurses and continued throughout the acute hospital stay.2,3 In other cases, mobilization refers simply to moving a patients’ limbs in bed or sitting them out of bed. The timing of commencement of activity is also highly variable and often hard to determine. As both what we do (intervention type, intensity, frequency, amount), and when we do it, may confer benefit or harm, we highlight variations in definition where relevant. We have focused our review on out-of-bed interventions commencing in the first 24 to 72 hours after stroke, as this is the period of greatest clinical uncertainty.Brief History of Early Mobilization After StrokeEarly mobilization was first discussed at a Swedish consensus conference on stroke care in the mid-1980s (Bo Norving and Bent Indredavik, personal communication, 2014) with several local guidelines in Norway and Sweden recommending the practice. Early mobilization became more prominent in the literature in the early 1990s when Indredavik and colleagues reported their clinical trial results showing marked reduction in death and disability in patients managed in a stroke unit with a focus on early rehabilitation and mobilization when compared with general medical ward care.4 This trial formed part of the seminal Cochrane review by Langhorne and colleagues5 that clearly demonstrated the benefits of organized stroke unit care. Early mobilization/rehabilitation was incorporated into subsequent discussions as an important component of stroke unit care1 and began to appear in national clinical guidelines around 1994.6 The first randomized controlled trial of early mobilization commenced in 2004 (A Very Early Rehabilitation Trial [AVERT] Phase II),3 testing the feasibility and safety of starting task-specific mobility training, in a stroke unit context, within 24 hours of stroke onset. The intervention protocol was informed by observational studies, which highlighted that in Indredaviks’ Norwegian stroke unit, a philosophy of early mobilization did result in patients, even those with severe stroke, being more physically active throughout the day.7 The sub-24 hour timing of intervention commencement in AVERT Phase II was selected to better inform clinical practice, given the uncertainty about the potential harms and benefits of starting activity, particularly in the first day(s) after stroke. This will be covered in more depth below.Trials EvidenceOnly 4 completed trials (AVERT Phase II, n=713; Very Early Rehabilitation or Intensive Telemetry After Stroke [VERITAS], n=328; Akershus Early Mobilization in Stroke Study [AKEMIS], n=569; and Lausanne trial, n=4210; total n=201) have tested an early mobilization intervention commenced within 24 to 72 hours of stroke, against a usual care (later mobilization±monitoring) control in a stroke unit environment (Table 1). Generally, inclusion criteria have been broad with a wide range in age, stroke severity, and stroke type (only the Lausanne trial exlcuded patients with intracerebral hemorrhage). The interventions tested have varied; with some focused on frequent and ongoing mobility training supported by therapist or nurse (AVERT, VERITAS), and others testing a graduated head-raising protocol in bed then out of bed after 52 hours (Lausanne trial). Importantly, no trial has demonstrated significant effect on complications, mortality, or global disability (modified Rankin Scale). AVERT Phase II investigators did report faster return to unassisted walking11 and reduced costs of care.12 An individual patient meta-analysis, including data from AVERT and VERITAS, suggested some significant improvement in function at 3 months poststroke13; however, sample size remains small (n=103). Increasingly, we are seeing early rehabilitation trials from China. A systematic review and meta-analysis of 37 Chinese trials (n=5916) found significant functional benefits with earlier, highly varied interventions compared with no/little intervention; however, trial quality was often low.14 A more recent early rehabilitation trial by Liu and colleagues15 included 243 patients with intracerebral hemorrhage, randomized to commence rehabilitation (no mobility component specified) within 48 hours of stroke onset, or receive usual care (rehabilitation starting >7 days of bed rest). A higher risk of death (hazard ratio, 4.44; 95% confidence interval, 1.24–12.87) and a 6-point deficit (95% confidence interval, 4.2–8.7) in SF-36 (physical) was reported in patients with prolonged bed rest. Uncertainty remains about how to integrate findings from Chinese rehabilitation trials into reviews, but improved reporting should help.14,16 Two large ongoing trials, AVERT Phase III (n=2104) and Ischemic Stroke and Early Vertical Positioning (SEVEL) (n=400) are currently underway (Table 2). These trials will substantially increase the evidence base in this field.Table 1. Completed Trials of Early and Very Early Mobilization After StrokePublication (Trial Name)Randomized SampleIntervention ProtocolTime (Hours) Between Stroke and MobilizationOutcome*Bernhardt et al 20083 (AVERT)71• Recruited within 24 h of stroke, goal to start mobilization within 24 h of stroke• Emphasis on patient being upright and out of bed (sitting or standing)• At least twice a day for first 14 days or until dischargeIntervention (n=38):Median=18.1,IQR=12.8–21.5Control (n=33):Median=30.8,IQR=23.0–39.9Complications/safetyDeaths: intervention=8/38, SC=3/33, absolute risk difference=12%, ns.Serious adverse events†: intervention =15, control=14, ns.Nonserious adverse events: intervention =61, control=76, P=0.04Falls: intervention=27, SC=28, ns.Functional outcomemRS 0–2: intervention=39.5%, control =30.3%, adjusted‡ OR=4.10, P=0.05Langhorne et al 20108 (VERITAS)32• Recruited within 24 h of admission, with goal to start mobilization within 24 h of stroke• Goal for patient to be sitting, standing or walking (adjusted to patient needs)• Continued at least four times a day, during the inpatient stay, or for one week after recruitmentIntervention (n=16):Mean=27.3Range=26–29Control (n=16):Mean=32.0Range=22.5–47.3Complications/safetyDeaths: EM=0%, control=6%Complications§: EM=8, control=17Complications (days 5–90): EM=8, control=8Complications of immobility (days 0–5): intervention=0, control=3Functional outcomemRS 0–2: intervention=75%, control =44%, adjusted‖ OR=2.3, (P=0.44)Diserens et al 201110 (Lausanne trial)50 (42 included in analysis)• Recruited within 12 h of admission, with protocol started 24 h after stroke• Patient’s head of the bed kept at 0° for first 24 h poststroke, followed by 45° for 24 h, then 90° for 4 h• At 52 h poststroke, patients were moved out of the bed to either sitting or standingIntervention (n=25):Not reportedControl (n=17):Not reportedComplications/safetyDeaths: intervention=0%, control=6%Severe complications including death¶ (during hospitalization): intervention=8%, control=47%Minor complications (during hospitalization): intervention=20%, control=0%, ns.Functional outcomemRS 0–2: intervention=40%, control=30%, ns.Sundseth et al 20129 (AKEMIS)65 (56 included in analysis)• Recruited if admitted to hospital within 24 h of stroke, with mobilization out of bed within 24 h of admission• No predefined mobilization protocol. Mobilization, defined as any out of bed activity, followed the stroke unit’s standard routine for mobilization, adjusted to patients’ needs• Mobilization occurred several times per dayIntervention (n=27):Median=13.1IQR=8.5–25.6Control (n=29):Median=33.3IQR=26.0–39.0Complications/safetyDeaths: intervention=7/27, control =2/29, adjusted# OR=5.26, ns.Patients who experienced ≥1 complication: intervention=67%, control=66%, ns.Functional outcomemRS 0–2: intervention=40%. control=60.7%, adjusted** OR=2.7, ns.IQR indicates interquartile range; mRS, modified Rankin Scale; ns, nonsignificant; and OR, odds ratio.*Outcome data are at 3 months, unless otherwise stated.†Serious adverse events included stroke progression, pneumonia, recurrent stroke, myocardial infarction, atrial fibrillation, and other (does not include death).‡Adjusted for age, baseline National Institutes of Health Stroke Scale (NIHSS), premorbid mRS (modified Rankin Scale).§Complications included chest infection, falls, fatigue, and stroke progression. Complication of immobility was a subset including deep vein thrombosis, urinary tract infection.‖Adjusted for age, baseline NIHSS, cointervention.¶Severe complications included hospital acquired pneumonia, acute coronary syndrome, and pulmonary embolism. Minor complications were those that did not affect the autonomy of the patient (eg, allergic reactions, bed sores).#Adjusted for age, NIHSS on admission, and mortality.**Adjusted for age and admission NIHSS.Table 2. Ongoing Trials of Early and Very Early Mobilization After StrokeEstimated EnrollmentInterventionsRecruitment Time FramePrimary OutcomeAVERT Phase III (Florey Institute of Neuroscience and Mental Health) NCT01846247Active, not recruiting2104Intervention: usual care + very early mobilizationPatient will receive standard stroke unit care with earlier and additional physiotherapy and nursing sessions as per an intervention protocolComparator: usual careRecruited <24 h of onset of stroke symptomsmRS at 3 months after strokeSEVEL Trial (Nates University Hospital) NCT01573299Terminated*400Intervention group 1: early vertical positioningThe patient can sit outside of the bed, the day after stroke onsetIntervention group 2: progressive vertical positioningThe patient is progressively verticalized and is allowed to sit outside of the bed on the third day after the stroke onsetNo time frame specifiedmRS at 3 months after strokeSource: http://www.clinicaltrials.gov (December 4, 2014). mRS indicates modified Rankin Scale.*Trial stopped because of slow enrollment rate (Fanny Herisson, personal communication, 2014).Early Mobilization Recommendations Now Common in GuidelinesDespite the limited evidence for early mobilization, guidelines have changed considerably in recent years. We reviewed 30 acute stroke guidelines worldwide for recommendations related to early mobilization (see Table I in the online-only Data Supplement). Although 22 (73%) contained recommendations to mobilize early, only 8 (36%) defined the intervention, with most recommending a start within 24 hours of stroke onset. The primary justification for mobilizing early was to prevent complications (13/22 guidelines), not promote recovery. The pool of evidence on which the recommendations are based is small. Variation in what is recommended (standing and sitting out of bed, walking, actively engaging in activities of daily living, etc) and when it should begin (after stabilization, within 24 hours, within 72 hours, etc) highlights the ongoing uncertainty surrounding best practice care in the first days after stroke.Potential Mechanisms: BenefitReducing ComplicationsThe main rationale given for early mobilization is to prevent or reduce immobility-related complications, such as infections, deep vein thromboembolism, and falls.17 Further secondary changes associated with stroke-related inactivity include loss of cardiovascular fitness, muscle atrophy, a shift in muscle fiber type to a greater predominance of fast-fatigable, insulin-resistant fibers, and increased intramuscular fat.18 Currently, there is no clear evidence that early mobilization reduces complications, and although physical activity and exercise can mediate secondary fitness and muscle changes in patients in the later phase of stroke,18 studies in the acute phase are lacking. We do know that patients managed in organized stroke units experience fewer immobility-related complications,19 and this has been attributed to earlier and better management, including early mobilization and early rehabilitation.1 Our group and others have used structured observation of patient activity over the active day (0800–1700 hours) and shown that patients managed in stroke units that practice early rehabilitation have a short time to first mobilization (median 18 hours poststroke; interquartile range, 7.3–43.0), spend only 30% of the day time in bed, 46% out of bed, and 20% standing and walking,20 which is different to patients managed in acute stroke units with a median inactive time of 65.5% (interquartile range, 6.3–87.8).21 Whether earlier activity can independently influence the incidence or severity of immobility-related or other complications is uncertain.Promoting Brain RecoveryPreclinical studies show a critical sensitive period of enhanced neuroplasticity early after stroke.22,23 Areas ipsilesional and contralesional to the ischemic infarct are hyperexcitable; genes responsible for axonal sprouting, dendritic spine formation, and synaptogenesis are upregulated in peri-infarct areas; long-term potentiation is enhanced; and maladaptive plasticity and excitotoxic processes are subdued.22,23 Rehabilitation efficacy declines over time,24 suggesting that earlier initiation of training may enhance recovery. Poststroke exercise reduces infarct volume in stroke-affected animals, particularly when exercise begins soon after stroke onset.25 These findings provide a biological basis for early training in humans, but when to start and how much training remains unclear.Potential Mechanisms: HarmSome clinicians have concerns that upright activity early might inhibit reperfusion of salvageable penumbral tissue. A systematic review of studies (n=57) in which transcranial Doppler was used to measure cerebral blood flow velocity during head position changes reported that elevating the head-of-the-bed to 30° in patients with large middle cerebral artery occlusion reduces cerebral blood flow velocity,26 particularly in those with ongoing occlusion. The associations between these changes and clinical outcomes were not examined. Horizontal positioning may support newly established leptomeningeal or transcortical collateral channels, which have been associated with preservation of the ischemic penumbra and smaller infarct size.27 A recent study found no significant differences in cerebral blood flow velocity between patients assessed at 6 to 16 hours postischemic middle cerebral artery stroke and controls in any tested position (70°, 45°, 0°, −15°), and there was no relationship between cerebral blood flow velocity and favorable neurological outcome.28 Nevertheless, these concerns underpin the Lausanne and SEVEL trials, which test gradual verticalization by day 3 (see Tables 1 and 2), as well as a new large cluster randomized controlled Head Position in Stroke Trial (HeadPoST, http://www.clinicaltrials.gov: NCT02162017) comparing a head flat or head elevated 30° protocol for the first 24 hours after diagnosis.Optimal blood pressure management in acute stroke (ischemic and hemorrhagic) remains uncertain.29 It is possible that the effect of mobilization on blood pressure concerns some clinicians. Physical activity does increase mean blood pressure, but this is transient, returning to baseline once activity has ceased.30 Blood pressure drop (postural hypotension) may also be a concern; however, even in patients with moderate to severe stroke, large drops (systolic blood pressure >30 mm Hg) seem to occur in <10% of patients.30 Clinicians seem more concerned about starting out of bed activity within 24 hours of onset in patients with hemorrhagic (59% concerned) than ischemic (23% concerned) stroke,31 possibly because of fear of further bleeding. Increased risk of symptomatic ICH also underlies concerns about early mobilization of patients treated with recombinant tissue-type plasminogen activator (r-tPA).32 Factors such as infection of uncertain cause, severe chest infection, severe stroke (National Institutes of Health Stroke Scale >20), drowsiness, and confusion also seem to influence clinicians’ decisions about the timing of mobilization after r-tPA.32 Evidence to guide early mobilization after r-tPA is scarce. A small trial of 29 patients mobilized 12 to 24 hours after r-tPA found 75% of patients had no adverse response to mobilization, with the remainder experiencing nonserious events.33 AVERT Phase III includes over 500 patients treated with r-tPA. These data should help inform clinical guidelines.Safety Criteria NeededEstablishing safety criteria for this intervention is an important next step. Clearly, not all patients admitted with stroke should start out-of-bed activity or training within hours, or even days, of stroke onset. Currently, there are no clear safety guidelines to guide initiation or progress of treatment. Within the context of a trial, trialists seem willing to include adults (no upper age limit) with ischemic or hemorrhagic stroke, who are rouseable, with no early deterioration, no signs of secondary intracerebral hemorrhage, acute coronary syndromes, or severe heart failure. Patients treated with r-tPA were excluded from earlier trials,3,8–10 but are included in AVERT Phase III. Additional broad physiological safety criteria include systolic blood pressure between 120 and 220 mm Hg and heart rate between 40 and 100 beats per minute.3 Physiological monitoring before each out-of-bed episode over the first 1 to 3 days, particularly targeting systolic blood pressure and consciousness, was practiced in several trials using a range of regimes. At present, cerebral imaging is not used to select patients or guide treatment.Early Mobilization in PracticeIncreasingly, we see early mobilization incorporated into stroke care audits as a quality indicator (see Table II in the online-only Data Supplement), although mobilization is largely undefined, and early is either not quantified or ranges from within 24 hours of stroke onset34 to day 2 of hospital admission.35,36 Only first mobilization is recorded. Reports on compliance vary and range from 49% in Spain37 to 85% in Australia.35 Does compliance with early mobilization have an effect on patient outcomes? Stroke audit data from Catalonia, Spain, showed that early mobilization was significantly associated with reduced mortality at 30 days (OR, 2.05; 95% confidence interval, 1.31–3.19) and 12 months (OR, 1.54; 95% confidence interval, 1.05–2.24).38 Others have found early mobilization to be independently associated with a lower risk of medical complications39 and shorter hospital stay.36 These indictors hinge on the timing of first mobilization. It is, therefore, likely that rather than indicate the effect of a specific intervention, the indicator serves as a proxy for how efficiently early team care is coordinated, which may include access to rehabilitation-focused staff.ConclusionsEarly mobilization has become a regular topic of workshops and scientific sessions at stroke conferences around the world. It seems like a simple activity that could improve the quality of acute care. This review highlights the early acceptance and, in some cases, adoption of starting some form of mobilization early, despite concerns. What is currently missing from the discussion is a shared understanding of what an early mobilization intervention is, evidence to help us decide which patients can safely start mobilizing early, and the mechanism by which early mobilization might aid recovery or cause harm as a treatment in the first hours or days of stroke. It was this uncertainty, on the back of promising Phase II results, that prompted us to conduct the soon to be completed international AVERT Phase III trial. On the sliding scale of mobilization interventions, our intervention protocol is firmly at the rehabilitation (training) end of the spectrum, and we will be able to examine who received treatment, how often and when over the course of their acute hospital stay. This large trial will help inform clinical guidelines about who might benefit or be harmed by starting mobility training within 24 hours of stroke onset and will provide a detailed evaluation of care costs. As with all research, it is also likely to generate new research questions. Increasingly, we are trying to understand how to enhance and accelerate recovery after stroke using multimodal approaches, including rehabilitation therapies. Improving rehabilitation interventions through better understanding of the neurobiology of recovery and when (and what) training may offer the most benefit, as well as determining who we should target, is an important part of this endeavor. Our new Center of Research Excellence in Rehabilitation and Recovery, funded by the National Health and Medical Research Council, is focused on these important goals. The model aligns with recommendations from the 2010 prioritizing a world agenda synergium,40 and we look forward to linking with other international rehabilitation and recovery research initiatives from around the globe.AcknowledgmentsWe thank Audrey Raffelt for sourcing guidelines, extracting recommendations, and preparing tables. We also thank the following people for their assistance in translating Good Clinical Practice Guidelines: Wen Wen Zhang, Atte Meretoja, Tina Kaffenberger, Thomas Linden, Torunn Askim, Marie Dagonnier, and Sharon Kramer. The Victorian State Government provides infrastructure support to The Florey for which we are grateful.Sources of FundingDr Bernhardt is funded by a National Health and Medical Research Council (NHMRC) Senior Research Fellowship. Dr English is funded by an NHMRC Early Career Fellowship.DisclosuresDr Bernhardt is Principal Investigator for AVERT Phase III. The other authors report no conflicts.FootnotesThe online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.114.007434/-/DC1.Correspondence to Julie Bernhardt, PhD, The Florey Institute of Neuroscience and Mental Health, 245 Burgundy St, Heidelberg, Victoria, Australia 3084. E-mail [email protected]