Skip to content
Back to search results

RMI: Rivermead Mobility Index

Introduced by
Frank Collen, Derick Wade, and Rivermead Rehabilitation Centre
Year
1991
Type
Clinician-observed performance test
Family
Stroke and neurological mobility assessment
Font size

What does it do?

The Rivermead Mobility Index (RMI) is a brief, clinician-observed performance test of basic mobility abilities developed for assessing stroke and neurological rehabilitation outcomes.

Published in 1991 by Frank Collen and colleagues at Rivermead Rehabilitation Centre (Oxford, UK), the 15-item index measures bed mobility, sitting/standing balance, transfers, and ambulation. The RMI is widely used in stroke units and rehabilitation settings to track functional recovery and predict discharge outcomes.

The intuition behind the method

Acute stroke and acquired brain injury produce motor deficits limiting independence in fundamental mobility tasks. Early ability to transfer from bed to chair, stand independently, or walk predicts faster rehabilitation progress and earlier discharge to home. The RMI captures hierarchical progression of mobility recovery through simple yes/no capability observation: patients who cannot sit independently require total assistance; those managing transfers may return home with community support; those achieving stair negotiation approach full independence. RMI scores correlate with EDSS, NIHSS severity, and functional outcome measures; the simple binary scoring (capable/unable) eliminates rater judgment inherent in graded scales.

History

The original Rivermead Motor Assessment (RMA) was developed in 1980s by Collen and Wade for detailed assessment of motor recovery after stroke, but was time-intensive (30+ minutes). In 1991, the abbreviated Rivermead Mobility Index was published (Collen et al.) as brief clinician-observed performance test suitable for routine rehabilitation ward use. RMI gained wide adoption in UK stroke units in 1990s and remained standard in many European and Australasian rehabilitation centers. However, adoption in North America was limited due to preference for NIHSS (emergency) and Functional Independence Measure (rehabilitation discharge). Modern practice increasingly integrates RMI with timed gait measures (10-meter walk, TUG) for comprehensive mobility assessment.

When to use it

Recommended for all acute stroke patients in inpatient rehabilitation units within first week post-stroke (baseline assessment) and at discharge or 4-week intervals during hospitalization. Useful in brain injury, Parkinson's disease with gait dysfunction, and spinal cord injury rehabilitation. Serial RMI tracking (weekly) predicts length of stay and discharge destination better than single admission score. Not appropriate for bedbound patients with severe consciousness impairment (NIHSS >30, GCS <8) unable to cooperate with testing. Preferred over EDSS in acute stroke settings (EDSS emphasizes chronic disability; RMI emphasizes functional tasks relevant to discharge planning). Less sensitive than timed gait measures (10-Meter Walk Test, Timed Up and Go) in patients with mild impairment.

Limitations and considerations

  • Binary yes/no format lacks granularity; distinguishes capable from unable but does not quantify effort, assistance level, or safety concern. Patient performing transfers with minimal use of therapist's hand scores 1 (same as independent); patient unable to maintain standing balance alone scores 0 (same as bedbound).
  • Does not assess speed, endurance, or quality of movement; patient may achieve all items but with excessive energy expenditure, compensation patterns, or pain not reflected in RMI.
  • No measurement of assistive device requirement distinction; walking with bilateral parallel bars scores same (yes) as walking with cane scores same as walking unsupported.
  • Ceiling effects in younger/mildly impaired stroke patients; RMI 15 achieved by majority, limiting discrimination within high-functioning group. Timed measures (10-meter walk, TUG) more sensitive in mild impairment.
  • Does not capture upper extremity function, hand dexterity, or cognitive/communication deficits; stroke affecting dominant hand or language may show intact RMI despite significant disability.
  • Requires adequate safety environment (rails, space for ambulation) not always available in acute care; testing deferred, complicating longitudinal tracking.

Source conditions do not replace independent scientific review or validation by an analysis engine.

Strengths

  • Brief administration (10-15 minutes) fits busy rehabilitation unit schedules; does not require specialist equipment or training beyond observational nursing assessment.
  • Binary yes/no scoring eliminates rater judgment variability; inter-rater reliability excellent (ICC 0.95+) even among untrained staff if protocol followed.
  • Hierarchical item structure aligns with natural recovery progression after stroke; early items (rolling, sitting) precede later items (stairs, outdoor walking), enabling prediction of ultimate functional trajectory.
  • Correlates strongly with length of rehabilitation stay, discharge destination (home vs facility), and 6-month functional outcome (Barthel Index, Modified Rankin Scale).
  • Applicable across diverse rehabilitation settings (acute stroke units, inpatient rehabilitation, community); no equipment cost.
  • Simple total score (0-15) intuitive for communication with patients, families, and interdisciplinary team.

Applications

  • Early stroke rehabilitation admission assessment: baseline RMI within 48-72 hours post-stroke predicts 90-day outcome and discharge destination (Spearman r=0.7-0.8 with modified Rankin Scale); informs length-of-stay estimation and therapy intensity.
  • Progress monitoring in inpatient rehabilitation: weekly RMI tracks functional recovery trajectory; expected progression 1-3 points per week post-stroke; plateau suggests plateau in recovery, identifies patients needing therapy modification or management of secondary complications.
  • Discharge planning: RMI ≥11 predicts home discharge feasibility; RMI <8 typically requires facility/assisted living. Families use RMI score to understand discharge readiness and home modification needs.
  • Rehabilitation length-of-stay prediction: RMI on admission combined with age and stroke severity (NIHSS) predicts length of rehabilitation stay; models enable resource allocation and early discharge planning.
  • Comparative rehabilitation effectiveness research: RMI serial measurement in observational cohorts and pragmatic trials comparing rehabilitation models (early discharge with community support vs extended inpatient stay); sensitive to functional trajectory differences.
  • Caregiver burden assessment: RMI coupled with caregiver interview identifies dependence level; RMI 4-7 requires substantial family time commitment vs RMI 11-15 requires minimal supervision, informing caregiver support needs and respite planning.

How it works

How it works

1

In safe clinical environment, clinician observes patient perform the following 15 tasks in sequence. Bed mobility: (1) Supine-to-prone roll (both sides); (2) Supine-to-sitting transfer (either side). Sitting: (3) Sit at side of bed unsupported for ≥10 seconds; (4) Sit at side of bed unsupported and perform reach forward; (5) Sit to standing from firm chair without armrests. Standing: (6) Stand unsupported for ≥10 seconds; (7) Stand unsupported and perform reach forward; (8) Stand with eyes closed for ≥3 seconds; (9) Stand and turn 360° (continuous turns); (10) Stand and pick up object from floor. Transfers: (11) Transfer from bed to chair; (12) Transfer from chair to chair (different height). Ambulation: (13) Walk 10 meters indoors with/without aid; (14) Walk up and down stairs (at least 3 steps); (15) Walk outdoors on rough ground or in garden. Clinician must ensure safe environment (grab bars, supervision) but does not provide physical assistance beyond what patient normally uses. Score yes if patient completes task safely; score no if unable or unsafe to attempt.

Formula
Observational assessment of patient mobility abilities conducted in clinical setting (hospital, rehabilitation center, or home). Clinician observes patient performing 15 mobility-related tasks: bed mobility (rolling supine to prone, lying to sitting), sitting balance, standing balance, transfers (sit-to-stand, chair-to-chair, bed-to-chair), walking (indoors with aid, stairs). Each task scored as yes/no (capable) based on observation. Administration 10-15 minutes including setup and observation.
2

Binary scoring (1 = yes capable, 0 = no unable) for each of 15 items. No intermediate grades (partial credit). Total score = sum of all items, range 0-15. Higher score indicates greater independence. Individual items can be reported separately to track specific functional domains (bed mobility 1-2, sitting balance 3-4, standing balance 6-9, transfers 11-12, ambulation 13-15) but RMI typically reported as single total score. Missing items (patient declines test due to fatigue or safety concern) leave score incomplete; scoring algorithm does not impute missing data.

Formula
Each of 15 items scored 1 (yes, capable) or 0 (no, unable). Total RMI score = sum of items, range 0-15. Higher score indicates better mobility function. No weighting or subscale scoring; all items contribute equally to total.
3

RMI stratification based on published receiver-operator characteristic analysis: RMI 0-3 indicates complete bed dependence, unable to sit or stand without assistance; these patients require 24-hour care and typically discharge to facility. RMI 4-7 suggests bed/chair dependence with minimal transfer ability; hospital care extended, discharge usually facility unless substantial family support. RMI 8-10 indicates transfer capability with moderate assistance and some standing balance; rehabilitation progress evident, home discharge feasible with community support. RMI 11-14 suggests near-independent mobility with minor difficulties (perhaps stairs or outdoor ambulation problematic); home discharge expected with outpatient rehabilitation. RMI 15 indicates full independent mobility without observable deficit; patient ready for discharge to community with routine follow-up.

Formula
RMI score stratification: 0-3 = confined to bed, unable to sit independently; 4-7 = bed/chair dependent, minimal transfer ability; 8-10 = transfers possible with minimal/moderate assistance; 11-14 = mostly independent in mobility with minor difficulties; 15 = independent mobility without limitations. Baseline RMI and change over rehabilitation predict discharge destination (home vs facility) and length of stay in acute stroke/brain injury rehabilitation.

Common pitfalls

  • Scoring task as 'yes' when patient requires stand-by assistance or verbal cuing; RMI definition 'capable' assumes safety without therapist intervention. Use detailed protocol to distinguish independent from assisted performance.
  • Administering RMI in unsafe environment (patient at fall risk, no grab bars) leading to refusal to attempt items scored as 'no'; test conditions must be consistent. Do not penalize patient refusal due to environment.
  • Missing items (patient declines stairs due to time constraint) left incomplete; do not impute or assume failure. Report missing items explicitly and note incomplete assessment.
  • Comparing RMI serial scores without controlling for environmental factors (different testing location, with/without brace, different assistive devices) affecting performance variability.
  • Over-interpreting stable RMI score as lack of progress in first 2 weeks post-stroke when motor recovery is rapid; daily RMI changes >0.5 points common; use weekly trend rather than single assessment.
  • Using RMI as outcome measure in highly selected trial populations without accounting for ceiling effects; mild stroke cohort will show minimal RMI change despite meaningful clinical improvement.

Frequently asked questions

Can RMI be scored if the patient declines to attempt some items?

Yes, clinician should document which items were attempted and which were declined. Declining due to fatigue or pain is common late in assessment (patient may fatigue after items 1-10). Score obtained items; do not assume refusal means unable (patient might achieve item if rested). Report total RMI from completed items with notation of missing items. If <10 items completed, reliability is reduced; document circumstances and retest when patient more alert.

How does RMI differ from Barthel Index?

RMI assesses basic mobility (sit, stand, transfer, walk) through direct observation; Barthel Index assesses activities of daily living (toileting, dressing, eating) through patient report or observation. RMI focuses on physical capacity; Barthel Index includes cognition and self-care. RMI applicable acutely; Barthel often used at discharge. They correlate (r=0.7-0.8) but measure different constructs. RMI preferred for predicting early rehabilitation progress; Barthel preferred for functional outcome at discharge or home assessment.

Is RMI valid for progressive neurological diseases like Parkinson's or MS?

RMI is less commonly used in progressive diseases but can be applied. In Parkinson's disease, RMI captures mobility decline related to rigidity/bradykinesia/postural instability. In MS, RMI reflects functional impairment from various disease phenotypes (motor, ataxic, cognitive). However, EDSS (MS) and MDS-UPDRS (Parkinson's) are more sensitive and widely used in these populations. RMI useful in progressive disease rehabilitation units to track functional change; less sensitive than disease-specific scales.

What is the difference between RMI and Timed Up and Go (TUG)?

RMI is binary observational checklist (yes/no capable of 15 tasks); TUG is single timed test (rise from chair, walk 3 meters, turn, walk back, sit). Both assess mobility but RMI broader (bed mobility, sitting balance, standing, transfers, stairs, outdoor); TUG narrower (functional mobility in real-time). RMI requires 10-15 minutes; TUG requires 2-3 minutes. RMI useful in rehabilitation with complex mobility deficits; TUG useful for fall risk screening in community-dwelling older adults. They are complementary; many settings use both.

What RMI score predicts independent home discharge?

Literature suggests RMI ≥11 associated with independent or supervised home discharge; RMI 8-10 requires assistance for ambulation and transfers; RMI <8 typically necessitates facility placement or 24-hour home care. However, individual variation is substantial: some RMI 9 patients achieve home discharge with community support and family help; some RMI 13 patients unable to access home due to architectural barriers. Use RMI as prognostic guide, not determinant; integrate with caregiver availability, home environment, and patient preference.

Related methods

Sources

  1. Collen, F. M., Wade, D. T., Robb, G. F., Bradshaw, C. M. (1991). The Rivermead Mobility Index: A further development of the Rivermead Motor Assessment. International Disability Studies, 13(2), 50-54.

    Open DOI reference
    Copied.