Modified Rankin Handicap Scale

Acute Stroke: (Table from Dromerick et al, 2003; n = 95, length of rehabilitation was 19.5 (8.3) days)

Acute Stroke: (Wolfe et al, 1991)

• Excellent test-retest reliability (Kappa w = 0.95)

Post-Stroke- 6 months: (Wilson et al, 2005)

• Excellent test-retest reliability
  (Rater 1: Kappa = 0.81; 0.94 and 
  Rater 2: Kappa = 0.95; 0.99)

Acute Stroke: 

(Wolfe et al, 1991; van Swieten et al, 1988; Wilson et al, 2002, 2005; Shinohara et al, 2006; Quin et al, 2008)

• Excellent intra-rater reliability (Kappa w = 0.95)
• Excellent inter-rater reliability (Kappa range 0.75 - 0.96)

(van Swieten et al, 1988):

• Excellent inter-rater reliability (Kappa = 0.82, out-patient)
• Adequate inter-rater reliability (Kappa = 0.51, in-patient)

(Wilson et al, 2002, 2005):

• Excellent inter-rater reliability (Kappa = 0.78)

(Shinohara et al, 2006)

• Excellent inter-rater reliability (ICC = 0.95 neurologists; 0.96 nurses)

(Quinn et al, 2008)

• Adequate inter-rater reliability (Kappa = 0.57, overall)

(Quinn et al, 2010)

• The longer the interview, the more variability in rater consistency

• With focused assessment, patients struggle to answer categorically leaving room for rater interpretation

• Further study of reliability is needed on the modified, simplified modified Rankin score and the standard mRS

(Bruno et al 2010; 50 patients post stroke having treatment in an outpatient setting were rated with the simplified mRS (smRS) (yes/no questions) by 9 mRS web-certified raters with repeat testing within 20 minutes of the first rating )

• Excellent test re-test reliability of the simplified mRS questionnaire with a non-structured interview (smRS) had a κ statistic of 0.72 (CI₉₅ 0.58-0.85) with 78% agreement between raters (compared to previous research on the mRS with a structured interview with adequate reliability r=0.62 with 73% agreement)( Quinn et al, 2009;Cincura et a, 2009, Shinohara et al, 2006)

• Administration time was1.67 minutes( less time than the structured interview format for the mRS (@15 minutes)

(Ghandehari, K., et al. 2012) Five raters administered the Asian Stroke Disability Scale (ASDS) and compared it to the interrater reliability of the modified Rankin Scale (mRS) for 25 subjects post stroke (65.5 years of age, 56% males).

• The Interrater variability for the mRS did not differ significantly(X² = 1.758, p = 0.780). The paired interrater variability of the mRS did also not vary significantly (X² = 0.553, p = 1.000)

(Fearon, et al 2012;Bruno et al 2013)

• In a study of 74 stroke survivors (median age = 72 years [IQR = 62-79]; median time since stroke = 5 days [IQR = 3-0]), prestroke measures of the mRS and the standard mRS with interview had only adequate inter rater reliability (κ_w= 0.55; with 55% agreement and κ_w=0.70 with 70% agreement respectively) (Fearon et al, 2012)

• The mRS cannot be used as a prestroke measure. It must only be used as a post stroke measure because some of the items cannot be scored as 0, 1 or 2 because the patient needs to compare post stroke function with pre stroke function; thus before the first stroke, there are no stroke symptoms. (Bruno et al letter, 2013)

(Bruno et al 2011) With 50 consecutive outpatients 4.83 (+3.00) months post stroke, 9 paired raters administered the msRSq by a standard interview twice (one 20 minutes after the first) and then repeated test administration over the telephone 1-3 days later .

• The average estimated time to administer the smRSq in person was 1.29 minutes (range, 0.50 to 2.25 minutes).

   

• The In-person raters agreed 78% of the time (k= 0.71; CI, 0.57 to 0.86 with weighted Kw_0.86; CI, 0.79 to 0.94).

• The first in-person and telephone raters agreed 82% (k= 0.76; CI, 0.63 to 0.90 and Kw= 0.87; CI, 0.79 to 0.95).

• The second in-person and telephone raters agreed 82% (k=0.77; CI, 0.63 to 0.90 and Kw= 0.89; CI, 0.82 to 0.96).

(Zhao, et al, 2010) 56 participants, mean age = 71 years old (± 13.6), less than 4 days post-stroke were administered the mRS by three trained rater groups (experienced (Exp), inexperienced (Inexp), and inexperienced with a decision tool (Inexp+DT).

• Adequate reliability was found across all groups (ICC = 0.675, 95% CI = 0.559-0.791)

• Adequate reliability was found between each of the groups

  • Exp vs. Inexp (K_w = 0.686, 95% CI = 0.541-0.819, p = 0.032)

  • Exp vs. Inexp+DT (K_w = 0.568, 95% CI = 0.359-0.724, p = 0.553)

  • Inexp vs. Inexp+DT (K_w = 0.736, 95% CI = 0.557-0.859, p = 0.121)

Sub-Acute Stroke:

(Saver, 2010) n = 50 patients, average 71.5 years old (range 43-93). Participants received the Rankin Focused Assessment to derive the mRS (administered by two coordinators) at their 90 day assessment as part of the Phase III National Institutes of Health Field Administration of Stroke Therapy-Magnesium Trial.

• Among all participants the percent agreement was 94% (Kw = 0.99, 95% CI = 0.99-1.00; K = 0.93, 95% CI = 0.85-1.00). Among the 43 participants living at the 90 day assessments the percent agreement was 93% (Kw = 0.99, 95% CI = 0.98-1.0, Kw = 0.91, 95% CI 0.82-1.00)

Telephone Administration: (Janssen, 2010; Dennis, 2011)

Acute

(Dennis, 2012) compared two postal versions of the mRS: one with a tick box for the five descriptions, the other was the simplified modified Rankin questionnaire (smRSq) that has yes/no responses and derives the mRS score. Both were followed-up with a telephone interview using the Rankin Focused Assessment

• Adequate reliability (Kappa = 0.44, Kappa w = 0.68) between the two postal versions of the mRS

• Adequate reliability (Kappa = 0.47, Kappa w = 0.66) between the tick box version of the mRS and the telephone interview

• Adequate reliability (Kappa = 0.55, Kappa w = 0.73) between the smRSq and the telephone interview

(Savio et al 2013) 131 patients post stroke were interviewed by 2 certified nurses. Half of the patients were randomized to be interviewed by telephone followed by face to face assessment and half the reverse order.

• Telephone assessment of stroke disability with the modified Rankin Scale is similar in reliability in comparison to direct face to-face assessment.

• The median value of the modified Rankin Scale score was 4 (IQR 3-5) by telephone as well as by face-to-face assessment (p = 0.8).

• Excellent reliability between the two methods was K_w = 0.82 (CI ₉₅:0.77-0.88).

• Sensitivity of the telephone assessment was lower for scores 2 and 3 (17% and 46%, respectively) than for scores 4-5 (range 67-90%).

Subacute

(Janssen, 2010) compared mRS scores for 83 participants a median of 5 months post-stroke comparing administration face to face to over the telephone.

• Adequate reliability (Kappa = 0.41, 95% CI 0.26-0.55; Kappa weighted = 0.71, 95% CI 0.59-0.82).

Acute Stroke: (Cup et al, 2003; n = 26; mean age 68 (15) years; Kwon et al, 2004; n = 459; mean age = 70 (11.4) years; Weimar et al, 2004; n = 4246; mean age = 67.1 (69) years)

Study1:

• Excellent concurrent validity with:
       Barthel Index (r = - 0.81)
       Frenchay Activities Index (r = - 0.80)
       EQ-56 (r = 0.68)

• Adequate concurrent validity with:
       SA-SIP-30 (r = 0.47)

Study 2:

Excellent concurrent validity with:
     Barthel Index (r = - 0.89)
     FIM-motor dimension (r = - 0.89)

Study 3:

• Excellent concurrent validity with:    

SF-36 Physical Function (r = 0.84)
Barthel Index (r = 0.82)

(Bruno et al, 2013) Forty subjects, documented in severity with the National Institutes of Health Stroke Scale (NIHSS) post treatment with intravenous tissue plasminogen activator in the ER Department were followed up to assess functional outcomes using the simplified Modified Rankin Scale (smRSq) . The participants averaged 69 years of age (±14), represented 45% males, had an average initial NIHSS score of 10 and were, on average, 3-7 months post stroke.

• Median follow up score on the smRSq was 3 (IQR = 1-6).

• The correlation of the NIHSS and the smRSq had an excellent and significant correlation (r=0.69; p<0001).

• The smRSq is simple and brief and can be administered over the phone and still preserve validity

(Bruno et al 2011)

• In a sample of patients receiving rehabilitation in an outpatient setting approximately 5 years post stroke, the smRSq correlated adequately with the physical component of the SF Short Form 12 (r= 0.50, P= 0.005) and the mental component (r=0.36, P= 0.048) .

(Gorelick et al, 2012)

This review of randomized clinical trials of stroke intravenous thrombolytic studies (rt-PA) was carried out to determine the relationships between baseline stroke severity associated with efficacy (based on functional outcomes of the NIH Stroke Scale [NIHSS] and the modified Rankin Scale [mRS] and safety.

Summary of Findings Acute Post Stroke (5 major RCT’s)

(Bruno et al, 2013)

Thirty two of 60 subjects post stroke enrolled in a reliability study met the criteria to participate in the current study (mean age 59.6 years (± 15), 53% men) to correlate stroke size (acute ischemic stroke) with smRSq scores .The patients had either computed tomography or magnetic resonance imaging to confirm an ischemic stroke and were scored on the smRSq at least 3 months post stroke. Stroke volume was computed and strokes were classified into 2 size categories: lacunar type measuring ≤ 6.28 cm³, which corresponds to a cylinder with a maximum diameter and height of 2.00 cm, or strokes >6.28 cm³. The Spearman correlation analysis compared the smRSq between the lacunar type and the larger strokes.

• Acute stroke size correlated well with the smRSq supporting the validity of the smRSq for assessing functional outcomes after stroke

• Lacunar stroke volumes (n = 17) ranged from 0.03 to 4.58 cm³, and the larger stroke volumes (n =15) ranged from 11.52 to 250.02 cm³.

• Lacunar strokes were associated with lower smRSq scores (median 1) compared to the larger strokes (median 4) ‘

• There was an adequate correlation of stroke size with the smRSq score (r=0.68; P<0.001)

(Tilson et al, 2010; n =283; mean age 63. 5 (± 12.5) years; assessed patients approximately 20 and 60 days post-stroke); data is from the Locomotor Experience Applied Post Stroke study, and the mRS was used as the anchor for determining the minimally clinically important difference. A poor, but substantial relationship was found between the mRS and comfortable gait speed. The MCID was estimated as an improvement in CGS of 0.16 m/s anchored to the mRS.

• The area under the curve was calculated as 0.69 (95% CI = 0.63-0.75).

(Liotta et al, 2013, to determine readmission of patients with ICF as

measured by the Rankin Scale ( mRS) at 14 days, 28 days and 3

months post stroke)

Retrospective evaluation of 246 patients readmitted 30 days following treatment of acute intracranial hemorrhage on the ICU. Of 246 patients with a mean age of 65 years:

• 78% survived discharge. Of these, 22 (11%) were re-admitted 4-15 days post ICU discharge.

• The most common reason for readmission was infection (n = 10) and vascular event (n = 6)

• Readmitted patients had similar mRS scores to those not readmitted at 14 days (median 4 [IQR = 4-5], median 5 [IQR = 4-5], respectively), but those who were readmitted had higher mRS scores at three months post stroke (median 5 [IQR = 3-6] for those readmitted and 3 [IQR = 1-4] for those not readmitted.

(Kerr et al, 2012)

Within the Virtual International Stroke Trials Archive, the rt-PA-treated patients were compared to the non rt-PA treated patients (10,000 samples).The end points of the mRS at 30 and 90 days and NIHSS at 7 and 90 days post stroke were studied.

• The smallest sample sizes to generate statistical power greater than 80 % were 620 and 480 respectively for the mRS and 370 and 420 respectively for the NIHSS

• The 7 day NIHSS was the most sensitive outcome measurement to detect differences in the rt-PA treated patients.

(Ali, 2013; n = 3787 with complete data available, n = 3872 Stroke Impact Scale data available at 3 months; mean age for those self-reporting 68 (58-76) years and those requiring proxy 75 (66-80). Authors examined outcomes data from the Virtual International Stroke Trials Archive.

• Excellent concurrent validity of the mRS at three months post-stroke with the European Quality of Life Scale-5D (EQ-5D) Weighted Scores (r=-0.75, p<0.0001), the EQ-5D Visual Analog Scale (r=-0.64, p<0.0001), Stroke Impact Scale Recovery (r=-0.73, p<0.0001), and Stroke Impact Scale-16 (r=-.083, p<0.0001).

• The mRS was used to examine the proportion of patients classified as having a good primary outcome (mRS ≤ 1), but a poorer quality of life and was found to be superior.

Subacute Stroke

(Goldie, 2014; n= 12, 370, mean age of 69.5 years old (range = 18-130), from a systematic search and review across 11 studies. Coefficients reported from assessments completed at 90 days post-stroke.

• The mRS explained 80.8% (p = 0.000) of the variance in the National Institutes of Health Stroke Scale (NIHSS), 87.6% (p = 0.000) of the variance in the Barthel Index (BI), 83.0% (p = 0.000) of the variance in the Stroke Impact Scale-16 (SIS-16), and 76.9% (p = 0.000) of the variance in the Scandanavian Stroke Scale(SSS) at 90 days post-stroke.

• Once the Baseline age and NIHSS were adjusted for, the mRS explained 56.6% (p < 0.001) of the variance in the NIHSS, 75.2% (p < 0.001) of the variance in the BI, and 80.5% (p< 0.001) of the variance in the SIS-16 at 90 days post-stroke.

Acute Stroke: (Tilley et al. 1996, n = 333)

Berzina, 2015; n= 266, Median age = 65 years old (IQR = 57.5-77), Median time post-stroke = 8 months (IQR = 1-12), participants were interviewed with the Comprehensive International Classification of Functioning, Disabilities, and Health (ICF) Core Stroke Set, and then this information was used to score the mRS and mRS-Systematic Interview (mRS-SI).

• Poor to Adequate significant correlations were found between the mRS and the sum of problems in ‘Body Function’ (ρ=0.59, p < 0.001), ‘Body Structures’ (ρ=0.23, p < 0.001), and ‘Activities and Participation’ (ρ=0.69, p < 0.001) of the ICF.

• Authors found that the mRS is mostly associated with problems under “Activities and Participation,’ specifically “Self-Care’ and ‘Walking.’

   

Owolabi, 2010; n= 100, mean age = 58.9 (± 10.7) at least one month post-stroke were administered the mRS, the health-related quality of life for stroke patients (HRQOLISP) and the SF-36.

• The physical domains of the HRQOLISP were found to differ among the mRS strata (p <0.006), however the domains of the spiritual dimension did not (0.217 ≤ p ≤ 0.906).

• All SF-36 subscales differed among mRS strata (p ≤ 0.023)

Golicki, 2015; n = 408, mean age = 69.0 (± 12.9), median time post-stroke 8 days; participants completed the mRS, EuroQol-5 dimensions-5 levels of severity (EQ-5D-5L), the EuroQol-5 dimensions- 3 levels of severity (EQ-5D-3L) as well as additional measures.

• Adequate to Excellent convergent validity was found between the mRS and the dimensions of the EQ-5D-5L (ρ=0.42-0.79) and the EQ-5D-3L (ρ=0.36-0.78)

Acute Stroke: (Dromerick et al, 2003)

• Adequate Floor Effect in 18% of stroke sample at admission to rehabilitation (see normative data for more information).

Acute Stroke: (Dromerick et al, 2003)

• Poor at detecting change compared to the FIM (C = 0.59)*

*see normative data for more information

(Batcho, C.S., et al. 2014)

68 subjects participated in a 2 month longitudinal study to evaluate responsiveness of patients receiving early versus late intervention post stroke.

• There were no significant changes in the mRS scores (mean score of 2 (with range of 2-3.5 pre and post; p=0.99) in patients who had delayed intervention post stroke ( mean 10.3 months [ + 7.7 months])

• There was a significant change in the mRS score (P<0.001) for the patients who received early intervention post stroke (average at 2.01 months [+2.6 month])

• There was a significant, adequate positive correlation (p=0.64; p<0.001) between the pre-post test change scores on the mRS and the ACTIVLIM-Stroke Score ( a 20 item self reported Rasch built questionnaire that measures the ability of stroke patients to perform daily living activities).

Close Head Injury (Schaefer et al, 2004)

• Excellent concurrent validity with: 

Signal-intensity abnormal volume on diffusion weighted images (r = 0.77) 

Number of lesions on images  (r = 0.66)

• Adequate concurrent validity with:

Lesion location in the corpus callosum (r = 0.51)