A Preliminary Comparison of the Katrak Hand Movement Scale With the Upper Extremity Fugl Meyer Assessment Shows a Strong Correlation After Stroke

Abstract

Objective

To investigate the concurrent validity of the KHMS with the FMA-UE.

Design

The FMA-UE and the KHMS were administered to 50 adults with stroke to evaluate their concurrent validity.

Setting

Three tertiary rehabilitation hospitals.

Participants

Participants were aged ≥18 years, receiving stroke or rehabilitation services from a participating hospital, and had a confirmed diagnosis of stroke (ischemic or hemorrhagic) with upper limb involvement. Fifty patients were recruited to the study (20 women, 30 men, N=50) with a mean age of 71 (SD 13.4, range 35-90) years. Time since stroke varied from 2 days to 187 months, with a median of 0.8 months.

Interventions

Not applicable.

Main Outcome Measures

Concurrent validity of the KHMS with the FMA-UE.

Results

A correlation of r=0.948 was found between the 2 scales (P=.0001). Moderate floor effects were noted in our sample (16%); however, significant ceiling effects were recorded (44%).

Conclusion

The KHMS demonstrated a statistically strong correlation with the FMA-UE and holds promise for use, particularly in the clinical setting, to evaluate upper limb motor impairment after stroke.

Stroke is a major cause of long-term disability, with impairments to upper limb movements as a result of stroke affecting around 40% of survivors¹ which significantly affects both functional hand use and quality of life. Upper limb impairments are not static, however, and recovery varies markedly between individuals.² Recommended rehabilitation treatments change over time to target assessed impairments, and so the ability to clinically assess and monitor recovery is vital to rehabilitation.² Understanding upper limb impairment after stroke is therefore essential to planning therapeutic efforts to restore function.

Multiple assessment tools have been used to evaluate upper limb movement after stroke, including the Action Research Arm Test, the Wolf Motor Function Test,¹ the Motor Assessment Scale,³ and the Fugl-Meyer Assessment for the Upper Extremity (FMA-UE).⁴ While there is no consensus regarding which assessment tool should be used to monitor upper limb recovery after stroke,^1,5 the FMA-UE is the most frequently used tool in research studies evaluating upper limb impairment.⁵ One of the acknowledged clinical limitations of these tools is that they are time-consuming to administer, leading to low clinical use. For example, the FMA-UE takes approximately 30 minutes to complete,⁶ the Motor Assessment Scale 20-30 minutes, the Wolf Motor Function Test 10-12 minutes, and the Action Research Arm test 10 minutes.¹ Of interest, therefore, is a 6-point scale, the Katrak Hand Movement Scale (KHMS), which was designed to assess hand movement and recovery after stroke in the clinical setting⁷ and can be administered in less than 1 minute.

The KHMS has been previously shown to predict recovery of hand movement and function in conjunction with the ability to shrug the hemiplegic shoulder at 5 weeks post onset of stroke.⁷ In a further study with a larger cohort, Katrak et al (1998) reported that early shoulder shrug and synergistic hand movements were useful predictors of hand function at 1, 2, and 3 months after stroke onset.⁸ Nijland et al further found that finger extension and shoulder abduction within 72 hours of stroke onset could help predict functional recovery of the upper limb at 6 months.⁹

Given the time-consuming nature of most tests to evaluate upper limb function after stroke, we were interested to investigate whether the much quicker KHMS would be of value in the clinical setting. Hence, the aim of this study was to investigate the concurrent validity of the KHMS to the FMA-UE in an inpatient population who had stroke. This study also aims to build on preexisting preliminary data on the KHMS.⁸ We hypothesize that the KHMS will correlate with the FMA-UE scale, and may therefore be of use in assessing upper limb recovery after stroke in the clinical setting.

Methods

Study design and participants

A prospective cohort study was undertaken. The study was approved by the Human Research Ethics Committee in accordance with the National Health and Medical Research Council (NHMRC) guidelines, and informed written consent was obtained from all participants.

The study was conducted at 3 tertiary rehabilitation hospitals. Potential participants were aged ≥18 years, receiving stroke or rehabilitation services from a participating hospital, and had a confirmed diagnosis of stroke (ischemic or hemorrhagic) with upper limb involvement. Patients excluded from the study included those who were unable to provide consent, those with a pre-existing upper limb impairment on the involved side, those with concurrent pathology (eg, fracture) of the involved limb that might restrict the ability to undertake tests, or those unable to comply with the study instructions. Potentially suitable patients were identified by stroke and rehabilitation clinicians and referred to study investigator who then screened them for eligibility. The KHMS and FMA-UE were administered to all eligible participants by 1 investigator on a single occasion.

Instruments

For the KHMS, hand movements are clinically observed then rated on a scale from 1 to 6,^7,8 where 1=no movement in digits, 2=synergistic active finger flexion only (no active finger extension), 3=synergistic active finger flexion and extension, without isolated finger movements, 4=able to extend the index finger in isolation (while maintaining the other fingers flexed), 5=able to bring the thumb into opposition to tip of the index finger but unable to oppose thumb to other digits, and 6=able to oppose the thumb to all fingertips (see fig 1).

Fig 1.

Katrak Hand Movement Scale. Reprinted from Arch Phys Med Rehabil, 79, Katrak et al, Predicting upper limb recovery after stroke: the place of early shoulder and hand movement, 758-761, Copyright (1998), with permission from Elsevier.

Given the wide, international use of the FMA-UE, it was chosen for testing the criterion-related validity of the KHMS. The FMA-UE includes 33 items, with a score of 0-2 for each item (0 for incomplete, 1 for partial completion, and 2 for smooth completion) with a total score ranging from 0 to 66.¹⁰ On both scales, a higher score indicates greater functional movement.

Statistical analyses

Descriptive statistics of demographic variables were calculated. Relations between KHMS and FMA-UE scores were assessed with the Spearman's rank correlation coefficient and their associated CIs. Spearman's rank correlation was chosen as it is a non-parametric test of association that is not overly influenced by outliers or skewed distributions. Interpretation of the Spearman's correlation coefficient was as follows: ρ<0.25, weak relation; 0.25–<0.50, fair relation; 0.50-0.75, moderate relation; and ρ>0.75, good relation.¹¹ Ceiling and floor effects for each scale were tested by running frequencies for the KHMS and documenting the percentage of respondents who scored at the top and bottom of each scale. Ceiling and floor effects less than 15% were considered minimal, while effects greater than 15% were considered moderate to substantial, according to criteria used by McHorney and Tarlov.¹² Statistical significance was set at P<.05. An a priori power calculation was completed to determine sample size. Power was valued at 0.80 (Beta=0.20) and alpha at 0.05 (2-tailed), assuming an expected correlation coefficient of 0.40 (fair). An estimated 47 participants were required as a minimum sample size.

Results

Fifty patients who had stroke were recruited to the study (20 women, 30 men) with a mean age of 71 (SD 13.4, range 35-90) years. Time since stroke varied (2 days to 187 months) and therefore medians are reported (table 1). Slightly greater than minimal floor effects were noted in our sample (16%); however, significant ceiling effects were recorded (44%) (table 2).

Table 1.

Characteristics of participants (n=50) and their descriptive data for KHMS and FMA-UE

Characteristic	(n=50)
Age (y), mean ± SD	71 (13.4)
Sex, number men (%)	30 (60)
Time since stroke (mo), mean ± SD, median (range)	5.7 (26.7), 0.8 (2 days to 187 mo)
KHMS, mean ± SD	4.4 (1.9)
FMA-UE mean ± SD	43.6 (23.1)

Table 2.

Descriptive data for KHMS for total sample (n=50) and subgroups according to cutoffs on the FMA-UE recommended by Hoonhorst et al¹³

	Total Sample, Number (%) n=50	FMA-UE Score 0-22, Number (%) n=13	FMA-UE Score 23-31, Number (%) n=5	FMA-UE Score 32-47, Number (%) n=2	FMA-UE Score 48-51, Number (%) n=2	FMA-UE Score 52-66, Number (%) n=28
KHMS Score 1	8 (16)	8 (61)	0	0	0	0
KHMS Score 2	4 (8)	4 (31)	0	0	0	0
KHMS Score 3	2 (4)	0	2 (40)	0	0	0
KHMS Score 4	5 (10)	1 (8)	3 (60)	1 (50)	0	0
KHMS Score 5	9 (18)	0	0	1 (50)	2 (100)	6 (21)
KHMS Score 6	22 (44)	0	0	0	0	22 (79)

We found a strong correlation between the 2 scales (r=0.948, P=.0001) (table 2). The KHMS was established to strongly correlate with the FMA-UE in adults with upper limb paresis after stroke and was faster to administer in the clinical setting to document upper limb motor recovery after stroke.

Discussion

We found that scores obtained using the KHMS achieved almost perfect correlation with those obtained using the FMA-UE for patients with upper limb impairment after stroke, although there was a ceiling effect observed with 44% of the sample scoring the highest rating (6/6). This correlation was most evident in the extremes of the scales (ie, very mild and very severe levels of impairment); however, the agreement was inconclusive in the middle ranges where there was overlap in some of the values.

Despite this, and given its rapid administration, the KHMS shows promise for use particularly in the clinical setting where it can evaluate upper limb impairment after stroke with good relation to the FMA-UE.

With respect to the time required to administer the FMA-UE test, we found this was variable depending on the participant's cognition and the severity of stroke, ranging from approximately 10 minutes (for those with a dense stroke) to 45 minutes (for those with a less dense stroke with cognitive impairment). Those with cognitive impairments often required the instructions to be repeated a number of times. This finding concurs with times previously reported in the literature, with some reporting that the full upper and lower limb assessment takes 45 minutes,¹⁴ and others stating that performance of the motor scale can be done in 20 minutes.⁶ In contrast, the time taken to complete the KHMS was approximately 30 seconds, ranging from 10 seconds to 1 minute.

Study limitations

A limitation of the study was that assessments were done on a single occasion only in a participant cohort drawn from a government-funded rehabilitation service and that other psychometric properties of the KHMS were not investigated. Further research is required to assess the KHMS for responsiveness, reliability, and predictive ability. Further, our results, with 22 (44%) participants scoring 6 on the KHMS, may suggest a ceiling effect. When we mapped KHMS categorical scores against the validated classifications of the FMA-UA, it would appear that our sample was simply over-represented with adults who only experienced mild upper limb impairments (FMA-UA score 52-66). Thus, replication of the current study with a larger sample is recommended, with patients sampled from various settings (inclusive of inpatient and community rehabilitation) and with a greater range of impairments.

Conclusions

In a stroke population, a statistically high degree of correlation was found between scores of upper limb function obtained with the KHMS and FMA-UE. The KHMS provides a rapid assessment of hand movements after stroke and may be suitable for use particularly in a clinical setting. A potential ceiling effect may be present, and therefore replication of our study with a larger population is recommended and with a more equal distribution of levels of impairment.