Clinimetric Properties of the Duruoz Hand Index in Patients With Stroke

Participants 

The study included 56 consecutive inpatients with hemiplegia after stroke, whom we recruited from 3 rehabilitation centers. Their mean ages and time since stroke ± standard deviation (SD) were 62.0±14.2 years and 84.0±52.6 days, respectively. The inclusion criteria were: having hemiplegia due to stroke, no previous pathology of the arm and/or hand, and cognitive ability to understand the questionnaire. Motor recovery level of the upper extremity was not an inclusion or an exclusion criterion. All participants received a conventional stroke rehabilitation program, 5 days a week, for 4 weeks. The conventional program is patient-specific and consists of neurodevelopmental facilitation techniques, physiotherapy, occupational therapy, and speech therapy (if needed). The baseline characteristics of the participants are summarized in table 1. Approval was obtained from the Medical Ethics Committee prior to the study.

Table 1. Baseline Characteristics of the Study Population (N=56)

	Characteristic	Value
	Age (y)	62.0±14.2
	Sex (women/men)	23/33
	Education
	Illiterate	9
	Elementary school	36
	High school, university	11
	Type of lesion (ischemia/hemorrhage)	46/10
	Paretic side (right/left)	23/33
	Time since stroke (d)	84.0±52.6
	Brunnstrom stages of arm	2.7±1.5
	Brunnstrom stages of hand	2.4±1.6
	MAS of elbow	1.5±0.9
	MAS of hand	1.5±0.9

NOTE. Values are n or mean ± SD. Abbreviation: MAS, Modified Barthel Index.

Sample Size 

We determined the required sample size by using Number Cruncher Statistical Systems.31,a Using an estimate of .90 for the lower-bound confidence interval (CI) (as per findings in patients with rheumatic conditions),27 a sample size of 56 participants with 2 observations per participant would achieve 84% power to detect an intraclass correlation coefficient (ICC) of .95.

Measures 

We assessed body functions and structures of the affected arm and hand of the patients in terms of motor recovery (Brunnstrom stages),32, 33 spasticity (Modified Ashworth Scale [MAS]),34 and sensation (light touch and joint position sense). The DHI and FIM self-care items were used to measure activity limitation of the hand.

Brunnstrom Stages 

Brunnstrom32, 33 defined 6 sequential stages of motor recovery and described how the hemiplegic arm, hand, and leg progress through these stages as an acceptable method for assessing recovery in patients with hemiplegia. The 6 stages of the Brunnstrom for the hand are: (1) flaccidity; (2) little or no active finger flexion; (3) mass grasp, use of hook grasp but no release, no voluntary finger extension, and possibly reflex extension of digits; (4) lateral prehension, release by thumb movement, semi-voluntary finger extension, with small range, (5) palmar prehension, possibly cylindrical and spherical grasp, awkwardly performed and with limited functional use, voluntary mass extension of digits, with variable range, and (6) all prehensile types under control, skills improving, full-range voluntary extension of digits; individual finger movements present, but less accurate than on the opposite side.

Modified Ashworth Scale 

We used the MAS to grade spasticity for arm and hand. The MAS is a 5-point ordinal rating scale with good interrater reliability designed to measure muscle tone. Higher MAS scores indicate worse spasticity.34

Sensory Evaluation 

Light touch and joint position sense of the paretic hand were recorded. We assessed light touch sense using a piece of cotton wool after the patient was familiarized to normal sensation on the anterior chest and forehead. The results were then categorized as normal, impaired, and absent. We assessed joint position sense by holding the patient’s index finger laterally at the distal interphalangeal joint and moving it up or down a few millimeters. The patient was asked to report the direction of movement relative to the last position and this was graded as normal, abnormal, or absent.9

FIM Instrument 

The FIM instrument is the functional status component of the Uniform Data System for Medical Rehabilitation. It is widely used in rehabilitation centers and has properties useful for stroke investigators. It contains 18 items that measure independent performance in self-care, sphincter control, transfers, locomotion, communication, and social cognition. FIM scores range from 1 to 7: a FIM item score of 7 is categorized as “complete independence,” whereas a score of 1 is “complete dependence” (performs <25% of task). Scores below 6 require another person for supervision or assistance.35, 36 The FIM self-care items were used in the present study.24, 37, 38, 39 The reliability and validity of the Turkish version of the FIM has been well documented.40 The second author (GY) has been trained in administration of the FIM and certified. The other assessors (NS, KS, PB) were trained by the second author. All assessors had at least 2 years of experience with the FIM in patients with stroke.

Duruoz Hand Index 

The DHI was developed as a self-report questionnaire that can be routinely used to assess hand-related activity limitation in patients with RA.27 It has been cross-validated for outcome assessment of hand-related activity in patients with OA,30 systemic sclerosis,20 and those receiving hemodialysis.11 Its reliability has been shown in patients with RA,27 OA,30 and systemic sclerosis.20 Its responsiveness has also been shown during the course of the disease41 and after hand surgery in patients with RA.29

It contains 18 items on hand ability in the kitchen, during dressing, while performing personal hygiene, office tasks, and other general items (appendix 1). In previous studies, 3 factors were extracted for the DHI. The first factor represents activities requiring force and rotation (questions 2, 3, 5, 6, 11, 12, 15, 18), the second factor represents activities requiring dexterity and precision (questions 1, 4, 7−10), and the third factor represents dynamic activities requiring flexibility of the first 3 fingers.11, 27 Scores for kitchen tasks range from 0 to 40. Scores for dressing, hygiene, and office tasks range from 0 to 10. Scores for the “other” category range from 0 to 20. Persons rate their ability from 0 (no difficulty) to 5 (impossible to do). The questionnaire yields a total score from 0 to 90, takes about 3 minutes to complete, and 6 levels of answers allow a more sensitive grading of hand-related activity limitation. A higher score indicates greater activity limitation or more difficulty.27, 28

Design 

We assessed the patients on 3 different occasions. Two baseline measurements were performed with a 24-hour interval in between; the study population was considered to be clinically stable during the period between these 2 baseline measurements. A third follow-up assessment measurement was made immediately after completing a 4-week inpatient rehabilitation program. Figure 1 shows the flow of patients from recruitment to completion of the study. Because we aimed to measure activity limitation, we asked the patient to assess their ability without any hand preference.42 Literate patients completed the questionnaires by themselves and interviewers read the questions to illiterate patients.

View Large Image Download to PowerPoint

Fig 1. Flow diagram from recruitment to completion of the study.

Data Analysis 

We analyzed the data using SPSSb for Windows. Our data showed normal distribution. We calculated the means and SDs for all measures. Reliability refers to the reproducibility (the degree to which the score is free from random error) and the internal consistency of the instrument. In the present study, test-retest reliability and internal consistency of the DHI was measured by ICCs and the Cronbach α coefficient, respectively. Test-retest reliability measures the stability over time by administering the same test to the same subjects at 2 points in time; it is commonly evaluated using correlation statistics such as ICC, Pearson or Spearman coefficients, and κ coefficients.6 Internal consistency assesses the homogeneity of the scale items. It is generally examined using Cronbach α statistics or split-half reliability. Item-to-item and item-to-scale correlations are also accepted methods.6

We calculated the minimal detectable change (MDC) between first and third assessment of the DHI. The MDC is a clinically relevant measure suggesting the change that might be expected because of an intervention rather than sampling error at a significance level of .05. The MDC is the 95% CI of the standard error (SE) of measurement multiplied by the square root of 2; that is, MDC = 1.96 × √2 × SE of measurement. From the error variance, the SE of measurement was calculated as its square root. The SE of measurement is the SD of the population of all possible measurement errors. The SE of measurement is an estimate of how much a score is likely to vary with repeated measurements.43, 44, 45

Validity is a term for how well an instrument measures what it intends to measure. Forms of validity include face, content, construct, and criterion. Concurrent, convergent, discriminative, and predictive validity are all considered to be forms of criterion validity.7

The Spearman correlation coefficient was used to show the association between DHI and FIM self-care items (construct validity).6 Discriminative power of the DHI was investigated by comparing the data of patients with dominant side paresis versus nondominant side paresis, using an independent sample t test.

Responsiveness is the sensitivity to change within a patient over time, which might be indicative of therapeutic effects.6 The usefulness of an outcome measurement is related to its sensitivity to change (responsiveness).29 Responsiveness is most commonly evaluated through correlation with other change scores, effect sizes, standardized response means, relative efficiency, sensitivity and specificity of change scores, and receiver operating characteristics analysis.6 We used effect size and paired t test to assess responsiveness.6, 30, 41, 46 The effect size is defined as the mean change in scores between the baseline and the follow-up visit divided by the SD of the baseline score. A higher effect size indicates greater responsiveness. A negative value indicates that the mean score at the baseline visit is smaller than the mean score at the follow-up visit. A paired t test was used to determine the statistical significance of change scores.

Reliability 

We determined the test-retest reliability of the DHI between the first and second assessments using the ICC. The accepted evaluation criteria and standards for ICC values are as follows: values of .75 or greater represent excellent reliability; values between .40 and .74 represent adequate reliability, and values of .40 or lower represent poor reliability.6 The ICCs for the test-retest reliability of the DHI were found to be excellent for both the individual section and total scores (table 3). Each section showed ICCs of .99. We assessed internal consistency of the DHI with the Cronbach α; it is suggested that the α should be .80 and over for excellent internal consistency.6 In our study, Cronbach α coefficients for the DHI sections ranged from .91 to .97, which shows excellent homogeneity of the items in the questionnaire (see table 3).

Table 3. Data on Test-Retest Reliability and Internal Consistency

	DHI Section	Test-Retest Reliability ICC (95% CI)	Internal Consistency (Cronbach α)
	Kitchen	.99(.93–.99)	.93
	Dress	.99(.94–.99)	.91
	Hygiene	.99(.94–.99)	.95
	Office	.99(.93–.99)	.96
	Other	.99(.95–.99)	.92
	Total	.99(.93–.99)	.97

Validity 

We assessed construct validity by Spearman correlation coefficients (table 4). Correlation coefficients of .60 or greater represent excellent correlation, values between .31 and .59 represent adequate correlation, and values of .30 or lower represent poor correlation.6 In the present study the correlation between the DHI and the FIM self-care items was excellent. An adequate correlation was found between the DHI and Brunnstrom stages, and between the sensation of the affected upper extremity. There was no correlation between the DHI and spasticity level of the affected upper extremity. Discriminant validity was shown between subgroups of participants defined according to the affected side; as expected, patients with nondominant side paresis showed a better performance on the DHI, and the difference between the 2 groups was statistically significant.

Table 4. Data on the Construct Validity of the DHI

	DHI	Spearman ρ	P
	FIM self-care	−.73	<.001
	Brunnstrom stages of arm	−.54	<.001
	Brunnstrom stages of hand	−.50	<.001
	Sensation	.31	<.05
	MAS of elbow	−.10	NS
	MAS of wrist	−.10	NS

Abbreviation: NS, not significant.

Responsiveness 

We assessed responsiveness of the DHI by calculation of the effect size and paired t test. There was a significant improvement in the DHI score after 4 weeks of inpatient rehabilitation program. The mean ± SD paired difference value was 6.5±18.65 (P<.05). The effect size was .24.

Study Limitations 

Stroke patients recruited in this study were referred from all over the country for inpatient rehabilitation. Generally, an estimated 50% of the stroke population is referred to a rehabilitation center if they cannot return home directly after dismissal from the hospital. A potential limitation of this study is the generalizability of the results. According to our inclusion criteria, our findings and conclusions are based on the population of subacute stroke inpatients without severe cognitive deficits but with severe motor impairment of hand and upper extremity. Most severe strokes with cognitive problems and least severe outpatient strokes were eliminated at the beginning of the study. Further studies evaluating chronic stroke outpatients with less motor impairment should be performed.