- Email: [email protected]
Validity and reliability of the Dynamic Gait Index in children with hemiplegic cerebral palsy
Gait & Posture 75 (2020) 28–33
Contents lists available at ScienceDirect
Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost
Full length article
Validity and reliability of the Dynamic Gait Index in children with hemiplegic cerebral palsy⋆
T
⁎
Ayca Evkayaa, , Evrim Karadag-Saygia, Duygu Karali Bingulb, Esra Giraya a Marmara University Medical School, Department of Physical Medicine and Rehabilitation, Fevzi Çakmak Mahallesi, Tepe Sokak, No: 41, Üst Kaynarca, Pendik, Istanbul, Turkey b Istanbul University Medical School, Department of Physical Medicine and Rehabilitation, Pain Medicine Division, Süleymaniye Esnaf Hastanesi Poliklinikleri, Takvimhane Caddesi, No: 19, Eminönü, Istanbul, Turkey
A R T I C LE I N FO
A B S T R A C T
Keywords: Balance Cerebral palsy Dynamic Gait Index Gait Postural control Validity and reliability
Background: Dynamic Gait Index (DGI) is a performance-based tool can be applied in a short time and evaluates dynamic balance and gait ability. Research question: Is the DGI valid and reliable for assessing gait and balance disorders in children with hemiplegic cerebral palsy (CP)? Methods: Sixteen children with hemiplegic CP (5 females, 11 males; mean age 10y 3mo, SD 2y 7mo; range 6–14y; Gross Motor Function Classification System (GMFCS) levels I [n = 9], II [n = 7]) and 16 age-matched typically developing (TD) (8 females, 8 males; mean age 9y 9mo, SD 2y 6mo; range 6–14y) participated. The relationship between the DGI, Four-Square Step Test (FSST), Timed Up and Go Test (TUG) and Pediatric Berg Balance Scale (PBS) was analyzed. To determine the test-retest reliability, the DGI was performed twice and; for the inter-rater reliability, only DGI was reapplied by a different rater on the same day. Internal consistency was obtained by Cronbach-α value. Validity was tested by Spearman correlation coefficient and reliability was calculated by Intraclass correlation coefficient (ICC). Results: There was a significant difference between hemiplegic CP and TD and between the children with GMFCS level I and II in the comparison of results of the DGI and other tests. All items on the DGI had appropriate internal consistency (Cronbach-α = 0.969). The test-retest (ICC = 0.970 CI(0.915- 0.990)) and inter-rater (ICC = 0.983 CI(0.882- 0.998)) reliabilities were found to be excellent. A negative, moderate correlation between FSST and DGI (rs = −0.673, p = 0.004); a positive, high correlation between PBS (rs = 0.724, p = 0.002) and DGI and a negative, high correlation between TUG and DGI (rs = −0.828, p < 0.001) was detected. Significance: DGI with features such as its feasibility in a short time, being simple but distinctive and not requiring heavy equipment is a valid and reliable method in children with hemiplegic CP.
1. Introduction Dynamic Gait Index (DGI) is a performance-based tool that quantifies the dynamic balance instability developed by Shumway-Cook and Woollacott, evaluates the ability of the individual to modify gait in response to changing functions during walking [1,2]. In adults, it has been shown that DGI allows comparisons between patient groups with various pathologies leading to balance problems and contains excellent psychometric properties [3–6] and the test is completed in approximately 10 min [1]. Although the incidence of balance disorders and high incidence of
falls in children with cerebral palsy (CP) are well known [7], in the literature there has been very little effort to quantify the dynamic gait stability of these children. On the contrary, most of the literature focuses on assessing the balance instability seen in these children with standing posture balance measurement methods [8,9]. The problem with this approach is; measurement methods of static stability and dynamic gait stability or predictions of falls do not show good correlation [10,11]. Therefore; standing posture balance measurement methods may not be the best measurement to assess the dynamic gait stability of children with CP. The Dynamic Gait Index is a sensitive and efficient tool for adults
⋆
The study results were presented as oral presentation at the 6th Medical Rehabilitation Congress. Corresponding author. E-mail addresses: [email protected] (A. Evkaya), [email protected] (E. Karadag-Saygi), [email protected] (D. Karali Bingul), [email protected] (E. Giray). ⁎
https://doi.org/10.1016/j.gaitpost.2019.09.024 Received 2 May 2019; Received in revised form 21 July 2019; Accepted 22 September 2019 0966-6362/ © 2019 Elsevier B.V. All rights reserved.
Gait & Posture 75 (2020) 28–33
A. Evkaya, et al.
Item 2 “When I tell you go, walk as fast as you can; when I tell you slow, walk as slowly as you can”, Item 3 “When I tell you to look right or left keep walking straight but turn your head to the right or left”, Item 4 “When I tell you to look up or down keep walking straight but turn your head to the up or down. Keep your head up or down until I tell you look straight, then keep walking straight, but return your head to the center”, Item 5 “Begin walking at your normal pace. When I tell you, turn and stop, turn as quickly as you can to face the opposite direction and stop”, Item 6 “Begin walking at your normal speed. When you come to the shoebox, step over it, not around it, and keep walking”, Item 7 “Begin walking at normal speed. When you come to the first cone, walk around the right side of it. When you come to the second cone, walk around it to the left” Item 8 “Walk up these stairs as you would at home, i.e., using the railing if necessary. At the top, turn around and walk down.” The test is completed in approximately 10 min [1]. The Four-Square Step Test evaluates the ability to change direction when performing clinical steps. The patient is asked to complete the whole sequence as soon as possible, provided that both feet touch the ground at the same time without touching the lines. The test is completed in approximately 5 min [13,14]. The Timed Up and Go is a test used to measure mobility and evaluates walking speed, posture control, functional mobility and balance. The test is completed in approximately 10 min [15]. The Pediatric Berg Balance Scale is a revised version of the Berg Balance Scale (BBS) for children by Franjoine et al. in order to assess the functions of daily life activities. The scale consists of 14 evaluations and each section is scored between 0–4; the highest score is 56. In PBS; the order of the sections in the standard BBS, is rearranged to be easy to difficult as functional sorting form, the time standards in the sections related to protection of static postures have been reduced to the pediatric population and the guidance has been simplified. It can be administered in approximately 20 min [16,17].
[1,3,5,6] and therefore may be a useful tool for children. The objective investigation of the movement is important to determine the appropriate treatment methods and to understand the effectiveness of these methods for children with CP. The DGI; although it has been widely used in postural control studies in adults [3,5,6,12], it has not yet been investigated to assess walking difficulties in children with CP. Postural control and dynamic balance are important in children, so these structures need precise and reliable measurements. Therefore, the aim of this study was to evaluate the reliability and validity of DGI in children with cerebral palsy. 2. Methods 2.1. Participants Potential participants consisted of 16 children with hemiplegic CP whose level of Gross Motor Function Classification System (GMFCS) I and II, ages 6 through 14 years, presented to the Pediatric Rehabilitation Special clinic of Department of Physical Medicine and Rehabilitation at the Marmara University Pendik Education and Research Hospital between April 2018 and June 2018. In addition, 16 age-gender matched typically developing (TD) children were included in the study as a control group The GPower V.3.1.9.3 program was used to determine appropriate sample size. When the average expected value in the patient group was 21.4 (with a standard deviation of 1.4) and the average expected value in the typically developing group was 23.1 (with a standard deviation of 1.4) based on the study results of Lubetzky-Vilnai et al. [13], it was found that 14 subjects for each group must had been enrolled to have 95% power with 5% type 1 error level. The inclusion criteria were: (1) being classified at GMFCS level I–II; (2) having a spasticity value 1, 2 or maximum 3 of the lower extremity muscles according to Modified Ashworth Scale (MAS); (3) being able to stand and walk without orthosis and any auxiliary equipment; and (4) being able to understand and execute instructions for assessment. The exclusion criteria were: (1) having undergone any intervention (e.g. botulinum toxin injection or orthopedic surgery) that may affect gait between the test and re-test assessment; (2) having visual impairments excluding refractive defects; (3) having any accompanying systemic disorder; (4) having uncontrolled epilepsy; (5) having the lower extremity contracture that affect the assessment; and (6) having the emergence of health problems that may affect the study. The study was conducted after obtaining approval from Marmara University Health Sciences Institute Ethics Committee in accordance with the Declaration of Helsinki (02.04.2018-113) and after the registry of Clinical Trials Registry. The trial is registered with ClinicalTrials.gov (NCT03662139). Informed consent was obtained from all legal guardians of the participating children. Our study is an observational methodological research study that evaluates validity and reliability.
2.3. Validity and reliability protocol Data collection was conducted in a quiet room where children can feel comfortable to minimize distractions. Some modifications for children in measurements were performed; these modifications included verbal redirects and repetitions. Some test items in which verbal expression caused confusion were taught to the children by showing movement. The area of assessments was a solid surface, free from any physical obstacles and has a large place with enough lighting. Participants wore daily comfortable clothes and shoes. Two raters evaluated participants’ performance. To examine testretest reliability the DGI was administered to each participant in the CP group by rater 1 in 2 testing sessions 7 days apart. To evaluate the interrater reliability the DGI was rated simultaneously by rater 2 in the latter testing session during the same day. Two raters were not allowed to consult each other during the test and had no access to previous test results. The FSST, TUG, and PBS were administered and rated by rater 1 immediately after the first session to test the concurrent validity of the DGI. The patient assessment form was accomplished, and demographic characteristics were collected at the first session. The whole testing protocol in the first session took approximately 45 min. Participants were rested for five minutes between the two evaluation to minimize fatigue. In the second session, only the DGI was performed. The participants in the CP group were determined by the polyclinic patients who met the required criteria. The control group was selected from TD healthy children who accompanied their families to hospital.
2.2. Measurements The Modified Ashworth Scale was used to evaluate spasticity, GMFCS was applied for functional level assessment. The Dynamic Gait Index, Four Square Step Test (FSST), Timed Up and Go (TUG) and Pediatric Berg Balance Scale (PBS) were applied for gait and balance assessment. The Dynamic Gait Index measures functionality and dynamic balance in walking and stair climbing. The DGI consists of 8 items including normal gait on flat ground, gait with speed changes, gait with horizontal head movements, gait with vertical head movements, gait and pivot turn, gait and step over obstacle, gait around obstacles and steps up and down stairs. The performance on each item is rated on a 4point scale (3, independent walking; 2, mild impairment; 1, moderate impairment; 0, severe disorder) with a maximum score of 24. Each item is indicated to children as follows: Item 1 “Walk at your normal speed”,
2.4. Statistical analysis "SPSS 20.0 Statistical package" program was used to evaluate the data of the groups. The data was presented with tables and graphs. Statistical significance was accepted as p < 0.05. Descriptive characteristics are provided as number-percent, mean, 29
Gait & Posture 75 (2020) 28–33
A. Evkaya, et al.
standard deviation, median, minimum-maximum values. Differences in these characteristics were tested using “Mann Whitney U" test and “Chisquare” test for continuous and categoric variables, respectively. The comparison of the applied methods for the patient and TD group was performed with the "Mann Whitney U" test. The internal consistency of DGI items was analyzed by the "Cronbach-α coefficient". To evaluate the test-retest and inter-rater reliability analysis of the DGI, Intraclass Correlation Coefficient (ICC) and 95% confidence intervals (CI) with the two-way random-effect model were calculated. Reliability estimates were interpreted as follows: > 0.90=excellent; 0.75-0.90=good; 0.50-0.75=medium; < 0.50=low [18–20]. The correlation of balance assessment methods was tested using the "Spearman correlation coefficient". The level of relationship was classified as follows < 0.30=small/negligible; 0.30–0.50=low; 0.50–0.70=moderate; 0.70-0.90=high; > 0.90=very high [21,22].
Table 2 Comparison of the DGI Scores in Hemiplegic CP and TD.
DGI DGI DGI DGI DGI DGI DGI DGI DGI
1 2 3 4 5 6 7 8 Total
Hemiplegic CP (n = 16) Median (Min–Max)
TD (n = 16) Median (Min–Max)
p
2 (1–3) 2.50 (1–3) 2 (0–3) 2 (0–3) 2 (2–3) 2 (0–3) 3 (1–3) 2.50 (1–3) 19 (8-23)
3 (3–3) 3 (3–3) 3 (3–3) 3 (3–3) 3 (2–3) 3 (3–3) 3 (3–3) 3 (3–3) 24 (23–24)
p < 0,001 0.001 0.001 0.001 0.005 0.001 0.008 0.001 p < 0.001
CP: Cerebral palsy, TD: Typically developing, DGI: Dynamic Gait Index, Min: Minimum, Max: Maximum, p: Statistical significance. *Mann-Whitney U test. Table 3 Test–retest Reliability of the DGI scores in Hemiplegic CP.
3. Results 3.1. Participants Sixteen patients; 5 girls (31.3%), 11 males (68.8%) with a mean age of 10.31 ± 2.62 years and 16 TD children; 8 girls (50%), 8 males (50%) with a mean age of 9.75 ± 2.51 years were included in this study and their characteristics are presented in Table 1. Most children were at level I (n = 9), which was the highest level for GMFCS. 56.3% (n = 9) of the children had right extremity and 43.8% (n = 7) had left extremity effected in the hemiplegic CP group. 3.2. DGI total scores
(n = 16)
1st assessment Median ((Min–Max)
2nd assessment Median(Min–Max)
ICC*
95% CI
DGI DGI DGI DGI DGI DGI DGI DGI DGI
2 (1–3) 2.50 (1–3) 2 (0–3) 2 (0–3) 2 (2–3) 2 (0–3) 3 (1–3) 2.50 (1–3) 19 (8–23)
2 (1–3) 2.50 (1–3) 2.50 (1–3) 2.50 (1-3) 2 (2–3) 2 (0–3) 3 (1–3) 2.50 (1–3) 18.50 (10–23)
0.805 0.913 0.861 0.889 0.906 0.965 0.913 0.972 0.970
0.429–0.933 0.749–0.970 0.613–0.951 0.690–0.961 0.737–0.967 0.898–0.988 0.749–0.970 0.922–0.990 0.915–0.990
1 2 3 4 5 6 7 8 Total
CP: Cerebral palsy, DGI: Dynamic Gait Index, Min: Minimum, Max: Maximum, ICC: Intraclass correlation coefficient, CI: Confidence interval.
The DGI total score was significantly lower for the 16 children with hemiplegic CP compared with 16 TD (p < 0.001). Children with CP demonstrated a median DGI score of 19 (range, 8–23) while heathy children had a median score of 23 (range, 23–24). As a result of DGI assessments, the items with the lowest scores in children with hemiplegic CP were identified as gait with horizontal and vertical head movements (items 3 and 4) and gait and step over obstacle (item 6). The item with the best scores was identified as taking gait around obstacles (item 7) (Table 2).
Table 4 Inter-rater Reliability of the DGI Scores in Hemiplegic CP.
3.3. Reliability The internal consistency of the DGI was high with a Cronbach's alpha coefficient of 0.969. The test-retest reliability of DGI was excellent (ICC = 0.970 Cl (0.915-0990)). Inter-rater reliability of DGI was excellent (ICC = 0.983 Cl(0.8820990)). The ICC values for each item was shown in Tables 3 and 4. The Bland-Altman plots for test-retest compliance and inter-rater compliance are shown in Fig. 1.
(n = 6)
1st assessment Median ((Min- Max)
2nd assessment Median (Min- Max)
ICC*
95% CI
DGI DGI DGI DGI DGI DGI DGI DGI DGI
2 (1–2) 2.50 (2–3) 2 (2–3) 2 (1–3) 2 (2–3) 2 (1–3) 3 (2–3) 3 (2–3) 18 (13–23)
2 (1–2) 2.50 (2–3) 2 (2–3) 2 (1–3) 2.50 (2–3) 2.50 (1–3) 3 (2–3) 3 (2–3) 18.50 (13–23)
1.00* 1.00* 1.00* 0.906 0.828 0.928 1.00* 1.00* 0.983
0.434–0.986 −0.038- 0.975 0.566–0.990
1 2 3 4 5 6 7 8 Total
0.882–0.998
CP: Cerebral palsy, DGI: Dynamic Gait Index, Min: Minimum, Max: Maximum, ICC: Intraclass correlation coefficient, CI: Confidence interval. * ICC = 1.00 due to 100% compliance between non-variability items and evaluators.
3.4. Validity Table 1 Comparison of Characteristics Data in Hemiplegic CP and TD.
Age (year) Height (cm) Body weight (kg) BMI (kg/m2) Gestational week Birth weight (kg)
Hemiplegic CP (n = 16) Mean ± SD
TD (n = 16) Mean ± SD
p
10.31 ± 2.62 142 ± 15 38.43 ± 10.85 18.56 ± 3.12 35.12 ± 4.22 2.52 ± 0.78
9.75 ± 2.51 140 ± 13 38.87 ± 11.63 19.40 ± 3.23 37.25 ± 2.38 3.03 ± 0.67
.519* .451* .940* .559* .155* .054*
In order to evaluate the validity of the DGI, the study used FSST, TUG and PBS scales were used. A negative, moderate correlation with FSST (rs = −0.673, p = 0.004); a positive, high correlation with PBS (rs = 0.724, p = 0.002) and a negative, high correlation with TUG (rs = −0.828, p < 0.001) were found (Table 5). 3.5. Comparison of the test scores according to GMFCS Levels The comparative values of the GMFCS levels and the gait and balance tests of the hemiplegic CP included in the study were examined. When the test scores were compared according to the GMFCS levels
CP: Cerebral palsy, TD: Typically developing, BMI: Body Mass Index, SD: Standard deviation, p: Statistical significance. * Mann-Whitney U test. 30
Gait & Posture 75 (2020) 28–33
A. Evkaya, et al.
4. Discussion There are many studies in the literature focusing on balance and postural control in children with CP [23–25]. In our study, we found that high levels of significant difference (p < 0.001) in favor of the TD in terms of DGI, FSST, TUG, and PBS results for gait and balance assessment. This indicates that children with hemiplegic CP have poor balance and postural control and gait disorders associated with these problems. Lubetzky-Vilnai et al., investigated the DGI in children in their study, 10 children aged between 8–15 years who were diagnosed with FASD (Foetal Alcohol Spectrum Disorder) who had postural control disorder and 10 age-gender matched TD were compared and DGI total score in children with FASD diagnosis was found to be significantly lower than TD (p = 0.01). The mean of the patients with FASD was 21.4 ± 1.4 and the mean of the TD 23.3 ± 1.06 [26]. Although these results are consistent with our study; it can be said that the mean values of the patient group were lower compared to the other study because of the difference in diagnosis and accompanying disorders in our study. The test-retest reliability was excellent for DGI total score (ICC = 0.970 Cl(0.915-0.990)) in our study. The Cronbach alpha coefficient for the internal consistency of the test was found to be very high with 0.969. The inter-rater reliability was excellent for DGI total score (ICC = 0.983 Cl(0.882-0.998)). Eight different DGI items were found to be good-excellent reliable (ICC = 0.805-0.972) and normal gait (item 1), gait with horizontal and vertical head movements (items 3 and 4) are good; other items had excellent reliability for test-retest reliability. Eight different DGI items were found to be good-fully compliant (ICC = 0.828–1.000) and normal gait (item 1), gait with speed changes (item 2), gait with horizontal head movements (item 3), gait around obstacles (item 7) and steps up and down stairs (item 8) were fully compliant items that received the same score from the two raters; the gait and pivot turn (item 5) was good, and the other two items had excellent reliability for inter-rater reliability. Lubetzky-Vilnai et al. the test-retest reliability was found to be medium (ICC = 0.710 Cl(0.260-0.910, p = 0.005) reliable and interrater reliability was found to be good (ICC = 0.820 Cl(0.490-0.940), p = 0.001) reliable [26]. When the results were compared, our study had higher test-retest and inter-rater reliability results. With these findings, it can be said that the tool can be used in children, but there is still a need in the literature for different studies investigating the usability of DGI in children. The relationship of DGI with other clinical tests is examined, a negative, moderate correlation with FSST (rs = −0.673, p = 0.004); a positive, high correlation with PBS (rs = 0.724, p = 0.002) and a negative, high correlation with TUG (rs = −0.828, p < 0.001) were found. In the correlations between the other clinical tests, a positive, high correlation between FSST and TUG (rs = 0.844, p < 0.001), a negative, high correlation between FSST and PBS (rs = −0.797, p < 0.001), and a negative, high correlation between TUG and PBS (rs = −0.763, p = 0.001) were found. When we look at the literature, the number of studies on children who study the relationship between the balance assessment methods used in our study is very small. Gan et al. found a strong negative relationship between PBS and TUG [27]. Bandong et al. examined the validity and reliability of the FSST which included children 5–12 years of age CP, DS. They analyzed the correlation between FSST and TUG and stated that the FSST was a valid and reliable test in both the CP and the DS [13]. These studies show parallels with our results. In the literature, we did not find a study that looks at the correlation between the tests in which all the clinical tests we use in our study are used together or in groups. More studies are needed to investigate the correlation with each other of balance assessment methods on children. Independent verticalization and mobilization are required to be able
Fig. 1. The Bland-Altman plots for test-retest and inter-rater compliance. *points where two circles overlap.
Table 5 Correlation analysis of the DGI total score with the FSST duration, the TUG duration and the PBS total score in Hemiplegic CP. (n = 16) DGI
rs* p*
FSST
TUG
PBS
−.673** .004
−.828** .000
.724** .002
CP: Cerebral palsy, DGI: Dynamic Gait Index, FSST: Four Square Step Test, TUG: Timed Up and Go, PBS: Pediatric Berg Balance Scale, rs: Correlation coefficient., p: Statistical significance. * Spearman correlation coefficient.
of the CP group; no significant difference was found except gait and step over obstacle (item 6), steps up and down stairs (item 8) and total score of DGI. There was a high level of difference between GMFCS I and GMFCS II children for FSST (p = 0.001) and TUG (p = 0.007) scores. There was no significant difference between GMFCS I and GMFCS II children for stand on one foot (item 9), stepping (item 13) and functional reach (item 14) items. A high level of significant difference was found in favour of GMFCS level I children (p = 0.005) for PBS total scores. Stand-sit (item 2), standing balance (item 4) and sitting balance (item 5) items in PBS were all items for which all children received full points.
31
Gait & Posture 75 (2020) 28–33
A. Evkaya, et al.
Also, all tests were found to have discriminative ability in terms of balance for CP patients with GMFCS levels I and II. The data obtained from the result of the study; for physiotherapists working in the pediatric area, with features such as easy and fast administration, being simple but distinctive and do not containing heavy equipment; DGI has been shown to be a valid and reliable method in independent walking hemiplegic CP patients.
to evaluate dynamic and static balance. According to GMFCS levels, the levels containing these characteristics are defined as I and II. There are studies in which children with GMFCS level I and II are included and compared to assess the balance in the literature. In our study, which included children with GMFCS level I and II, there was a significant difference between the two levels for the total score of the DGI (p = 0.21). The items with significant differences between the two levels gait and step over obstacle (item 6) steps up and down stairs (item 8). When we think that the level of distinctive functionality for GMFCS levels I and II are the use of a railing to walk up and down stairs; we can say that it was an expected result. No study evaluating DGI in relation to different GMFCS levels could be found in the literature. In the comparison of two GMFCS levels for FSST, there was a significant difference between level I and II (p = 0.001). Similar studies were not found in the literature. Therefore, in our study, we thought that we could contribute to the literature by comparing the results of FSST according to the GMFCS levels. For TUG, there was a high level of significant difference between children with GMFCS level I and II (p = 0.007). Carey et al. found similar results for GMFCS I and II (p = 0.000), GMFCS I and III (p = 0.000), and GMFCS II and III (p = 0.003) [28]. Gross Motor Function Classification System GMFCS is the recommended lower extremity classification by SCPE [29,30]. Also, Winters and Gage classification system is specific to hemiplegic cerebral palsy which reflects the gait abnormalities in detail. It has four categories. We could not investigate the relationship; because we lack enough children with hemiplegic cerebral palsy in each category. The relationship between Winters and Gage classification categories and DGI should be Winters and Gage classification enlightened in further studies. When the children with GMFCS levels I and II were compared for PBS, the balance was significantly better in favor of level I patients (p = 0.005). In terms of PBS items, there was no significant difference except stand on one foot (item 9), stepping (item 13) and functional reach (item 14). In two similar studies evaluating the difference between GMFCS levels with PBS, similar results were observed in terms of items and total score [27,31]. Not performing sensitivity analysis of DGI can be viewed as the limitation of the study. We could not analyze sensitivity of DGI because the tools which we used to validate DGI do not have exact cut-off values that differentiate children with impaired and intact balance. Sensitivity of DGI should be investigated in further studies. The inclusion of a higher number of children in future studies may lead to better analyzes. In addition, there is a need for more comprehensive studies in containing children with CP in different levels of GMFCS and consisting of different clinical types such as spastic diplegic, dyskinetic, ataxic type CP. We have encountered a single study in the literature on the use of the DGI tool in children. In future studies, it may be useful to investigate the use of DGI in children and to determine the cut-off values in different age groups. The number of studies in which balance assessment methods are associated with GMFCS levels of children with CP is low. Looking forward, the relationship between the results of the balance assessment methods and the GMFCS levels can be investigated.
Funding source There is no funding for this study. Clinical trial registration number The study was conducted after approval from the Ethics Committee of Marmara University Health Sciences Institute in accordance with the Declaration of Helsinki (approval number: 02.04.2018-113). The trial is registered with ClinicalTrials.gov, number NCT03662139. Declaration of Competing Interest The authors declare no conflict of interest. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Acknowledgements We wish to thank Anne Shumway-Cook for giving us permission to use the DGI. We also thank all the parents and children who participated in this study. References [1] A. Shumway-Cook, M. Woollacott, Motor Control: Theory and Practical Applications, first ed., Williams and Wilkins, Baltimore, 1995. [2] T. Herman, et al., The Dynamic Gait Index in healthy older adults: the role of stair climbing, fear of falling and gender, Gait Posture 29 (2) (2009) 237–241. [3] D. Cattaneo, J. Jonsdottir, S. Repetti, Reliability of four scales on balance disorders in persons with multiple sclerosis, Disabil. Rehabil. 29 (24) (2007) 1920–1925. [4] Y.-P. Chiu, et al., Use of item response analysis to investigate measurement properties and clinical validity of data for the dynamic gait index, Phys. Ther. 86 (6) (2006) 778–787. [5] C.D. Hall, S.J. Herdman, Reliability of clinical measures used to with peripheral vestibular disorders, J. Neurol. Phys. Ther. 30 (2) (2006) 74–81. [6] J. Jonsdottir, D. Cattaneo, Reliability and validity of the dynamic gait index in persons with chronic stroke, Arch. Phys. Med. Rehabil. 88 (11) (2007) 1410–1415. [7] M. Bottos, C. Gericke, Ambulatory capacity in cerebral palsy: prognostic criteria and consequences for intervention, Dev. Med. Child Neurol. 45 (11) (2003) 786–790. [8] P.A. Burtner, et al., The capacity to adapt to changing balance threats: a comparison of children with cerebral palsy and typically developing children, Dev. Neurorehabil. 10 (3) (2007) 249–260. [9] J. Chen, M.H. Woollacott, Lower extremity kinetics for balance control in children with cerebral palsy, J. Mot. Behav. 39 (4) (2007) 306–316. [10] H.G. Kang, J.B. Dingwell, A direct comparison of local dynamic stability during unperturbed standing and walking, Exp. Brain Res. 172 (1) (2006) 35–48. [11] T.M. Owings, et al., Measures of postural stability are not predictors of recovery from large postural disturbances in healthy older adults, J. Am. Geriatr. Soc. 48 (1) (2000) 42–50. [12] A. Shumway-Cook, et al., Predicting the probability for falls in community-dwelling older adults, Phys. Ther. 77 (8) (1997) 812–819. [13] A.N. Bandong, G.O. Madriaga, E.J. Gorgon, Reliability and validity of the Four Square Step Test in children with cerebral palsy and Down syndrome, Res. Dev. Disabil. 47 (2015) 39–47. [14] R.P. Duncan, G.M. Earhart, Four square step test performance in people with Parkinson disease, J. Neurol. Phys. Ther. 37 (1) (2013) 2–8. [15] E.N. Williams, et al., Investigation of the timed ‘up & go’ test in children, Dev. Med. Child Neurol. 47 (8) (2005) 518–524. [16] M.R. Franjoine, J.S. Gunther, M.J. Taylor, Pediatric balance scale: a modified version of the berg balance scale for the school-age child with mild to moderate motor impairment, Pediatr. Phys. Ther. 15 (2) (2003) 114–128. [17] S.H. Yi, et al., Validity of pediatric balance scales in children with spastic cerebral palsy, Neuropediatrics 43 (6) (2012) 307–313. [18] D.V. Cicchetti, Guidelines, criteria, and rules of thumb for evaluating normed and standardized assessment instruments in psychology, Psychol. Assess. 6 (4) (1994) 284.
5. Conclusion As a result of this study investigating the validity and reliability of DGI in children with hemiplegic CP, DGI was also found to be a valid and reliable measurement method such as FSST, TUG, and PBS for the balance assessment of children with independent walking hemiplegic CP. The Dynamic Gait Index, FSST, TUG and PBS were found to have discriminative ability in terms of balance for the hemiplegic CP and TD group. 32
Gait & Posture 75 (2020) 28–33
A. Evkaya, et al.
[25] D.R. Wolff, et al., Postural balance measurements for children and adolescents, J. Orthop. Res. 16 (2) (1998) 271–275. [26] A. Lubetzky-Vilnai, T.L. Jirikowic, S.W. McCoy, Investigation of the Dynamic Gait Index in children: a pilot study, Pediatr. Phys. Ther. 23 (3) (2011) 268–273. [27] S.M. Gan, et al., Psychometric properties of functional balance assessment in children with cerebral palsy, Neurorehabil. Neural Repair 22 (6) (2008) 745–753. [28] H. Carey, et al., Reliability and responsiveness of the timed up and go test in children with cerebral palsy, Pediatr. Phys. Ther. 28 (4) (2016) 401–408. [29] S. Kingsnorth, et al., Chronic pain assessment tools for cerebral palsy: a systematic review, Pediatrics 136 (4) (2015) e947–60. [30] S.L. Pavao, et al., Discriminant ability and criterion validity of the Trunk Impairment Scale for cerebral palsy, Disabil. Rehabil. 41 (18) (2019) 2199–2205. [31] H. Lim, Correlation between the selective control assessment of lower extremity and pediatric balance scale scores in children with spastic cerebral palsy, J. Phys. Ther. Sci. 27 (12) (2015) 3645–3649.
[19] T.K. Koo, M.Y. Li, A guideline of selecting and reporting intraclass correlation coefficients for reliability research, J. Chiropr. Med. 15 (2) (2016) 155–163. [20] L. Portney, M. Watkins, Foundations of Clinical Research: Application to Clinical Practice, second ed., Prentice-Hall, Upper Saddle River, New Jersey, 2000. [21] D.E. Hinkle, W. Wiersma, S.G. Jurs, Applied Statistics for the Behavioral Sciences, second ed., Houghton Mifflin, Boston, 1988. [22] M.M. Mukaka, A guide to appropriate use of correlation coefficient in medical research, Malawi Med. J. 24 (3) (2012) 69–71. [23] P. Crenna, M. Inverno, Objective detection of pathophysiological factors contributing to gait disturbance in supraspinal lesions, in: E. Fedrizzi, G. Avanzini, P. Crenna (Eds.), Motor Development in Children, John Libbey & Company Ltd., London, 1994, pp. 103–118. [24] M.J. Kurz, D.J. Arpin, B. Corr, Differences in the dynamic gait stability of children with cerebral palsy and typically developing children, Gait Posture 36 (3) (2012) 600–604.
33
Contents lists available at ScienceDirect
Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost
Full length article
Validity and reliability of the Dynamic Gait Index in children with hemiplegic cerebral palsy⋆
T
⁎
Ayca Evkayaa, , Evrim Karadag-Saygia, Duygu Karali Bingulb, Esra Giraya a Marmara University Medical School, Department of Physical Medicine and Rehabilitation, Fevzi Çakmak Mahallesi, Tepe Sokak, No: 41, Üst Kaynarca, Pendik, Istanbul, Turkey b Istanbul University Medical School, Department of Physical Medicine and Rehabilitation, Pain Medicine Division, Süleymaniye Esnaf Hastanesi Poliklinikleri, Takvimhane Caddesi, No: 19, Eminönü, Istanbul, Turkey
A R T I C LE I N FO
A B S T R A C T
Keywords: Balance Cerebral palsy Dynamic Gait Index Gait Postural control Validity and reliability
Background: Dynamic Gait Index (DGI) is a performance-based tool can be applied in a short time and evaluates dynamic balance and gait ability. Research question: Is the DGI valid and reliable for assessing gait and balance disorders in children with hemiplegic cerebral palsy (CP)? Methods: Sixteen children with hemiplegic CP (5 females, 11 males; mean age 10y 3mo, SD 2y 7mo; range 6–14y; Gross Motor Function Classification System (GMFCS) levels I [n = 9], II [n = 7]) and 16 age-matched typically developing (TD) (8 females, 8 males; mean age 9y 9mo, SD 2y 6mo; range 6–14y) participated. The relationship between the DGI, Four-Square Step Test (FSST), Timed Up and Go Test (TUG) and Pediatric Berg Balance Scale (PBS) was analyzed. To determine the test-retest reliability, the DGI was performed twice and; for the inter-rater reliability, only DGI was reapplied by a different rater on the same day. Internal consistency was obtained by Cronbach-α value. Validity was tested by Spearman correlation coefficient and reliability was calculated by Intraclass correlation coefficient (ICC). Results: There was a significant difference between hemiplegic CP and TD and between the children with GMFCS level I and II in the comparison of results of the DGI and other tests. All items on the DGI had appropriate internal consistency (Cronbach-α = 0.969). The test-retest (ICC = 0.970 CI(0.915- 0.990)) and inter-rater (ICC = 0.983 CI(0.882- 0.998)) reliabilities were found to be excellent. A negative, moderate correlation between FSST and DGI (rs = −0.673, p = 0.004); a positive, high correlation between PBS (rs = 0.724, p = 0.002) and DGI and a negative, high correlation between TUG and DGI (rs = −0.828, p < 0.001) was detected. Significance: DGI with features such as its feasibility in a short time, being simple but distinctive and not requiring heavy equipment is a valid and reliable method in children with hemiplegic CP.
1. Introduction Dynamic Gait Index (DGI) is a performance-based tool that quantifies the dynamic balance instability developed by Shumway-Cook and Woollacott, evaluates the ability of the individual to modify gait in response to changing functions during walking [1,2]. In adults, it has been shown that DGI allows comparisons between patient groups with various pathologies leading to balance problems and contains excellent psychometric properties [3–6] and the test is completed in approximately 10 min [1]. Although the incidence of balance disorders and high incidence of
falls in children with cerebral palsy (CP) are well known [7], in the literature there has been very little effort to quantify the dynamic gait stability of these children. On the contrary, most of the literature focuses on assessing the balance instability seen in these children with standing posture balance measurement methods [8,9]. The problem with this approach is; measurement methods of static stability and dynamic gait stability or predictions of falls do not show good correlation [10,11]. Therefore; standing posture balance measurement methods may not be the best measurement to assess the dynamic gait stability of children with CP. The Dynamic Gait Index is a sensitive and efficient tool for adults
⋆
The study results were presented as oral presentation at the 6th Medical Rehabilitation Congress. Corresponding author. E-mail addresses: [email protected] (A. Evkaya), [email protected] (E. Karadag-Saygi), [email protected] (D. Karali Bingul), [email protected] (E. Giray). ⁎
https://doi.org/10.1016/j.gaitpost.2019.09.024 Received 2 May 2019; Received in revised form 21 July 2019; Accepted 22 September 2019 0966-6362/ © 2019 Elsevier B.V. All rights reserved.
Gait & Posture 75 (2020) 28–33
A. Evkaya, et al.
Item 2 “When I tell you go, walk as fast as you can; when I tell you slow, walk as slowly as you can”, Item 3 “When I tell you to look right or left keep walking straight but turn your head to the right or left”, Item 4 “When I tell you to look up or down keep walking straight but turn your head to the up or down. Keep your head up or down until I tell you look straight, then keep walking straight, but return your head to the center”, Item 5 “Begin walking at your normal pace. When I tell you, turn and stop, turn as quickly as you can to face the opposite direction and stop”, Item 6 “Begin walking at your normal speed. When you come to the shoebox, step over it, not around it, and keep walking”, Item 7 “Begin walking at normal speed. When you come to the first cone, walk around the right side of it. When you come to the second cone, walk around it to the left” Item 8 “Walk up these stairs as you would at home, i.e., using the railing if necessary. At the top, turn around and walk down.” The test is completed in approximately 10 min [1]. The Four-Square Step Test evaluates the ability to change direction when performing clinical steps. The patient is asked to complete the whole sequence as soon as possible, provided that both feet touch the ground at the same time without touching the lines. The test is completed in approximately 5 min [13,14]. The Timed Up and Go is a test used to measure mobility and evaluates walking speed, posture control, functional mobility and balance. The test is completed in approximately 10 min [15]. The Pediatric Berg Balance Scale is a revised version of the Berg Balance Scale (BBS) for children by Franjoine et al. in order to assess the functions of daily life activities. The scale consists of 14 evaluations and each section is scored between 0–4; the highest score is 56. In PBS; the order of the sections in the standard BBS, is rearranged to be easy to difficult as functional sorting form, the time standards in the sections related to protection of static postures have been reduced to the pediatric population and the guidance has been simplified. It can be administered in approximately 20 min [16,17].
[1,3,5,6] and therefore may be a useful tool for children. The objective investigation of the movement is important to determine the appropriate treatment methods and to understand the effectiveness of these methods for children with CP. The DGI; although it has been widely used in postural control studies in adults [3,5,6,12], it has not yet been investigated to assess walking difficulties in children with CP. Postural control and dynamic balance are important in children, so these structures need precise and reliable measurements. Therefore, the aim of this study was to evaluate the reliability and validity of DGI in children with cerebral palsy. 2. Methods 2.1. Participants Potential participants consisted of 16 children with hemiplegic CP whose level of Gross Motor Function Classification System (GMFCS) I and II, ages 6 through 14 years, presented to the Pediatric Rehabilitation Special clinic of Department of Physical Medicine and Rehabilitation at the Marmara University Pendik Education and Research Hospital between April 2018 and June 2018. In addition, 16 age-gender matched typically developing (TD) children were included in the study as a control group The GPower V.3.1.9.3 program was used to determine appropriate sample size. When the average expected value in the patient group was 21.4 (with a standard deviation of 1.4) and the average expected value in the typically developing group was 23.1 (with a standard deviation of 1.4) based on the study results of Lubetzky-Vilnai et al. [13], it was found that 14 subjects for each group must had been enrolled to have 95% power with 5% type 1 error level. The inclusion criteria were: (1) being classified at GMFCS level I–II; (2) having a spasticity value 1, 2 or maximum 3 of the lower extremity muscles according to Modified Ashworth Scale (MAS); (3) being able to stand and walk without orthosis and any auxiliary equipment; and (4) being able to understand and execute instructions for assessment. The exclusion criteria were: (1) having undergone any intervention (e.g. botulinum toxin injection or orthopedic surgery) that may affect gait between the test and re-test assessment; (2) having visual impairments excluding refractive defects; (3) having any accompanying systemic disorder; (4) having uncontrolled epilepsy; (5) having the lower extremity contracture that affect the assessment; and (6) having the emergence of health problems that may affect the study. The study was conducted after obtaining approval from Marmara University Health Sciences Institute Ethics Committee in accordance with the Declaration of Helsinki (02.04.2018-113) and after the registry of Clinical Trials Registry. The trial is registered with ClinicalTrials.gov (NCT03662139). Informed consent was obtained from all legal guardians of the participating children. Our study is an observational methodological research study that evaluates validity and reliability.
2.3. Validity and reliability protocol Data collection was conducted in a quiet room where children can feel comfortable to minimize distractions. Some modifications for children in measurements were performed; these modifications included verbal redirects and repetitions. Some test items in which verbal expression caused confusion were taught to the children by showing movement. The area of assessments was a solid surface, free from any physical obstacles and has a large place with enough lighting. Participants wore daily comfortable clothes and shoes. Two raters evaluated participants’ performance. To examine testretest reliability the DGI was administered to each participant in the CP group by rater 1 in 2 testing sessions 7 days apart. To evaluate the interrater reliability the DGI was rated simultaneously by rater 2 in the latter testing session during the same day. Two raters were not allowed to consult each other during the test and had no access to previous test results. The FSST, TUG, and PBS were administered and rated by rater 1 immediately after the first session to test the concurrent validity of the DGI. The patient assessment form was accomplished, and demographic characteristics were collected at the first session. The whole testing protocol in the first session took approximately 45 min. Participants were rested for five minutes between the two evaluation to minimize fatigue. In the second session, only the DGI was performed. The participants in the CP group were determined by the polyclinic patients who met the required criteria. The control group was selected from TD healthy children who accompanied their families to hospital.
2.2. Measurements The Modified Ashworth Scale was used to evaluate spasticity, GMFCS was applied for functional level assessment. The Dynamic Gait Index, Four Square Step Test (FSST), Timed Up and Go (TUG) and Pediatric Berg Balance Scale (PBS) were applied for gait and balance assessment. The Dynamic Gait Index measures functionality and dynamic balance in walking and stair climbing. The DGI consists of 8 items including normal gait on flat ground, gait with speed changes, gait with horizontal head movements, gait with vertical head movements, gait and pivot turn, gait and step over obstacle, gait around obstacles and steps up and down stairs. The performance on each item is rated on a 4point scale (3, independent walking; 2, mild impairment; 1, moderate impairment; 0, severe disorder) with a maximum score of 24. Each item is indicated to children as follows: Item 1 “Walk at your normal speed”,
2.4. Statistical analysis "SPSS 20.0 Statistical package" program was used to evaluate the data of the groups. The data was presented with tables and graphs. Statistical significance was accepted as p < 0.05. Descriptive characteristics are provided as number-percent, mean, 29
Gait & Posture 75 (2020) 28–33
A. Evkaya, et al.
standard deviation, median, minimum-maximum values. Differences in these characteristics were tested using “Mann Whitney U" test and “Chisquare” test for continuous and categoric variables, respectively. The comparison of the applied methods for the patient and TD group was performed with the "Mann Whitney U" test. The internal consistency of DGI items was analyzed by the "Cronbach-α coefficient". To evaluate the test-retest and inter-rater reliability analysis of the DGI, Intraclass Correlation Coefficient (ICC) and 95% confidence intervals (CI) with the two-way random-effect model were calculated. Reliability estimates were interpreted as follows: > 0.90=excellent; 0.75-0.90=good; 0.50-0.75=medium; < 0.50=low [18–20]. The correlation of balance assessment methods was tested using the "Spearman correlation coefficient". The level of relationship was classified as follows < 0.30=small/negligible; 0.30–0.50=low; 0.50–0.70=moderate; 0.70-0.90=high; > 0.90=very high [21,22].
Table 2 Comparison of the DGI Scores in Hemiplegic CP and TD.
DGI DGI DGI DGI DGI DGI DGI DGI DGI
1 2 3 4 5 6 7 8 Total
Hemiplegic CP (n = 16) Median (Min–Max)
TD (n = 16) Median (Min–Max)
p
2 (1–3) 2.50 (1–3) 2 (0–3) 2 (0–3) 2 (2–3) 2 (0–3) 3 (1–3) 2.50 (1–3) 19 (8-23)
3 (3–3) 3 (3–3) 3 (3–3) 3 (3–3) 3 (2–3) 3 (3–3) 3 (3–3) 3 (3–3) 24 (23–24)
p < 0,001 0.001 0.001 0.001 0.005 0.001 0.008 0.001 p < 0.001
CP: Cerebral palsy, TD: Typically developing, DGI: Dynamic Gait Index, Min: Minimum, Max: Maximum, p: Statistical significance. *Mann-Whitney U test. Table 3 Test–retest Reliability of the DGI scores in Hemiplegic CP.
3. Results 3.1. Participants Sixteen patients; 5 girls (31.3%), 11 males (68.8%) with a mean age of 10.31 ± 2.62 years and 16 TD children; 8 girls (50%), 8 males (50%) with a mean age of 9.75 ± 2.51 years were included in this study and their characteristics are presented in Table 1. Most children were at level I (n = 9), which was the highest level for GMFCS. 56.3% (n = 9) of the children had right extremity and 43.8% (n = 7) had left extremity effected in the hemiplegic CP group. 3.2. DGI total scores
(n = 16)
1st assessment Median ((Min–Max)
2nd assessment Median(Min–Max)
ICC*
95% CI
DGI DGI DGI DGI DGI DGI DGI DGI DGI
2 (1–3) 2.50 (1–3) 2 (0–3) 2 (0–3) 2 (2–3) 2 (0–3) 3 (1–3) 2.50 (1–3) 19 (8–23)
2 (1–3) 2.50 (1–3) 2.50 (1–3) 2.50 (1-3) 2 (2–3) 2 (0–3) 3 (1–3) 2.50 (1–3) 18.50 (10–23)
0.805 0.913 0.861 0.889 0.906 0.965 0.913 0.972 0.970
0.429–0.933 0.749–0.970 0.613–0.951 0.690–0.961 0.737–0.967 0.898–0.988 0.749–0.970 0.922–0.990 0.915–0.990
1 2 3 4 5 6 7 8 Total
CP: Cerebral palsy, DGI: Dynamic Gait Index, Min: Minimum, Max: Maximum, ICC: Intraclass correlation coefficient, CI: Confidence interval.
The DGI total score was significantly lower for the 16 children with hemiplegic CP compared with 16 TD (p < 0.001). Children with CP demonstrated a median DGI score of 19 (range, 8–23) while heathy children had a median score of 23 (range, 23–24). As a result of DGI assessments, the items with the lowest scores in children with hemiplegic CP were identified as gait with horizontal and vertical head movements (items 3 and 4) and gait and step over obstacle (item 6). The item with the best scores was identified as taking gait around obstacles (item 7) (Table 2).
Table 4 Inter-rater Reliability of the DGI Scores in Hemiplegic CP.
3.3. Reliability The internal consistency of the DGI was high with a Cronbach's alpha coefficient of 0.969. The test-retest reliability of DGI was excellent (ICC = 0.970 Cl (0.915-0990)). Inter-rater reliability of DGI was excellent (ICC = 0.983 Cl(0.8820990)). The ICC values for each item was shown in Tables 3 and 4. The Bland-Altman plots for test-retest compliance and inter-rater compliance are shown in Fig. 1.
(n = 6)
1st assessment Median ((Min- Max)
2nd assessment Median (Min- Max)
ICC*
95% CI
DGI DGI DGI DGI DGI DGI DGI DGI DGI
2 (1–2) 2.50 (2–3) 2 (2–3) 2 (1–3) 2 (2–3) 2 (1–3) 3 (2–3) 3 (2–3) 18 (13–23)
2 (1–2) 2.50 (2–3) 2 (2–3) 2 (1–3) 2.50 (2–3) 2.50 (1–3) 3 (2–3) 3 (2–3) 18.50 (13–23)
1.00* 1.00* 1.00* 0.906 0.828 0.928 1.00* 1.00* 0.983
0.434–0.986 −0.038- 0.975 0.566–0.990
1 2 3 4 5 6 7 8 Total
0.882–0.998
CP: Cerebral palsy, DGI: Dynamic Gait Index, Min: Minimum, Max: Maximum, ICC: Intraclass correlation coefficient, CI: Confidence interval. * ICC = 1.00 due to 100% compliance between non-variability items and evaluators.
3.4. Validity Table 1 Comparison of Characteristics Data in Hemiplegic CP and TD.
Age (year) Height (cm) Body weight (kg) BMI (kg/m2) Gestational week Birth weight (kg)
Hemiplegic CP (n = 16) Mean ± SD
TD (n = 16) Mean ± SD
p
10.31 ± 2.62 142 ± 15 38.43 ± 10.85 18.56 ± 3.12 35.12 ± 4.22 2.52 ± 0.78
9.75 ± 2.51 140 ± 13 38.87 ± 11.63 19.40 ± 3.23 37.25 ± 2.38 3.03 ± 0.67
.519* .451* .940* .559* .155* .054*
In order to evaluate the validity of the DGI, the study used FSST, TUG and PBS scales were used. A negative, moderate correlation with FSST (rs = −0.673, p = 0.004); a positive, high correlation with PBS (rs = 0.724, p = 0.002) and a negative, high correlation with TUG (rs = −0.828, p < 0.001) were found (Table 5). 3.5. Comparison of the test scores according to GMFCS Levels The comparative values of the GMFCS levels and the gait and balance tests of the hemiplegic CP included in the study were examined. When the test scores were compared according to the GMFCS levels
CP: Cerebral palsy, TD: Typically developing, BMI: Body Mass Index, SD: Standard deviation, p: Statistical significance. * Mann-Whitney U test. 30
Gait & Posture 75 (2020) 28–33
A. Evkaya, et al.
4. Discussion There are many studies in the literature focusing on balance and postural control in children with CP [23–25]. In our study, we found that high levels of significant difference (p < 0.001) in favor of the TD in terms of DGI, FSST, TUG, and PBS results for gait and balance assessment. This indicates that children with hemiplegic CP have poor balance and postural control and gait disorders associated with these problems. Lubetzky-Vilnai et al., investigated the DGI in children in their study, 10 children aged between 8–15 years who were diagnosed with FASD (Foetal Alcohol Spectrum Disorder) who had postural control disorder and 10 age-gender matched TD were compared and DGI total score in children with FASD diagnosis was found to be significantly lower than TD (p = 0.01). The mean of the patients with FASD was 21.4 ± 1.4 and the mean of the TD 23.3 ± 1.06 [26]. Although these results are consistent with our study; it can be said that the mean values of the patient group were lower compared to the other study because of the difference in diagnosis and accompanying disorders in our study. The test-retest reliability was excellent for DGI total score (ICC = 0.970 Cl(0.915-0.990)) in our study. The Cronbach alpha coefficient for the internal consistency of the test was found to be very high with 0.969. The inter-rater reliability was excellent for DGI total score (ICC = 0.983 Cl(0.882-0.998)). Eight different DGI items were found to be good-excellent reliable (ICC = 0.805-0.972) and normal gait (item 1), gait with horizontal and vertical head movements (items 3 and 4) are good; other items had excellent reliability for test-retest reliability. Eight different DGI items were found to be good-fully compliant (ICC = 0.828–1.000) and normal gait (item 1), gait with speed changes (item 2), gait with horizontal head movements (item 3), gait around obstacles (item 7) and steps up and down stairs (item 8) were fully compliant items that received the same score from the two raters; the gait and pivot turn (item 5) was good, and the other two items had excellent reliability for inter-rater reliability. Lubetzky-Vilnai et al. the test-retest reliability was found to be medium (ICC = 0.710 Cl(0.260-0.910, p = 0.005) reliable and interrater reliability was found to be good (ICC = 0.820 Cl(0.490-0.940), p = 0.001) reliable [26]. When the results were compared, our study had higher test-retest and inter-rater reliability results. With these findings, it can be said that the tool can be used in children, but there is still a need in the literature for different studies investigating the usability of DGI in children. The relationship of DGI with other clinical tests is examined, a negative, moderate correlation with FSST (rs = −0.673, p = 0.004); a positive, high correlation with PBS (rs = 0.724, p = 0.002) and a negative, high correlation with TUG (rs = −0.828, p < 0.001) were found. In the correlations between the other clinical tests, a positive, high correlation between FSST and TUG (rs = 0.844, p < 0.001), a negative, high correlation between FSST and PBS (rs = −0.797, p < 0.001), and a negative, high correlation between TUG and PBS (rs = −0.763, p = 0.001) were found. When we look at the literature, the number of studies on children who study the relationship between the balance assessment methods used in our study is very small. Gan et al. found a strong negative relationship between PBS and TUG [27]. Bandong et al. examined the validity and reliability of the FSST which included children 5–12 years of age CP, DS. They analyzed the correlation between FSST and TUG and stated that the FSST was a valid and reliable test in both the CP and the DS [13]. These studies show parallels with our results. In the literature, we did not find a study that looks at the correlation between the tests in which all the clinical tests we use in our study are used together or in groups. More studies are needed to investigate the correlation with each other of balance assessment methods on children. Independent verticalization and mobilization are required to be able
Fig. 1. The Bland-Altman plots for test-retest and inter-rater compliance. *points where two circles overlap.
Table 5 Correlation analysis of the DGI total score with the FSST duration, the TUG duration and the PBS total score in Hemiplegic CP. (n = 16) DGI
rs* p*
FSST
TUG
PBS
−.673** .004
−.828** .000
.724** .002
CP: Cerebral palsy, DGI: Dynamic Gait Index, FSST: Four Square Step Test, TUG: Timed Up and Go, PBS: Pediatric Berg Balance Scale, rs: Correlation coefficient., p: Statistical significance. * Spearman correlation coefficient.
of the CP group; no significant difference was found except gait and step over obstacle (item 6), steps up and down stairs (item 8) and total score of DGI. There was a high level of difference between GMFCS I and GMFCS II children for FSST (p = 0.001) and TUG (p = 0.007) scores. There was no significant difference between GMFCS I and GMFCS II children for stand on one foot (item 9), stepping (item 13) and functional reach (item 14) items. A high level of significant difference was found in favour of GMFCS level I children (p = 0.005) for PBS total scores. Stand-sit (item 2), standing balance (item 4) and sitting balance (item 5) items in PBS were all items for which all children received full points.
31
Gait & Posture 75 (2020) 28–33
A. Evkaya, et al.
Also, all tests were found to have discriminative ability in terms of balance for CP patients with GMFCS levels I and II. The data obtained from the result of the study; for physiotherapists working in the pediatric area, with features such as easy and fast administration, being simple but distinctive and do not containing heavy equipment; DGI has been shown to be a valid and reliable method in independent walking hemiplegic CP patients.
to evaluate dynamic and static balance. According to GMFCS levels, the levels containing these characteristics are defined as I and II. There are studies in which children with GMFCS level I and II are included and compared to assess the balance in the literature. In our study, which included children with GMFCS level I and II, there was a significant difference between the two levels for the total score of the DGI (p = 0.21). The items with significant differences between the two levels gait and step over obstacle (item 6) steps up and down stairs (item 8). When we think that the level of distinctive functionality for GMFCS levels I and II are the use of a railing to walk up and down stairs; we can say that it was an expected result. No study evaluating DGI in relation to different GMFCS levels could be found in the literature. In the comparison of two GMFCS levels for FSST, there was a significant difference between level I and II (p = 0.001). Similar studies were not found in the literature. Therefore, in our study, we thought that we could contribute to the literature by comparing the results of FSST according to the GMFCS levels. For TUG, there was a high level of significant difference between children with GMFCS level I and II (p = 0.007). Carey et al. found similar results for GMFCS I and II (p = 0.000), GMFCS I and III (p = 0.000), and GMFCS II and III (p = 0.003) [28]. Gross Motor Function Classification System GMFCS is the recommended lower extremity classification by SCPE [29,30]. Also, Winters and Gage classification system is specific to hemiplegic cerebral palsy which reflects the gait abnormalities in detail. It has four categories. We could not investigate the relationship; because we lack enough children with hemiplegic cerebral palsy in each category. The relationship between Winters and Gage classification categories and DGI should be Winters and Gage classification enlightened in further studies. When the children with GMFCS levels I and II were compared for PBS, the balance was significantly better in favor of level I patients (p = 0.005). In terms of PBS items, there was no significant difference except stand on one foot (item 9), stepping (item 13) and functional reach (item 14). In two similar studies evaluating the difference between GMFCS levels with PBS, similar results were observed in terms of items and total score [27,31]. Not performing sensitivity analysis of DGI can be viewed as the limitation of the study. We could not analyze sensitivity of DGI because the tools which we used to validate DGI do not have exact cut-off values that differentiate children with impaired and intact balance. Sensitivity of DGI should be investigated in further studies. The inclusion of a higher number of children in future studies may lead to better analyzes. In addition, there is a need for more comprehensive studies in containing children with CP in different levels of GMFCS and consisting of different clinical types such as spastic diplegic, dyskinetic, ataxic type CP. We have encountered a single study in the literature on the use of the DGI tool in children. In future studies, it may be useful to investigate the use of DGI in children and to determine the cut-off values in different age groups. The number of studies in which balance assessment methods are associated with GMFCS levels of children with CP is low. Looking forward, the relationship between the results of the balance assessment methods and the GMFCS levels can be investigated.
Funding source There is no funding for this study. Clinical trial registration number The study was conducted after approval from the Ethics Committee of Marmara University Health Sciences Institute in accordance with the Declaration of Helsinki (approval number: 02.04.2018-113). The trial is registered with ClinicalTrials.gov, number NCT03662139. Declaration of Competing Interest The authors declare no conflict of interest. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Acknowledgements We wish to thank Anne Shumway-Cook for giving us permission to use the DGI. We also thank all the parents and children who participated in this study. References [1] A. Shumway-Cook, M. Woollacott, Motor Control: Theory and Practical Applications, first ed., Williams and Wilkins, Baltimore, 1995. [2] T. Herman, et al., The Dynamic Gait Index in healthy older adults: the role of stair climbing, fear of falling and gender, Gait Posture 29 (2) (2009) 237–241. [3] D. Cattaneo, J. Jonsdottir, S. Repetti, Reliability of four scales on balance disorders in persons with multiple sclerosis, Disabil. Rehabil. 29 (24) (2007) 1920–1925. [4] Y.-P. Chiu, et al., Use of item response analysis to investigate measurement properties and clinical validity of data for the dynamic gait index, Phys. Ther. 86 (6) (2006) 778–787. [5] C.D. Hall, S.J. Herdman, Reliability of clinical measures used to with peripheral vestibular disorders, J. Neurol. Phys. Ther. 30 (2) (2006) 74–81. [6] J. Jonsdottir, D. Cattaneo, Reliability and validity of the dynamic gait index in persons with chronic stroke, Arch. Phys. Med. Rehabil. 88 (11) (2007) 1410–1415. [7] M. Bottos, C. Gericke, Ambulatory capacity in cerebral palsy: prognostic criteria and consequences for intervention, Dev. Med. Child Neurol. 45 (11) (2003) 786–790. [8] P.A. Burtner, et al., The capacity to adapt to changing balance threats: a comparison of children with cerebral palsy and typically developing children, Dev. Neurorehabil. 10 (3) (2007) 249–260. [9] J. Chen, M.H. Woollacott, Lower extremity kinetics for balance control in children with cerebral palsy, J. Mot. Behav. 39 (4) (2007) 306–316. [10] H.G. Kang, J.B. Dingwell, A direct comparison of local dynamic stability during unperturbed standing and walking, Exp. Brain Res. 172 (1) (2006) 35–48. [11] T.M. Owings, et al., Measures of postural stability are not predictors of recovery from large postural disturbances in healthy older adults, J. Am. Geriatr. Soc. 48 (1) (2000) 42–50. [12] A. Shumway-Cook, et al., Predicting the probability for falls in community-dwelling older adults, Phys. Ther. 77 (8) (1997) 812–819. [13] A.N. Bandong, G.O. Madriaga, E.J. Gorgon, Reliability and validity of the Four Square Step Test in children with cerebral palsy and Down syndrome, Res. Dev. Disabil. 47 (2015) 39–47. [14] R.P. Duncan, G.M. Earhart, Four square step test performance in people with Parkinson disease, J. Neurol. Phys. Ther. 37 (1) (2013) 2–8. [15] E.N. Williams, et al., Investigation of the timed ‘up & go’ test in children, Dev. Med. Child Neurol. 47 (8) (2005) 518–524. [16] M.R. Franjoine, J.S. Gunther, M.J. Taylor, Pediatric balance scale: a modified version of the berg balance scale for the school-age child with mild to moderate motor impairment, Pediatr. Phys. Ther. 15 (2) (2003) 114–128. [17] S.H. Yi, et al., Validity of pediatric balance scales in children with spastic cerebral palsy, Neuropediatrics 43 (6) (2012) 307–313. [18] D.V. Cicchetti, Guidelines, criteria, and rules of thumb for evaluating normed and standardized assessment instruments in psychology, Psychol. Assess. 6 (4) (1994) 284.
5. Conclusion As a result of this study investigating the validity and reliability of DGI in children with hemiplegic CP, DGI was also found to be a valid and reliable measurement method such as FSST, TUG, and PBS for the balance assessment of children with independent walking hemiplegic CP. The Dynamic Gait Index, FSST, TUG and PBS were found to have discriminative ability in terms of balance for the hemiplegic CP and TD group. 32
Gait & Posture 75 (2020) 28–33
A. Evkaya, et al.
[25] D.R. Wolff, et al., Postural balance measurements for children and adolescents, J. Orthop. Res. 16 (2) (1998) 271–275. [26] A. Lubetzky-Vilnai, T.L. Jirikowic, S.W. McCoy, Investigation of the Dynamic Gait Index in children: a pilot study, Pediatr. Phys. Ther. 23 (3) (2011) 268–273. [27] S.M. Gan, et al., Psychometric properties of functional balance assessment in children with cerebral palsy, Neurorehabil. Neural Repair 22 (6) (2008) 745–753. [28] H. Carey, et al., Reliability and responsiveness of the timed up and go test in children with cerebral palsy, Pediatr. Phys. Ther. 28 (4) (2016) 401–408. [29] S. Kingsnorth, et al., Chronic pain assessment tools for cerebral palsy: a systematic review, Pediatrics 136 (4) (2015) e947–60. [30] S.L. Pavao, et al., Discriminant ability and criterion validity of the Trunk Impairment Scale for cerebral palsy, Disabil. Rehabil. 41 (18) (2019) 2199–2205. [31] H. Lim, Correlation between the selective control assessment of lower extremity and pediatric balance scale scores in children with spastic cerebral palsy, J. Phys. Ther. Sci. 27 (12) (2015) 3645–3649.
[19] T.K. Koo, M.Y. Li, A guideline of selecting and reporting intraclass correlation coefficients for reliability research, J. Chiropr. Med. 15 (2) (2016) 155–163. [20] L. Portney, M. Watkins, Foundations of Clinical Research: Application to Clinical Practice, second ed., Prentice-Hall, Upper Saddle River, New Jersey, 2000. [21] D.E. Hinkle, W. Wiersma, S.G. Jurs, Applied Statistics for the Behavioral Sciences, second ed., Houghton Mifflin, Boston, 1988. [22] M.M. Mukaka, A guide to appropriate use of correlation coefficient in medical research, Malawi Med. J. 24 (3) (2012) 69–71. [23] P. Crenna, M. Inverno, Objective detection of pathophysiological factors contributing to gait disturbance in supraspinal lesions, in: E. Fedrizzi, G. Avanzini, P. Crenna (Eds.), Motor Development in Children, John Libbey & Company Ltd., London, 1994, pp. 103–118. [24] M.J. Kurz, D.J. Arpin, B. Corr, Differences in the dynamic gait stability of children with cerebral palsy and typically developing children, Gait Posture 36 (3) (2012) 600–604.
33