EVALUATION OF HUMAN PERFORMANCE
THE STAR EXCURSION BALANCE TEST
Sarah Crymble, Caroline Glennie, Mary Leech, Susan Mullen, Cormac Ryan, Nicola Wallace
Star Excursion Balance Test
Introduction
Balance is an integral component of almost every activity of daily living (ADL). A decrement in balance can result from musculoskeletal injury, head trauma, disease or ageing (Kinsey et al. 1998) (Olmsted et al. 2002). As a result of balance decrements/inadequacies, everyday functioning may be impaired. Thus balance is of key clinical relevance to any rehabilitation/prophylactic physiotherapy program.
There are many outcome measures, which assess balance e.g. stabliometer, which are both valid and reliable (Era et al. 1997) (Ageberg et al. 1998). However such measures assess static balance, whereas many ADL’s are classified as dynamic activities because they cause the centre of gravity to move in response to muscular activity (Kinzey et al. 1998). The majority of dynamic balance assessment tools e.g. functional reach tests and the berg balance scale, were developed specifically for paediatrics (Donahoe et al. 1994), geriatrics (Berg et al. 1995) and neurological patients (Weiner et al. 1993). There exist few practical tools for measuring dynamic balance in adult, non-neurologically impaired patients. Thus there is an urgent need for a valid, reliable, and practical measure of dynamic balance.
Recently, the Star Excursion Balance Test (SEBT) has demonstrated potential in this area (Hertel et al. 2000). However, much work needs to be done on the validity and reliability of this measure before it should be used clinically. The purpose of this study is to assess the intra- and inter-day reliability of the SEBT. It is important to perform this investigation, because clinically assessment of balance will occur prior to rehabilitation and post rehabilitation a number of days after the original measure. In order for any change in SEBT score to be attributable to an improvement in balance the test must be shown to be reliable. Additionally, the SEBT does not produce a single measure of balance but a number of values, which are assessed individually, leading to possible confusion over interpretation of results. This study will also attempt to combine the individual scores of the SEBT to give one overall measure of balance.
Aims of study
1. To assess the intra- and inter- day reliability of the SEBT.
2. To combine the individual scores obtained from the SEBT, so as one single measure of balance is produced.
Scientific Basis of the Techniques of Measurement
There has been little research done on the use of the SEBT. A total of four articles and one abstract have been published in peer review journals indicating the need for research on this measurement tool.
The test will be described in detail in the methodology; however, briefly, the patient stands at the centre of a grid with 8 lines extending from it. On one leg, the patient reaches as far as possible with the other leg. Theoretically, the further the reach the better the dynamic balance.
Hertel et al. 2000 investigated intra- and inter-tester reliability of the SEBT on two separate days in 16 healthy participants. Participants performed 2 bouts of the 8 directions on each leg on two separate trial days 1 week apart. Two examiners were used and a different examiner assessed each trial on the same day. Intra-tester reliability ranged from 0.78-0.96 (ICC) on day 1 and 0.82-0.96 (ICC) on day 2. Inter-tester reliability ranged from 0.35-0.84 (ICC) on day 1 and 0.81-0.93 (ICC) on day 2 (See Tables 1 & 2). According to Portney and Watkins et al. (1993) an ICC of 0.75, is considered good reliability, but it should exceed 0.90 in order to be able to evaluate individual patients longitudinally. Based on these results only some directions can be used reliably questioning the overall reliability of the measure and illustrating the need for a universal measurement to be gained from this test so that reliability of the test as a whole can be assessed. Significant learning effects (P<0.05) were found for 4 of the 8 directions of reach, the lateral (L), postero-lateral (PL), posterior (P) and postero-medial (PM) directions. Possibly because visual feedback is impaired for these directions thus there is more reliance on the vestibular and somatosensory systems. The authors suggested a minimum of 6 practice trials in each direction should be performed before recording any baseline data for clinical trials. This is in contrast to earlier studies by Kinzey and Armstrong, which estimated that 6 practice sessions of 5 trials in each direction would be necessary, to increase intra-tester reliability to levels above 0.86, suggesting that large volumes of practice should be performed to eradicate any learning effects. It should be noted however that the form of SEBT used in the Kinzey and Armstrong (1998) paper was a four directional test, as opposed to an eight directional test used in the Hertel et al. (2000) paper and in the current study. Thus the results of the Hertel et al. (2000) paper are more applicable to the current study, and the rest of the literature as all other research literature uses the eight directional SEBT.
Table 1: Intratester reliability for examiners 1 and 2 on days 1 and 2*Direction / Examiner 1, Day 1 / Examiner 2, Day 1 / Examiner 1, Day 2 / Examiner 2, Day 2
Anterolateral
Right / .95 (1.84) / .92 (2.53) / .95 (1.78) / .95 (2.09)
Left / .88 (2.53) / .90 (2.91) / .93 (2.27) / .91 (2.44)
Anterior
Right / .95 (1.77) / .94 (2.53) / .93 (2.18) / .95 (1.97)
Left / .91 (2.15) / .89 (3.31) / .91 (2.25) / .95 (2.10)
Anteromedial
Right / .91 (2.33) / .93 (2.53) / .91 (2.34) / .95 (1.98)
Left / .95 (1.60) / .90 (3.61) / .94 (1.88) / .92 (2.67)
Medial
Right / .96 (1.92) / .94 (2.07) / .93 (2.32) / .95 (2.12)
Left / .92 (2.40) / .94 (2.06) / .93 (2.32) / .96 (1.93)
Posteromedial
Right / .92 (2.64) / .93 (1.88) / .90 (2.38) / .96 (1.93)
Left / .89 (2.31) / .88 (2.60) / .96 (1.70) / .94 (2.28)
Posterior
Right / .93 (2.57) / .89 (2.92) / .91 (3.22) / .92 (3.24)
Left / .91 (2.40) / .83 (3.35) / .93 (2.13) / .95 (2.26)
Posterolateral
Right / .89 (2.83) / .92 (2.84) / .94 (2.66) / .96 (2.15)
Left / .86 (3.21) / .85 (2.49) / .85 (3.03) / .90 (2.76)
Lateral
Right / .84 (3.18) / .93 (2.29) / .95 (1.83) / .82 (3.28)
Left / .78 (3.38) / .85 (2.49) / .87 (2.44) / .87 (2.33)
Table 1: *Intraclass correlation coefficients are listed first; SEMs are in cm and listed in parentheses (Table taken from Hertel et al 2000).
Table 2: Intertester Reliability Estimates for Days 1 and 2*Direction / Day 1 / Day 2
Anterolateral
Right / .80 (3.52) / .93 (2.27)
Left / .78 (3.66) / .86 (2.96)
Anterior
Right / .78 (3.99) / .88 (2.87)
Left / .76 (3.95) / .89 (2.75)
Anteromedial
Right / .84 (3.40) / .87 (2.90)
Left / .76 (3.92) / .89 (2.78)
Medial
Right / .69 (4.41) / .88 (2.80)
Left / .69 (4.15) / .93 (2.33)
Posteromedial
Right / .78 (3.57) / .89 (2.80)
Left / .80 (3.08) / .93 (2.33)
Posterior
Right / .66 (4.72) / .88 (2.80)
Left / .76 (3.68) / .91 (2.69)
Posterolateral
Right / .61 (4.96) / .86 (3.87)
Left / .58 (4.16) / .88 (2.83
Lateral
Right / .35 (4.73) / .81 (3.31)
Left / .53 (4.46) / .85 (2.77)
Table 2: *Intraclass correlation coefficients are listed first; SEMs are in cm and listed in parentheses (Table taken from Hertel et al 2000).
The validity of the SEBT has never been assessed directly. However there is some literature to support its validity (Hertel et al. 2000, Olmsted et al. 2002). In order for maximal performance to occur, firstly the centre of gravity (COG) must be maintained within the limits of the base of support, and secondly eccentric and isometric neuromuscular control of the joints of the stance leg must be intact. Hertel et al. (2000) suggests that these two paradigms give the SEBT excellent logical/construct validity. Assessing criterion validity is a much more difficult task because no ‘gold standard’ measure exists for assessing dynamic balance. However the SEBT has been shown to demonstrate predictive validity (Olmsted et al. 2002). Olmsted et al. (2002) investigated if the SEBT could detect balance deficits in patients with chronic ankle stability (n = 20 injured and 20 uninjured). The group with chronic ankle instability had a significantly decreased reach while standing on the injured ankle when compared to the matched limb of the uninjured group. Also the patients with chronic ankle instability reached significantly less when standing on their injured limbs as compared with their uninjured limbs (P<0.05) (Olmsted et al 2002)
In conclusion, there is a paucity of scientific literature on the SEBT. The limited evidence has demonstrated variable inter- and intra-rater reliability as well as both construct and predictive validity (Hertel et al. 2000, Olmsted et al. 2002). However there is no literature on the inter day reliability of the SEBT. Additionally the fact that the SEBT does not give a single measure of balance makes interpretation of results difficult for both the scientist and the clinician. Based on the literature available on the SEBT, the inter day reliability of this measure needs to be assessed, and the various direction results need to be combined to give a single result for balance to decrease confusion when interpreting the results of the SEBT. These points highlight the relevance of the current study for the scientist and the clinician.
Methodology
Subjects
9 healthy, young adults (7 female, 2 male) aged 25 ± 2.87 (mean ± st.dev) participated in this study. All participants met the inclusion criteria (see Figure 1).
· No Cerebral Concussions
· No Vestibular Disorders
· No ankle injury sustained in the past 2 years (minor or major)
· No prior balance training
· No upper respiratory or ear infections
Figure 1: Inclusion criteria for participating in the current study.
All participants prior to the study signed a voluntary consent form that was approved by the University’s review board, which included an explanation of the study and any possible risks or discomfort the subject may encounter.
SEBT Test Description
The Star Excursion Balance Test (SEBT) is a functional test that incorporates a single-leg stance on one leg (e.g. right leg) whilst trying to reach as far as possible with the opposite leg (e.g. left leg). The participants stand in a square at the centre of the grid with 8 lines extending from the centre at 45° increments (see Figure 2).
Figure 2: The Star Excursion Balance Test Layout Plan (SEBT)
Each of the 8 lines extending represent the individual directions which each subject are required to reach out with the most distal part of their reach foot. The eight directions consist of antero-lateral (AL), anterior (A), antero-medial (AM), medial (M), postero-medial (PM), posterior (P), postero-lateral (PL) and lateral (L). A standard tape measure (cm) was used to quantify the distance the subject had reached from the centre of the grid (see Figure 2) to the point that the subject managed to reach along each diagonal line. Set guidelines for each trial were adhered to (Figure 3).
Trials were discarded if the examiner felt that
(1) The subject lifted the stance foot from the centre of the grid
(2) Subject lost his/her balance
(3) Subject did not touch the line with the reach foot while continuing to fully weight bear on the stance leg.
Figure 3: Guidelines for each trial
Protocol
The test was explained to each subject verbally, allowing the subject to ask any questions regarding the test. The test was performed with the subject maintaining a single-leg stance while reaching with the contra lateral leg i.e. the reach leg. The aim is to reach as far along the 8 directions as possible to touch the furthest point on the line as lightly as possible so as to avoid using the reach leg for support. The subject then returned to the centre of the grid on both feet whilst maintaining balance. Each subject performed 3 circuits of the SEBT. Each circuit consisted of 3 reaches (trials) in each of the 8 directions. Subjects were given a 5 second rest between each reach. The protocol was repeated 1 week after the initial test day. The laboratory was quiet at the time of testing to minimise distraction.
Statistical Analysis
The Shrout and Fleiss (2,1) model was used to calculate the intra-circuit reliability whilst the (2,k) model was used to calculate intra-circuit reliability (Portney and Watkins 2000). Reliability was assessed using interclass correlation coefficients (ICCs). An ICC value above 0.75 was regarded as good whilst a value greater than 0.9 was regarded as excellent reliability.
Results and Discussion
Part 1: Within circuit reliability for each of the 8 directions for all 6 circuits
Circuit No. / Direction of reachA / AL / L / M / AM / PL / P / PM
DAY 1
C1
/ 0.75 / 0.95 / 0.96 / 0.62 / 0.62 / 0.94 / 0.90 / 0.83C2 / 0.95 / 0.90 / 0.91 / 0.67 / 0.74 / 0.95 / 0.93 / 0.94
C3 / 0.95 / 0.89 / 0.90 / 0.86 / 0.82 / 0.93 / 0.81 / 0.91
DAY 2
C4 / 0.93 / 0.93 / 0.96 / 0.82 / 0.81 / 0.96 / 0.91 / 0.96
C5 / 0.91 / 0.94 / 0.92 / 0.89 / 0.80 / 0.98 / 0.98 / 0.81
C6 / 0.80 / 0.93 / 0.98 / 0.87 / 0.89 / 0.95 / 0.97 / 0.95
Table 3: Intra-circuit reliability for each of the 8 directions on Day 1 and Day 2.
These results in table 3 Indicate that estimates of intra-circuit reliability for all directions were good–excellent, except circuits 1 and 2 in medial (M) and antero-medial (AM) directions (Range: 0.62-0.74).