Does the use of a template board imply new normative data for the Jebsen-Taylor Hand Function Test?

Abstract

Introduction

Recent research reports the necessity of a template board to ensure accuracy when administering the Jebsen-Taylor Hand Function Test. However, the original studies that collected normative data did not use a template board. This study collected a sample of data for one female age group of the Jebsen-Taylor Hand Function Test using a template board.

Methods

Thirty female participants with no history of hand trauma or a condition affecting the hand were recruited. Each participant carried out subtests 2, 3, 5, 6 and 7 of the Jebsen-Taylor Hand Function Test, which was compared to the normative data available for this test.

Results

Results show a significant difference in the mean performance for subtest 2 (non-dominant) and subtest 3 (dominant and non-dominant) with p-values < 0.05 for a difference in means t-test. The remaining subtests 5, 6 and 7 were not statistically significant.

Conclusions

The findings reported in this study add to the empirical evidence investigating the normative data for the Jebsen-Taylor Hand Function Test. The statistically significant difference in mean times for two subtests suggests that new normative data may need to be established if clinicians intend to use a template board when administering the test. Resolving this issue is a natural avenue for further research in this area.

Keywords

Jebsen-Taylor outcome measures service evaluation audit

Introduction

The availability of quantitative normative data allows clinicians to interpret how patients present in an area of performance in relation to the normal population,¹ assisting in identifying areas of concern and setting treatment goals. The Jebsen-Taylor Hand Function Test is an outcome measure used to assess a range of clinical conditions. Normative data for the test are available for paediatric, adult and elderly populations^2–4 and a modified version for the Indian population.⁵ The test consists of seven subtests. These include writing, turning over cards, lifting everyday objects (coins, paper clips and bottle tops), lifting kidney beans onto a spoon, lifting checkers and, heavy and light cans. Performance in each of these subtests is timed and can be compared with the normative data.

Based on the International Classification of Functioning, Disability, and Health (ICF) in 2001, the assessment of function has been determined to consist of three components. These are Body Functions and Structures, Activities and Participation.⁶ Within this framework, 1400 descriptors of function have been defined. In 2009, experts used this framework to identify 117 descriptors relevant to the clinical management of hand conditions.⁷ This is referred to as the Comprehensive International Classification of Functioning, Disability, and Health Core Set for Hand Conditions. A brief version with 23 descriptors was also devised. The Jebsen-Taylor Hand Function Test has been classed as an assessment tool that measures Activity.⁸ Examples of similar categorised tests include the Nine Hole Peg Test and the Purdue Pegboard Test. Examples of Participation assessments include patient-related outcome measures such as the Disability of Arm, Shoulder and Hand (DASH) and the Patient-Related Wrist Hand Evaluation (PRWHE). Assessments that address Body Functions and Structures include goniometry and manual muscle testing.⁸ A recent study has demonstrated that while accuracy is a prerequisite to administering the Jebsen-Taylor Hand Function Test, it is difficult for therapists to achieve in practice without a standardised template board to guide test set up.⁹ A template board guides therapists in laying out the test items in an accurate manner: a strategy reported by previous authors^10,11 and used in other standardised tests for the same purpose.¹² The authors also point out that the template impacted on the time required for set up of each subtest. Of note, studies that collected normative data for the Jebsen-Taylor Hand Function Test did not use a template board to aid with the set up of test items. Therefore, the original normative data may not longer be appropriate as a benchmark when using a template board. A template board made to a precise specification would also ensure that the table surface used for assessment was standardised. A previous study¹³ reported that normal subjects were significantly faster at lifting wooden checkers compared to plastic checkers, therefore demonstrating the importance of texture (wooden checkers were used in the normative data studies for the Jebsen-Taylor Hand Function Test). The Jebsen-Taylor Hand Function Test is widely used in clinical practice, and it has been an outcome measure in more than 135 research studies.¹⁴

This study collects a sample of performance scores using a template board when administering the Jebsen-Taylor Hand Function Test. Its aim is to investigate if there is evidence of a significant difference between the sample and normative data. A statistically significant difference would question the use of the original data as a benchmark when the test is completed with the aid of a template board.

Methods

Design

A t-test for difference of means was employed to test the null hypothesis that the sample mean and the mean reported for the normative data for each subtest are the same. The t-test assumes that the sample data are normally distributed with equal variance, with the central limit theorem stating that as the sample size gets large enough the sampling distribution of the mean is approximately normally distributed regardless of the shape of the distribution of the individual values in the population. As a general rule, 30 is commonly employed as a sample size benchmark to rely on the central limit theorem as statistical research has found that when the size is 30 the sampling distribution of the mean is approximately normal.¹⁵ Consequently, 30 participants were recruited from the Southern Health and Social Care Trust to produce the sample data. This also replicates the sample size used in each age band for the original data. The original data comprised four age bands (20–29; 30–39; 40–49; 50–59), each with 30 subjects. The authors reported that there was no statistically significant difference in the performance times of each age band, and therefore the datasets were merged to present an overall normative dataset for a 20–59 age band.³ The study information sheet was distributed to departments within the hospital. Interested participants contacted the principal investigator to participate in the study. The inclusion criteria were female staff, aged of 20 to 59. This corresponded with a strand of the original normative data, and there was an expectation that staff in this demographic would be easier to recruit in a female-dominated workforce. The exclusion criteria were a significant hand injury or condition affecting hand function and any staff member familiar with the administration of the Jebsen-Taylor Hand Function Test.

Ethics

The study complied with the Research Governance Framework for Health and Social Care and Good Clinical Practice. The Southern HSC Trust Research Governance Committee approved the study. Approval through a Research Ethics Committee was not deemed necessary as the study was non-invasive, involved staff only and did not entail any form of intervention or gathering of sensitive information. Each participant provided written, informed consent.

Materials

Medium-density fibreboard was used to construct a 1365 mm × 413 mm × 10 mm template board. Markings for the positioning of test items used coloured permanent markers, guided using a transparent metric grid (Figure 1).

Figure 1.

Layout of a standardised template for the JTT (not to scale).

Procedure

Eligible staff were invited to participate in the study. Each participant was provided with written information on the study. The new sample group completed the selected subtests on the template board. For subtests 5, 6 and 7, an additional wooden board (as described in the original study) was placed on top of the template. Hand dominance, age and that there was no history of a hand condition affecting the function of each participant was ascertained.

Subtests 2, 3, 5, 6 and 7 were the focus of this study. Subtests 1 (writing) and 4 (feeding) were not included because there is no requirement to lay out test items accurately on a table surface. Each participant was provided with the Jebsen-Taylor Hand Function Test standardised verbal instructions before each subtest was completed. In each subtest, both hands were tested. The time taken (in seconds) to complete each test was recorded.

Subtest 2: Simulated page turning

Five rectangles (3″ × ″5″), 5″ from the edge of the table/template border and 2″ apart indicated where to position the five index cards in this subtest (Figure 1). The instructions were read to each participant as detailed in the original article.³ Timing commenced on, “Go”. The testing of the non-dominant hand was first. An additional instruction was provided to ensure the subtest was completed in the same method as intended: participants were instructed that they did not have to set the pages accurately back on top of the rectangles on the template board.

Subtest 3: Small common objects

Seven shapes (outline of small common objects) on the board indicated where to place the coffee jar, the 1″ paper clips, the bottle caps and the pennies for this subtest. The coffee jar was central and 5″ from the edge of the board and the other objects were positioned 2″ apart from the each side of the coffee jar (Figure 1). Timing commenced on, “Go.” On the second trial, the items were set up as a mirror image of the first trial.

Subtest 5: Stacking checkers

Four small circles and a dotted line on the template board indicated where to position the checkers and the additional board, respectively (Figure 1). Timing commenced on, “Go.” Again, the testing of the non-dominant hand was first.

Subtest 6/7: Lifting large light/heavy objects

Five large circles 5″ from the edge of the table/template board and 2″ apart indicated where to place the large cans. The set-up and procedure were the same for both subtests only with the substitution of heavier cans in Subtest 7 (Figure 1). Figure 2 illustrates the additional board positioned on top of the template board. The testing of the non-dominant was first and timing commenced on, “Go.”

Figure 2.

Wooden board placed on top of template in subtest 6.

Results

Thirty female participants with a mean age of 33.87 years (SD = 9.81) completed each of the subtests. Twenty-eight of the participants were right-hand dominant. Table 1 reports the mean performance (in seconds) for the sample data as well as the maximum and minimum scores of each subtest. Normality p-values for the new dataset are also reported (Shapiro-Wilks test). Two subtests rejected the null hypothesis of normality. Subtest 2: simulated page turning (dominant) had a p-value of 0.02, while subtest 7: lifting large heavy objects (non-dominant) also had a p-value of 0.02. In this situation, an alternative approach would be to conduct a non-parametric test which makes no distributional assumption. This would have required both the raw sample data presented here and the raw normative data. As this was unavailable, the two subtests which violate the normality assumption are not reported in Table 2. The remaining eight subtests fail to reject the null hypothesis which implies that they meet the normality assumption of the t-test. Table 1 also reports the original Jebsen-Taylor Hand Function Test means scores and standard deviations.

Table 1.

New and original data on Jebsen-Taylor hand function test (mean performance in each subtest in seconds).

Subtest	New data on mean performance (in seconds)				Normality testing of new data^a	Original data on mean performance (in seconds)
Subtest	N	Mean (SD)	Minimum	Maximum	p	N	Mean (SD)
Subtest 2: Simulated page turning (non-dominant)	30	4.34 (1.05)	2.62	6.70	0.71	120	4.8 (1.1)
Subtest 2: Simulated page turning (dominant)	30	3.92 (0.92)	2.58	5.84	0.02	120	4.3 (1.4)
Subtest 3: Small common objects (non-dominant)	30	6.61 (1.02)	4.64	8.63	0.82	120	6 (1)
Subtest 3: Small common objects (dominant)	30	6.38 (1.35)	3.81	10.26	0.08	120	5.5 (0.8)
Subtest 5: Stacking checkers (non-dominant)	30	3.95 (0.84)	2.40	6.02	0.85	120	3.8 (0.7)
Subtest 5: Stacking checkers (dominant)	30	3.27 (0.56)	2.28	4.64	0.81	120	3.3 (0.6)
Subtest 6: Lifting large light objects (non-dominant)	30	3.12 (0.59)	2.06	4.56	0.45	120	3.3 (0.6)
Subtest 6: Lifting large light objects (dominant)	30	2.91 (0.54)	2.06	4.05	0.40	120	3.1 (0.5)
subtest 7: lifting large heavy objects (non-dominant)	30	3.25 (0.53)	2.44	4.64	0.02	120	3.3 (0.5)
Subtest 7: Lifting large heavy objects (dominant)	30	3.03 (0.56)	1.95	4.47	0.31	120	3.2 (0.5)

Shaprio-Wilk Test.

Table 2.

Comparing new mean scores with original mean scores.

Subtest	New performance times (in seconds)	Original performance times (in seconds)
	Mean (SD)	Mean (SD)	F-test p	t-test p
Subtest 2: Simulated page turning (non-dominant)	4.34 (1.05)	4.8 (1.1)	0.800	0.041
Subtest 3: Small common objects (non-dominant)	6.61 (1.02)	6 (1)	0.846	0.003
Subtest 3: Small common objects (dominant)	6.38 (1.35)	5.5 (0.8)	p < 0.001	0.002
Subtest 5: Stacking checkers (non-dominant)	3.95 (0.84)	3.8 (0.7)	0.179	0.315
Subtest 5: Stacking checkers (dominant)	3.27 (0.56)	3.3 (0.6)	0.687	0.804
Subtest 6: Lifting large light objects (non-dominant)	3.12 (0.59)	3.3 (0.6)	0.975	0.143
Subtest 6: Lifting large light objects (dominant)	2.91 (0.54)	3.1 (0.5)	0.555	0.070
Subtest 7: Lifting large heavy objects (dominant)	3.03 (0.56)	3.2 (0.5)	0.397	0.106

Equality of variances implied that a pooled-variance t-test is appropriate, whereas samples with different variances would warrant a separate variance t-test. Table 2 reports the results for an F-test which tests the null hypothesis that the variances are equal. The null hypothesis is rejected for subtest 3: small common objects (dominant). Consequently, the p-values reported for the t-tests in the final column are for a separate variance t-test for this subtest, while it is a pooled variance t-test for the remaining subtests. At the 5% cut-off, three out of the eight t-tests for difference in mean have p-values > 0.05. These are subtest 2: simulated page turning (non-dominant) (p = 0.041), subtest 3: small common objects (non-dominant) (p = 0.003) and subtest 3: small common objects (dominant) (p = 0.002).

Discussion

This study presents new data for the Jebsen-Taylor Hand Function Test using a template board for a sample of female participants between the ages of 20 and 59. Of the 10 datasets investigated, two did not follow a normal distribution, five replicated the original normative data, and there were discrepancies between the new and original datasets in three datasets: subtest 2 (simulated page turning) for non-dominant hands and subtest 3 (small common objects) for both dominant and non-dominant hands.

One common variable distinguishes the set-up of subtests 2 and 3 from subtests 5, 6 and 7. Subtests 2 and 3 do not use the additional board for set-up (Figure 2), which means that when a template board is not used, the clinician administering the test must position the test items accurately so that they are at the correct distance from the table edge as well as in relation to each other. The additional board in subtests 5, 6 and 7 assists the process of set up because the board edge helps create a uniform distance from the table edge. Therefore, the clinician’s only concern is to ensure accurate positioning of the test items in relation to each other. Harte et al.⁹ previously observed that accurate set up of subtest 3 was the most difficult to do without a template board. This suggests that reproducibility of these subtests was affected in the original data and may have influenced overall performance times. These findings highlight the importance of using the template board when administering subtests 2 and 3. Slight variances in the positioning of test items may be reasoned as improbable to impact on performance times. However, as already reported, characteristics of the test items such as the material of the checkers can affect performance times.¹³ Similarly, using the template board ensures that the table surface is consistent across different environments when the test is administered.

A broad range of performance times is noted in subtest 3 for both dominant and non-dominant hands (Table 1). In both the original and new data values, this subtest takes the longest on average to perform compared to other subtests explored in this study. This subtest requires six items to be lifted while subtests 2, 6 and 7 require five items to be lifted and subtest 5 only requires four items to be lifted. The participant also has to transport the items the longest distance (i.e. from the table to the coffee jar) compared to the other subtests. Finally, there are three different objects used in subtest 3, and these items are the smallest when compared to the other subtests included in this study. The broad range of performance times may have been influenced by the observation that in performing this subtest, it was more common for participants to frequently drop items compared to the other subtests. This would obviously be due to the use of the smallest objects in subtest 3. As the raw data for the original study are unavailable, it is not known if there was also a wide range in the performance times.

In this study, subtests 5, 6 and 7 produced similar mean values observed in the original dataset. More accurate positioning can be achieved in these subtests due to the physical layout of these tests, e.g. use of additional wooden board. However, caution should be given in using the original norms as a reference point when using the template for the male population and other female age ranges. Assumptions cannot be made that all the performance scores will follow a similar trend as in this study.

Performance-based tests, such as the Jebsen-Taylor Hand Function Test, can be compared with normative data, but clinicians may also use these to observe first-hand the quality of a patient’s hand function or to measure their progress across time, i.e. comparing repeated performance times. In this study, while mean performance times did differ in three subtests, these were each a difference of under one second. Clinically, this difference may not be material. Therefore, therapists are advised to refer to all the available reference points (i.e. the original and new normative data, the patient’s subjective experience, previous test performances and quality of movement) when using a template board before interpreting scores.

This study adds to the body of knowledge on the clinimetric properties of performance-based hand function tests as recommended by Schonveld et al.¹⁶ It provides clinicians with a reference point for one female age group of the population if they intend to incorporate a template board in their clinical practice when administering the Jebsen-Taylor Hand Function Test. It must be noted that for this to be viable, clinicians should make the template board in the same fashion as described in the methodology to ensure reproducibility. A limitation of this study is that it only focused on one portion of the population whereby normative data (without a template board) exists for both male and female adults, children and the elderly. Also, the finding that data for subtest 2 for the dominant hand and subtest 7 for the non-dominant hand did not follow a normal distribution was unexpected as the sample size used in this study replicated the original study sample size and motivated by statistical theory (central limit theorem). Therefore, a parametric analysis could not be performed for these two subtests.

In contemporary hand therapy practice, there appears to be growth in the use of patient-reported outcome measures for function¹⁷ with 92% of the American Society of Hand Therapy members using these in clinical practice.¹⁸ As discussed, they are categorised as Participation assessments. While this evolution in practice is welcome, a comprehensive functional assessment can only be obtained by also incorporating Activity and Body Functions and Structures tests in clinical activity as recommended by ICF⁶ The Jebsen-Taylor Hand Function Test has been categorised as an assessment of Activity.⁸ Continued research to improve and ensure such tests are as robust as possible is necessary so clinicians have access to dependable outcome measures. Historically, the Jebsen-Taylor Hand Function Test has had contrasting reports on it uses as an outcome measure. There are reported poor correlations with the Michigan Hand Questionnaire¹⁴ and patients with hemiplegia,¹⁹ though strong correlations were found with the Klein-Bell scale in a cohort of patients with spinal injuries.²⁰ It is reportedly sensitive to differences between the normal population and people with rheumatoid arthritis.²¹ Test–retest reliability has been reportedly weak²² and strong.^3,11,23

Evidence of reliability, validity and responsiveness is a requirement of all outcome measures.²⁴ The introduction of a template board to the Jebsen-Taylor Hand Function Test has previously shown to aid clinicians in setting up the test efficiently and in achieving reproducibility by ensuring accurate positioning of test items.⁶ This study shows that when using a template board in the test, the normative data could be replicated in all the subtests (examined in this study) except subtests 2 and 3. A statistically significant difference was observed between subtests 2 and 3 and the original normative data for one age range (females between ages of 20 and 59) of the Jebsen-Taylor Hand Function Test. A natural extension of this study would be to complete analysis across all age ranges and gender to assess if there are other systematic deviations from the normative data across all subtests and if these deviations follow a similar pattern. Further research should also consider the sample size to ensure a normative dataset is available for robust analysis, given the unexpected findings for two of the subtests in this study.

Footnotes

Acknowledgements

The authors wish to thank all the staff who assisted in the collection of the data for this study.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The Southern Health and Social Care Trust provided funding for this research project.

References

Hoffman

Fitness and health assessment. In: Hoffman

(ed.) Norms for fitness, performance, and health, United States of America: Human Kinetics, Inc, 2006, pp. 2–10.

Taylor

Sand

Jebsen

. Evaluation of hand function in children. Arch Phys Med Rehabil 1973; 54: 129–135.

Jebsen

Taylor

Trieschmann

. An objective and standardized test of hand function. Arch Phys Med Rehabil 1969; 50: 311–319.

Hackel

Wolfe

Bang

. Changes in hand function in the aging adult as determined by the Jebsen Taylor test of hand function. Phys Ther 1992; 72: 373–377.

Priya

Desai

. Normative data of Jebsen-Taylor Hand Function Test (modified version) on Indian population. Indian J Physiother Occup Ther 2011; 5: 24–26.

World Health Organisation. ICF: international classification of functioning disability and health, Geneva: World Health Organisation, 2001.

Rudolf

Chung

. Development of the International Classification of Functioning, Disability and Health core sets for hand conditions – results of the World Health Organisation International Consensus process. Disabil Rehabil 2012; 34: 681–693.

Weinstock-Zlotnick

Bear-Lehman

. How therapists specializing in hand therapy evaluate the ability of patients to participate in their daily lives: an exploratory study. J Hand Ther 2015; 28: 261–268.

Harte

Curran

Hamill

. Using a template to improve the accuracy and efficiency of the Jebsen-Taylor Hand Function Test: a comparative study. Hand Ther 2014; 19: 11–16.

10.

Agnew

Maas

. An interim Australian version of the Jebsen test of hand function. Aust J Physiother 1982; 28: 23–29.

11.

Bovend’Eerdt

TJH

Dawes

Johansen-Berg

. Evaluation of the modified Jebsen-Taylor Hand Function and the University of Maryland Arm Questionnaire for stroke. Clin Rehabil 2004; 18: 195–202.

12.

Whiting

Lincoln

Bhavani

. Rivermead perceptual assessment battery. Occup Ther Health Care 1986; 3: 209–210.

13.

Rider

Linden

. Comparison of standardized and non-standardised administration of the Jebsen Hand Function Test. J Hand Ther 1988; 2: 121–123.

14.

Sears

Chung

. Validity and responsiveness of the Jebsen-Taylor Hand Function Test. J Hand Surg 2010; 35: 30–37.

15.

Berenson

Levine

Krehbiel

. Basic business statistics: concepts and applications, 10th ed. New Jersey, NJ: Pearson Prentice Hall, 2006, pp. 235–235.

16.

Schonveld

Wittink

Takken

. Clinimetric evaluation of measurement tools used in hand therapy to assess activity and participation. J Hand Ther 2009; 22: 221–236.

17.

Coenen

Kus

Rudolf

. Do patient-reported outcome measures capture functioning aspects and environmental factors important to individuals with injuries or disorders of the hand? J Hand Ther 2013; 26: 332–342.

18.

Valdes

Algar

Cyr

. Hand therapist use of patient report outcome (PRO) in practice: a survey study. J Hand Ther 2014; 27: 299–308.

19.

Sandi

Spaulding

McPherson

. Jebsen Hand Function Test: performance of the uninvolved hand in hemiplegia and of right-handed, right and left hemiplegia persons. Arch Phys Rehabil 1988; 69: 419–422.

20.

Lynch

Bridle

. Validity of the Jebsen-Taylor Hand Function Test in predicting activities of daily living. Occup Ther J Res 1989; 9: 316–318.

21.

Sharma

Schumacher

McLellan

. Evaluation of the Jebsen-Taylor Hand Function Test for use in patients with rheumatoid arthritis. Arth Care Res 1994; 7: 16–19.

22.

Stern

. Stability of the Jebsen-Taylor Hand Function Test across three test sessions. Am J Occup Ther 1992; 46: 647–649.

23.

Mak

MKY

Lau

ETL

Tam

VWK

. Use of Jebsen Taylor hand function test in evaluating the hand dexterity in people with Parkinson's disease. J Hand Ther 2015; 28: 389–396.

24.

Badalamente

Coffelt

Elfar

. Measurement scales in clinical research of the upper extremity, part 1: general principles, measures of general health, pain and patient satisfaction. J Hand Surg 2013; 38A: 401–406.