Evaluation of hand function in healthy individuals and patients undergoing uninstrumented central corpectomy for cervical spondylotic myelopathy using nine-hole peg test
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/neuroindia.NI_12_17
Source of Support: None, Conflict of Interest: None
Objectives: To evaluate the hand function in healthy individuals and in patients with cervical spondylotic myelopathy (CSM) undergoing central corpectomy using the nine-hole peg test (NHPT).
Keywords: Cervical spondylotic myelopathy, modified JOA scale, nine hole peg test, Nurick grade
Cervical spondylotic myelopathy (CSM) is the most common cause of acquired spastic quadriparesis or paraparesis that results in significant disability in individuals around 50 years of age. The progressive disability seen in these patients is caused by a combination of muscle spasticity, weakness, and sensory deficits. Surgical decompression for CSM has been reported to improve the functional status even in poor-grade patients.,,,,,,, The outcomes following surgery for CSM have been measured as changes in various functional scales such as Nurick grade and the modified-Japanese Orthopedic Association (mJOA) scale.,, Although these scales report overall changes in function, there is a certain degree of subjectivity in some of the measurements and measurement of these indices do not provide information on the functional change that may have occurred within the first one week to ten days following surgical decompression of the spinal cord. Tests of hand dexterity like rapid hand flick test did not show significant improvement in scores on day 7 compared to day 1 postoperatively.
The nine-hole peg test (NHPT) has been reported to be a tool for quick assessment of finger dexterity., This test measures finger dexterity as a measure of the time in seconds taken to transfer nine plastic or wooden pegs from a bowl into a slotted board. Normative data for this test has been reported for the Western population but not for the Indian population. Hence, we derived normative data for NHPT among the subjects representing the Indian population and assessed its performance as a tool to measure hand function in a cohort of patients with CSM who underwent uninstrumented central corpectomy.
The study protocol was approved by the Institutional Review Board.
Nine-hole peg test
Commercially available nine-hole peg board and pegs (Rolyan 9-Hole Peg Test Kit, model A851-5, Patterson Medical Holdings, Warrenville, IL) was used to perform the test. The apparatus consisted of a rectangular plastic board (26 × 13 × 2.75 cm) with three rows of three holes each on one half (each hole being separated from another by 2.4 cm) and a peg container with 9 plastic pegs on the other half. Each peg measured 0.6 cm in diameter.
For performing the test, the patient was seated in front of a table that had the peg board and pegs on it. On being given a verbal cue, the patient took the pegs one by one and placed them in the slots provided. The test was terminated at the instant the last peg was placed into the slot. The time taken from the patient touching the first peg to the time when the last peg was placed into the slot was recorded in seconds. After an initial practice trial, the test was performed five times with each hand and the mean time was calculated in seconds as the time taken to perform the NHPT with that hand. These values were recorded separately for the right and left hand. The mean value of the right hand and left hand for each patient was compared and the higher of the two values was recorded as the “worst score” and the lower of the two values was recorded as the “best score” for that patient. The average of the values for the right and left hand for each patient was calculated and this was recorded as the median time (in seconds) taken to perform the NHPT.
After obtaining informed consent, normative data was collected from 202 healthy adults (118 male and 84 female subjects) in different age groups (stratified into five-year intervals from 20 to 70 years of age). These subjects were relatives of the patients admitted in the neurosurgical wards and those accompanying patients to the outpatient clinic. The test was explained in detail to them and was done in a uniform fashion by the same examiner (SJ).
Patients with CSM
Informed consent was obtained from the patients who were recruited. All patients with CSM of Nurick grade 1 to 5, who underwent uninstrumented central corpectomy from 2011 to 2013, were recruited. The operative procedure has been described in detail in other publications., Patients with Nurick grade 0 CSM, co-existing cerebral or cerebellar disease or injuries or deformities of hands that would affect hand dexterity, as well as those patients who had undergone previous surgery of the cervical spine were excluded. In all those patients who were eligible for inclusion and consented for participation, a detailed neurological examination was done and the Nurick grade as well as mJOA score for cervical myelopathy were recorded pre-operatively, at one week postoperatively, and at six months or more follow up after surgery. During the study period, 47 patients with CSM performed the preoperative NHPT. There were 43 male and 4 female patients, with their mean age at the time of presentation being 51.2 years (range 33 to 77 years). Three of these 47 patients were unable to perform the NHPT at one week after surgery. Twenty-one of the remaining 44 patients were available for assessment of the NHPT at six months or later.
The mean time in seconds taken to complete the NHPT with either hand as well as the median NHPT were measured preoperatively; at one week; and, at a follow up of six months or more postoperatively following uninstrumented central corpectomy for CSM. Correlation between the preoperative NHPT score, the Nurick grade, and, the Total mJOA (TmJOA) and upper limb component of the mJOA score (UlmJOA) was studied. The difference in the NHPT score at follow up postoperatively was correlated with the UlmJOA score.
All data were entered into an Excel spreadsheet (Microsoft, 2003) and statistical analysis was performed using Statistical Package for the Social Sciences (SPSS) software (Version 16.0). From the age-, side-, and sex-matched normative data collected from healthy individuals, the proportion of patients who had abnormal prolongation of the NHPT score for each hand was determined preoperatively, at one week, and at more than a six-month follow up. An abnormally prolonged NHPT score was defined as a value that was more than two standard deviations added to the mean/median NHPT score for that age group.
Comparisons of the mean NHPT score of both the hands, and the median values of preoperative NHPT and one week postoperative NHPT, as well as that at a longer term follow-up postoperatively were made using the Student's t-test. Correlation of the mean NHPT of either hand and median NHPT score at each time point with the corresponding Nurick grade and UlmJOA score was determined. A Pearson correlation coefficient (r) of 0.4 or more was considered to be significant correlation. The change in the NHPT score at follow up of six months or more was correlated with the change in UlmJOA score. A P value of < 0.05 was considered statistically significant.
NHPT scores in healthy individuals
Normative data obtained from the healthy volunteers were stratified into ten groups with age range of five years as summarized in [Table 1]. The median NHPT score for male and female subjects for each age group was found to be significantly different with a mean difference of 0.71 s (P = 0.002) [Figure 1]. The distribution of values in the age groups of less than 45 years and more than 45 years followed the pattern of a normal binomial distribution, as shown in [Figure 1]. The median NHPT score for age ≤45 years was 11.9 ± 0.9 s (95% confidence interval [CI], 11.7 to 12 s) and that for age >45 years was 13.9 ± 1.6 s (95% CI,13.6 to 14.2 s). As this followed the pattern of normal distribution, a value that was two standard deviations higher than the mean was deemed to be abnormally prolonged. For each age group, the time taken to perform the NHPT with the right hand was significantly lower than the time taken to perform the test with the left hand (P < 0.001) [Table 2].
Preoperative and postoperative NHPT scores in patients with CSM
The preoperative median NHPT score (n = 47) was 24.8 ± 15.6 s (range: 10.8–80.3 s; NHPT score in the right hand was 20.6s, and the left hand was 27.9s) in the patient cohort and it was 24.6 ± 16.6 s (range: 12–88.3 s, NHPT score in right hand was 21.7s and left hand was 25.6s) at one week postoperatively. The preoperative mean NHPT score was 20.6 ± 12.3 s (range: 9.9–80.3 s) for the right hand, and 27.9 ± 21.8 s (range: 11.7–108.7 s) for the left hand. In 36 of the 47 patients (76.6%), the median NHPT score was above the normal level for age- and sex-matched individuals, indicating hand dysfunction.
At one week postoperatively, the median NHPT score persisted to be above the normal level in 33 of these 36 patients. The other three patients in this subgroup were unable to perform the test (two of them declined to perform the test as they were on bed rest after repositioning of a slipped graft and the other patient had deterioration in his neurological status and was unable to sit long enough to complete the test). None of the 11 patients whose NHPT score preoperatively was within the normal range had worsening of their score at one week postoperatively.
Of the 21 patients available for a long-term follow-up of 6 months or more (mean follow-up duration 17.6 months, range: 9 to 37 months) postoperatively, 14 had an abnormally prolonged median NHPT score preoperatively. In five (35.7%) of these 14 patients, the NHPT score improved to normal and in the other nine patients, the NHPT score remained prolonged. All seven patients in whom the preoperative NHPT score was within normal range maintained this status at follow up.
The median NHPT score at a long-term follow-up of more than six months was 16.2 ± 3.7 s and this was significantly lower than the preoperative median NHPT score of 22.4 ± 13.5 s in these 21 patients (P = 0.04). The average change in the median NHPT score was an improvement of 6.2 s, which accounted for a 27.7% decrease in the time taken to perform the test. The average improvement in the NHPT score of the right hand was 3 s (17%; range: 0.6 to 12.1 s) and the left hand was 9.5 s (35%; range: 1.38 to 47.2 s) compared to the preoperative score.
Correlation between the preoperative NHPT score, preoperative Nurick grade, and mJOA
There was a strong negative correlation between the preoperative NHPT scores of either hand and the median NHPT score and preoperative UlmJOA scores (the correlation coefficients being −0.71 and −0.77) when analyzed with the best score and worst score respectively [Figure 2]. There was a good negative correlation between the preoperative median NHPT score and the preoperative TmJOA (r =−0.4, P = 0.008) as well as the UlmJOA score. (r = −0.6, p = <0.0001). Of the 11 patients with a preoperative UlmJOA score of 5 (indicating normal hand function), the NHPT score was prolonged in three patients. There was a poor correlation between the preoperative Nurick grade and the preoperative median NHPT score. The results are summarized in [Table 3]. Thus, the preoperative NHPT score is a good indicator of hand dysfunction, but does not necessarily indicate the presence of myelopathy or functional dysfunction secondary to lower limb spasticity.
Changes in functional grades at follow up
The preoperative mean Nurick grade was 3.2 ± 1 (range: 1 to 5) and it improved to 1.6 ± 1.0 (range: 0 to 4) at six months or longer follow up (P < 0.001; n = 21). The mean TmJOA scale score was 13 ± 2.2 (range: 9 to 17) preoperatively and it improved to 16.0 ± 2.2 at six months or longer follow up (range: 10 to 18; P < 0.001). The UlmJOA score was 4.1 ± 0.9 (range: 2 to 5) preoperatively and it improved to 4.7 ± 0.5 (range: 4 to 5) at follow up (P = 0.007).
Change in the NHPT score and the UlmJOA score at follow up
No significant change was detected in the median NHPT score at one week postoperatively in the 44 patients in whom this data could be recorded (24.6 ± 16.6 s) as compared to the preoperative NHPT score (24.8 ± 15.6 s). No significant change was noted in the UlmJOA score at one week postoperatively.
In five out of the 14 patients with a prolonged NHPT score, the score improved to normal at a longer follow up. In this group, the mean improvement in the UlmJOA score was 1 (range: 0 to 2). There was an improvement in the UlmJOA score in four of these five patients and the a preoperative UlmJOA score of 5 was maintained in one patient. In the remaining 9 patients with a prolonged preoperative NHPT score, the follow-up NHPT score remained prolonged. In these patients, the mean improvement in the UlmJOA score was 1 (range: 1 to 3). There was no change in the UlmJOA score in two of these 9 patients (one of whom had a preoperative score of 5), the UlmJOA score showed improvement in 6 patients, and deteriorated in one patient. In the 7 patients who maintained normal preoperative and follow-up NHPT scores, there was no change in the UlmJOA score in five patients who had a preoperative UlmJOA score of 5 and the remaining two had an improvement in the UlmJOA score from 4 to 5. There was no significant correlation between the change in the NHPT score and the change in the UlmJOA score at follow up.
NHPT in Indian population
We measured the performance of healthy individuals in the Indian population using the NHPT. A significant difference between the scores of right hand and the left hand was detected; with the right hand performing the test faster than the left hand. This is a reflection of the cultural practices as well as the handedness of our cohort of normal individuals whose right hand is trained from childhood to perform more tasks that require a higher degree of dexterity.
Previously reported normative data on NHPT have utilized a slightly different version of performing the test with only two trials. These do not report the values for the right and left hand separately and have calculated a mean value of the time taken by either hand to perform this test. In our assessment, values from the right hand and left hand were recorded separately as this could have an implication in assessing the effect of surgical decompression, particularly in patients who have predominant unilateral symptoms.
Surgical decompression for CSM has been reported to improve the functional status even in poor-grade patients. Hand function is affected in patients with CSM and it can be attributable to pyramidal tract as well as posterior column involvement. In our patient cohort, 66% of the patients had unilateral or bilateral hand grip weakness as a manifestation of upper limb involvement. Various grading systems such as the Nurick and the mJOA scale exist to assess the functional outcome in patients with CSM., While these report overall changes in function, there is a certain degree of subjectivity in some of the measurements. While the JOA scale has a subsection to evaluate the upper limb function, it does not optimally evaluate minimal disability of hand function. Hence we used the NHPT, which has a higher test-retest reliability and lesser inter observer variability, to objectively assess the hand function.,
Preoperative NHPT score and functional grades
The preoperative NHPT score aided in identifying hand dysfunction in our cohort of patients with CSM with 36 of the 47 patients having scores 2 standard deviations above that for the age-matched normal individuals. Of the 11 patients who had a preoperative UlmJOA score of 5 (classified as normal function), 3 had prolonged NHPT scores. Thus, NHPT aided in detecting hand dysfunction in three patients who would have been classified to have “no hand dysfunction” as per the UlmJOA score. There was a good correlation between the preoperative NHPT score and the UlmJOA scale score as well as the TmJOA scale score. The correlation with the Nurick grade was however poor and this further underscores that a single functional grade may not be adequate to indicate disruption in various facets of function in patients with CSM. Nurick grade is a measure of the extent to which spasticity of the lower limbs affects function in patients with CSM, and hence did not correlate with the NHPT score. The Nurick grade has been shown previously to have poor correlation with the UlmJOA score.
In our cohort, 34 of the 47 patients (72%) had better scores in the right hand, thus indicating that even in the presence of myelopathy, the right hand seems to be performing better than the left hand. Out of the 13 patients who had better NHPT scores on performing the test with the left hand, 7 had asymmetrical involvement of the right hand alone. This also underscores the importance of reporting data regarding the NHPT with either hand or as the best/worst score rather than as the median NHPT score. These observations were similar to the rapid hand flick test, where the right hand (dominant hand) performed better compared to the left.
Postoperative NHPT score and functional grade
No significant change was observed in the NHPT scores at a one week follow up after surgery. Thus, NHPT could not identify an early improvement in hand function, although some patients may subjectively report improvement in hand function during the early postoperative period. There was no significant change in the UlmJOA score at one week following surgery. The rapid hand flick test, however, has been reported to be able to detect significant change in patients as early as one day following decompression surgery for CSM. The performance of the rapid hand flick test essentially tests the function of finger flexors and extensors. NHPT tests the fine dexterous movements such as the pincer grasp that is more relevant in day-to-day functioning of an individual. It is possible that while gross movements of finger flexion and extension may improve early after decompressive surgery for CSM, movements that require more dexterity, as is assessed by NHPT, may improve later.
At a long-term follow up of six months or more, a significant improvement in the Nurick grade as well as the mJOA scale score was observed in the 21 patients who were available for follow up. This is in concurrence with the previously reported data from our institution., A total of fourteen patients had their preoperative median NHPT scores above the normal range, and in 5 of them, there was improvement of the scores to within normal range. Although there was 27.7% improvement in the median NHPT score at a long term follow up when compared to the preoperative score, this did not correlate with the change in the UlmJOA score.
It was also observed that the NHPT score tended to return to the normal level in 35.7% of the patients in whom it was prolonged preoperatively, at a longer follow up after surgery. It has been reported previously that the NHPT scores could detect a significant change in the hand function at one month following the posterior decompression for CSM. It has also been reported that in patients with stroke, NHPT could detect further improvement once the patients had reached the maximum scores with other tests, and that it was more sensitive than other tests in detecting improvement at three months. Thus, the NHPT could aid in detecting improvement in function at one month following surgery but is not a good indicator of the degree of improvement in hand function as early as one week following surgery. The reason for this delayed improvement could be the latency required for the reversible changes within the cord to improve even though the cord has been adequately decompressed.
Normative data of the NHPT scores in Indian population indicates better scores among female subjects and the right hand score was significantly better than left hand score. The NHPT is a good quantitative test to evaluate hand function in patients with CSM and correlated well with the UlmJOA score. Although this test could detect subtle hand dysfunction preoperatively, it cannot replace the UlmJOA scoring system.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2]
[Table 1], [Table 2], [Table 3]