Modification of Nurick scale and Japanese Orthopedic Association score for Indian population with cervical spondylotic myelopathy
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.152627
Source of Support: None, Conflict of Interest: None
Aims: Existing scales for functional grading of patients with cervical spondylotic myelopathy (CSM), such as the Nurick scale and modified Japanese Orthopedic Association (mJOA) scale, do not address certain culture-specific activities of the Indian population while grading patients with CSM.
Keywords: Cervical spondylotic myelopathy; functional grading; Nurick grade; modified Japanese Orthopedic Association scale
Nurick grading  [Table 1], Japanese Orthopedic Association (JOA) score and its modified version (mJOA)  [Table 2] are the most widely used disease-specific severity scales for grading the degree of functional impairment in cervical spondylotic myelopathy (CSM). Nurick grading, being simpler to administer, is more popular. Nurick grading assesses the employability of the person; and in that sense, is not strictly speaking, disease specific. It has been observed that there is discordance between improvement in lower limb function and regaining employment.  The functional disability arising out of myelopathy not only affects the employability of the patient but also affects a host of his/her daily activities. These daily activities are so culture-dependent that it is almost impossible to devise a generic trans-cultural grading scale that can be applied across all cultures.
We modified the Nurick grade and mJOA scale by incorporating activities that are pertinent to our population, termed as the Indian modification of Nurick grade (imNurick grade) and the Indian modification of JOA scale (imJOA scale). We then evaluated these modified scales in Indian patients with CSM to determine whether these modifications had a meaningful impact on the functional scores of these patients. These modifications might also be relevant to the cultural practices of other Asian populations such as those in the South Asian region, Middle East, and South East Asia.
Modifications of the functional scales
We modified the Nurick grading (imNurick grade) [Table 3] and mJOA scale (imJOA scale) [Table 4] and [Table 5] by incorporating some of the most frequent daily activities of the average Indian which include sitting cross-legged, squatting, and rising up from squatting position with/without vertical support and eating with the hand rather than with spoons or chopsticks. We excluded employability in our modification of Nurick grading. We modified the upper limb component of JOA scale (imulmJOA) by incorporating the ability to tear chapatti/roti (flat, round Indian bread) with one hand and scoop rice from the plate without letting it slip through the fingers.
We applied the imNurick grade, imulmJOA scale and the original Nurick grade and mJOA scale prospectively to 93 patients with CSM who were managed in our unit between 2010 and 2012. There were two females among the 93 patients. The mean age was 51.3 ± 11.5 years (range 23-80 years). All the patients were assessed by two independent observers (RKM and KKR) who were blinded from the other's findings. One observer had completed training more than 5 years ago, and the other observer was in the latter half of his post-graduate training.
The Nurick grade, imNurick grade, llmJOA score, ulmJOA score, and imulmJOA score for each patient were noted and tabulated. We excluded patients with joint pain secondary to other causes (such as arthritis of the hip and/or knee joint) limiting the ability to get up from squatting/sitting positions.
After entering the data into an Excel spreadsheet, analysis was done using commercially available software [IBM Statistical Package of Social Sciences (SPSS) Statistics version 19.0, SPSS Inc., Chicago]. Construct validity of the modified scales was assessed using Cuzick nonparametric trend test, Spearman's rho, and Cronbach's alpha by comparing it with the components of mJOA score. Agreement between the various scales was determined using the kappa test. 
There was good agreement between the assessments (Nurick grade, imNurick grade, mJOA scores, and imJOA scores and its subdivisions) among the two observers (kappa = 1).
Construct validity of imNurick scale and imulmJOA score
The validity of the construct of any new rating scale has to be measured against an existing gold standard. In the absence of a gold standard scale for disease severity in patients with CSM, we used the components of the mJOA score for comparison. Cuzick nonparametric trend test showed that both Nurick scale (z = 4.4, P = 0.00) and imNurick scale (z = 5.5, P = 0.00) have a valid construct when tested against llmJOA score. The imulmJOA score too had a good construct with ulmJOA score (z = 2.5, P = 0.01). As seen in [Table 6], Spearman's rho value confirmed the progressive trends across various scales. Scatter plot of the scores of Nurick and imNurick against llmJOA [Figure 1] and ulmJOA against imulmJOA [Figure 2] further validate the modified scales. Both Nurick and imNurick scales had excellent internal consistency as a construct with llmJOA score (Cronbach's alpha 0.95 and 0.92, respectively). Similarly, the imulmJOA had excellent internal consistency with ulmJOA (Cronbach's alpha 0.96) and good internal consistency with llmJOA score (Cronbach's alpha 0.75).
Agreement between Nurick grade and imNurick grade
Though there was substantial agreement between Nurick and the modified grade when taken as a whole group (weighted kappa of 0.75), there was significant disagreement between the two scales in patients who were Nurick grade 2 and 3 (kappa = 0.07), suggesting that the modified grade was different in quality compared to the existing grade.
Agreement between ulmJOA score and imulmJOA score
There was substantial agreement between ulmJOA score and imulmJOA score (weighted kappa of 0.75) [Figure 2]. Of the 41 patients who did not have any hand dysfunction on the original mJOA score (a score of 5), 38 had no hand dysfunction on imJOA score (a score of 6). Three of them were unable to tear chapattis/roti and mix rice while eating, although the mJOA score did not detect any hand dysfunction. Interestingly, two patients who had only symptoms on the left side, with an ulmJOA of 3 and 4, respectively, had no difficulty in tearing chapattis/roti and mixing rice while eating with the right hand, and thus had a score of 6 on our modified scale.
Spondylotic myelopathy and functional disability
Motor dysfunction in compressive myelopathy results from decreased activation of anterior horn cells by descending corticospinal pathways, spasticity, segmental inhibition of agonist anterior horn cells by afferent flux from spindles of antagonist muscles, disuse atrophy with subsequent reduction in the total motor units available for recruitment, and joint contractures. ,,,, The initial insult resulting in the reduced number of the corticospinal fibers conveying impulses to the anterior horn motor neuron pool, along with plastic re-organization of synaptic connections in the cerebral cortex and spinal motor neuronal pool, together contribute to what is termed as spastic paresis below the level of the compression. In addition, segmental loss of anterior horn cells at the level of compression results in hand dysfunction. As the severity of compression progresses, more than one of the above-postulated mechanisms come into play resulting in progressive functional deficit, some of which may be reversible with decompression.
Do our modified Nurick and JOA scales measure something different/more?
Our proposed modification in the Nurick grade was able to discriminate a subset of patients who were scored as Nurick grade 2 in the original scale, by testing their ability to rise up from sitting/squatting position as this is an important function that needs to be measured when assessing patients from the Indian subcontinent. There was significant disagreement between the two scales in patients who were Nurick grade 2 and 3 (kappa = 0.07). The incorporated change in the upper limb component of the mJOA scale was, however, found to be in agreement with the existing ulmJOA scale indicating that hand function assessment with culture-specific questions can be used to report outcomes using the mJOA scale.
Is a modification of existing established myelopathy grading systems necessary?
The severity of disease in CSM has been traditionally gauged by assessing the degree to which it affects the person's functioning as such, including his daily activities and employability. Nurick grading, the oldest of all rating scales for CSM, relates the lower limb dysfunction to employability.  The degree to which severity of functional impairment in the lower limbs affects the employability of a person varies, the variables being the nature of work itself, ethnicity of the person, cultural habits, and of course compensation arising secondary to the loss/lack of job. It has been observed that there is discordance between improvement in lower limb function and regaining employment. In other words, employability may not be dependent on lower limb function in certain occupations, and hence, it may not truly reflect the disease severity in myelopathy. As the daily activities and practices vary in different cultures, one should take into consideration these factors while devising scoring systems for that specific population.
In the majority of Indians, daily activities ranging from attending nature's call, eating food, to various religious rituals involve squatting and sitting cross-legged on the floor. Similar practice is also followed by those living in other South Asian countries, Middle East, and South East Asia. Thus, squatting and sitting cross-legged and getting up from these positions constitute an important daily activity for them. A limitation in these activities resulting from spasticity reflects an additional facet to the disease severity apart from employability. Similarly, majority of Indians eat with their hand, making the ability to break chapatti/roti and scoop/mix rice from the plate an important daily function. It may be noted that the original JOA scoring system incorporated a culture-specific activity of being able to eat with chopsticks to assess the upper limb function. Thus, a scoring system taking into consideration these common ethnic practices would be more appropriate for the Indian population. Benzel  modified the JOA score in 1991 to suit the Western population eating with spoons. It is time to modify the same to suit the Indian population.
In the only previously reported Indian modification of the JOA scale, the authors have classified upper limb function based on the ability to eat with a spoon, eat rice with fingers, and ability to write.  The authors have not provided data regarding the interobserver agreement or validation of their scale vis-a-vis the original JOA scale.  In their modification, the ability to eat rice has been placed at a lower grade of function compared to the ability to grip a pen and write. In our proposed modification, the ability to scoop rice with fingers and to tear rotis/chapathis has been placed at a higher level than the ability to button shirts. This was based on our observation that patients who could button shirts without difficulty needed assistance to scoop rice with fingers or tear rotis/chapathis. It may be difficult to use the previous Indian modification of JOA in the subset of patients who have had no formal education as it requires assessment of an ability to write.  Our proposed modification also has the advantage that the scoring can be performed through a telephonic interview without direct observation of the patient. The previously reported modification  has not been validated while our proposed modification has been compared with previously validated Nurick and modified JOA scales. 
Our proposed modification also has near-perfect interobserver reliability with a kappa value of 1 observed between two independent observers at different levels of experience, thus obviating the need for more observers.  Previous validation studies for different instruments have typically used two observers. ,,
The adoption of any grading scale requires the scale to reflect disease severity, ease of administration, and ability of the scale to capture changes following intervention or over a period of time, changes that the patient reports to be significant. The last two conditions require a longitudinal intervention study which is a limitation of our cross-sectional study. Our validated scale could then be compared with the previously proposed Indian modification of JOA in patients prior to surgery as well as at follow-up.
Where do our modifications fit into the existing scales?
Twenty-four of 37 patients (65%) who were able to get up from squatting (imNurick 2) required the handrail to climb stairs (llmJOA 4). This means ability to climb stairs without the support of handrails indicates a less severe disease than to get up from a squat without support. We propose to incorporate this function as part of the lower limb function assessment in our proposed modification of the JOA scale, as indicated in [Table 5]. In the lower limb component of the mJOA score, the scores of 5 and 6 do not seem to affect any specific activity, rendering them redundant. Hence, we propose that in our modification, the scores of 5 and 6 from original mJOA score can be merged into a single score of 6. We advocate exclusion of employability from the Nurick scale and to incorporate the ability to get up unaided from squat. The inclusion of modified hand function in imJOA scale as seen in [Table 5] seems appropriate.
Modified Nurick (imNurick) grade and JOA (imJOA) score address the ethnic practices of the Indian population and of those residing in other South Asian countries, Middle East, and South East Asia more optimally than the original Nurick grade and mJOA scale, and therefore, are a better discriminator of different levels of functional disability while reporting outcomes in this population.
[Figure 1], [Figure 2]
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5], [Table 6]