Segment 1: A patient with LID performing the finger tap during off and on medication conditions. TAP OFF score was 2/4 on the right and 3/4 on the left, whereas TAP ON was 4/4 on both the right and left sides. Segment 2: A patient without LID performing the finger tap during off and on medication conditions. TAP OFF score was 1/4 on both the right and left sides, whereas TAP ON was 3/4 on both the right and left sides.
Marina Picillo MD, Gustavo B. Vincos MD, Drew S. Kern MD, MSc, Susan H. Fox MRCP(UK), PhD, Anthony E. Lang MD and Alfonso Fasano MD, PhD
Article first published online: 27 OCT 2015 | DOI: 10.1002/mdc3.12246
The finger tapping test evaluates bradykinesia, focusing on decrement in rate, amplitude, or both with repetitive action. Vertical positioning of the hands during this task may also be clinically relevant. We developed a “TAP score,” measuring the vertical level above the lap where the patient performs the finger tapping ranging from 1 (task performed with the hand close to the lap) to 4 (above the head). In this pilot study, we retrospectively applied the TAP score in addition to usual motor scales during acute levodopa challenge in 123 PD patients (of whom 88 presented l-dopa-induced dyskinesia [LID]). TAP ON was higher than TAP OFF. Patients with LID presented higher TAP ON. TAP ON was related to LID severity, whereas TAP OFF was inversely related to the OFF motor symptoms. The TAP score may be a measure of proximal movement amplitude representing an easy method to evaluate defective or excessive motor output in patients with advanced disease.
Bradykinesia is defined as the progressive reduction in speed, amplitude, or both of repetitive actions and is an important diagnostic feature of Parkinson's disease (PD). It is assessed by the patient performing repetitive movements including the finger tapping task of the UPDRS-III. Advanced PD is also characterized by motor fluctuations and levodopa-induced dyskinesias (LIDs). LIDs are difficult to objectively rate and only a few rating scales have been validated. These scales are time-consuming and require extensive training and expertise in movement disorders to administer.
The aim of the present pilot study was to explore the clinical meaning of a specific aspect of performance of the finger tapping item of the UPDRS-III. We have observed that patients with PD tend to perform the finger tapping in the off condition with their arm and hand lower and closer in approximation to their lap, whereas they raise their arm at least to head level in the on condition with dyskinesia. We sought to use this phenomenon in developing a simple and easily administrated scale to score off/on/on with dyskinesia state in patients with PD. To this aim, we developed a “TAP score” evaluating the vertical level where the patient performs the finger tapping. By retrospectively applying the TAP score to videos of PD patients undergoing a l-dopa challenge, we tested the following hypotheses: (1) TAP score during the on phases (TAP ON) is higher than the score during the off phases (TAP OFF); (2) TAP ON is positively correlated with the severity of LID in the whole body and specifically in the upper limbs; and (3) TAP OFF is negatively correlated with the severity of bradykinesia in the whole body and specifically in the upper limbs.
Patients and Methods
At the Toronto Western Hospital (TWH), all patients with PD being evaluated for DBS undergo a videotaped l-dopa challenge taking a l-dopa dose corresponding to their morning dose of dopaminergic medications according to the Core Assessment Program for Surgical Interventional Therapies in Parkinson's Disease (CAPSIT-PD). The videotaped l-dopa challenge is performed by experienced and trained nurses, who consistently demonstrate the finger tapping task at the level of the trunk for two to three times, then the patient is asked to perform the task on his or her own. One hundred twenty-three l-dopa challenge videos recorded from 1 August 2012 to 30 June 2014 and stored in the archive of the DBS center of the TWH were assessed. Eighty-eight presented LID whereas 35 did not. Patients with a history of orthopedic issues affecting the mobility of upper limbs (e.g., frozen shoulder or contractures) were excluded.
The TAP score was preliminary tested in a smaller number of patients in which different heights were measured; finally, a version ranging from 1 (finger tapping performed with the hand close to the lap) to 4 (finger tapping performed above the head) was developed, with higher scores representing a higher position of the hand (Fig. 1A).
Figure 1. (A) TAP scoring system. Line C corresponds to the top of the head, line B corresponds to the ears, and line A corresponds to the top of the shoulders. For the finger tapping performed above the line C, a score of 4 was applied; for the finger tapping performed above the line B, but under the line C, a score of 3 was applied; for the finger tapping performed above the line A, but under the line B, a score of 2 was applied; for the finger tapping performed under the line A, a score of 1 was applied. (B) Change in TAP scores between off (triangles) and on (circles) medication conditions in patients with or without l-dopa-induced dyskinesias. (C) A paradigmatic patient performing the finger tap during the off (left panel, TAP OFF: 2/4) and on medication condition (right panel, TAP ON: 4/4).
The TAP score was applied for both hands in both off and on conditions (TAP OFF and TAP ON, respectively) and rated by a reviewer blinded to the patient's medications and condition. Twenty-three randomly selected videos (i.e., the first video recorded each month of the time frame) were reevaluated by the same rater after 4 weeks and by another author as well in order to assess intra- and inter-rater reliability, respectively.
The patients' demographic and clinical data previously scored during the l-dopa challenge were retrieved, including UPDRS-III in both off and on conditions (at the time of the l-dopa challenges the new MDS-UPDRS was used yet at the center) and the Clinical Dyskinesia Rating Scale (CDRS), a validated scale to assess LID in PD. Both the total scores and the score related to the upper limbs were considered. Written informed consent for the videotape and permission to publish were obtained by all the patients at the time of the l-dopa challenge. The study was approved by the local ethical committee.
Because the Mann–Whitney test did not detect any significant differences of motor scores (including TAP ON and OFF, UPDRS-III OFF bradykinesia score, and CDRS upper limbs score), data from right and left hands were pooled to reduce the number of multiple comparisons (thus, the database consisted of a total of 246 observations). Inter- and intrarater reliability were evaluated by the intraclass correlation coefficient (ICC) and 95% confidence intervals (CIs). ICC values above 0.90 were considered as excellent and between 0.75 and 0.90 as good. Because data distribution was found to be non-normal, medians (25th–75th percentiles, interquartile range) were reported. Wilcoxon's sign-rank and Mann–Whitney's tests were used to perform within- and between-group comparisons, respectively. Spearman's rank-correlation coefficient was used to correlate TAP ON with CDRS total and upper limb scores, TAP OFF with UPDRS-III OFF, and UPDRS-III OFF bradykinesia score. The predetermined significance level for all analyses was 0.05. Multiple comparisons were corrected for with Bonferroni's test, and the significance level was set at P < 0.007. All analyses were performed using SPSS software (v.17; SPSS, Inc., Chicago, IL).
Table 1 shows the demographic and clinical features of the 123 patients included. All fulfilled the diagnostic criteria for PD, had response to l-dopa greater than 50%, and had motor fluctuations.
Table 1. Demographic and clinical characteristics of the 123 PD patients included, divided in two groups according to the presence of l-dopa-induced dyskinesia
||Total Cohort (=123)
||PD With LID (=88)
||PD Without LID (=35)
Data expressed in median (25th–75th percentiles), except where otherwise specified.
aSignificant difference between UPDRS-III off and UPDRS III on within each group (P < 0.001).
bSum of the UPDRS III scores for finger tapping, pronation-supination, and hand grips tasks (items 23, 24, and 25).
cSignificant difference between TAP OFF and TAP ON within each group (P < 0.001); ΔUPDRS-III: the difference between the UPDRS-III in off and in on conditions evaluating l-dopa responsiveness.
R, right; L, left; N, number; NA, not applicable.
|Age at l-dopa challenge, years
|Disease duration, years
|Gender, women/men (N)
|Side of onset, R/L (N)
|Bradykinesia score offb
|CDRS total score on
|CDRS upper limbs score
Intra-rater reliability was good for the TAP OFF (ICC = 0.87; 95% CI: 0.78–0.93) and excellent for the TAP ON (ICC = 0.93; 95% CI: 0.89–0.96). Inter-rater reliability was good for both TAP OFF and TAP ON (ICC = 0.75; 95% CI: 0.49–0.86 and ICC = 0.85; 95% CI: 0.73–0.91, respectively).
TAP ON was significantly higher than TAP OFF (P < 0.001). Spearman's rank test showed a positive correlation between TAP ON and both CDRS total score (ρ = 0.486; P < 0.001) and CDRS upper limbs score (ρ = 0.533; P < 0.001) and an inverse correlation between TAP OFF and UPDRS-III off (ρ = −0.263; P = 0.001) and UPDRS-III off bradykinesia score (ρ = −0.249; P < 0.001).
Patients with LID presented higher TAP ON, but similar TAP OFF, as compared to patients without (P < 0.001 and P = 0.3, respectively).
Figure 1B shows the change in TAP scores between on and off medication conditions in patients with and without LID. Figure 1C shows a paradigmatic patient with LID with different TAP scores during the off and on condition. Video 1 displays 2 other example patients.
By categorizing the height at which the finger tapping is performed by 123 PD patients, we introduced a simple score able to detect the effects of l-dopa. Specifically, we demonstrated that the TAP ON is higher than TAP OFF regardless of the occurrence of LID; in addition, TAP ON was found to be higher in patients with LID, as compared to those without, and, accordingly, it also correlated to the severity of dyskinesia. Indeed, these results support a relationship between the height of the finger tapping performance and the on condition as well as the presence and severity of LID. The inverse relationship between the TAP OFF and both the total UPDRS-III and the UPDRS-III off bradykinesia score suggests that patients tend to perform the finger tapping lower and closer to the lap while in off.
In keeping with our initial hypothesis, these findings confirm that the height at which the patient performs the finger tapping may represent a measure of the proximal amplitude of voluntary movements. Several studies have focused their attention on the assessment of the amplitude of distal movements, and occasionally of proximal movements, in patients with PD with sophisticated devices.[8-11] However, no previous studies focused on the amplitude of the proximal movements in PD and their response to l-dopa by means of an easy and straightforward scoring of a task already implemented in clinical practice. In keeping with previous data[8, 9] and supporting previous observations, the association between the TAP OFF and the UPDRS-III OFF bradykinesia score suggests that the amplitude of the proximal and distal movements are related and the TAP OFF may represent a measure of the upper-limb bradykinesia. Indeed, we found that TAP ON is significantly higher as compared to the TAP OFF in all patients, suggesting that the amplitude of proximal movements is significantly increased by l-dopa irrespective of LID. Alternatively or complementary, the hand elevation in patients on medication may reflect an improvement of the postural motor adjustments during voluntary movements of the upper limbs. A lack of awareness of arm position may also play a part in this phenomenon, which is more pronounced in those with dyskinesia who have reduced awareness of involuntary movements.
We acknowledge that our study has limitations. First, when LID were present, a true blind assessment was not feasible; however, given that the TAP score is based on the objective position of the hand during the finger tapping task in relation to specific body site of reference, we believe that an unblinded assessment would have not influenced the scoring. Second, selection of the videos was not randomized. Third, the physical distance between the heights of the scoring system varies and patients can seldom perform the task at different levels (especially when they repeatedly cross the ears' plane) or—more rarely—exactly at the level of one reference plane; in order to partly overcome this limitation, we instructed raters to score the patient on the basis of the level most frequently occupied during the task. Finally, our study included only patients under evaluation for DBS, thus representing a fairly homogeneous group in terms of severity of disease, irrespective of the presence of dyskinesia. We acknowledge that our results are not easily generalized to all PD patients and further studies are needed to confirm the relationship between TAP OFF and bradykinesia in other subjects with a wider range of motor impairment and without motor fluctuations.
On the other hand, though a pilot study, the relatively large sample size (n = 123) and the use of standardized videotaped l-dopa challenges performed by experienced and trained personnel ensures reliability in the administration of the finger tap task. Finally, we have demonstrated that the TAP score is sensitive to change, that is, in detecting on/off differences and has adequate intra- and inter-rater reliability. In order to confirm the present results, the next step is a prospective application of our score in a large, heterogeneous cohort of PD patients, possibly also measuring the repeatability of TAP scores in individual patients.
In conclusion, our pilot study suggests that the height at which the finger tapping is performed closely mirrors the severity of both bradykinesia and hyperkinesia in patients with advanced disease and is an easy, feasible, and reliable method to obtain further information from the finger tapping. Interestingly, the TAP score can be retrospectively applied to existing patient video recordings; therefore, we believe that it could be easily implemented in virtually every research protocol involving PD patients.