Introduction

The functional benefits of upper limb reconstructive surgery in people with tetraplegia have been well documented in the literature. Therapists consider the International Classification of Functioning, Disability and Health, a biopsychosocial model of health, as the conceptual framework for interpreting outcomes measures following this reconstructive surgery [1, 2]. It emphasizes the importance of the dimension of participation, which is known to be correlated with the quality of life [3]. Based on this model, rehabilitation teams have reached consensus on the choice of measurement tools. However, it is still necessary to have a thorough knowledge of the effects of this surgery, regardless of the influence of environmental factors [4]. The purpose of an “analytical” approach is to help surgeons make appropriate choices and adapt and modify their surgical strategy. This approach is mainly based on force measurements using dynamometers and ranges of motion using goniometers. Although their reproducibility and internal consistency have been widely acknowledged, most of these devices are too insensitive to document small but meaningful functional gains [5]. They do not provide therapists with a current functional evaluation of the operated arm and/or hand, nor do they provide patients with motivating feedback on their progress, as most of the measures remain abstract for them [6].

In 1994, Wuolle stressed an important issue in arm and hand function assessment in patients with tetraplegia, namely the difficulty in distinguishing between what relates directly to the hand and what relates to the participation of the entire upper limb and trunk [7]. Lo et al. also pointed out that tests assessing hand function separately were more sensitive to change than tests including hand function in the assessment of the whole upper limb and trunk [8]. Following the same principles that guided the design of the MCS, the Grasp Release Test (GRT) was also presented as a manual function test dedicated to the evaluation of basic tasks. It was originally designed to assess the use of a hand neuroprosthesis [4]. People were asked to grasp, move and release three objects with the palmar grip and three with the lateral grasp. Time pressure was applied as they were instructed to do so as many times as possible in three 30-s trials for each object. However, a disadvantage of the GRT was that it required formatted objects, marketed in a dedicated set for evaluation. Unfortunately, this set is no longer marketed. In 2006, Spooren published a metrological study on the Van Lieshout Test (VLT) designed to assess skilled upper limb performance in patients with cervical SCI. Although its responsiveness was significantly correlated with that of the GRT, the VLT has never been published in the context of upper limb reconstructive surgery [9].

Based on these findings, we developed the Motor Capacities Scale (MCS) a few years ago in order to meet the need for a specific assessment tool to evaluate the abilities of a patient to perform basic functional tasks regardless of contextual influences (environmental and personal factors). Activities of daily living (ADL) are widely evaluated in clinical settings by dedicated tools [4, 9]. As the initial article pointed out, the MCS does not address the basic ADL such as eating, dressing, bathing etc. The purpose of the MCS is to explicitly focus on elementary motor skills required for ADLs that are typical of people with SCI. We also assumed that the assessment of hand function requires that body movement and trunk stability be taken into consideration.

The MCS is an arm/hand function test with 31 basic tasks to be completed prior to and following upper limb reconstructive surgery (Table 1). The process of selecting the tasks was already reported [10, 11]. All tasks involve the motor skills of the hand and upper extremity. The grid includes four sub-categories: A, B, C and D. Each task takes into account the possible benefits of hand surgery by individually assessing the motor skills of the hand and upper extremity in performing basic tasks. MCS A tests transfers, repositioning and locomotion; MCS B tests the motor capacities for spatial exploration; MCS C tests these capacities for grasping and gripping of the right hand; MCS D assesses them for grasping and gripping of the left hand. “Reference” objects were selected from objects that patients are likely to use in their daily lives. The objects are similar in each assessment phase, before and after surgery. They can thus be used to analyze the two types of grips most often performed by the patient with tetraplegia, the key-pinch and grasp.

Table 1 Motor Capacities Scale (please refer to Appendix 1 for further details on administration and scoring).
Full size table

Notably the psychometric study of the MCS [10, 11] showed that it had excellent construct validity and repeatability. Its convergent and divergent validity was demonstrated and its interrater reliability was extensive with ICC = 0.99. However, the responsiveness to change has never been studied and the standardized instructions for the administration have never been published.

The main objective of this study was to investigate the sensitivity to change and responsiveness of the MCS in individuals with tetraplegia who have undergone upper limb reconstructive surgery. Three secondary objectives were as follows: (a) to determine the ability of the MCS to discriminate those who had undergone a unilateral program and those who had benefited from a bilateral program; (b) to determine the correlation between postoperative MCS scores and patient satisfaction; (c) to estimate the minimal clinically important differences (MCIDs) of the MCS.

Methods

Study participants

An observational study was conducted with a sample of 27 individuals from a cohort of 131 persons with tetraplegia who had undergone functional upper limb surgery between 1 January 1990 and 31 April 2014. Since the first publication on the MCS was in 2004, only patients who had had this surgery from 2005 onwards were invited to join the study on the basis of the following inclusion criteria: adults who were medically and neurologically stable, people who had completed their surgical program. None declined the invitation. They were all informed verbally and in writing and were also asked to sign a written informed consent form.

The “surgical” history included the following data: age at first surgery, time interval between onset of SCI and first surgery, single neurological level and AIS grade in reference to the International Standards for Neurological and Functional Classification of Spinal Cord Injury [12], time lapse between last surgical time and follow-up examination, age at the postoperative assessment. The upper limbs were classified according to the International Classification for Surgery of the Hand in Tetraplegia (ICSHT) which takes into consideration the remaining active muscles graded at least 4/5 MRC in the forearm and the hand [13].

All patients benefited from the surgical procedures and proven techniques described in the literature (Table 2). Restoration of elbow and/or wrist extension was performed if it was prerequisite for hand surgery. Hand surgery included a two-stage approach: (1) finger extension restoration and finger joints stabilization; (2) finger flexion reconstruction.

Table 2 Surgical procedures according to the ICHST classification.
Full size table

Outcome data

The MCS and the Functional Independence Measure (FIM) were administered preoperatively in the month preceding the first intervention and following full completion of the surgical program. Data were collected through direct observation. Examiners were occupational therapists employed at two rehabilitation centers.

The MCS is subdivided in four sub-categories of items (Table 1). For each sub-category, a subscore for A, B, C, and D is determined. The execution of each task is scored differently, depending on the sub-category: a 5-point scale for the 17 items of the sub-category A, a 2-point scale for both items of the sub-category B, a 4-point scale for the 12 items of the sub-categories C and D. A total score is calculated by summing the subscores of each category. The minimum score is 55 and the maximum score is 233 (see Supplementary Appendix 1 for the guide for use).

The FIM is still extensively used for people with SCI but it is not specific to this population. It is a generic functional scale mainly used to track changes in a person’s ability to carry out an activity in an independent manner [14]. It evaluates motor (13 items) and cognitive (5 items) impairments. Each item of the FIM scale is scored from 1 to 7 (where 1 indicates total assistance and 7 complete independence). The scale ranges from 18 to 126. It was demonstrated that the FIM and, more specifically, the “motor” items are able to detect changes over time in patients with tetraplegia [15].

Finally, patients were also asked to provide a satisfaction score for the overall surgical result using a numeric scale from 0: very dissatisfied to 10: very satisfied. The change was considered significant when the satisfaction rate was >6.

Data analysis

Customary descriptive statistics were used to describe the demographics and the outcome data: the total MCS score, MCS A, B, C, and D subscores, and the FIM score before and after surgery, as well as the satisfaction index after surgery. The D’Agostino-Pearson Normality Test [16] was applied on the MCS data and revealed nonnormally distributed data. The correlations between pre- and postoperative data were therefore calculated using the nonparametric Spearman’s rank-correlation coefficient (r) [17]. Hypotheses were confirmed when p was <0.05. Correlation was considered to be excellent if r > 0.91, good if 0.90 < r < 0.71, moderate if 0.5 < r < 0.70, low if 0.31 < r < 0.50 and null if r < 0.30 [18].

To detect the sensitivity to change between preoperative and postoperative scores, we used the Wilcoxon matched pairs signed-rank test. An alpha level of 0.05 (95% confidence interval) was considered significant and boxplots using the Tukey method were built [19].

To calculate standard response mean (SRM) and effect size (ES) for two pairs of scores –for the total MCS score, for each subscore: MCS A, MCS B, MCS C, and MCS D and for the FIM score, we used the original definitions. SRM corresponds to the mean change in scores divided by the standard deviation (SD) of the patients’ change scores. Cohen’s ES was calculated by dividing the mean change score by the SD of the baseline score. An absolute value of 0.2–0.4 was considered as a small degree of responsiveness, a value of 0.5–0.7 was considered as a moderate degree and a value ≥0.80 as a large effect of an intervention [20].

The ESs calculated from the difference between pre- and postoperative data were used to investigate correlations with the satisfaction score. The nonparametric Spearman correlation test computed the p values and correlation coefficients, with a 95% confidence interval by an approximation. We then separated the total sample of participants into two groups: “unilateral surgery” vs. “bilateral surgery”. The Mann–Whitney test was applied to compare ranks and detect differences between groups, considering a 95% confidence interval represented by boxplots (Tukey method).

Last, we estimated the MCID defined as the “smallest difference in score in the domain of interest which patients perceive as beneficial” [21, 22]. The method that we applied to calculate the MCIDs was the anchor-based method combined with the distribution-based method. The “anchor” question was “how much are you satisfied with the overall surgical result” as a reference to determine if the patient is satisfied after completing his surgical program [23]. A score >6 was considered to be at least a good level of satisfaction (7–8) and at best a very good level of satisfaction [9, 10].

Results

Twenty-seven individuals were enrolled in the study. Demographic and surgical data are reported in Table 3. The surgical program was unilateral in 14 people and bilateral for the other 13 people. All patients from group 6 (ICSHT) had function in the triceps. The largest ICHST groups were groups 3, 4, and 5. When the program was bilateral, both elbows were operated in most cases. A complete surgical program (both elbows and both hands) was carried out for four patients. MCS, FIM, and satisfaction scores are presented in Table 4. (See Supplementary Appendix 2).

Table 3 Demographic and surgical data.
Full size table
Table 4 Outcome data: MCS, FIM and satisfaction scores before and after surgery.
Full size table

Correlations and associations between scores

Correlations studies showed excellent correlations between pre- and postoperative total MCS score, MCS A subscore and FIM (Table 5). These findings were corroborated by the scatter plots showing a strong association between the postoperative FIM score with both the postoperative total MCS score (r = 0.83) and the MCS subscore A (r = 0.88) (Fig. 1).

Table 5 Correlation coefficients between post and preoperative data (Spearman Correlation Test).
Full size table
Fig. 1: Correlations and associations between scores.
figure 1

Associations between the postoperative FIM score and the postoperative total MCS score (Graph 1), the MCS A, B, and C subscores (Graph 2), the MCS D subscore (Graph 3).

Full size image

Responsiveness of the different scores

Boxplots showed significant differences for each score before and after surgery (p < 0.05) (Fig. 2). The ES and SRM were moderate or good for MCS B, MCS C, and MCS D (Table 6). For MCS A and the FIM, both SRM and ES showed a small degree of responsiveness. For the total MCS score, the ES value indicated a moderate degree of responsiveness and the SRM was excellent.

Fig. 2: Changes of pre- and postoperative data.
figure 2

Changes are indicated by a superior horizontal connection dotted line when detected by the Wilcoxon matched pairs signed rank test (significant differences p < 0.05).

Full size image
Table 6 Effect size and standardized response mean.
Full size table

Associations with satisfaction scores

The median satisfaction index following surgery was 8 with a 25th percentile of 6 and a 75th percentile of 8.1. Only variations of total MCS score, MCS A subscore and FIM score showed significant associations with satisfaction but with low magnitude (Fig. 3). However, the total MCS score, MCS C subscore and MCS D subscore showed significantly higher ES values in the bilateral surgery group than in the unilateral group (Fig. 4).

Fig. 3: Associations between satisfaction scores (x axis) and the differences pre- and postoperative (y axis) for the total MCS score, MCS A, B, C, and D subscores FIM score.
figure 3

When significant (p < 0.05), the correlation coefficient (r) showing the magnitude of the association was described in the body of the scatted graphs.

Full size image
Fig. 4: Boxplots built by Tukey method showing the differences between pre- and postoperative records for groups separated by unilateral (14 patients) vs bilateral surgery (13 patients) and for total MCS score and MCS A, B, C, and D subscores as well for the Functional Independence Measure (FIM) score.
figure 4

The significant differences (p < 0.05) are indicated by a superior horizontal connection line (Mann–Whitney test).

Full size image

Minimal clinically important differences

The MCIDs revealed low thresholds of change in MCS scores beyond which a patient or a therapist would consider changes as significant and positive (satisfaction rate >6). The corresponding values for the MCIDs are reported in Table 7.

Table 7 Minimal clinically important differences (MCID) defined by satisfaction as a meaningful external anchor combined with distribution-based approaches.
Full size table

Discussion

The high correlation between MCS A and FIM underlines the conceptual link between the two assessments regarding the overall contribution of upper limb reconstructive surgery for patients with tetraplegia. Conversely, the low correlation observed between the MCS B, C, and D subscores and the FIM score reflects the importance that these sub-categories give to the motor functions of the operated upper limb. Importantly, in the original 2004 study [10], a strong relationship was established between MCS and the Sollerman test, known to assess basic motor skills not subject to the “influence” of contextual factors. This emphasized the external validity of the MCS within the applied theoretical framework.

With regard to sensitivity to change, the results of this study show that the MCS as a whole and sub-categories MCS B, C, and D are responsive to the changes provided by upper limb reconstructive surgery. Only MCS A showed low responsiveness consistent with the low sensitivity of the FIM. This is crucial because it indicates that the B, C, and D subscores are the main determinants of sensitivity to change. Yet all scores showed a positive and significant variation as presented in Fig. 2.

The bilateral surgical group (as opposed to unilateral) was associated with greater variations in the total MCS score and the MCS C and D subscores. These results suggest the potential power of the MCS to discriminate between the two groups.

To our knowledge, MCS is the only evaluation scale that attempts to control the influence of environmental factors while addressing the space exploration function and gripping function in its three facets: grasp, hold, and release. By breaking down the evaluation into different sequences, it enables the rehabilitation team and the surgeon to better assess the adverse effects of certain procedures. Last, it should be stressed that MCS does not require any specific material. The objects used for the evaluation come from everyday life. It can be administered in a common environment in a rehabilitation center.

Given the good responsiveness of each sub-category, the MCS is likely to provide relevant information on changes over time. SRM and MCIDs values confirm the ability of the MCS to reflect significant changes.

Limitations and future perspectives

The development of the MCS was prompted by a gap in outcome assessment of upper limb reconstructive surgery in patients with tetraplegia. The findings of this study are limited to people with tetraplegia. For a more complete demonstration of the psychometric properties of the MCS, a control group of nonoperated patients seems unavoidable; the ceiling and floor effects of the MCS must be investigated in both groups and a Rasch analysis would help determine if there are any obsolete elements in the MCS.

Given the inconclusive results for sub-category A, the question of whether it should remain in the grid is relevant. At this stage of the psychometric analysis, it seems premature to remove it. The subscore A appears to be a good measure of the subject’s overall level of motor skills, and it is not impossible that this score will increase with hindsight. It also provides a measure of the subject’s overall disability status.

Moreover, to better demonstrate that MCS is sensitive to small changes, future research should include the evaluation of subscores after each surgery depending on the extent of the surgical program (right elbow, left elbow, right hand, and left hand). This would also provide additional information on the strategies used by patients to perform manual tasks after each surgical stage and also make it possible to question the capacity of each subscore to increase over the course of training and rehabilitation.

The estimation of MCID refers to a significant relationship between changes in MCS scores and patient satisfaction. In our study, satisfaction was assessed only in the postoperative period. Monitoring changes in satisfaction over time from a preoperative baseline to the postoperative period would have been preferred. Moreover, incorporating health and quality of life assessment tools in the MCID estimation would be fruitful to demonstrate the real added value in the daily life of the patient.

Although the MCS is primarily dedicated to the therapists who need to refer to a tool that can value small meaningful changes and generate satisfaction, it should provide the patient with the opportunity to monitor his or her own progress.