Controversies surrounding Patient-Reported Outcome Measures (PROMs) and the
cumbersome-nature of movement analysis-based (MAB) methods for shoulder function
evaluation make the exploration of alternatives needed. Research aimed at the
simplification of MAB outcome measures had demonstrated previously that the B-B
Score, which relies on two movements only, was valid for out-of-laboratory evaluations
of shoulder function. Nevertheless, further investigations were needed to optimise
testing procedures, test the B-B Score’s capability of acquisition using a user-friendly
device, and critically evaluate its measurement properties in comparison to current
The aim of this thesis was to develop and assess the simplest possible MAB shoulder
function scoring procedure for clinical measurement.
The research included four steps: 1) Optimisation of the B-B Score testing procedure
(Phase 1 study [data-driven]), 2) Comparison of measurements using a smartphone or
an inertial sensor system (Phase 2 study [data-driven]), 3) Validation in frequentlyoccurring
pathologies (rotator cuff conditions, instability, fracture, capsulitis) (Phase 3
study [data-driven]), 4) Benchmarking of the new approach with concurrent MAB
outcome measures and PROMs (literature review).
Amongst the tested methods, the B-B score was optimised by using the mean of three
replicates in the computation of the range of accelerations by angular velocities. The
comparison of easily-used smartphone and reference device showed non-significant
differences and excellent relationships between measurements (Intraclass Correlation
Coefficient [ICC=0.97]). The smartphone’s B-B Score intra-rater and inter-rater
reliability was excellent (ICC=0.92), but limits of agreement could reach up to ±19.4%.
The score was responsive (area under the curve [AUC≥0.70]) and demonstrated
excellent discriminative power between patients and controls (AUC≥0.90), except for
shoulder instability (AUC=0.67). The correlations with PROMs were moderate to high.
The benchmarking established that the measurement properties of the B-B Score compared equivalently with those of PROMs and MAB outcome measures, except for
Shoulder function can be efficiently evaluated using a simple scoring procedure
performed with a smartphone, which facilitates its objective assessment. Further
research is needed to understand how best to reduce the effects of variability
associated with single measurements in order to optimise clinical applicability and to
explore the B-B Score’s properties in other situations requiring functional assessments
of the shoulder.
Keywords: shoulder, shoulder function; outcome assessment; validation studies,
reliability and validity; inertial sensors; smartphone sensors; body-worn sensors;
kinematics; sensitivity and specificity.||en