Exploratory Person-Centred Competence Profiles Among Italian Physiotherapists: Overall Competence Gradients and Occupational Correlates

Background/Objectives: Competence in physiotherapy is multidimensional, but person-centred analyses may either reveal distinct configurations or primarily reproduce differences in overall level. This secondary analysis examined whether reproducible competence groupings could be identified among Italian physiotherapists and whether they provided occupational information beyond the continuous MSPC total score. Methods: The source dataset comprised 481 physiotherapists. The recurrent eight-person response-pattern subgroup identified in higher-k solutions was excluded a priori from the primary person-centred analysis after case-level review, yielding a primary analytic sample of 473 participants; the full 481-participant analysis was retained as a sensitivity analysis. Five varimax-rotated PCA components were used for k-means clustering (k = 2–5). Cluster stability was assessed by the bootstrap-adjusted Rand index, and sensitivity analyses compared clustering with a median split of the MSPC total score and tested incremental occupational information. Results: In the primary n = 473 analysis, the retained two-group solution comprised 284 and 189 participants, showed modest separation (silhouette = 0.304; Calinski–Harabasz = 223.0; Davies–Bouldin = 1.23) and high bootstrap stability (mean ARI = 0.905; median = 0.917). The groups were broadly parallel and primarily reflected an overall competence gradient. Concordance with a strict median split at the MSPC median (108) was 91.5% (Cohen’s κ = 0.829). Self-employment remained the strongest adjusted correlate of Higher Relative Competence membership (OR = 3.33, 95% CI 2.18–5.09), while each one-category increase in job satisfaction was associated with higher odds of membership (OR = 1.37, 95% CI 1.12–1.68). For self-employment, adding cluster membership to a model already containing the continuous MSPC total score improved fit (likelihood-ratio χ2 = 20.25, p < 0.001), whereas adding the total score after cluster membership did not (χ2 = 1.15, p = 0.284). The full-sample sensitivity analysis reproduced the same substantive pattern. Conclusions: The solution is best interpreted as an exploratory, level-based segmentation of self-perceived competence rather than as qualitatively distinct professional types. Its incremental value appears strongest for employment status. Excluding the eight wording-sensitive cases from the primary analysis did not alter the central occupational finding.