All 34 categories tested in a proposed ICF framework for documenting musculoskeletal problems in Spanish primary care physical therapy cleared the study's retention threshold, according to a multicenter analysis. The framework groups categories around body functions, activities and participation, and environmental factors. Their reported prevalence ranged from 23.3% to 94.3%, and every category remained retained in the disability and comorbidity subgroups the researchers examined.
Patients and physical therapists rated relevance on a 100-millimetre visual analogue scale, or VAS, scored from 0 to 100 millimetres. For Body Functions and Activities and Participation, a score below 5 millimetres counted as no problem, while a score from 5 to 100 millimetres counted as a problem. Environmental Factors were recorded as relevant context, and a category was retained when at least 5% of valid observations met the rule.
How the categories were tested
The specific ICF Core Set itself comprised 42 categories, but body-structure categories were excluded from the standardized form, so the analysis assessed 34 categories. Adults were eligible if they had a physician-diagnosed musculoskeletal disorder, had been referred from primary care and were sufficiently proficient in Spanish. Non-musculoskeletal cases and hospital-based specialist referrals were excluded, and participation required written informed consent.
The analysis included 387 patients, whose mean age was 57.8 years, with a range from 20 to 90 years; 67.8% were women. Low back pain accounted for 24.3% of referrals and upper-limb tendinopathies for 16.5%. The dataset also contained 387 patient-physical therapist dyads, with recruitment involving 71 participating therapists and 47 primary care physical therapy units.
The sample-size calculation set a target of 385 adults using 95% confidence and a plus-or-minus 5% margin of error. Recruitment was consecutive but non-probability based, so the calculation was not intended to establish that the participants represented the wider population.
This was a multicenter, cross-sectional study in the public SACYL healthcare system in Castilla y León. Patients and physical therapists rated the categories independently, creating paired views of the same cases. The protocol received ethics approval on February 27, 2024, under reference CEIm 3096, and was prospectively registered as NCT06326229.
The pattern behind the result
Pain dominated the patient-reported results. It was identified by 94.3% of patients and had the highest average VAS score, 55.7 millimetres with a standard deviation of 29.0 millimetres. Joint mobility was also common, at 81.4%. For environmental context, health services, systems and policies were relevant in 80.8% of cases, and health professionals in 79.8%.
The researchers also compared the category ratings with WHODAS 2.0 and SCQ scores. WHODAS showed moderate positive associations with Body Functions, Activities and Participation, and their combined summary, with reported correlation values from 0.329 to 0.357. SCQ associations were weaker, from 0.178 to 0.202. Neither measure was significantly associated with Environmental Factors.
Patient and therapist ratings also corresponded reasonably closely overall. Intraclass correlation coefficients, a statistic for absolute agreement between paired scores, ranged from 0.697 to 0.958, a span the study described as moderate to excellent. The authors treated these figures as correspondence between complementary perspectives rather than strict inter-rater reliability. Twenty comparisons remained statistically significant after a correction for multiple testing, compared with 27 before adjustment, but effect sizes were generally small. The largest average difference was 7.83 millimetres, while 5.7% to 29.3% of dyads exceeded the study's 15-millimetre benchmark.
One category showed how group-level agreement can coexist with more noticeable differences for individual patients. For mobility of joint functions, patients rated the category 7.83 millimetres higher on average than therapists did. Among 386 valid dyads, 113, or 29.3%, differed by more than 15 millimetres. That 15-millimetre mark was an interpretive benchmark, not a validated clinical cutoff, so the study does not establish that these individual differences were clinically meaningful.
Why the finding remains preliminary
The findings should be read as local and preliminary. Because the design was cross-sectional, it could not evaluate longitudinal measurement properties. All participants came through one public provider, and non-probability sampling limited representativeness and transferability to other services.
The study was supported by SACYL grant GRS/3378/C/2025 and by a University of Valladolid research stay under MOVILIDAD INVESTIGADORES E INVESTIGADORAS UVa 2026; the funders reportedly had no role in the study's conduct or manuscript. The authors said raw data supporting the conclusions would be made available without undue reservation.
Paper data and sources
Original title: Empirical evaluation of an ICF core set for managing musculoskeletal conditions in primary care physical therapy services in Spain: a multicenter study
Authors: Hernández-Lázaro H, Jiménez-del-Barrio S, Mingo-Gómez M et al.
Journal/Repository: Frontiers in rehabilitation sciences
Status: Peer-reviewed
First online: 2026-10-07
DOI: Not available
Original paper · Full text