Validity of the Danish version of Multimorbidity Treatment Burden Questionnaire in patients with persistent spinal
Jacob S Gandløse1,2, Jonathan Vela2,3, Thorvaldur S Palsson1,2
1Department of Physiotherapy and Occupational Therapy, Aalborg University Hospital, Aalborg, Denmark.
Abstract:
BackgroundPatients with persistent spinal pain often have comorbidities, increasing treatment burden. The Danish MTBQ-DK measures this burden, but its validity in this population is unknown.ObjectiveEvaluate the construct validity, internal consistency, convergent validity, and discriminative ability of the MTBQ-DK in patients with persistent spinal pain.MethodsPatients (n = 360) referred to the Department of Rheumatology, Aalborg University Hospital were included and completed self-report questionnaires, including the MTBQ-DK, PHQ-2, GAD-2, EQ-5D-5L (incl. EQ-VAS), and demographic items such as education and BMI. Clinical data were extracted from patient records, including ICD-10 diagnosis codes (used to determine number of chronic conditions and healthcare utilization) and use of pain medication. Construct validity was assessed using confirmatory factor analysis, and ordered logistic regression examined associations across treatment burden categories. Convergent validity, internal consistency and questionnaire properties were examined as well.ResultsA modified CFA supported the construct validity of the MTBQ-DK. In the ordered logistic regression, individuals with higher BMI (OR = 1.28, p = 0.02) and a greater number of ICD-10 diagnosis codes (OR = 1.07, p = 0.01) had significantly increased odds of being classified into a higher treatment burden category, compared to those with no reported treatment burden. Conversely, individuals not using pain medication had lower odds of being in a higher burden group (OR = 0.26, p < 0.01). These results indicate that the MTBQ-DK burden categories meaningfully differentiate between patient subgroups. Convergent validity showed fair correlations with PHQ-2 (ρ = 0.36), GAD-2 (ρ = 0.40), and chronic disease count (ρ = 0.28) (all p < 0.001). Negative correlations were found with EQ-5D-5L (ρ = -0.44) and EQ-VAS (ρ = -0.46) (both p < 0.001). Internal consistency was acceptable (Cronbach's α = 0.85). A floor effect ranged from 33.2% to 70.6%.ConclusionThe MTBQ-DK demonstrates acceptable validity for assessing treatment burden in patients with persistent spinal pain. Construct validity was supported by confirmatory factor analysis, internal consistency was acceptable, and convergent validity showed expected correlations. Ordered logistic regression confirmed meaningful differentiation between treatment burden groups, highlighting its potential as a tool to identify patients needing additional support.
