Developing a rehabilitation intervention difficulty index: A mixed-methods study using NASA-TLX and Borg RPE in a tertiary clinical setting.

backgroundRehabilitation difficulty varies substantially across clinical areas and intervention types, yet few brief, standardised measures exist to quantify difficulty from the therapist's perspective. Understanding rehabilitation difficulty is important for workforce planning, resource allocation, and identifying contexts where therapist support is most needed.

objectiveTo develop and preliminarily evaluate the Rehabilitation Intervention Difficulty Index (RIDI), a mixed-methods measure combining cognitive/affective and physiological dimensions of rehabilitation difficulty, and to identify contextual factors associated with high difficulty.

methodsQuantitative component: 441 rehabilitation sessions from 28 therapists across 12 clinical areas were assessed using the NASA Task Load Index (NASA-TLX; four items: mental demand, temporal demand, effort, frustration) and Borg Rating of Perceived Exertion (RPE). The RIDI was constructed as the mean of the NASA composite score and normalised Borg RPE (0-10 scale). Psychometric properties were evaluated using Cronbach's alpha and exploratory factor analysis. Descriptive statistics, one-way ANOVA, Pearson correlations, and intervention-level summaries were performed. Qualitative component: Semi-structured interviews and focus groups with 20 therapists (10 individual interviews, 10 focus-group participants; 11 physiotherapists, 9 occupational therapists) were conducted across selected clinical areas in the same hospital. Thematic analysis identified key sources of rehabilitation difficulty. Qualitative and quantitative findings were integrated through joint display analysis.

resultsQuantitative findings: RIDI showed good internal consistency for the NASA items (α = 0.80, 95% CI [0.77, 0.83]) and acceptable consistency for the five-item set (α = 0.75, 95% CI [0.72, 0.78]), with a unidimensional factor structure explaining 52% of the variance. RIDI scores differed significantly across clinical areas (F = 9.18, p < 0.001, η² = 0.19), with higher scores in neurorehabilitation and acute neuro settings (both 5.80) and lower scores in outpatient physical therapy (3.63). Among frequently used interventions (n ≥ 5), transfer training and splinting had the highest RIDI scores (6.78 and 6.71), while passive range of motion had the lowest (2.68). RIDI was moderately correlated with session duration (r = 0.28, p < 0.001) but not with therapist experience (r = 0.01, p = 0.78).Qualitative findings: Six themes described sources of difficulty: time demands (78 coded segments, 18 participants), cognitive demands (72, 18), physical demands (65, 17), patient-related factors (57, 16), environmental constraints (44, 15), and coping strategies (33, 14). No substantially new themes were identified after approximately 70% of the data, suggesting thematic saturation.Mixed-methods integration: High RIDI scores in neuro and acute inpatient areas converged with qualitative descriptions of heavy time, cognitive, and physical demands. Coping strategies were discussed much less often in high-RIDI areas (8 segments) than in lower-RIDI areas (23 segments), suggesting that in the most demanding contexts therapists may have fewer opportunities to apply deliberate coping strategies.

conclusionsRIDI showed acceptable reliability and preliminary evidence of construct validity as a therapist-reported measure of rehabilitation difficulty that captures both cognitive/affective and physiological aspects of workload. Variation in RIDI scores across clinical areas and interventions was consistent with meaningful differences in perceived difficulty. The under-representation of coping strategies in high-RIDI contexts suggests that high rehabilitation difficulty may require structural and organisational responses-such as staffing, workflow redesign, and environmental modifications-rather than relying primarily on individual coping. RIDI may provide a practical tool for identifying high-demand areas, informing workforce planning, and targeting support where it is most needed.