Self-reported outcomes and work participation before and during co-design of a municipal rehabilitation initiative for individuals with long-lasting back problems – a case series

Introduction: The `My Back´ research project aimed to develop, test, and evaluate a co-designed municipal back rehabilitation initiative, involving micro-changes in a complex organization. Translating guideline recommendations into the context of municipal rehabilitation is challenging due to the intricate nature of patients' problems and the system’s inadequate interdepartmental communication. Furthermore, little is known about outcomes for people with back pain in municipal rehabilitation.

Purpose: The purpose of this study was to compare outcomes of self-management, quality of life, disability, and work participation before and during iterative testing of the co-designed actions.

Material and Methods: Individuals referred for rehabilitation for back problems between November 2020 and December 2023 received questionnaires at baseline and at 3-, 6- and 12-months. This case series compared changes in outcomes from baseline to 3- and 6-month follow-ups in the periods before and during iterative testing of actions. Outcome measures were the General Self-Efficacy Scale (GSE), Quality of Life (EQ5D), and Oswestry Disability Index (ODI). Raw data from each individual are shown as spaghetti plots in Figure 1. Unadjusted mixed-linear models generated beta coefficients with 95% confidence intervals (CI).

Work participation was defined by the Danish Register for Evaluation of Marginalisation (DREAM) based on the individuals’ financial support and their employment rate. Data were extracted for Svendborg and four comparable municipalities. Data were dichotomized to participating in work or not at 14 weeks after referral, and from logistic regression, an odds ratio for work participation (estimated). The model included time-phase (before and during), municipality (Svendborg versus other), sex, age, and educational level.

Results: Questionnaires were answered by 27% of all participants at baseline, with 33 individuals in the before period (mean age = 58, SD 13, 49% female) and 92 during the period of iterative testing (mean age = 62, SD 14.0, 57% female). At 3 months, 16 and 57 individuals responded, and at 6 months, 27 and 45, respectively. No statistically significant differences were shown between the two periods in GSE: 0.3 [95% CI -2.4;3.1]; QoL: -3.1 [-14.5;6.9]; ODI: 3.7 [-2.9;10.2]. However, a statistically significant effect over time was found for QoL and ODI (p< 0.03 and < 0.000) but not for GSE (p=0.156 and 0.289).

Data from registries were available on work participation for 1415 people between 18 and 70 years of age (n=296 before and 72 during (interim analyses) in Svendborg) and other municipalities (n=816 before and 231 during). The odds ratio for work participation for people in ‘My Back’ in Svendborg during iterative testing was 1.88 [0.97;3.61] (Figure 2).

Conclusion: No short-term between-group differences were observed; however, positive changes were observed within both periods in two of the three outcomes. Work participation did not show statistically significant differences in the intervention municipality, either when comparing the before and during periods or in comparisons with other municipalities. Future analyses will estimate the long-term effects of co-designed back rehabilitation on people with long-lasting back problems, emphasizing the necessity of testing and monitoring over a longer period when dealing with co-design of micro-changes in complex organizations.