Abstract
Objective: To apply system dynamics methodology to study the evolution of temporary work disability in the context of pain-related musculoskeletal disorders [MSD-TWD].
Methods: Data were obtained from the MSD-TWD program records on 13 077 patients with acute disability [7805 in the control group [CG]; 5272 in the intervention group [IG]] who suffered 16 297 episodes of temporary work disability. Samples were randomized into two homogeneous sub-samples for validation purposes. The dynamic model developed with system dynamics methodology included 20 variables [five levels, seven rates, two auxiliary variables, six parameters], five differential equations, and eight algebraic equations. A sensitivity analysis of various scenarios was carried out.
Results: Episodes were described according to their duration; short-term and long-term MSD-TWD. By tuning the model parameters, the actual survival curves of both groups in the two sub-samples were almost exactly reproduced. An explicit temporal expression of the survival curve was used in solving the equations of the dynamic model. The mean duration of short-term episodes was 18 days [d] in the CG and 14 d in the IG, while the mean duration of long-term episodes was 98 and 57 d, respectively. The conversion rate from short-term to long-term work disability was 7.3% in the CG compared with 3.5% in the IG. The model was cross-validated. Sensitivity analysis showed no overlap between the CG and IC curves.
Conclusion: The dynamic model proposed is an excellent approach to the generic temporary work disability process, and is also able to explain the effects of intervention on the process.