Exploratory follow-up-based classification of potential medico-social care needs in patients with multiple sclerosis and mild-to-moderate baseline disability

Abstract Background Multiple sclerosis (MS) is associated with heterogeneous clinical trajectories, disability progression, healthcare resource use, impaired health-related quality of life (HRQL), and varying medico-social needs. This study explored whether early inpatient resource use, longitudinal disability progression, and end-of-follow-up physical HRQL could be integrated into an exploratory, cohort-derived, follow-up-based classification of potential medico-social care needs in patients with mild-to-moderate baseline disability. The classification was intended for descriptive, hypothesis-generating purposes rather than clinical decision-making or resource allocation. This single-center secondary longitudinal complete-case analysis included 238 patients selected from a prospectively followed source cohort of 1,796 patients. Eligible patients had hospitalization data for the first 12 months after diagnosis, baseline and final Expanded Disability Status Scale (EDSS) scores, and end-of-follow-up 36-Item Short Form Health Survey (SF-36) data. Associations were assessed using Spearman’s rank correlation and adjusted partial correlation, with multivariable binary logistic regression performed as a sensitivity analysis. Results Change in EDSS score (ΔEDSS) correlated negatively with the SF-36 Physical Component Summary (PCS; ρ=−0.394; p < 0.001) and Mental Component Summary (MCS; ρ=−0.269; p < 0.001), with a stronger association for PCS. Hospitalization days during the first 12 months showed a weak positive correlation with ΔEDSS (ρ = 0.150; p = 0.020), which remained significant after adjustment (partial r = 0.164; p = 0.013). High inpatient resource-use risk was not independently associated with high disability progression in the sensitivity analysis. The three-domain classification categorized 52 patients (21.8%) as low risk, 109 (45.8%) as moderate risk, and 77 (32.4%) as high risk. High inpatient resource-use risk was more common in patients with moderate than mild baseline disability (51.0% vs. 18.0%; odds ratio [OR] = 4.749; 95% confidence interval [CI]: 2.425–9.298; p < 0.001). Conclusions Early inpatient resource use, longitudinal disability progression, and end-of-follow-up physical HRQL could be integrated into an exploratory follow-up-based classification of potential medico-social care needs. Because the classification was developed and applied within the same complete-case cohort without internal or external validation, its risk-group distribution should be regarded as sample-specific and hypothesis-generating. It should not be used as a prognostic or clinical decision-support tool before prospective validation in independent cohorts.