Table.

Comparison of DMT Use and Time to DMT Initiation Overall and by Type Among pwMS Who Resided in Rural vs Urban Areas

	DMT use during the 1-y postindex observation period (2019/2020) (odds ratio [95% CI])		Time from MS incident date to DMT initiation between 2008 and 2021 (hazard ratio [95% CI])
	Base model	Explaining model	Base model	Explaining model
Residence (rural vs urban)
Overall	0.83 (0.69–0.99)	0.89 (0.74–1.07)	0.89 (0.79–1.01)	N/A
Induction/higher-efficacy therapy	0.74 (0.57–0.95)	0.74 (0.57–0.97)	0.84 (0.69–1.02)	N/A
Older/lower-efficacy therapy	1.01 (0.83–1.23)	N/A	0.91 (0.80–1.03)	N/A

Abbreviations: DMT = disease-modifying therapy; N/A = not applicable; pwMS = people living with multiple sclerosis.

Bolded numbers indicate a statistical significance at p < 0.05. The base logistic and Cox models included the confounders of age, sex, the Charlson Comorbidity Index score, years living with MS (logistic model), and year of the MS incident date (Cox model); comorbidities were not included to avoid collinearity with the Charlson Comorbidity Index score. When statistical significance was observed between rural and urban residence, an additional model (‘explaining model’) added sociodemographic status (mediator). See eTable 7 for the full logistic and Cox models.