Table 1.
Study cohort characteristics
| Feature | Comparison between main and subcohort | |
|---|---|---|
| Main cohorta (N = 99,403) | Subcohortb (N = 25,932) | |
| Age, year (SD) | 52.1 (13.8) | 52.8 (12.7)) |
| Gender, n (%) | ||
| Female | 75,002 (75.3) | 19,499 (75.2) |
| Male | 24,541 (24.7) | 6429 (24.8) |
| Unknown | 0 (0.0) | 4 (0.0) |
| Race, n (%) | ||
| Asian | 1302 (1.3) | 380 (1.5) |
| Black | 10,064 (10.1) | 2578 (9.9) |
| Hispanic | 6554 (6.6) | 1887 (7.3) |
| White | 67,237 (67.5) | 18,303 (70.6) |
| Unknown | 14,396 (14.5) | 2784 (10.7) |
| Feature | Comparison among different DMT group | ||
|---|---|---|---|
| No DMT (N = 16,487) | Standard efficacy (N = 12,419) | High efficacy (N = 109) | |
| Age, year (SD) | 57.6 (14.5) | 43.9 (12.0) | 54.0 (11.9) |
| Gender, n (%) | |||
| Female | 12,427 (75.4) | 9380 (75.6) | 83 (76.2) |
| Male | 4058 (24.6) | 3036 (24.4) | 26 (23.8) |
| Unknown | 2 (0.0) | 3 (0.0) | 0 (0.0) |
| Race, n (%) | |||
| Asian | 228 (1.4) | 194 (1.6) | 0 (0.0) |
| Black | 1512 (9.2) | 1400 (11.2) | 14 (12.8) |
| Hispanic | 1116 (6.8) | 1009 (8.1) | 14 (12.8) |
| White | 11,642 (70.6) | 8831 (71.1) | 59 (54.2) |
| Unknown | 1989 (12.0) | 985 (8.0) | 22 (20.2) |
| Cohort duration, days (SD)c | 1342.1 (847.2) | 1490.7 (929.4) | 1828.5 (1040.9) |
| Monotherapy DMT, n (%) | 0 (0.0) | 6856 (55.1) | 51 (48.1) |
| Elixhauser Comorbidity Index, median (IQR)d | 11 (15) | 6 (9) | 10 (12) |
aWe classified an individual as having MS if the person had at least three separate MS-related claims on three separate days within a 1-year period, in any combination of inpatient, outpatient, or DMT-specific pharmacy claims (MS main cohort)
bTo analyze healthcare utilization and expenditure, we then identified from the MS main cohort a subset of 25,932 patients who additionally had continuous enrollment for both medical and pharmacy benefits for at least 3 months preceding and at least 12 months after the index date (MS subcohort). For patients on more than one DMT, the claims attributed to a given DMT treatment were counted only after the treatment was started in analysis of healthcare utilization and costs
cCohort duration was defined as time from the index date until either the termination of insurance coverage at United Healthcare or the end of the study period, whichever occurred first
dElixhauser Comorbidity Index was computed using the R package “comorbidity” and all ICD-9 and ICD-10 codes collected during the follow-up period