Table 3. EHR-derived MS severity score (MSSS) captures the difference between progressive MS and relapsing-remitting MS patients.
| Discovery Setb (n = 329) | Validation Setb (n = 213) | |||||
| Outcomea | PPMS/SPMS (n = 34) Mean (SE) | RRMS (n = 295) Mean (SE) | p-value | PPMS/SPMS(n = 25) Mean (SE) | RRMS (n = 188) Mean (SE) | p-value |
| Observed MSSS | 3.86 (0.27) | 0.86 (0.10) | 1.55E-23 | 4.36 (0.29) | 0.73 (0.11) | 3.32E-26 |
| Derived MSSS | 2.90 (0.18) | 0.98 (0.06) | 9.42E-23 | 3.22 (0.18) | 1.10 (0.07) | 1.56E-12 |
Observed MSSS is based on actual data from MS patients who receive care at the Partners MS Center. Derived MSSS is based on algorithm with 40% frequency cut-off for EHR variables.
Patients with known MS disease category were divided into a discovery set (n = 329, including 34 PPMS/SPMS patients and 295 RRMS patients) and a validation set (n = 213, including 25 PPMS/SPMS and RRMS patients). For the observed measure of MSSS, ANOVA was performed and the comparison was adjusted for sex, age of symptom onset and disease duration as covariates. For derived surrogate measure of MSSS, t-test was performed. The effects of sex, age of symptom onset, and disease duration are accounted for in the derivation of the surrogate measure of MSSS.
Abbreviations: BPF, brain parenchymal fraction; MSSS, multiple sclerosis severity score; PPMS, primary progressive multiple sclerosis; RRMS, relapsing-remitting multiple sclerosis; SPMS, secondary progressive multiple sclerosis.