Table 2.
Sex and age-specific differences in health-care use for any reason between multiple sclerosis cases and controls: 5 years before the index date in the health administrative and clinical cohorts.
| Administrative cohort | Clinical cohort | ||||||
|---|---|---|---|---|---|---|---|
| MS casesa | Controlsa | MS casesa | Controlsa | ||||
|
Crude rate (per person-year)/ crude proportion (%) |
Crude rate (per person-year)/ crude proportion (%) |
aRR (95% CI) |
Crude rate (per person-year)/ crude proportion (%) |
Crude rate (per person-year)/ crude proportion (%) |
aRR (95% CI) | ||
| Physician visits | Males | 11.4 | 6.83 | 1.67 (1.57–1.77) b | 6.70 | 6.19 | 1.09 (0.91–1.30) |
| Females | 14.9 | 10.2 | 1.45 (1.40–1.51) b | 12.4 | 10.5 | 1.17 (1.06–1.30) | |
| < 30 | 10.5 | 7.72 | 1.39 (1.27–1.52) b | 9.34 | 8.41 | 1.13 (0.94–1.35) | |
| 30–49 | 13.1 | 8.87 | 1.49 (1.43–1.55) b | 11.1 | 9.69 | 1.14 (1.02–1.27) | |
| ≥50 | 16.9 | 10.7 | 1.58 (1.49–1.68) b | 13.9 | 10.4 | 1.28 (0.97–1.68) | |
| Hospital admissions | Males | 0.202 | 0.117 | 1.73 (1.54–1.95) b | 0.0832 | 0.0896 | 0.93 (0.61–1.45) |
| Females | 0.249 | 0.171 | 1.43 (1.34–1.54) b | 0.194 | 0.181 | 1.07 (0.88–1.30) | |
| < 30 | 0.173 | 0.130 | 1.35 (1.13–1.62) b | 0.144 | 0.143 | 1.02 (0.70–1.48) | |
| 30–49 | 0.207 | 0.145 | 1.43 (1.32–1.55) b | 0.161 | 0.161 | 1.00 (0.80–1.23) | |
| ≥50 | 0.312 | 0.189 | 1.65 (1.49–1.82) b | 0.261 | 0.174 | 1.40 (0.84–2.34) | |
| Any prescription filled | Males | 94.0 | 87.3 | 1.08 (1.06–1.09) b | 82.8 | 80.9 | 1.02 (0.96–1.09) |
| Females | 97.2 | 93.0 | 1.04 (1.04–1.05) b | 93.8 | 89.9 | 1.04 (1.02–1.07) | |
| < 30 | 96.7 | 91.9 | 1.05 (1.04–1.07) | 91.8 | 89.6 | 1.02 (0.98–1.07) | |
| 30–49 | 95.6 | 90.5 | 1.06 (1.05–1.06) | 90.8 | 86.3 | 1.05 (1.02–1.08) | |
| ≥50 | 97.6 | 93.1 | 1.05 (1.04–1.06) | 91.6 | 91.7 | 1.00 (0.93–1.07) | |
aRR, adjusted rate ratio; CI, confidence interval.
bold = P < 0.05 for sub-group.
aphysician visits and hospital admissions represented as crude rates; “any” prescription filled represented as crude proportions.
bBased on the likelihood ratio test, the rate ratios statistically differed (P < 0.05) between males and females or between the age categories: < 30, 30–49 and ≥50 years.
The post-hoc Bonferroni adjusted P-values derived when each age-group was directly compared against each other in the administrative cohort were as follows: physician visits (< 30 vs. 30–49: P = 0.54; < 30 vs. ≥50: P = 0.05; 30-49 vs. ≥50: P = 0.27), hospitalizations (< 30 vs. 30-49: P = 1; < 30 vs. ≥50: P = 0.14; 30-49 vs. ≥50: P = 0.10), prescriptions-filled (< 30 vs. 30–49: P = 1.00; < 30 vs. ≥50: P = 1.00; 30–49 vs. ≥50: P = 0.58).