Table 3.
Distribution of 1 + hour and 8 + hour outages by quartile of SVI and prevalence of Medicare DME users per 1000
| Metric | Quartile | Quartile values | Total count, N | Median (IQR) count, N | Max count, N | |||
|---|---|---|---|---|---|---|---|---|
| 1+ hour | 8+ hour | 1+ hour | 8+ hour | 1+ hour | 8+ hour | |||
| SVI | Q1 | [0–0.28] | 22,468 | 1185 | 40 (73) | 1 (4) | 195 | 25 |
| Q2 | (0.28–0.53] | 28,228 | 1405 | 57 (92) | 2 (4) | 217 | 22 | |
| Q3 | (0.53–0.77] | 34,911 | 1678 | 68 (99) | 2 (5) | 330 | 26 | |
| Q4 | (0.77–1] | 40,600 | 1888 | 80 (83) | 3 (5)a | 414 | 35 | |
| DME use, per 1000 Medicare enrollees | Q1 | [0–45] | 32,211 | 1616 | 62 (95) | 2 (5) | 234 | 26 |
| Q2 | (45–58] | 33,565 | 1704 | 63 (98) | 2 (5) | 394 | 35 | |
| Q3 | (58–74) | 35,276 | 1765 | 68 (98) | 2 (5) | 414 | 25 | |
| Q4 | (74–478] | 25,155 | 1071 | 42 (82) | 1 (3)a | 217 | 22 | |
Analysis includes 2038 counties with 2+ years of reliable data.
aThe Wilcoxon rank sum test p-value < 0.01 when comparing the median of 8+ hour power outages in (1) counties belonging to the highest SVI (social vulnerability index) quartile versus all other counties and (2) counties belonging to the highest DME (durable medical equipment) use quartile versus all other counties. Power outage data was purchased from PowerOutage.us.