Table 3 FEV1 and FVC quality assessment in primary care units
From: Deep learning for spirometry quality assurance with spirometric indices and curves
GPs with regular practice | GPs with AI-assistance | |||
|---|---|---|---|---|
Task, n (%) | Month 0 | Month 1 | Month 2 | P value |
Acceptable | N = 171 curves | N = 431 curves | N = 840 curves | |
FEV1 maneuvers | 140 (81.9%) | 359 (83.3%) | 771 (91.8%) | < .0001 |
FVC maneuvers | 120 (70.2%) | 343 (79.6%) | 751 (89.4%) | < .0001 |
Usable (including acceptable) | N = 171 curves | N = 431 curves | N = 840 curves | |
FEV1 maneuvers | 151 (88.3%) | 396 (91.9%) | 833 (99.2%) | < .0001 |
FVC maneuvers | 152 (88.9%) | 398 (92.3%) | 833 (99.2%) | < .0001 |
Good quality ratings (A, B, or C grades) | N = 72 files | N = 148 files | N = 281 files | |
FEV1 tests | 51 (70.8%) | 117 (79.1%) | 258 (91.8%) | < .0001 |
FVC tests | 38 (52.8%) | 107 (72.3%) | 250 (89.0%) | < .0001 |