Table 4.
Radiologist characteristics by categories of screening and diagnostic specificity.1
| Low to average performance in screening or diagnostic performance | Top performers | ||||
|---|---|---|---|---|---|
| Adjusted Screening Performance: | Low to Average | High | Low to Average | High | P value comparing highest performers to all others combined |
| Adjusted Diagnostic Performance: | Low to Average | Low to Average | High | High | |
| N=90 | N=31 | N=37 | N=37 | ||
| Gender, n (%) | 0.047 | ||||
| Female | 34 (37.8%) | 7 (22.6%) | 12 (32.4%) | 6 (16.2%) | |
| Male | 56 (62.2%) | 24 (77.4%) | 25 (67.6%) | 31 (83.8%) | |
| Age group, n (%) | 0.144 | ||||
| 32 – 44 | 26 (28.9%) | 4 (12.9%) | 11 (29.7%) | 6 (16.2%) | |
| 45 – 54 | 36 (40%) | 10 (32.3%) | 14 (37.8%) | 11 (29.7%) | |
| 55 or older | 28 (31.1%) | 17 (54.8%) | 12 (32.4%) | 20 (54.1%) | |
| Academic Medical Center Affiliation, n (%) | 0.325 | ||||
| None | 73 (82%) | 28 (90.3%) | 30 (83.3%) | 30 (83.3%) | |
| Adjunct | 5 (5.62%) | 0 (0%) | 3 (8.33%) | 4 (11.1%) | |
| Primary | 11 (12.4%) | 3 (9.68%) | 3 (8.33%) | 2 (5.56%) | |
| Fellowship training, n (%) | 0.311 | ||||
| Not-fellowship trained | 81 (90%) | 29 (93.5%) | 34 (91.9%) | 36 (97.3%) | |
| Fellowship trained | 9 (10%) | 2 (6.45%) | 3 (8.11%) | 1 (2.7%) | |
| Years of mammography experience, n (%) | 0.106 | ||||
| 0 – 9 | 28 (31.1%) | 5 (16.1%) | 11 (31.4%) | 5 (13.9%) | |
| 10 – 19 | 32 (35.6%) | 8 (25.8%) | 12 (34.3%) | 11 (30.6%) | |
| 20 or more | 30 (33.3%) | 18 (58.1%) | 12 (34.3%) | 20 (55.6%) | |
| Percent of time spent in breast imaging, n (%) | 0.124 | ||||
| 0 – 19% | 19 (21.8%) | 6 (19.4%) | 12 (32.4%) | 9 (27.3%) | |
| 20 – 39% | 23 (26.4%) | 8 (25.8%) | 8 (21.6%) | 14 (42.4%) | |
| 40 – 79% | 15 (17.2%) | 4 (12.9%) | 8 (21.6%) | 2 (6.06%) | |
| 80 – 100% | 30 (34.5%) | 13 (41.9%) | 9 (24.3%) | 8 (24.2%) | |
| Hours per week spent in breast imaging, n (%) | 0.432 | ||||
| 0 – 8 | 16 (18.8%) | 4 (13.3%) | 13 (35.1%) | 10 (30.3%) | |
| 9 – 16 | 27 (31.8%) | 11 (36.7%) | 13 (35.1%) | 12 (36.4%) | |
| 17 – 32 | 17 (20%) | 5 (16.7%) | 4 (10.8%) | 6 (18.2%) | |
| 33 or more | 25 (29.4%) | 10 (33.3%) | 7 (18.9%) | 5 (15.2%) | |
1
Percentages shown correspond to non-missing observations. P-values compare the 37 top performers (4th column) to the 158 others combined (columns 1-3) using Fisher's Exact Test. Low, Average, and High performance categories were assigned relative to the 33rd and 67th tertiles of the distribution of radiologist-specific random effects for screening specificity (−0.324, 0.210) and diagnostic specificity (−0.192, 0.183). Missing data on survey questions: Academic medical center affiliation n=3; Years of mammography experience n=3; Percentage of time spent in breast imaging n=7; Hours per week spent in breast imaging n=10.