Table 2.
Radiologist demographic and clinical practice characteristics and recall rate (N = 124 radiologists)*
Characteristic | No. | % | Recall rate | Recall OR | 95% CI | P |
---|---|---|---|---|---|---|
Radiologist demographics | ||||||
Age, y | ||||||
35–44 | 37 | 29.8 | 12.1 | 1.00 | (referent) | .001 |
45–54 | 48 | 38.7 | 10.6 | 0.88 | 0.86 to 0.9 | |
≥55 | 39 | 31.5 | 8.5 | 0.78 | 0.76 to 0.8 | |
Sex | ||||||
Male | 96 | 77.4 | 9.8 | 1.00 | (referent) | .47 |
Female | 28 | 22.6 | 11.4 | 1.09 | 0.86 to 1.37 | |
Practice type | ||||||
Work full-time | ||||||
No | 32 | 26.0 | 10.9 | 1.00 | (referent) | .11 |
Yes | 91 | 74.0 | 9.9 | 0.95 | 0.93 to 1.04 | |
Affiliation with an academic medical center | ||||||
Yes | 20 | 16.1 | 9.8 | 1.00 | (referent) | .42 |
No | 104 | 83.9 | 10.3 | 1.13 | 0.83 to 1.54 | |
General experience in breast imaging | ||||||
Years of mammography interpretation | ||||||
<10 | 28 | 22.7 | 11.8 | 1.00 | (referent) | .005 |
10–19 | 57 | 46.3 | 10.8 | 0.81 | 0.64 to 1.02 | |
≥20 | 38 | 30.9 | 8.6 | 0.66 | 0.51 to 0.85 | |
% of time spent working in breast imaging | ||||||
<20 | 55 | 45.1 | 9.2 | 1.0 | (referent) | .015 |
20–39 | 52 | 42.6 | 11.3 | 1.30 | 1.06 to 1.61 | |
≥40 | 15 | 12.3 | 9.5 | 0.92 | 0.68 to 1.44 | |
No. of mammograms interpreted in the past year | ||||||
≤1000 | 31 | 25.2 | 7.6 | 1.0 | (referent) | .002 |
1001–2000 | 46 | 37.4 | 11.1 | 1.51 | 1.12 to 1.82 | |
>2000 | 46 | 37.4 | 10.4 | 1.29 | 0.97 to 1.35 | |
% of mammograms interpreted in the past year that were screening mammograms | ||||||
≤50 | 10 | 8.1 | 9.5 | 1.0 | (referent) | .77 |
51–75 | 52 | 42.3 | 10.0 | 1.06 | 0.73 to 1.55 | |
76–100 | 61 | 49.6 | 10.7 | 0.99 | 0.68 to 1.44 |
Frequencies may not add up to 124 due to missing responses. CI = confidence interval. OR = odds ratio. Recall rate is the rate of calling a mammogram positive. Odds ratios, confidence intervals, and omnibus P values (two-sided) were calculated using logistic regression adjusting for patient age, breast density, and prior mammography.