Table 5.
Multivariable models adjusting for patient characteristics, radiologist characteristics, facility characteristics, and a facility random effect (20 019 mammograms, 27 facilities)*
| Characteristic | Odds of having a positive mammogram given no cancer diagnosis (false-positive rate),† OR (95% CI) | Odds of having a positive mammogram given a cancer diagnosis (sensitivity),‡ OR (95% CI) | Odds of having a cancer diagnosis given a positive mammogram (PPV2),§ OR (95% CI) |
| Facility structure and organization | |||
| Clinical services | |||
| Interventional services offered (FNA, core or vacuum-assisted biopsy, cyst aspirations, needle localization, or other procedures) | |||
| No | 1.0 (referent) | 1.0 (referent) | 1.0 (referent) |
| Yes | 1.97 (0.94 to 4.10) | 1.75 (0.47 to 6.45) | 1.41 (0.50 to 4.00) |
| Facility is currently short staffed (ie, not enough radiologists) | |||
| Strongly disagree/disagree/neutral | 1.0 (referent) | 1.0 (referent) | 1.0 (referent) |
| Agree/strongly agree | 1.21 (0.92 to 1.58) | 1.36 (0.79 to 2.33) | 0.83 (0.62 to 1.11) |
| Financial and malpractice | |||
| Profit status | |||
| Nonprofit | 1.0 (referent) | 1.0 (referent) | 1.0 (referent) |
| For profit | 0.97 (0.72 to 1.32) | 0.91 (0.52 to 1.59) | 0.87 (0.63 to 1.20) |
| How have medical malpractice concerns influenced recommendations of diagnostic mammograms, ultrasounds, or breast biopsies at your facility following screening mammograms? | |||
| Not changed | 1.0 (referent) | 1.0 (referent) | 1.0 (referent) |
| Moderately increased/greatly increased | 1.48 (1.09 to 2.01)‖ | 1.74 (0.94 to 3.23) | 0.81 (0.57 to 1.15) |
| Scheduling process | |||
| Average wait time to schedule diagnostic mammogram | |||
| ≤3 d | 1.0 (referent) | 1.0 (referent) | 1.0 (referent) |
| >3 d | 1.09 (0.76 to 1.57) | 1.17 (0.57 to 2.41) | 0.99 (0.66 to 1.47) |
| Do women wait for interpretation of diagnostic mammogram? | |||
| No, never/yes, some of the time | 1.0 (referent) | 1.0 (referent) | 1.0 (referent) |
| Yes, all of the time | 1.02 (0.72 to 1.45) | 1.17 (0.58 to 2.35) | 0.88 (0.62 to 1.26) |
| Interpretation and audit processes | |||
| Interpretive process | |||
| What percentage of diagnostic mammograms are interpreted on-site? | |||
| 0 | 1.0 (referent) | 1.0 (referent) | 1.0 (referent) |
| 100 | 0.74 (0.30 to 1.85) | 3.12 (0.54 to 18.09) | 0.74 (0.19 to 2.86) |
PPV2 = positive predictive value of biopsy; OR = odds ratio; CI = confidence interval; FNA = fine-needle aspiration.
False-positive rate is defined as the percentage of mammograms without a cancer diagnoses within 1 year that also have a BI-RADS assessment 4, 5, 0 or 3 with a recommendation for biopsy, surgical consultation, or FNA.
Sensitivity is defined as the percentage of mammograms with a cancer diagnoses within 1 year that also have a BI-RADS assessment 4, 5, 0 or 3 with a recommendation for biopsy, surgical consultation, or FNA.
PPV2 is defined as the percentage of mammograms with either a BI-RADS assessment 4, 5, 0 or 3 with a recommendation for biopsy, surgical consultation, or FNA that result in a cancer diagnosis within 1 year.
Value is statistically significant, P = .01 (two-sided F test).