In Canada, the site of the current study, practice recommendations suggest prompt additional work-up for highly suspicious screening abnormalities and "early recall" within 6 to 12 months of low-risk abnormalities for repeat screening. A recent analysis of multiple studies suggested that a delay in treatment of 3 to 6 months for women with symptomatic breast carcinoma is associated with increased rates of recurrence and mortality. However, there are no data to indicate whether a delay in diagnosis alters prognosis for women with screen-detected, asymptomatic breast carcinoma.

In the current, retrospective study investigators analyzed data in the Canadian Breast Cancer Screening Database of 4465 women with invasive breast cancer diagnosed within 3 years of a screening abnormality found between 1990 and 1998. Data were collected on the prognostic indicators of tumor size and axillary lymph node metastases, time interval between abnormal screening and pathological diagnosis of invasive breast cancer, and other information including clinical index of suspicion, family history of breast cancer, and age at diagnosis. Data were analyzed by chi-square tests and logistic regression to evaluate the effect of suspicion level on delay and prognostic factors.

After controlling for suspicion, delays greater than 20 weeks were significant for poorer prognosis. Most women were diagnosed between 4 to 12 weeks after the abnormal screen. Compared with these women, the risk that a tumor was greater than 10-mm in size increased from 0.9-fold waiting 12 to 20 weeks to 1.2-fold for delays 20 to 52 weeks, 1.5-fold for delays 52 to 104 weeks, and 2.1-fold for delays 104 to 156 weeks.

Compared with women diagnosed between 4 to 12 weeks after an abnormal breast screen, the odds that lymph nodes were positive were 1.0 for women waiting 12 to 20 weeks but increased 1.2-fold for delays 20 to 52 weeks, 2.2-fold for delays 52 to 104 weeks, and 3.2-fold for delays 104 to 156 weeks.

Comparison of high-suspicion versus low-suspicion screens indicated that women with highly suspicious screening abnormalities were diagnosed sooner than women without highly suspicious screening abnormalities, (median 31 days versus 47 days, respectively). Highly suspicious screening abnormalities were more likely to have tumors greater than 10-mm in size compared with all other screens (79.4% versus 55.9%, respectively) and more likely to have axillary lymph node metastases (33.9% versus 17.3%, respectively).

On analysis, these data created a "J-shaped" relation between delay to diagnosis and prognostic indicators, and it "reflects a tendency for physicians to more promptly refer women with high-suspicion, worse prognosis screens." The authors called this "suspicion bias" and commented, "This precludes the data from being used to accurately evaluate the effect of delay in the first 3-4 months after an abnormal screen."

The authors conclude, "we have demonstrated that, for delays longer than 12 to 20 weeks from an abnormal breast screen to diagnosis of invasive breast carcinoma, a pattern of increased likelihood of lymph node metastases and increased tumor size began to emerge." The authors add, "irrespective of the cause, reducing delays may improve the prognosis for women and the effectiveness of breast screening programs."