Risk Factors Associated with Delayed Breast Cancer Diagnoses: What Are We Missing?

Background

Although timely diagnosis is essential to ensuring equitable treatment outcomes, little is known about factors that contribute to delayed breast cancer diagnoses.

Objective

To examine the association of sociodemographic, clinical, and health system factors with delayed breast cancer diagnoses.

Design

This retrospective cohort study used cancer registry data to identify all women with newly diagnosed breast cancers in 2024 within a health system serving a demographically diverse population of > 1 million patients. Time from initial presentation to pathological diagnosis was ascertained via medical chart review and corroborated by physician reviewers.

Participants

Women aged ≥ 18 years with adequate medical records and an incident breast cancer diagnosis in 2024 were included. Individuals assigned male at birth or those with recurrent or progressive disease were excluded.

Main Measures

The primary outcome was a delayed diagnosis defined as ≥ 90 days from first indication to pathological confirmation. Odds ratios (OR) were calculated using univariate and multivariable logistic regression models.

Key Results

Of 431 eligible patients, 93 (22%) experienced a delayed breast cancer diagnosis. Patients with and without delays were comparable with respect to age, race, language, and employment status. Screening mammography was the most common first indication of breast cancer (67%), followed by patient self-identification (20%). Patients with delays more often self-identified masses (34% vs 16%, p < 0.001). In adjusted analyses, delays were associated with self-detected breast masses (OR 2.31; 95% CI, 1.27–4.19), Medicaid insurance (OR 3.03; 95% CI 1.19–7.70), and a greater number of missed or cancelled appointments in the previous 6 months (OR 1.20; 95% CI, 1.07–1.36). Attending more medical appointments was associated with lower odds of delay (OR 0.93; 95% CI, 0.86–0.995).

Conclusion

Delayed breast cancer diagnoses were common in this cohort. However, sociodemographic factors explained little of the variation, suggesting that unmeasured factors play a prominent role.

Graphical Abstract