Trends in Breast Cancer Screening in a Safety-Net Hospital During the COVID-19 Pandemic

cross-sectional in at We obtained per during 2019 from after the implementation of a new EHR. The number of screening mammograms per was plotted against the 2019 baseline. Proportions of completed tests by phase of the pandemic (pre–COVID-19, first stay-at-home order, reopening, and second stay-at-home order) were compared by race/ethnicity and age with 2-sided, 2-sample proportion tests. Race/ethnicity was used as a proxy for the disproportionate burden of COVID-19 and experiences of individual and systemic racism experienced by minority communities. Analyses were conducted with Stata, version 16 (StataCorp LLC). P < .05 was used to determine significance. We followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for cross-sectional studies. 5 Deidentified data collected for quality improvement activities study review.


Introduction
National estimates project COVID-19 negatively influenced cancer screening, leading to an estimated deficit of 3.9 million breast cancer (BC) screenings among US adults. 1,2 In San Francisco, California, low-income neighborhoods disproportionately affected by COVID-19 bear the burden of higher BC stage at diagnosis. 3,4 We sought to evaluate the association of COVID-19 and BC screening in a safety-net hospital in San Francisco.

Methods
This cross-sectional study evaluated trends in BC screening at an urban integrated health system's safety-net hospital. We obtained the number of screening mammograms per month during 2019 from electronic health record (EHR) data, and aggregate numbers between September 1, 2019, and January 31, 2021, after the implementation of a new EHR. The number of screening mammograms per month was plotted against the 2019 baseline. Proportions of completed tests by phase of the pandemic (pre-COVID-19, first stay-at-home order, reopening, and second stay-at-home order) were compared by race/ethnicity and age with 2-sided, 2-sample proportion tests. Race/ethnicity was used as a proxy for the disproportionate burden of COVID-19 and experiences of individual and systemic racism experienced by minority communities. Analyses were conducted with Stata, version 16 (StataCorp LLC). P < .05 was used to determine significance. We followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for cross-sectional studies. 5 Deidentified data collected for quality improvement activities does not require approval from The University of California, San Francisco institutional review board; this study was therefore exempted from review.

Results
A total of 9291 screening mammograms were performed from January 1, 2019, to January 31, 2021: The term x represents the mean number of total appointments or missed appointments during pre-COVID-19 (September 2019 to January 2020) and the reopening phase (June to November 2020). a The pre-COVID-19 phase was defined as September 2019 to January 2020.  (Figure 2). The proportion of completed mammograms was lowest among Black women at all time points, younger women during the first stay-at-home order, and women aged 70 years or older during the second stay-at-home order.

Discussion
The reduction in the cumulative number of mammograms suggests a substantial deficit of missed BC screening, which may worsen preexisting disparities. Our results are consistent with those of reports that found discontinuation of BC screening in April 2020. 2,6 In contrast to reports showing recovery of screening volumes, 2,6 our data highlight persistent low BC screening volumes and an absolute decrease in the proportion of completed mammograms among Latinx and Black women. We  Phases of the COVID-19 pandemic were defined to concur with the implementation of local regulations: pre-COVID-19 from September 2019 to January 2020, stay-at-home order 1 from February to May 2020, reopening from June to November 2020, and stayat-home order 2 from December 2020 to January 2021. Proportions were calculated with the number of completed tests as numerator over the sum of the number of missed appointments and completed tests as denominator.
a Statistically significant difference in proportions when compared with pre-COVID-19 for the same racial/ethnic or age group (P < .05 for the test of proportions).