Health Equity Within Inequity: Timing of Diagnostic Breast Cancer Care in an Underserved Medical Population

Discussion

In our retrospective cohort study, after implementation of a multidisciplinary breast program and mammographic database, there was no change in timeliness of breast cancer treatment for patients at our safety-net hospital. There were no statistically significant differences in stage at diagnosis and timing of breast cancer workup or treatment between different racial groups. Although protocols were established and patients had equitable care, patients at our safety-net hospital continued to experience prolonged times for workup and treatment compared to other settings with more economic resources (11).

Delays or incomplete diagnostic evaluations and treatment after an imaging cancer-related abnormality is discovered are worse for African-American and Hispanic patients in comparison to non-Hispanic Caucasians (12). Our data show that time to treatment extended to over a month. Previous studies have also noted similar time frames of greater than 1 month of delay in treatment of African-Americans when compared to non-Hispanic Caucasians (13, 14). Interestingly, in our study, all patients, regardless of race, received equitable care. There were no differences in timing of diagnostic work up and treatment between ethnic minorities, such as Black and Hispanic patients, and White patients. However, our patients overall experienced an increase in time of diagnostic work up and treatment compared to patient populations of higher socioeconomic status (11). Delays in care ultimately lead to a decrease in survival (15). By effectively targeting each interval during the diagnostic and treatment process, it is predicted that the time from diagnosis to initiation of treatment will likely improve as well. Although some quality improvement measures were initiated at our Institution, these were not sufficient to result in improvement in the timeliness of breast cancer care. This highlights the importance of resources and interventions needed at each and every interval of breast cancer care and, ultimately, the necessity of funding needed to change the infrastructure in order to improve breast cancer care for patients of low socioeconomic status.

Based upon our data, screening and early detection continue to be an issue in underserved communities. Rates of presentation of breast cancer by palpable detection in the United States have been cited as 43% in 2000, reaching 34.6% in 2006, falling to 28.9% in 2010 (16, 17). In contrast, in our underserved community hospital population, in the majority of patients, breast cancer was detected via clinical symptoms rather than screening, at a rate of 68%. It is noteworthy that such a small percentage of breast cancers at our institution were screen-detected rather than clinically detected, which is consistent with other safety-net institutions in the United States and with patterns of presentation of patients with breast cancer in low- and middle-income countries (18, 19). This in part may be due to reduced access to screening (20), and the fact that Non-Hispanic Black and Hispanics are more likely than non-Hispanic White patients to have interval cancer despite a recent screening mammogram (21). Additionally, ductal carcinoma in situ was a relatively uncommon diagnosis at our Institution. This reiterates that regular screening for breast cancer, while uniformly recommended per United States Preventive Task Force and American Cancer Society guidelines, is not equally distributed across all segments of the population and therefore a universal screening model should be revisited with consideration of racial and socioeconomic status (22). Our study further illuminates the issue that despite advancement in mammography and public awareness of recommendations, disparities continue to exist in low-income and underserved communities. Access to care remains an issue as reflected by recent studies that show that insurance differences account for one-third of deaths in Black women with breast cancer (23).

In our study, there was no improvement in the time to completion of diagnostic imaging, time from diagnostic imaging to biopsy, and time from pathological diagnosis to initiation of treatment. For time to biopsy, barriers at our Institution included the need for healthcare providers to place biopsy orders after review of positive imaging results in addition to issues with approval by insurance companies. This highlights an issue that we did not initially anticipate. To expedite this process, we proposed a change in protocols to have biopsies completed at the time of diagnostic mammogram for patients with lesions of Breast Imaging-Reporting and Data System categories 4 or 5. In addition, an improved multidisciplinary program was initiated during the beginning of 2018 to improve the time from pathological diagnosis to the initiation of treatment. However, no improvement was noted. This may be secondary to the initiation of new protocols that require time for implementation, as well as adjustments, before subsequent effects are seen. Longer term data may reflect more effective changes.

In our population, our FP rate was 4.8%, which is lower than the average rate of 6-10% in the United States (24). In contrast, a large retrospective study which evaluated 168,251 mammograms found that facilities serving vulnerable women had higher FP rates than those serving nonvulnerable women (25). One reason for this may be that our facilities are not completing enough screening mammograms for our patient population. Higher screening frequencies are associated with higher cumulative FP risk for a patient (26). This suggests that more interventions at the provider and patient level within a healthcare system serving a low socioeconomic population is needed to increase adherence to mammographic screening guidelines.

Our results highlighted issues in breast cancer care within our safety-net hospital and was brought to the hospital leadership’s attention in order to provide adequate resources to improve the time to biopsy. This has subsequently led to an ongoing quality improvement project that has been initiated to shorten the median time from completion of imaging to biopsy from 29 days to 1 week. Our Institution was recently awarded a grant to reduce the time from an abnormal mammogram to biopsy consistent with the goals of minimizing racial disparities in cancer care, with dedicated physician time to tackle the problem highlighted by this study. The process is being deconstructed with recommendations made by the administration to increase the efficiency of the process. Achieving improvement in health outcomes ultimately requires a prospective quality improvement approach but this study thoughtfully highlights the nature of the problem such that it was possible to initiate the process of a solution.

The main limitation of our study was that data were obtained retrospectively at a single collective institution without randomization. Secondly, the total number of patients included in this study was small, with variability by year, and may not be powered for statistical significance. Finally, the initial years had fewer data points during the initiation of the program and mammographic screening protocols, which may be secondary to the transition and implementation of a new electronic health record as well as improved databases.

We propose that future prospective studies establish protocols in order to expedite and reduce temporal delays during a patient’s course, whether clinically or screen-detected, from first contact until treatment. This is now an ongoing major goal of our Institution. Reducing delays at underserved communities will reduce morbidity and mortality of minority and low-income patients with breast cancer (27, 28). Additional studies would include collecting more prospective data to evaluate implementation of adjuncts such as a clinical coordinator, multidisciplinary meetings, streamlining authorization, radiographic orders, and referrals to evaluate outcomes in patient care. Resources and interventions are needed during the entire care of patients of a low socioeconomic background in order to overcome the disparities seen in breast cancer. At the same time, investment in quality improvement specifically targeted at underserved patient populations is likely to achieve a higher return on investment than small incremental changes in wealthier patient populations – indicating that cancer health economics remains a zero-sum game with an asset allocation problem.

In summary, establishing screening protocols and a multidisciplinary breast program at a safety-net hospital with limited economic resources did not improve the time to completion of diagnostic imaging, time from diagnostic imaging to biopsy, nor time from pathological diagnosis to initiation of treatment. Within the same socioeconomic status, there was no differential delivery of screening, work-up, or treatment for breast cancer. Our study identified multiple time periods after screening as limiting factors during the screening and treatment process for our patient population, including time to biopsy as well as initiation of treatment. If the time between diagnostic imaging and biopsy can be reduced, it would result in quicker medical and surgical treatment, potentially leading to improved survival (27, 28). Future protocols aimed at specific interventions within each provider and system level is prudent to improve access and efficiency of breast cancer treatment and reduce disparities for patients in underserved communities.