Are all the critical factors necessary for the ED physician at the time of suspected diagnosis? Maybe yes, maybe no

We thank the authors for their response to our article “Retrospective review of patients with lung cancer presenting emergently” [ ]. Their comments and feedback are insightful and will contribute to meaningful future work. We appreciate and acknowledge that this work includes several limitations therefore limiting generalizability. Emergency presentations (EPs) of cancer are global problem as many patients with undiagnosed cancer receive a likely diagnosis of cancer mediated through an emergency department (ED) visit [ ]. However, quality data from US-based studies is lacking as it pertains to the ED’s role in EPs and subsequent cancer-related outcomes [ ]. The primary objective of our publication was to expose US-based emergency physicians to the concept of EPs and how EPs can impact their clinical practice.

In regard to mortality, the authors are correct that many factors alter the mortality rate with lung cancer, including subtypes, stage, molecular markers, and access to state-of-the art anti-cancer therapies. In our cohort, 76% of patients had tissue confirmed non-small cell lung cancer, with approximately 50% of those being adenocarcinoma. While the availability of treatment, response to treatment, molecular markers, and cancer-related mortality are important factors to consider, they were not collected in this study, as these are generally determined well after the index ED visit. In many cases the diagnostic and treatment intervals can exceed months after the first documented symptom. Stage at diagnosis predicts eventual prognosis, with later stages at time of diagnosis being associated with worse mortality and poorer outcomes. With the understanding that many patients present with later-staged disease, the most impactful clinical and prognostic characteristic for ED physicians is assessing the likely stage at time of suspected diagnosis [ ]. Time to treatment initiation for many cancer types is associated with lower mortality, suggesting that decreasing the diagnostic interval of undiagnosed cancer may reduce cancer mortality [ ]. Our work demonstrates that ED physicians are often the first physicians that suspect a cancer diagnosis, and our work provides meaningful data for future prospective studies to assess the impact that ED physicians have on time to treatment initiation.

Meanwhile, no standardized or internationally accepted definition of EP of cancer exists, largely due to the differences in the available cancer care pathways that exist between countries [ ]. In our study we limited our cohort to patients with overt cardiopulmonary related issues as their presenting concern in the index ED visit, meanwhile excluding those with incidental discoveries of a lung mass (Section 2.4 Definition of Emergency Presentation). Robust previous work has demonstrated the clinical characteristics and outcomes for both incidentally detected lung cancer and symptomatic lung cancer. For example, Quadrelli et al. demonstrated that about 10% of lung cancer cases present with “other symptoms” outside of what we included in our study’s inclusion criteria, thus we have certainly underestimated the overall sample size in our study [ ]. Due to the limitations of our study type, retrospective cohort study, we chose our inclusion criteria carefully as it would be relatively easy to understand the clinicians’ thought processes upon ordering chest/thorax imaging. Future prospective work assessing the comprehensive list of confounding factors unrelated to their lung cancer diagnosis is crucial and this work is ongoing.

Significant cancer health disparities have a demonstrable impact on all facets across the cancer care continuum including cancer prevention, cancer screening availability, treatment availability, survivorship, and care at end of life [ ]. We agree with the authors that considerations based on patients’ income and health literacy would be useful for subsequent analysis, however those data were not available retrospectively. Furthermore, we have recently published a much larger epidemiological assessment of cancer diagnoses that were made in proximity to an ED visit, of which we notably demonstrated that outcomes were mediated by socioeconomic status, as defined by their home zip code [ ]. These conclusions and issues are prime for a future, prospective, multi-site study as our current site limits the generalizability of our findings since nearly all medical patients are of low socioeconomic status and are indigent.

Lastly, we greatly appreciate the comments the authors make about the reduction of unnecessary medical examinations and the cost and potential harm of over diagnosis. Many studies have demonstrated mixed success by deploying interventions to reduce ED use for non-emergent conditions, as well as efforts to increase primary care uptake [ ]. Low-value care and services along with unnecessary emergent testing are an ongoing issues in the emergency medicine literature. For example, the American College of Emergency Physicians has a policy statement to help curb unnecessary imaging for atraumatic low back pain [ ]. Such policy statements are one of many ways by which we can curb unnecessary testing and health care costs. While we agree that it is imperative to consider cost-effectiveness in addition to reducing unnecessary medical examinations, current discussions about unnecessary testing have not accounted for the millions of Americans that rely on the ED for primary, urgent, and emergent care. This is evident by the millions of patients in need of recommended cancer screenings that rely on the ED for care [ ]. Current discussions about unnecessary testing do not address ongoing health disparities, inequities in access to primary care, subspecialist care, and state of the art treatments. Additionally, current discussions about unnecessary testing don’t address the litigious nature of our society which clearly results in defensive medicine [ ]. Until concerted efforts demonstrate uniform success at increasing the uptake and availability of primary care and it is important to recognize that a paradox exists. The current state of ED care is that we provide care for anyone who needs it, and the ED may be the only practical (and therefore, best) way for many patients to be evaluated for cancer.

We greatly appreciate the response to our recent article, as the authors’ suggestions and feedback will be impactful for future prospective studies. Given the limitations mentioned in our current study, namely the retrospective nature by which the cohort study was conducted, future prospective studies are needed. However, irrespective of the many characteristics not presented in our recent publication (socioeconomic status, health literacy, cancer treatment status), the diagnosis of cancer when made by proxy of an ED visit represents a common route of cancer diagnosis. Future efforts should focus on the ED’s role in cancer prevention, cancer screening, early detection, and the management of the suspected cancer cases to ensure all people have equitable access to rapid diagnostics and cancer treatment.

CRediT authorship contribution statement

Nicholas R. Pettit: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Writing – original draft, Writing – review & editing. Andrea Noriega: Data curation, Writing – original draft, Writing – review & editing. Marissa Rose Vander Missen: Data curation, Writing – original draft, Writing – review & editing.