Background: A large percentage of cancer diagnoses emerge from the emergency setting yet limited literature is published describing this phenomenon in detail. Our study aims to identify the number of patients who present to a large county hospital with symptoms suspicious for malignancy and quantify their compliance with our internal referral clinic (IRC) system for further malignancy work-up. Furthermore, we aim to identify whether having a primary care provider (PCP) prior to diagnosis has a noticeable impact on preventing cancers with national screening guidelines in place.
Methods: A retrospective chart review at a county hospital with 182,099 annual visits in 2014 was conducted between August 18th, 2014 and May 23rd, 2015. Patients who presented to the ED and were suspected of having an underlying malignancy with subsequent referral to the IRC were included in our analysis. Descriptive statistics were computed for patient demographic characteristics and two-tailed Fisher’s exact test was used to calculate patient outcomes.
Results: Overall, 122 patients were referred from the ED to the IRC for a suspected underlying malignancy and were included in our final analysis. The mean age at presentation was 51 (range 18-81) and 52.5% were women. Of all patients referred to the IRC, 103 (84.4%) presented for further work-up of their suspected malignancy and 87 (71.3%) were successfully referred to further specialists. Mean time between ED presentation and IRC visit was 11.4 days (range 1-48). Sixty-nine (56.6%) patients were ultimately diagnosed with cancer; of these diagnoses, 33 (47.8%) were metastatic at time of presentation. The most common malignancies diagnosed were colorectal (14), breast (12), and pancreatic cancer (10). The most common malignancy diagnosed in patients with and without a PCP was pancreatic cancer and colorectal cancer, respectively. Patients without a PCP were more likely to be diagnosed with a malignancy than those with a PCP (63.8% vs. 42.9%; OR 2.34 [CI 1.09,5.03]; p=0.035). However, the proportion of patients diagnosed with metastatic cancer on presentation did not significantly vary with PCP status (21.4% with PCP vs. 30.0% without PCP; OR 1.57 [CI 0.65,3.78]; p=0.393).
Conclusions: This study found that patients without a PCP were twice as likely to be diagnosed with a malignancy versus those with a PCP. In this sample, there was no significant difference in percentage of cancers that were metastatic on presentation. The natural course of cancer varies depending on primary site, and several of the malignancies observed in this study often have metastases on presentation. The most common malignancies diagnosed among patients without a PCP were ones for which there is a national screening protocol (i.e. colorectal or breast cancer). This aligns with our hypothesis that patients without a PCP are more likely to present to the ED with a malignancy that would otherwise have been detected in a primary care setting. Savage et. al found colorectal, lung, and CNS malignancies to be the most common cancers presenting in a London-based ED. Our study population excludes patients who were direct hospital admissions. As a result, CNS and hematologic malignancies may not be appropriately represented in our study. Interestingly, over half of patients referred to the IRC were ultimately found to have a malignancy, suggesting that the IRC is an important diagnostic resource.