Background: Clinical experience suggests that cancer is frequently diagnosed in the emergency department (ED). Some studies have demonstrated that patients diagnosed with cancer in the ED have more advanced disease and poorer outcomes when compared to outpatient diagnoses. In this work, we use a statewide database to examine the rate of recent visit to an ED in patients with newly diagnosed cancer as an estimate of the frequency with which cancer was diagnosed in the ED.
Methods: We queried two large Indiana databases (The ICD-based Indiana Health Information Exchange [IHIE] and a cancer-specific registry; both include records of any ED visit in the state) from 2012-2017 for new cancer diagnoses based upon first appearance of a cancer-defining ICD code. To provide an inference of diagnostic opportunity in the ED, we determined the frequency with which patients with new cancer diagnosis had an ED visit in the past 6 months. Their malignancy types were obtained as well patient characteristics.
Results: From IHIE, we identified 187,036 patients with new cancer ICD code over this 5-year period, of whom, 41,144 (22%) visited the ED within 6 months of their codified diagnosis. Comparatively, the cancer registry revealed 161,203 patients with new cancer diagnoses, of whom 18,436 (11.4%) visited the ED within 6 months of that diagnoses. Among patients with new cancer and a recent ED visit, the most frequent cancer types in IHIE were skin (5.6%), lung (3.2%), and then breast cancer (2.4%). From the Cancer registry, the most common types were lung (4.8%), breast (2.7%), and colon cancer (2.6%). Forthcoming data will assess the frequency of diagnostic evidence from a recent ED visit that clearly suggested new cancer, rate of transition of care to a qualified medical specialist, and survival rate.
Conclusion: Between 11 and 22% of patients were cared for in an ED within six months prior to appearance of an ICD code indicating new cancer. These novel data imply a high rate of opportunity to discover undiagnosed cancer in the ED.