For patients with cancer, the emergency department (ED) is both an entry point for acute illness and frequently the site of the first clinical encounter preceding a confirmed diagnosis. What follows that encounter has direct implications for stage at diagnosis. For example, a patient presenting with unexplained weight loss or abdominal pain who is discharged with a nonspecific finding may be weeks or months from a confirmed cancer diagnosis. This analysis examines the pre-diagnosis pathways of patients with confirmed cancer diagnoses to quantify diagnostic delay among those whose cancer journey originated in the ED.
Delayed cancer diagnosis is a persistent challenge in the healthcare system with meaningful implications for disease progression, mortality and treatment costs. Patients discharged from the ED with a nonspecific diagnosis (i.e., ICD-10-CM diagnosis in the signs, symptoms and abnormal findings chapter (R00-R99)) frequently do not receive a referral, imaging order or follow-up plan. Prior research found that 20% of patients with a confirmed cancer diagnosis had at least one ED visit in the six to 12 months prior to diagnosis.1 Non-specific symptoms are present in 22% of cancer cases and are frequently associated with and attributed to non-cancerous conditions including heart failure, chronic obstructive pulmonary disease and depression, which share symptoms such as loss of appetite and fatigue.2 Patients presenting with non-specific symptoms may accumulate multiple ED visits, outpatient encounters and imaging studies before a cancer diagnosis is confirmed.3
Demographic differences contribute to time to diagnosis and treatment. Diagnosis of cancer as part of an ED visit is more common among patients who are older, Black, unmarried, poorer and lacking a consistent source of care.4 Uninsured patients are more likely to present with advanced-stage disease and experience lower overall survival compared to those with commercial insurance.5 Patients with early-onset colorectal cancer (i.e., diagnosis before age 50) are more likely to experience diagnostic delay and to be diagnosed at later stages than older patients, in part because clinical suspicion for malignancy is lower among younger patients.6 Adolescents and young adults who present to their primary care provider or to the ED with non-specific symptoms are frequently dismissed as too young or unlikely to have cancer, which can lead to delays particularly among the uninsured or underinsured.7 These patterns are particularly relevant given rising incidence rates across several early-onset cancer types.8,9
National all-payer claims data were used to examine adults age 18 or older with a confirmed cancer diagnosis across select cancers between 2018 and 2024. A 12-month look-back window applied to each patient's index diagnosis date was used to identify ED visits with qualifying non-specific finding codes, spanning symptom-level diagnoses such as unexplained weight loss, hematuria, new-onset anemia and unspecified abdominal pain, as well as abnormal laboratory and imaging findings without a confirmed etiology.
For patients with at least one qualifying pre-diagnosis ED visit, the analysis characterizes the time elapsed between that encounter, testing, confirmed diagnosis and treatment, along with the care setting transitions that followed the initial ED discharge. All measures are reported by cancer type.
Across select cancer diagnoses between 2018 and 2024, 24.6% of patients had a qualifying ED visit with a nonspecific finding in the 12 months preceding confirmed diagnosis (Figure 1). Brain cancer patients had the highest rate at 41.9%, followed by rectal cancer (34.9%), liver cancer (34.1%), colon cancer (33.4%) and stomach cancer (32.4%). Lymphoma patients had the lowest rates, at 13.0% for non-Hodgkin lymphoma and 11.6% for non-follicular lymphoma. Lung and bladder cancer patients fell below the all-cancer average at 21.5% and 18.2%, respectively.
Among patients with a qualifying ED visit with a nonspecific finding preceding their cancer diagnosis, the median time from that visit to a confirmed diagnosis ranged from 31 days for non-Hodgkin lymphoma to 123 days for rectal cancer (Figure 2). Kidney cancer (103 days), colon cancer (101 days) and lung cancer (92 days) also had prolonged intervals between the qualifying ED visit and confirmed diagnosis. Time from qualifying ED visit to biopsy ranged from six days for non-follicular lymphoma to 41 days for kidney cancer. Once a diagnosis was confirmed, time to treatment initiation was comparatively shorter across all cancer types, ranging from 19 days for non-follicular lymphoma to 48 days for kidney cancer.
The care pathway following the qualifying ED visit varied by cancer type, but repeat ED utilization prior to outpatient follow-up was the most common across nearly all cancers examined. Stomach cancer (46.5%) and liver cancer (46.1%) had the highest rates of repeat ED visits before outpatient follow-up was initiated, followed by rectal cancer (44.6%), lung cancer (42.6%), pancreatic cancer (42.4%) and colon cancer (42.2%) (Figure 3). Bladder cancer had the highest rate of discharge with outpatient follow-up within 30 days (41.8%), followed by kidney cancer (39.6%) and non-follicular lymphoma (39.5%). Discharge without any outpatient follow-up within 30 days ranged from 7.0% for lung cancer and non-follicular lymphoma to 10.8% for brain cancer. Inpatient admission following the qualifying ED visit was most common for non-follicular lymphoma (22.6%) and brain cancer (22.3%), reflecting the acute presentation patterns associated with those diagnoses.
Follow-up failure rates after the qualifying ED visit were highest among the youngest patients and declined with age, though failure patterns were consistent across all age groups (Figure 4). Among patients ages 18 to 44, 24.7% did not have an outpatient follow-up within 30 days of ED discharge and 45.5% had a repeat ED visit before outpatient follow-up was initiated – both the highest rates across age groups. Among patients 75 and older, those figures were 19.0% and 40.1%, respectively. The share of patients whose time from qualifying ED visit to a confirmed diagnosis exceeded 180 days was effectively flat across age groups, ranging from 30.2% among patients ages 65 to 74 to 31.2% among those ages 18 to 44.
Among patients with a qualifying pre-diagnosis ED visit, the median number of pre-diagnosis encounters vary by cancer type (Figure 5). Median outpatient visits ranged from three for lymphoma types to seven for lung, kidney and rectal cancer patients, with pancreatic cancer patients at a median of four. The median number of distinct specialties seen in the pre-diagnosis window ranged from five to eight across cancer types, compared to a range of seven to eleven distinct providers.
Among patients with a qualifying pre-diagnosis ED visit, the median number of total pre-diagnosis encounters varied by cancer type (Figure 6). Rectal cancer patients had the highest median at nine encounters, followed by lung and kidney cancer patients at eight. Most solid tumor types registered a median of seven encounters, while pancreatic cancer patients had a median of six. Lymphoma types had the lowest median encounter volume at four.
Across select cancer types, nearly one in four patients had a qualifying ED visit with a nonspecific finding in the 12 months preceding confirmed diagnosis. For most of those patients, the ED encounter did not initiate an efficient diagnostic pathway. Notably, repeat ED utilization before outpatient follow-up was the most common post-discharge pathway across nearly all cancer types examined.
Delay between the qualifying ED visit and confirmed diagnosis was substantial and varied by cancer type. Rectal cancer patients had the longest median interval at 123 days, followed by kidney (103 days), colon (101 days) and lung (92 days) cancers. The comparatively short time from confirmed diagnosis to treatment initiation across all cancer types suggests that delay accumulates in the diagnostic interval rather than the treatment interval.
Variation in time to treatment initiation across cancer types – ranging from 19 days for non-follicular lymphoma to 48 days for kidney cancer – likely reflects differences in oncology subspecialty supply rather than differences in clinical urgency. Radiation oncologists, urologic oncologists and hepatobiliary surgeons are highly concentrated in academic and urban settings, and access constraints in those subspecialties may contribute to the post-diagnosis interval. Variation in treatment delay warrants separate examination, particularly as overall cancer incidence grows and demand for oncology services increases across a provider workforce that is not expanding at pace.10
Stage at diagnosis is a key driver of both mortality and cost. Five-year survival for colorectal cancer drops from 91% at localized stage to 15% at distant stage; for ovarian cancer, from 92% to 32%. Per-patient per-month treatment costs for lung cancer range from $7,239 at stage I to $21,441 at stage IV.11 Cancer spending in the U.S. reached $183B in 2015 and is projected to exceed $245B by 2030.12 Diagnostic delay that results in later stage presentation compounds that burden.
Diagnostic delay exceeding 180 days was observed in approximately 30% of patients across all age groups, regardless of intervening outpatient contact volume. This points to the absence of a standardized post-ED follow-up protocol rather than a pattern attributable to patient disengagement. A standardized clinical response to qualifying nonspecific findings at ED discharge – including systematic referral protocols and a directed laboratory workup as a minimum threshold – would address the point in the care continuum where delay most consistently accumulates. The evidence also supports greater investment in provider education around evolving cancer incidence patterns, particularly for early-onset cancers in which clinical suspicion remains low and existing screening criteria do not yet reflect documented shifts in disease burden.