Oesophagogastric cancer: Timeliness of diagnosis in patients with pre-existing disease

Talk Code: 
Sarah Price
Sarah Price, Willie Hamilton
Author institutions: 
University of Exeter


Pre-existing conditions may complicate cancer diagnosis by presenting diagnostic alternatives for cancer symptoms or by competing for clinical attention. This study investigated associations between time to oesophagogastric cancer diagnosis and (a) pre-existing conditions providing plausible diagnostic alternatives to cancer; and (b) number of pre-existing conditions. Oesophagogastric cancer was studied because of its poor 1-year survival (42%) and the unknown impact of pre-existing conditions on the diagnostic process.


This primary care study used Clinical Practice Research Datalink data with English cancer registry linkage. Patients were aged ≥55 with an incident diagnosis of oesophagogastric cancer between 01/01/2012 and 31/12/2017. Diagnosis date was the earliest recorded CPRD or Registry diagnostic code. The index date was assigned as when patients met criteria for investigation in National Institute for Health and Care Excellence referral guidelines for suspected oesophagogastric cancer. It was determined from the patient’s age and the presence of codes for dysphagia, haematemesis, upper abdominal pain, weight loss, nausea, vomiting, dyspepsia, or reflux. CPRD records in the 2 years before the index date were searched for diagnostic codes for pre-existing conditions listed in the Quality and Outcomes Framework, or which provided plausible diagnostic explanations for the cancer symptoms. Time from index date to diagnosis was modelled using an accelerated failure time model, reporting time ratios (TR) with 95% confidence interval (CI). Predictors included cancer site (stomach or oesophagus), age, sex, pre-existing condition count, an “alternative-explanations” variable where pre-existing condition(s) offered diagnostic alternatives at the index date. Interaction terms were sought by separate cancer site.


3,806 (69.0% male) patients were studied, of whom 1,982 (68.8% male) met NICE criteria for investigation. Mean (standard deviation) age at diagnosis was 74.6 (9.5) and 72.9 (10.0) years, respectively, for stomach (n=526/1,982, 26.5%) and oesophageal (n=1,456/1,982, 73.5%) cancers. At least one pre-existing condition was diagnosed in 317/526 (60.3%) stomach and in 794/1,456 (54.5%) oesophageal cancer patients, of which 73/526 (13.9%) and 143/1,456 (9.8%), respectively, provided plausible diagnostic alternatives. Having 2 or more pre-existing conditions was associated with longer time to diagnosis for stomach (TR 1.55, CI 1.14 to 2.10, p=0.005), and with shorter times for oesophageal (0.64, 0.45 to 0.91, p=0.012) cancer. Having alternative explanations was moderately associated with longer time to diagnosis for oesophageal cancer (1.52, 1.01 to 2.30, p=0.045), and with shorter times for stomach cancer (0.64, 0.45 to 0.92, p=0.015).


The results suggest that pre-existing conditions complicate the diagnostic process for oesophagogastric cancer. The diagnostic process may be lengthened for stomach cancer by increasing count of conditions, and for oesophageal cancer by the presence of plausible diagnostic alternatives. The interpretation of shortening times to diagnosis may not reflect improved patient outcomes where patients present very late in their disease (“sick-quick”).

Submitted by: 
Sarah Price
Funding acknowledgement: 
Sarah Price is funded by National Institutes of Health Research Grant PRU2 [205304]. The dataset had been used for a separate study on time to diagnosis for many cancers in a study funded by Cancer Research UK (Grant Reference Number 21550). The views expressed are those of the author(s) and not necessarily those of the CRUK, NHS, the Department of Health and Social Care, other government departments or arm’s length bodies. This research is also linked to the CanTest Collaborative, which is funded by Cancer Research UK [C8640/A23385], of which Willie Hamilton is co-Director.