Introduction: Oesophago-gastric (OG) cancers remain a worldwide challenge with little sign of major improvements in survival rates. Modern guidelines focus on alarm (or ‘red flag’) symptoms as key triggers for gastroscopy and recommend empirical symptomatic treatment and non-invasive H. pylori testing in those with simple dyspepsia. However, the early symptoms of OG cancer are very common and non-specific, and the traditional alarm symptoms have poor sensitivity or specificity for malignancy. Diagnosis therefore necessitates investigation of symptoms though upper GI endoscopy in a relatively large group of patients, most of whom do not have malignant disease. This has fuelled considerable controversy regarding the role for gastroscopy in detecting cancer at a treatable stage. Objectives: Firstly, to develop data extraction and linkage methods for studying OG cancer outcome, and General Practice population rates of elective diagnostic gastroscopy, using administrative data for English hospitals (Hospital Episode Statistics). Secondly, to confirm the face-validity of the methodology using external sources of information and local audit data. Thirdly, to test the hypothesis that variations in rates of gastroscopy in English General Practice (GP) populations are associated with inequalities in OG cancer outcome. Fourthly, to explore whether practices with lower rates of gastroscopy exhibit a higher yield of serious pathology, consistent with more selective referral practice. Fifthly, to confirm the existence of wide variation in gastroscopy rate between practices in close geographical proximity, Design and methods: Analysis of Hospital Episode Statistics (2006-8) linked to death registry and practice population data. General practices with new cases of OG cancer were included, grouped into tertiles according to standardised elective gastroscopy rate per capita (low, medium or high). Outcome measures for cancer cases were: emergency admission during diagnostic pathway; major surgical resection and mortality at 1 year. Co-variates were age, gender, co-morbidity, and deprivation. Associations between the gastroscopy rate at the patient’s general practice and cancer outcomes were tested in binary logistic regression models, with extensive sensitivity testing of gastroscopy rate ‘exposure’ variable. An algorithm was developed to analyse coded diagnoses for all first elective gastroscopies, using both national and local audit data. Practices were mapped based on postal code. Results: 22,488 incident cases of OG cancer from 6,513 general practices. Mean OGD rate for Low, Middle, High practices: 4.4 vs 8.1 vs 12.9 per 1,000. Mean age of patients undergoing OGD was highest for low tertile practices (60.2 vs 59.5 vs 58.4 yrs; p<0.001). OG cancer cases registered with practices in the lowest tertile had the lowest rate of surgery (15.4% v 16.3% v 17.4%; p=0.004) with the highest rate of emergency admission (34% v 26% v 25%; p<0.001), and the highest mortality (61.2% v 58.9% v 58.0%; p<0.001). After adjustment for co-variates in logistic regression, the gastroscopy rate at the patient’s general practice was an independent predictor of all three outcomes. Practices with low rates of gastroscopy tend to have a higher “diagnostic yield” of serious disease: (15.3% vs 13.9% vs 13.1%; p<0.001). Low tertile practices also showed a relatively lower referral rate for suspected cancer in general based on analysis of rates of ‘fast-track’ referrals under the two week wait pathway (17.3 vs 17.9 vs 19.3 per 1,000). Conclusions: Wide variation exists in gastroscopy rate among general practice populations in England. OG cancer patients belonging to practices with the lowest gastroscopy rates are at greater risk of poor outcome. Low referring practices achieve a higher yield of serious disease but may have increased risk of referral at a later stage in the disease process. This association is more apparent among the most socially deprived practices. These findings suggest that initiatives or current guidelines aimed at limiting the use of gastroscopy may adversely affect cancer outcomes.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:617477 |
| Date | January 2013 |
| Creators | Shawihdi, Mustafa |
| Contributors | Bodger, Keith |
| Publisher | University of Liverpool |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://livrepository.liverpool.ac.uk/17695/ |
Page generated in 0.0017 seconds