Low iron levels (iron deficiency) are found in 1 in 14 adults with routine blood tests and sometimes can lead to low blood counts (anaemia). Blood loss from the gastrointestinal tract is a common cause, is often asymptomatic, and may be due to cancer. For these, endoscopies (invasive camera tests) are usually needed to make a rapid diagnosis.
Current UK medical guidelines provide recommendations to help doctors decide if endoscopy is needed for their patients with iron deficiency and, if so, how urgently it needs to be arranged. Our research has shown that risk of cancer is not clear in traditional “lower risk” patient groups (particularly younger patients). Doctors and patients need to be clear about the risk of gastrointestinal cancer so that more informed decisions can be made about the need for, and timeliness of, endoscopy. It is important for patients at the highest risk to receive urgent tests, and for patients at lower risk to be managed safely and avoid unnecessary invasive investigations.
This study will investigate the feasibility of developing a clinical risk prediction tool for gastrointestinal cancer for patients with iron deficiency. We will calculate the number of patients with iron deficiency who could contribute this future research (in groups according to age and sex), the numbers who develop cancer in each group and the availability of information which could contribute to a risk prediction tool.
This feasibility study is funded by Norfolk and Waveney ICB Research Capability Funding.
We aim to use CPRD to explore the feasibility of developing a risk prediction model for malignant gastrointestinal (GI) disease in patients with iron deficiency in primary care. The future risk prediction tool, subject to performance, could be used in the future to help clinicians and patients consider the pre-test probability of malignant GI disease and determine the urgency of onward gastrointestinal investigation. All analyses will be descriptive using counts and percentages, mean (SD) or median (IQR). We will not report counts of <5.
Objectives are:
1. Determine methodology to identify iron deficiency with and without anaemia in primary care.
2. Determine the number of eligible patients contributing to CPRD from 2010 to present with iron deficiency and iron deficiency anaemia, overall and stratified by sex and 10-year age bands
3. For each stratum, determine the number and proportion diagnosed with GI cancer overall (and by site – upper GI and lower GI separately) within 5 years of diagnosis of iron deficiency
4. For each stratum present number of patients with a faecal immunochemical test result and determine the distribution of values.
5. For broader strata (defined by sex and age) determine the number and proportion with variables expected to contribute to a multivariable prediction tool for GI cancer in patients with iron deficiency.
6. Describe whether we can match eligible patients with iron deficiency anaemia and iron deficiency without anaemia to 1) non-anaemic controls (regardless with of iron status), and 2) non-anaemic, non-iron deficient controls. Matching criteria will include age, sex, GP practice, and alive and under follow-up at index date.
Gastrointestinal malignancy diagnoses within 5 years of diagnosis of iron deficiency (with and without anaemia), including oesophageal; gastric; small bowel and colorectal cancer. Linkage will not be required for this feasibility study.
Leo Alexandre - Chief Investigator - University of East Anglia
Leo Alexandre - Corresponding Applicant - University of East Anglia
Kathryn Richardson - Collaborator - University of East Anglia
Njabulo Ncube - Collaborator - University of East Anglia