Jump to comment:
- Page navigation anchor for Generalisability of a new colorectal cancer risk assessment tool to clinical practice worldwideGeneralisability of a new colorectal cancer risk assessment tool to clinical practice worldwideWe would like to thank Ekta Sharma for the constructive criticism of our paper and address the specific issues that were raised.When we designed a risk assessment tool (RAT) for symptom combinations for non-metastatic colorectal cancer we also had data and analysis for metastatic cancer. Patients with metastatic cancer had a different diagnostic pattern both regarding the clinical features and their risk scores. However, as the aim of the study was to identify the features of non-metastatic patients we only presented a RAT for this group of patients. Furthermore, this study was designed to identify new onset cancers and not recurring colorectal cancers.It is well known that socio-economic status, smoking and hereditary disposition are influencing the risk of many cancers among them colorectal cancer. However, the design of our study did not have that approach. Weight loss is actually incorporated in the risk assessment tool: when reported a second time and in combination with anaemia.Merging more than 6000 different diagnostic codes into 575, made it possible for us to analyse the data according to clinical relevance. Also, the merging of the final variables into fewer relevant variables made this tool easier to use. GI bleeding from both the upper and lower tract were merged into variable bleeding, which of course could be considered a limitation.Competing Interests: None declared.
- Page navigation anchor for Generalisability of a new colorectal cancer risk assessment tool to clinical practice worldwideGeneralisability of a new colorectal cancer risk assessment tool to clinical practice worldwideI read this paper by Ewing et al with interest. With bowel cancer as the fourth most common cancer in the UK, this study is both timely and important.1 Ewing et al have contributed useful objective data which could further our evidence base for diagnosing colorectal cancer. However, close analysis of the study design and inclusion criteria highlighted potential limitations in the conclusions made and certain clarifications should be sought before using their risk assessment tool in clinical practice.Firstly, we are unable to conclude whether the symptom combinations identified are specific to non-metastatic colorectal cancer, as data concerning symptomatic metastatic patients were not considered. Caution should also be taken when applying these results to patients with recurring non-metastatic colorectal cancer, as only previously colorectal cancer naïve patients were included.Potential confounding factors from patient demographics such as smoking status, socio-economic class and genetic susceptibility may have been useful to decipher the generalisability of these results to patients in other areas.2,3 Furthermore, weight loss; an important factor indicating malignancy, could not be incorporated in to the risk assessment tool due to too few patients.4 This is perhaps a consequence of using a regional database.Moreover, so...Show MoreCompeting Interests: None declared.