CORRESPONDENCE

Response to: Post-colonoscopy colorectal cancer incidence

To the Editor: We, the authors, would like to thank the members of the SAGES executive committee for their interest in our article. We would additionally use this opportunity to applaud SAGES for their ongoing efforts to improve the training of endoscopists, and hence the quality of flexible endoscopy in South Africa (SA).

We note your queries and comments, and we respond as follows: Only the endoscopist performing the index colonoscopy was considered in our article. The procedure that diagnosed the cancer was not considered, given that this is not the endoscopy that may have missed a lesion. Cancers were diagnosed according to ICD-10 codes, and may have been diagnosed by other means, e.g. at laparotomy.

We excluded any confounding disease that may increase the risk of colorectal cancer (to reduce bias). The exclusions included patients with a history of polyps, not considering the number of polyps. We note that the rate of post-colonoscopy colorectal cancer (PCCRC) will be higher in patients with shorter adenoma to carcinoma time. We wish to emphasise that many patients were excluded based on membership criteria (they had to have been members for 5 years consecutively prior to the diagnosis of the cancer). We applied our exclusion criteria strictly.

Our study reported on the data as captured by the medical aid provider in identifying specialists. This limitation is addressed in the article.

The data that we had did not allow us to analyse other quality indicators of colonoscopy. We accepted this because this is generally not considered in studies of this nature. We also consider PCCRC as the final arbiter of the quality of a colonoscopy. We agree wholeheartedly that quality indicators and data capturing are essential in SA to improve outcomes after colonoscopy. We welcome further research in this area.

Structured competency-based training is very important, and this is an excellent point, yet these data are not captured by the medical aid provider.

Finally, and as mentioned in the article, the patients seen by gastroenterologists were younger patients and prone to aggressive disease. We have therefore subsequently run a subgroup analysis, showing that this younger group of patients seen by gastroenterologists had significantly more advanced stage cancer (38.7% stage 4 disease) than their older counterparts seen by other specialists (20.6%; p=0.038), confirming our assumption stated in the manuscript.

Thank you.

R L Fourie

Department of Surgery, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa leonyfourie@gmail.com

D B Bizos

Department of Surgery, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa

D Kruger

Department of Surgery, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa