Rates of upper gastrointestinal endoscopy and gastro-oesophageal malignancy outcomes

We read the excellent paper by Shawihdi et al with interest.1 The authors demonstrated in a retrospective study, using hospital episode statistics, that outcomes from gastro-oesophageal malignancy were worse among patients from practices with a low referral rate for upper gastrointestinal (GI) endoscopy. While the authors address some weaknesses in their data, we believe they have not highlighted all problems with their approach. As with all observational data, there is the problem of confounding, which often cannot be completely corrected in analyses. However, the major concern is that they looked at aggregate data, not individual patient data. This is open to the ecological fallacy; what is seen in populations may not apply to the individual patient. The term was coined when it was observed that illiteracy rates were lower in US states with higher proportions of immigrants.2 The logical conclusion from this was that immigrants...
Source: Gut - Category: Gastroenterology Authors: Tags: PostScript Source Type: research