|Home | About | Journals | Submit | Contact Us | Français|
The study by Roberts et al of mortality among patients in hospital for inflammatory bowel disease (IBD) has several limitations.1
Firstly, the observation that patients with IBD undergoing elective colectomy had lower mortality than those treated medically or needing emergency colectomy is not “strong evidence suggesting that the threshold for elective colectomy is too high.” Patients who electively undergo colectomy usually have chronic relapsing disease or risk of malignancy, whereas those admitted to hospital for medical management or emergency colectomy are usually far sicker with severe acute or fulminant disease.2 These subgroups of patients have entirely different indications for colectomy, so lowering the threshold for elective surgery would not necessarily reduce numbers being admitted with severe acute disease.
Secondly, the authors' method of risk adjustment for comorbidity is flawed given the poor accuracy and completeness of secondary medical diagnostic coding in the hospital episode statistics database.3 Furthermore, their risk model did not include well established predictors of the need for colectomy, such as extent of IBD and race,4 5 or take the severity of comorbid disease into account. It therefore assumed that patients with mild and severe comorbid disease have the same risk of mortality. Patients treated medically may have been high risk surgical candidates with poor prognosis who were appropriately not offered surgery. This contention is supported by the authors' observation that patients managed conservatively during their index hospital admission had high mortality irrespective of whether they subsequently underwent colectomy.
In conclusion, no evidence exists to change the current practice of consigning surgery in IBD to the treatment of last resort.2
Competing interests: None declared.