Mortality risks associated with Barrett’s oesophagus

Abstract

tality ratio (SMR) = 1.22, 95% CI = 0.79‐1.80] or myocardial infarction (SMR = 1.15, 95% CI = 0.80‐1.61), compared with the higher estimates of Moayyedi et al. for ischaemic heart disease compared with left ventricular failure (male SMR = 1.86, 95% CI = 0.97‐2.75; female SMR = 2.05, 95% CI = 1.05‐3.06). Regarding all-cause mortality, our study found only a small but statistically significant increased risk (SMR = 1.21, 95% CI = 1.06‐1.37), which is in quantitative agreement with other estimates from the UK 3 (hazard ratio = 1.37, 95% CI = 1.12‐1.66) and the Netherlands 4 (SMR = 1.46, 95% CI = 1.16‐1.82). The much higher all-cause mortality estimates of Moayyedi et al. (male = 5.52, 95% CI = 4.66‐6.38; female = 4.55, 95% CI = 3.57‐5.52), are in further contrast to an Irish study which found the risk unaltered. 5 Although we prospectively followed fewer BO patients compared with Moayyedi et al., it is ultimately the number of deaths which determines the statistical power of a SMR analysis, 6 for which our studies have approximately equal numbers. We also find that the conclusions of Moayyedi et al. require some clarification. They conclude that: ‘Patients with Barrett’s oesophagus die more commonly of bronchopneumonia and ischaemic heart disease compared with oesophageal adenocarcinoma’. Given the absolute numbers of deaths in these categories, this is true, as it would be for cohorts of almost every other precancerous lesion; mortality risks associated specifically with BO are indicated by SMRs, while the absolute numbers of deaths merely reflect the underlying mortality risks of the general population. Whilst it is important to acknowledge the relatively low absolute risk of death from oesophageal adenocarcinoma in BO, we would emphasize that the overall evidence currently available suggests that BO has little, if any, effect on the risk of other specific causes of death when compared with the population.