TABLE 1

Immortal time bias in cohort studies investigating the effects of anti-acid therapy on all-cause mortality in idiopathic pulmonary fibrosis (IPF)

First author, year [reference]	Sample size n	Data source	Exposure^#	Adjusted hazard ratio (95% CI)	Duration of follow-up
Lee, 2011 [13]^¶	204	Two centres	Anti-acid therapy versus non-use at diagnosis	0.47 (0.24–0.93)	6 years
Ho, 2013 [30]	132	Single centre	PPIs ≥12 months versus non-use	0.56 (0.34–0.92)^§	5 years
Ghebremariam, 2015 [18]	215	Two centres	PPIs ≥12 months versus non-use	0.58 (0.39–0.87)	5 years
Lee, 2016 [22]⁺	786	Single centre	PPIs ≥4 months versus <4 months	0.51 (0.21–1.22)	5 years
Liu, 2017 [24]	69	Single centre	Anti-acid therapy ≥6 months versus <6 months	0.23 (0.12–0.44)	5 years

PPI: proton pump inhibitor. ^#: exposure to anti-acid therapy included PPIs and H₂-receptor blockers; ^¶: suspected immortal time bias; ⁺: outcome defined as IPF-related mortality; ^§: we calculated approximations of hazard ratios and/or 95% confidence intervals based on reported median survival times and p-values.