Skip to main content
. 2021 Jul 12;125(6):893–900. doi: 10.1038/s41416-021-01480-0

Table 3.

Multivariable time-dependent Cox analyses for risk of all-cause mortality related to post-diagnosis PPI use in different risk groups.

PPI nonusersa PPI usersa aHRb P for interaction
No. of deaths/No. of person-years
Sex 0.007
Male 4548/36 324 2508/16,802 1.44 (1.36, 1.52)
Female 3891/31 691 2238/17,060 1.32 (1.24, 1.40)
Tumour stage <0.001
0 + I 351/10 830 299/4811 1.61 (1.36, 1.90)
II 1010/18 413 718/8666 1.44 (1.30, 1.60)
III 1414/14 966 1130/8339 1.53 (1.41, 1.66)
IV 3474/6419 1482/2886 1.20 (1.12, 1.27)
Tumour site <0.001
Colon 5741/43 720 3147/22,399 1.29 (1.23, 1.36)
Rectum 2698/24 295 1599/11,472 1.63 (1.50, 1.75)
CRC surgery <0.001
Yes 4939/59 767 3657/30,735 1.51 (1.44, 1.59)
No 3500/8248 1089/3127 1.06 (0.97, 1.16)

PPI proton pump inhibitors, CRC colorectal cancer, aHR multivariable-adjusted hazard ratio.

aPPI users were patients who had at least one PPI dispensation after the diagnosis of CRC; PPI nonusers were patients not collecting PPI dispensation after the diagnosis of CRC.

baHRs were adjusted for age, sex, pre-diagnosis PPI use, tumour site, tumour stage, CRC surgery, maintenance use of low-dose aspirin, maintenance use of non-aspirin NSAIDs, and the Charlson Comorbidity score.