Paperonium

Infections · eLife · 2024

Acid-reducing pills and COVID-19: the summary’s risk figures belong to other nose, throat and chest infections

Summary: hospitalizedfor COVID-19Table: 1,118 nose/throatinfections, identical to summaryResults: 138 hospitalizedfor COVID-19Summary: died of COVID-191.37 timesTable: 1,486 chest infections,identical to summaryResults: 48 died of COVID-191.46 times1.19×1.111.271.33×1.091.611.291.461.052.031×1.2×1.4×1.6×2×same risk as non-users
Summary:hospitalizedfor COVID-19Table:1,118 nose/throatinfections, identicalto summaryResults:138 hospitalizedfor COVID-19Summary: diedof COVID-191.37 timesTable: 1,486 chestinfections, identicalto summaryResults: 48 diedof COVID-191.46 times1.19×1.11 to 1.271.33×1.091.611.29 to 1.461.052.031×1.5×2×same risk as non-users

Times as likely as non-users; lines show ranges; grey rows are other infections

The paper says

The summary of this UK study says people on acid-reducing pills were 1.19 times as likely as non-users to be hospitalized for COVID-19 and 1.37 times as likely to die of it, each pinned to a narrow range: 1.11 to 1.27 and 1.29 to 1.46.

Its own data show

Those numbers and ranges match, digit for digit, the paper’s table figures for catching other airway infections: nose and throat (1,118 infections among pill users) and chest (1,486). Its results section gives the actual COVID-19 figures, based on 138 pill users hospitalized and 48 deaths: 1.33 times (1.09 to 1.61) and 1.46 (1.05 to 2.03).

A range spans the values the data can’t rule out; fewer deaths or illnesses widen it. The summary’s risks are smaller (1.19, not 1.33; 1.37, not 1.46) and far too certain: 48 deaths allow anything from 1.05 to 2.03. Most readers use the summary’s numbers; combined with other studies, its death figure would count about 28 times too heavily.

“The summary” means the paper’s abstract: the short summary printed at its top, which is the part most readers see.

Read the paper →Full technical write-up →

Associations of proton pump inhibitors with susceptibility to influenza, pneumonia, and COVID-19: Evidence from a large population-based cohort study · eLife (2024) · doi:10.7554/eLife.94973