Paperonium

Tuberculosis · Lancet Reg Health Am · 2025

“HIV plus drug use: biggest extra risk, 225%.” Its table: 172%; the 225% is another pair’s.

The summary: HIV + drug usecalled the biggestTable: not in prison + treatedbefore, same 3 numbersTable: HIV + drug usesecond biggest225%174%304%172%128%228%0%100%200%300%no extra from having both
The summary:HIV + drug usecalled the biggestTable: not inprison + treatedbefore,same 3 numbersTable: HIV +drug usesecond biggest225%174%304%172%128%228%0%200%400%no extra from having both

Extra risk from having both, beyond the two added; lines: values the data allow

The paper says

Some risk factors raise Brazilian tuberculosis patients’ risk of a poor result (death, dropping out, failed treatment) more together than their separate effects added up. The paper’s opening summary says the biggest such extra is from HIV plus drug use: 225% (range 174% to 304%).

Its own data show

The paper’s table and results text give HIV plus drug use 1.72, that is 172% (range 128% to 228%). The summary’s 225%, 174% and 304% are exactly the table’s 2.25, 1.74 and 3.04 for another pair: not being in prison plus earlier tuberculosis treatment. That pair, not HIV plus drug use, has the biggest extra.

Against patients with neither factor, HIV alone raises the odds of a poor result (poor results per good one) by 199%, drug use by 142%: 341% if simply added. With both, the odds are 513% higher, 172% beyond that sum. The summary instead gives another pair’s 225%, overstating this extra and wrongly ranking it biggest for poor results overall.

“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 →

Interplay between biological & social vulnerability and poor tuberculosis treatment outcome in Brazil: a nationwide study using multivariate modelling with excess risk · Lancet Reg Health Am (2025) · doi:10.1016/j.lana.2025.101305