Paperonium

Antibiotic resistance · PLoS Pathog · 2026

A paper’s summary says 84% of a superbug strain’s samples came from infections. Its count: 60%.

60%Its results andits spreadsheet98 of 16484%The summarytoo highThe most the paper’s totals allow82%135 of 16450%60708090
60%Its results andits spreadsheet98 of 16484%The summarytoo highThe most the paper’s totals allow82%135 of 16450%60708090

Share of this strain’s 164 samples that came from infections

The paper says

A Cape Town hospital kept samples of a superbug, a germ that survives the antibiotics doctors save for when others fail. One of its two most common strains (genetic families) turned up in 164 samples. The paper’s summary says 84% of those came from infections.

Its own data show

The paper’s results count 98 of the 164 (60%), and the spreadsheet of every sample published with it agrees. 84% isn’t even possible: the paper lists 186 infection samples in all, 51 of them (36 + 15) from two other strains, so this strain can have at most 135 (82%).

A person can carry this germ without being ill, so hospitals take two kinds of sample: from infections, and from checks for carriers. The paper suggests a strain found mostly in infections may cause more disease. At 60% that pattern survives (the other common strain: 22%), but the summary, the part most people read, makes the gap look far bigger.

“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 →

Carbapenem resistant Klebsiella pneumoniae isolates at a tertiary hospital in Cape Town, South Africa, are dominated by specific local clones rather than previously described international lineages · PLoS Pathog (2026) · doi:10.1371/journal.ppat.1013859