Rheumatology · Ann Rheum Dis · 2021
Summary: 0.10% of patients with inflammatory joint disease died of Covid. Its own counts: 0.146%.
Share of the patients who died of Covid, March to September 2020
The paper says
A study followed 110,567 patients with rheumatoid arthritis or a related inflammatory joint disease from 1 March to 1 September 2020. Its summary says 0.10% of them died of Covid. Its results section, its results table and its discussion all repeat 0.10%.
Its own data show
The results table prints the count beside the share: “161 (0.10%)”. 161 deaths among 110,567 patients is 0.146%, which rounds to 0.15%. The 161 holds up: it is the table’s deaths among rheumatoid arthritis patients (134) plus those among the rest (27). Rounding cannot give 0.10%: that would take 106 to 116 deaths.
A death risk here is just deaths divided by patients: the paper’s methods define it that way, and nearly every other risk in the same table follows that rule. It matters because the paper offers this figure for counselling patients, as about 1 in 1,000. Its own counts give about 1.5 in 1,000.