Arthritis drugs · JAMA Netw Open · 2024
The paper’s summary: JAK-blocker drugs clearly raise cancer risk. Its own numbers allow for no difference.
Cancer risk, as a multiple of the risk on TNF blockers
The paper says
The paper compares cancer risk in people with rheumatoid arthritis starting various immune-calming drugs against people starting one kind, TNF blockers. Its summary says JAK blockers, like two other drugs (rituximab and abatacept), had a “significantly” (clearly) higher risk: 1.36 times as likely, 36% more.
Its own data show
The 1.36 is reported correctly; “significantly” is not. The summary itself prints the range of values its data can’t rule out: 0.94 to 1.96 times, from 6% lower risk to 96% higher. That includes 1×, no difference. Its discussion and closing conclusions agree: only rituximab and abatacept were clearly higher.
Researchers call a risk “significantly” higher only when its whole range sits above 1×, ruling out “no difference”. This range covers both no difference and nearly double: not proof of harm, nor of safety. It matters because an earlier trial found more cancer on a JAK blocker than on TNF blockers; the summary presents this study as clear backing.