Health insurance · PLoS Med · 2026
The summary swaps insured and uninsured people’s numbers, so they say the opposite of its words.
How many times higher the doctor-visit odds: all three kinds of problem vs none
The paper says
People with physical, mental, and memory or thinking problems together are far likelier than people with none to visit a doctor or clinic. The opening summary says insurance shrinks this gap, yet gives insured people the bigger number: 6.22 times the odds, uninsured 3.40.
Its own data show
The paper’s results section prints the same two numbers the other way round: 6.22 for uninsured people, 3.40 for insured. The data behind its chart and its discussion agree. So the summary’s words (insurance shrinks the gap) are right; only its group labels are swapped, putting each number on the wrong group.
Each number compares people with all three kinds of problem to people with none, within one group (odds of 3 mean three visit a doctor for every one who doesn’t). Bigger means a wider gap, so “insurance shrinks the gap” needs the smaller number on insured people. Anyone copying the summary’s numbers would record insurance as widening it.