One well-written piece a month — and why it works.

Why it's worth your time. "Only a tiny sliver of psychosis cases are due to autoimmune causes — most experts cap it at 1 to 2 percent of patients." Michelman states the small number plainly, then makes the case that the sliver still matters: patients with psychosis-like symptoms whose brains are under antibody attack, and who may need a different treatment. It's a medicine feature that holds two risks — missed diagnoses and overdiagnosis — in one frame, and is honest that the trials that could settle it haven't reported yet.

• Put the study design next to every number. The ~5 percent antibody-positive rate comes from a screening programme of around 5,000 psychiatric patients; the "just over a quarter" misdiagnosis figure comes from a 2022 analysis of diagnoses. Readers always know how much weight a figure can bear.
• Give both failure modes a voice. One expert calls the field "an unexplored tip of the iceberg"; another worries it is tilting toward overdiagnosis. Neither side wins by omission.
• End on the evidence gap. No tidy moral: the piece closes with a source admitting "We don't have that gold standard evidence from a trial." Results from the SINAPPS2 trial are expected in 2027.