Scott criticizes interpretations of a study suggesting psychiatric care might cause suicide, arguing this ignores the more likely explanation of selection bias.
Longer summary
Scott critiques a study and subsequent editorial suggesting psychiatric interventions might increase suicide risk. He argues that the correlation between psychiatric treatment and suicide doesn't necessarily imply causation, using analogies like police contact and murder rates to illustrate the flawed reasoning. Scott points out that the original study authors acknowledged this limitation, but subsequent interpretations ignored it. He argues that while psychiatric interventions may have negative effects worth studying, this particular study doesn't provide useful evidence for that claim. The post ends by acknowledging that the relationship between psychiatric care and suicide is complex and deserves careful, high-quality research.
Shorter summary