If I'm reading that first study correctly, "Ideation" and "Ideation with plan" are lower but "Ideation with plan and attempt" is higher and "Attempt resulting in inpatient care" is almost twice as high.

Neither the discussion nor the conclusion mention this, so maybe I'm misinterpreting something?

The results for attempts are underpowered and they acknowledge this (note the p-values too), but not for ideation. From the discussion:

> We did not detect a difference in the odds of lifetime or past-year suicide attempts or attempts resulting in hospitalization. It is possible that we were underpowered to detect these differences given that suicide attempt items were less frequently endorsed than suicidal ideation items (Table 3). Given this study’s retrospective self-report survey design, we were unable to capture information regarding completed suicides, which may have also reduced the number of suicide attempts we were able to account for. Given that suicidal ideation alone is a known predictor of future suicide attempts and deaths from suicide, the current results warrant particular concern.