@columbiapsych.bsky.social thanks to @almabitran.bsky.social @prof-nick-allen.bsky.social and others!!

critically, most teens reported ⬇️ stress with peers; teens with a suicide risk history showed smaller social buffering of stress vs. when alone
Though suicide attempts and hospitalizations were thankfully rare, they were more common for teens reporting ⬆️ stress in the 1st ESM week

We had a large group of teens complete ESM for 4x week periods (4-7x/day). Focusing on stress based on recent social context, teens with a history of suicide ideation and/or attempts, reported:
-Spending more time alone
-Worse/more stress

@auerbachlab.bsky.social @columbiapsych.bsky.social @columbiauniversity.bsky.social #MedSky
#PsychSky

Suicide risk is worrying common among young kids ending up in the psych ED. Some critical risk factors are logged routinely in EHR but we identify other factors that could be more regularly tracked/assessed with checkboxes or flowsheets to enhance use as data and flag risk in EHR

Some may not expect in young kids, but we often see ⬆️ risk for STB among girls. LGBTQ+ disparities in risk also start this young. Increasingly, we see mention of social media or tech; parent revoking access to📱💻🎮 was a notable source of conflict precipitating attempts

1/5 of kids returned to our ED for STB, typically w/i 90 days. Despite the young age of this sample, this was similar to rates for teens. Girls were more likely to return as were kids on antidep. meds, with family stress, sleep concerns, and esp. with hx of prior hospitalization

We coded multiple domains of risk factors in EHR notes from treating & psych providers. Nearly all were more frequent in STB cases-including abuse, irritability, impulsivity, bullying, etc. Concerningly, 93% of cases that mentioned LGBTQ+ identity reported STB.

32 kids were brought to the ED following an attempt, mainly girls, mainly ingestion/overdose. Most had a prior history of suicide behavior before this event. 56% reported an interpersonal stress trigger, i.e. family conflict or bullying.

We identified 920 8-12yo children in our ED over a 3 year period presenting with behavioral health and/or suicide concerns. Suicide thoughts and behaviors (STB) were common in this population (65%) though thankfully acute suicide attempts were rare.

Thank to @auerbachlab.bsky.social, Jackie Kirshenbaum, & @epikerrykeyes.bsky.social

Unfortunately, many kids will end up back in the ER w/i a few months for additional care (similar to older teens). Early suicide thoughts or behaviors are the biggest risk factor in these data. More nuanced data can help us better understand risk and ways to help these families.

Overall, we see that many children seek mental health care in the ER and suicide risk is commonly observed even in young children. These rates are increasing over time. EHR are powerful but do likely underestimate the issue with potential undercoding in ICD diagnoses.

18% of kids had multiple mental health ER visits. 10% had multiple suicide-related visits. Mostly often returns were w/i 3 months of discharge. ER return w/i 1 year was most common for kids with suicide risk, more mental health diagnoses, without private insurance, and for girls.

Injury due to suicide behavior was present but not highly common–this was almost exclusively ingestion (overdose) or object (e.g., cutting) injuries. Suicide risk in the ER shows temporal patterns, most common during the day (vs. night) and school days (vs. weekends and summers).

Mental health and suicide risk are on the rise in the ER, accounting for 5% of all ER visits in 2010 to 10% in 2020. This also comes with an increase in patient related costs year over year $$.
Increasing % of suicide risk / all mental health visits by year x state.

library(tidyverse)
library(tidytext)
library(readtext)
library(openxlsx)

nsf <- str_replace_all(readtext('NSF_keywords.docx')$text,"\n","|")
doc <- as_tibble(readtext("Grant.docx")$text)
df <- unnest_tokens(doc,"sentences",out,value)
df <- df %>% filter(str_detect(df$out,nsf))
df$replace <- str_match(df$out,nsf)
write.xlsx(df,"check.xlsx")