SPARC: support calls helped adults early, caring texts helped adolescents later

Design

Pragmatic comparative effectiveness randomised trial, stratified by age

Population

723 adolescents and 797 adults screening at risk for suicide in US emergency and primary care

Primary outcome

Columbia suicidal ideation and behaviour score at 6 months

Effect

Adults 6 mo: texts vs calls +0.50 (0.12–0.88); adolescents 12 mo: −0.64 (−1.11 to −0.18)

SPARC was a pragmatic comparative effectiveness trial in nine emergency departments and 36 primary care clinics in the US. It randomised 723 adolescents aged 12–17 and 797 adults who screened at risk for suicide to one of two add-ons to usual care: instrumental support calls, which focus on practical help and linking to care, or two-way Caring Contacts text messages, which focus on ongoing connection. The primary outcome was the Columbia scale suicidal ideation and behaviour score at 6 months.

At 6 months adolescents did similarly in both arms, but adults receiving texts had higher scores than those receiving calls (adjusted difference 0.50, 95% CI 0.12–0.88). At 12 months, a secondary outcome, adolescents receiving texts had lower scores than those receiving calls (−0.64, −1.11 to −0.18), while adults were similar.

Both arms were active, so the trial shows which follow-up suited which age group, not whether either beats usual care alone. The participants were mostly white and female, and the differences are modest on the scale. Still, it gives services a reason to tailor follow-up by age rather than run one model for everyone.

• For adults after a positive screen, consider early, structured phone support focused on practical needs and linking to care.
• For adolescents, consider sustained two-way text contact over months rather than a short burst of calls.
• Make sure any text service can respond when a young person replies.
• Keep usual care, including safety planning, in place; both approaches were add-ons.

Why it matters

Most services run one follow-up model for every age; this suggests the right one differs.

Don't overread it

Both arms were active add-ons — the trial does not show either is better than usual care alone.

The statistics, in plain English

An adjusted difference of 0.50 points on the Columbia scale is modest. Both confidence intervals exclude zero, so the differences are unlikely to be chance, but the adolescent 12-month result was a secondary outcome and carries less weight than the 6-month primary.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.