← All posts

What happens after a positive screen: the part of school mental health measurement nobody plans for

April 22, 2026 · Program design · by Heardify

Schools are screening more students for mental health than they ever have. A 2025 JAMA Network Open paper from Cantor and colleagues at RAND, Screening for Mental Health Problems in US Public Schools, put a number on it: 30.5% of US public school districts have a policy mandating mental health screening. Illinois became the first state to mandate universal screening statewide in 2025, with annual screening for grades 3 through 12 starting in the 2027-28 school year (NWEF overview).

That's a lot of new screens being run. And a question that keeps coming up in field conversations: what happens after a student screens positive?

The honest answer, from the evidence: a lot of variation, and a lot of rough edges. Schools that have thought carefully about the pathway get reasonable outcomes. Schools that bolt a screener onto existing infrastructure without a clear next step often get the worst of both worlds, higher identification without a proportional increase in actual support. The screener isn't the problem. The pathway is.

What the Cantor paper actually found

The RAND / JAMA Network Open paper surveyed a nationally representative sample of US public school principals. The headline number is that 30.5% of districts require mental health screening. Drill into the follow-up practices and the picture gets more interesting.

When a student is identified as having anxiety or depression, schools report they:

  • Notify parents: 79.3%
  • Provide in-person treatment at school: 72.3%
  • Refer to a community mental health professional: 53.0%

Those numbers are self-reported by principals, so they're probably generous. The more telling figure is the barrier question. About 41% of principals said connecting screened students to appropriate services is "very hard or somewhat hard." Only 38% said it's easy. In other words, roughly as many schools find the follow-up step painful as find it manageable.

A couple of things stand out. Districts serving primarily students of color had higher screening rates than predominantly white districts. Larger schools (450+ students) screened more than smaller ones. The paper doesn't specify which screening tools are in use across the sample, which is itself a signal: screening policy has outrun measurement-tool standardization.

The SHIELD trial, and what universal screening actually does

The best causal evidence on this question is still the SHIELD trial (Screening in High Schools to Identify, Evaluate, and Lower Depression). Sekhar and colleagues randomized Pennsylvania public high schools by grade to either universal PHQ-9 screening or targeted screening via their existing Student Assistance Program. Roughly 14 schools, grades 9 through 12, followed for outcomes including depression treatment initiation.

The main result from the SHIELD trial: students in the universal screening arm were about twice as likely to initiate depression treatment compared to students in the targeted screening arm. That's a real effect. Qualitative follow-up work from the same team (A qualitative analysis of participant experiences) added useful texture. School staff generally welcomed the screening. The single biggest operational challenge they named was inadequate staffing to manage the referrals that came out of it.

Read those two findings together and you get the core tension: universal screening works, but only to the extent that the school can actually absorb what it produces. A screener that identifies 15% of a student body as meeting the cut-off only translates into better outcomes if there's a pathway for that 15% to go somewhere useful.

Where pathways break

In practice, the pathway breaks at a few predictable points.

Point 1: No designated triage role. A school counsellor with a caseload of 400 students can't be the person who personally meets with every PHQ-9 positive student within 48 hours. If the screening workflow doesn't name who does the first contact and on what timeline, nobody does it.

Point 2: Follow-up capacity mismatched to prevalence. In adolescent populations, baseline rates of a positive PHQ-9 screen (score over 10) sit around 10 to 20% depending on the population and cut-point. A school of 800 is looking at 80 to 160 positive screens in a given cycle. "We'll refer them to the school psychologist" doesn't hold up against that math if there's one school psychologist shared across multiple buildings.

Point 3: Unclear parent communication policy. The Cantor paper reports 79.3% of schools notify parents on a positive screen, but the timing, format, and content of that notification vary wildly. A form letter sent home a week later lands differently than a phone call the same day. The policy detail matters for both safeguarding and parent trust.

Point 4: No follow-up measurement. Most schools administer a screener at one point in time and treat a positive score as a binary referral trigger. They don't re-administer. So when a student who screened positive in October is doing better in February because of external factors (a family situation resolving, a sports season starting, a medication change), the school can't see that change without a second data point. This is the gap Heardify's own product work is focused on closing, but the broader observation applies regardless of tool.

Point 5: Tier 2 absence. The Cantor paper and a parallel 2025 review, Teacher-Delivered Mental Health Interventions: Promises, Challenges, and Recommendations for Future Directions by Lawson and colleagues in Administration and Policy in Mental Health, both land on the same structural issue: schools have universal (tier 1) supports and sometimes intensive (tier 3) clinical referrals, but the middle layer (tier 2, targeted group interventions for students showing early signs) is the thinnest. Lawson et al. report small effects (d = 0.22) for tier 1 teacher-delivered programs and essentially null effects for teacher-delivered tier 2 / 3 programs in recent meta-analytic data. Teachers, reasonably, aren't the right staff for the targeted layer. Which means if the school doesn't have someone else for that layer, the pathway from positive screen to support is just "refer out," and the community mental health system has its own capacity problem.

A framework for thinking about the pathway

Before standing up a new screening program (or auditing an existing one), it's worth walking through the pathway end to end. Five questions, in order:

1. Who sees the raw score first, and when?

"Automatically flagged in a dashboard that nobody has time to check" is a common failure mode. A designated staff member should see elevated scores within 24 to 48 hours of administration. That includes elevated item 9 of the PHQ-9 (or the equivalent on the PHQ-A, the adolescent-adapted version, which schools with under-15 populations should be using instead). The suicide-risk item requires same-day response under most safeguarding policies.

2. What's the first contact?

This isn't optional. The first-contact protocol should specify: who does it (not just "counsellor," a named role), what format (in-person check-in versus phone call versus referral letter), what's said (scripted prompts help, even for experienced staff), and what's logged (for audit and continuity).

3. What's the parent-communication rule?

Depending on jurisdiction, this ranges from mandatory same-day notification to optional delayed communication pending student consent (for older adolescents). Know the rule. Write it down. Train staff on it. The 2025 Mental Health Screening Toolkit from Colorado is one of the clearer state-level documents on this if you need a reference.

4. What's the next-step options menu, and what's the capacity for each?

List them. For each, note how many students can move through per term. Typical menu: brief check-in with a counsellor, referral to a group-based intervention (tier 2), referral to a school psychologist for assessment (tier 3), referral to a community clinician (tier 3 external), or monitoring with a re-screen at 6 to 8 weeks. If the menu has five options and the capacity adds up to 20 students a term while your positive-screen count is 80, you have a problem the screener cannot solve.

5. What's the re-measurement plan?

A single screen at a single point in time is a snapshot. A screen plus a follow-up screen at 6 to 8 weeks is a trajectory. Trajectories are what tell you whether the student is getting better, stable, or worsening, and whether the intervention you offered is doing anything. Re-measurement is the step that separates a screening program from a monitoring program, and it's the step most schools skip. Tools like the RCADS and the Strengths and Difficulties Questionnaire are both designed with repeated administration in mind.

The false-positive question

One critique that comes up in the public debate around school screening is that screeners produce false positives, and false positives either pathologize normal adolescent experience or expose students to unnecessary clinical contact (Manhattan Institute critical assessment). This critique has more and less serious forms.

The less serious form treats any positive score as equivalent to a diagnosis. It isn't. A PHQ-9 score above 10 is a signal to look more carefully, not a verdict. Good pathway design handles this by making the next step a conversation with a trained staff member, not a referral letter.

The more serious form is about pathway design specifically: if a school has a 15% positive-screen rate and refers all 15% for clinical assessment without first running a confirmatory conversation or a second-screen, the system is converting a screening signal into a clinical contact rate it can't back up. That's a legitimate design concern. The Cantor paper's finding that 41% of principals find follow-through "hard" is consistent with the pattern of schools running screeners without matching triage capacity.

The practical response isn't to stop screening. It's to triage more carefully. A positive PHQ-9 followed by a structured 15-minute conversation with a counsellor, followed (in some cases) by a re-administration a week later, filters a lot of transient symptoms from a clinically meaningful signal.

The gaps in what we know

A few things the current evidence doesn't settle:

  • Long-term outcomes of universal screening. The SHIELD trial showed treatment initiation effects at the end of the screening cycle. Whether those treatment initiations translate into symptom reduction two years out, fewer crises, or changed academic trajectories is less established.
  • The right cadence. Annual? Quarterly? On transitions (grade change, family disruption)? Evidence on optimal cadence for adolescent populations is thin.
  • Digital screeners vs paper. Administration mode affects disclosure rates, particularly on stigmatized items. The evidence base is growing but not settled.
  • Proportionality. Whether schools should screen the whole student body or focus resources on populations with elevated baseline risk (transitions, prior history, recent bereavement) is a real policy question with limited empirical guidance.

One more thing: vendor pathway claims

If a platform is selling a school a screening product, the question to ask (before the validation question, before the cost question) is: what's your pathway? Not "what does your tool do," which is usually fine. What does your tool expect the school to do after a positive screen? How many students can that pathway absorb? What's your guidance when a school's positive-screen volume exceeds its pathway capacity?

A vendor who's thought about this will have an answer. A vendor who hasn't will talk about dashboards.

Further reading

Heardify publishes twice-weekly on student wellbeing, measurement, and school mental health programs. Back to heardify.me or read more from the blog.