Clinical Alerts: Why Your Software Should Warn You Before the Session
A client's PHQ-9 jumped from 9 to 17 between sessions. Their GAD-7 crossed into the severe range for the first time. They endorsed item 9 — thoughts of being better off dead — after months of scoring zero on it.
Would you know before they sat down?
If you're tracking outcomes manually — reviewing scores when you remember to, scanning spreadsheets between clients — the honest answer is probably not. Not because you don't care, but because you're managing 25 caseloads and there are only so many numbers you can hold in your head.
Clinical alerts solve this. They're automated flags that fire when a client's assessment data crosses a threshold you care about. And they're one of the most underappreciated features in modern therapy practice tools.
What clinical alerts actually do
A clinical alert is a rule that monitors assessment scores and triggers a notification when something specific happens. The most common types:
Severity threshold alerts. A client's score crosses from one severity band into another — mild to moderate, moderate to severe. This is the most broadly useful alert because severity band changes correspond to clinical decision points.
Absolute score alerts. A score exceeds a specific number. PHQ-9 above 15. GAD-7 above 14. PCL-5 above 33. You set the threshold based on your clinical judgment and the alert fires when it's crossed.
Item-level alerts. A specific item gets endorsed. The most critical example: item 9 on the PHQ-9 (suicidal ideation). Any endorsement of that item — even "several days" — warrants immediate clinical attention. An alert ensures you never miss it, even when you're reviewing scores quickly.
Change-based alerts. A score changes by more than a certain amount since the last administration. A 5-point jump on the PHQ-9 might not cross a severity threshold, but it represents a meaningful shift that's worth investigating.
Inactivity alerts. A client hasn't completed an assessment in a specified timeframe. This isn't a clinical deterioration signal — it's an engagement signal. Clients who stop completing assessments may be disengaging from treatment, and catching that early gives you a chance to intervene.
Why therapists miss deterioration without alerts
The research on this is sobering. Therapists are not good at detecting client deterioration through session observation alone.
The reasons are structural, not personal. You see a client for 50 minutes once a week. Their presentation in session is influenced by their mood that day, their relationship with you, their desire to appear well, and a dozen other factors that have nothing to do with their overall trajectory. A client can be worsening steadily over months while appearing "about the same" in each individual session.
Assessment data captures something sessions can't: a longitudinal signal that's independent of any single interaction. But that signal is only useful if someone is watching it — and if you're watching 25 caseloads' worth of scores, some signals will inevitably get missed.
Alerts are the solution because they don't get tired, don't get distracted, and don't have 24 other clients to think about. They watch every score for every client against every threshold you've set, and they flag only the ones that need your attention.
Setting up alerts that actually help
The key to useful alerts is calibration. Too sensitive and you're drowning in noise. Too conservative and you miss the signals that matter.
Start with the critical items. Item 9 on the PHQ-9 should always trigger an alert on any endorsement. This is non-negotiable. No therapist should have to manually check every PHQ-9 for suicidality endorsement across their entire caseload.
Set severity band crossings. An alert when any client crosses from moderate to severe on any instrument gives you a manageable signal that something needs attention. You're not alerted on every score — only on clinically meaningful shifts.
Be conservative with change-based alerts at first. A 5-point change threshold catches meaningful shifts without flooding you with noise from normal score fluctuation. You can tighten it later if you want more sensitivity.
Inactivity alerts at 2-3 weeks. If a client hasn't completed an assigned assessment in two to three weeks, that's worth knowing. It might mean nothing — or it might mean they're avoiding, disengaging, or struggling.
Alerts in the context of session prep
The highest-value moment for clinical alerts is before the session — during your preparation.
A therapist who reviews their dashboard 10 minutes before a client arrives and sees a red flag on that client's latest PHQ-9 walks into the session differently. They know to ask about mood. They know to probe on the suicidality item. They know to assess whether the treatment plan needs adjustment.
A therapist without that information walks in cold, relies on the client's self-report ("I'm fine, I guess"), and may not discover the shift until it's too late to address it in that session — or at all.
The clinical value of alerts isn't the notification itself. It's the preparation time it buys you.
What alerts don't replace
Alerts are a safety net, not a substitute for clinical judgment.
A client can be in crisis with a PHQ-9 of 12. Another client can be stable with a PHQ-9 of 18. The numbers inform your judgment — they don't replace it. An alert tells you something changed; it doesn't tell you what to do about it.
Alerts also don't replace the therapeutic relationship. A flag on item 9 doesn't mean you read the score to the client and ask "are you suicidal?" It means you're prepared to have a nuanced, clinically informed conversation about what the client endorsed and what it means in their context.
The best use of alerts is as a starting point for clinical thinking, not an endpoint.
The bottom line
Every therapy practice tool should have clinical alerts. The question isn't whether you need them — it's how you calibrate them.
Start with the critical flags: suicidality endorsement, severity band crossings, and significant score changes. Add inactivity monitoring for engagement. Review your alert settings quarterly and adjust based on what's generating useful signals versus noise.
The therapists who use alerts consistently report the same thing: they catch things they would have missed. Not because they're inattentive, but because the volume of data in a full caseload exceeds what any human can reliably monitor.
Let the software watch the numbers. You watch the client.
Theracharts includes clinical alerts for severity thresholds, item-level flags, score changes, and assessment completion — built into every plan. Get started free.
