The alert queue

Every reading that arrives from a paired device is checked against the patient's vital configuration. If the value falls outside the configured range, Vantrexia creates a Vital Alert and adds it to the queue at Monitoring -> Patient Monitoring. The queue is the first screen most triage nurses open in the morning, and it is the single source of truth for what needs attention right now.

Severity levels

Vital alerts are categorized by severity:

Critical: a value outside the patient's hard safety limits. For blood pressure, this typically means a systolic reading at or above 180 mmHg, or any value flagged by the patient's care plan as a same-day-callback threshold. Critical alerts are highlighted in red in the Patient Monitoring table and appear under the Critical filter chip.
Warning: a value outside the patient's normal range but inside the safety boundary. These are the bread-and-butter of RPM: a systolic of 152 in a patient whose target is 130, a fasting glucose of 165 in a patient targeting 120, a weight gain of 3 lb overnight in a CHF patient.
Information: a soft signal that the patient may need reinforcement (e.g., readings missed for the last 48 hours, or a trend approaching threshold without crossing it).

A Critical alert requires a documented response within 15 minutes during business hours and within 60 minutes after-hours. The portal timestamps the moment the alert was created. Use that as your clock, not the moment you opened the queue.

What triggers each level

Thresholds live on the patient's Vital Configuration, set during enrollment by the care team. The defaults Vantrexia ships with are clinically conservative; you can tighten them per patient at Patients -> select patient -> Vitals -> Configure thresholds.

The most common triggers in production:

Blood pressure (high): systolic above max or diastolic above max for the patient. Two consecutive readings above threshold promote a Warning to a Critical in most CHF and stroke protocols.
Blood pressure (low): systolic under 90 or diastolic under 60 in a patient on antihypertensives is almost always Critical.
Heart rate (high/low): sinus tachycardia above 110 sustained, or bradycardia under 50 with reported symptoms.
Weight gain/loss: a 3-lb gain in 24 hours or a 5-lb gain in 7 days for CHF patients fires a Critical.
Glucose (high/low): fasting > 250 mg/dL or any reading < 70 mg/dL for diabetic patients is Critical.
SpO2 (low): anything below 90% in a non-COPD patient or below 88% in a COPD patient is Critical.

Each alert stores both the threshold value that was breached and the actual value measured, so you can see the gap at a glance without re-reading the chart.

Reading the queue

Open Monitoring -> Patient Monitoring. The page groups readings by patient: one row per patient, with their latest BP, HR, weight, glucose, and SpO2 columns plus a relative timestamp ("12m", "3h"). A red Critical pill on a row means at least one observation in the patient's recent set is out of safety range. Sort by clicking the column header to bring the worst readings to the top.

Three filter chips run across the top of the table:

All: every patient with recent readings.
Critical: only patients with at least one critical observation. Start your day here.
Pending: readings the system has received but not yet finished processing. Usually empty; if it stays full for more than 10 minutes, notify your administrator.

The KPI strip above the chips gives you the real-time state of the program: Critical count, Pending count, Today (readings processed since midnight), Total Patients, and With Data (patients who transmitted at least one reading in the period). When Critical sits at zero by lunch, the morning round is done.

Acknowledging an alert

Acknowledging tells the system "a clinician has seen this and is handling it". It does not resolve the alert: that comes when you either escalate, close it with a clinical note, or the patient's next in-range reading auto-resolves it.

From Patient Monitoring, click the patient's row, then click Charts to expand the inline chart panel. This shows the trend and the threshold zones so you can see whether the alert is a one-off spike or part of a pattern.

Open the patient's chart in a new tab at Patients -> select patient -> Alerts. Find the alert by timestamp.

Click Acknowledge. Add a one-line note ("Called patient, medication taken late this morning, repeat reading scheduled 2pm") and save. The note becomes part of the alert's audit trail and is visible to anyone reviewing the patient later.

If the alert needs more than a phone call (medication change, same-day visit, ED referral), promote it to an escalation. See Escalations.

Filtering the queue effectively

A busy clinic can have hundreds of alerts in flight. Combine these filters to keep your worklist sharp:

Critical first. Click the Critical chip the moment you log in. Work this list to zero before you touch anything else.
Search by name or MRN. The search box on Patient Monitoring matches both fields. Useful when a provider calls asking about a specific patient.
Pending before lunch. A growing Pending count signals an ingestion problem; raise it to your platform admin so they can check the device sync.
Group by acuity. Patients on the same care plan (CHF, post-MI, uncontrolled DM) tend to fire the same alert types. After a couple of weeks you'll know which patients to expect in the Critical list each Monday and can preempt with a Sunday-evening check-in call.

Auto-resolution

If a patient's next reading comes back inside their threshold band, the original alert is auto-resolved and disappears from the active queue. This is intentional: chasing every spike that the patient self-corrects 30 minutes later wastes triage time. The original alert is preserved in Patients -> Alerts -> Resolved for trend review.

Critical alerts do not auto-resolve. Even if the next reading is normal, a critical alert stays on the queue until a clinician acknowledges it with a note. This is the safety net. It forces a human to confirm a critical event was addressed, not just that the patient happened to feel better an hour later.

Common patterns triage nurses learn fast

A morning systolic spike that resolves by 10am is almost always delayed-medication. Confirm by phone, document, move on.
Three consecutive missed readings in a previously compliant patient, call the same day. Lost device, hospitalization, and skipped doses are all reasons to investigate.
A glucose alert paired with a weight loss alert in the same week warrants a same-day phone visit; this is the classic uncontrolled-DM pattern.
A weight alert without a corresponding BP alert in a CHF patient often precedes a decompensation by 24–48 hours. Don't wait for the BP to confirm.