Case study · Remote monitoring · Test and learn

Turning remote health alerts into clinical decisions

I led a test-and-learn sprint for a virtual ward in Greenwich, where patients take readings at home and a monitoring team flags concerns to NHS clinicians. The sprint showed the problem wasn't the alerts. It was what happened after them.

6
service improvements shipped during the sprint
5
organisations and teams aligned on one pathway
12
patients enrolled within the first 8 weeks
Go
decision to continue to Phase 2, with a clear 50-patient scaling gate
My role
Sprint lead, day to day
Teams
Two NHS clinical teams (JET and Frailty), a 24/7 monitoring and response team
Duration
Jan 2026 – present
Technology partners
Feebris (monitoring platform), Alcove (devices and onboarding)

The question

How might we make remote readings help clinicians decide before they see a patient, rather than interrupt them after an alert fires?

Context

Remote Health Monitoring is a virtual ward within DHACT. Patients take readings at home on a connected device. A monitoring team watches the readings around the clock and escalates concerns to two NHS clinical teams.

Five organisations and teams had to work as one pathway, each with its own systems, schedules and working culture. Rather than a full rollout, we ran a test-and-learn sprint: surface pathway issues fast, using real cases, without judging individual performance.

  • Separate systems and schedules
  • No shared case view
  • Manual referral handoffs
  • Patients new to the technology

My role and scope

  • Led the sprint day to day across three clinical and operational teams and three data platforms.
  • Built the sprint's tracking tools, weekly metrics framework and data collection process.
  • Analysed alert cases end to end, which reframed the problem from "too many alerts" to "no closed loop".
  • Redesigned governance so frontline issues reached decision-makers in days, not weeks.
  • Ran a co-design workshop with both technology partners and clinicians to agree interim referral, device setup and escalation processes.

Scope

Co-led with the agency delivery partner, accountable to the head of service. Covered referral, device onboarding, alert escalation, staff training and the evidence for the continue-or-stop decision.

The insight

The monitoring team was escalating the right alerts. But once an alert left them, nobody could see what happened next.

Duplicate work

Alerts fired for missed readings that a scheduled clinical visit already covered.

A broken feedback loop

The monitoring team had to phone clinicians repeatedly to find out if an alert was handled.

No proof of value

Without recorded outcomes, no one could say whether an alert changed a clinical decision.

From insight to action

  1. Context before the visit, not interruptions after it

    Proposed a pre-visit view of each patient's recent readings and trends, in the place clinicians already look: their daily schedule.

  2. Thresholds as clinical decisions

    Proposed that clinicians set alert thresholds at enrolment and discharge, so each alert reflects a deliberate clinical choice and noise drops.

  3. Close the loop in the case notes

    Proposed recording how each alert was resolved in the case notes, replacing repeated phone calls and giving the service evidence of what each alert achieved.

  4. Governance with one job per meeting

    Replaced fortnightly strategy meetings with a weekly operational huddle and a monthly decision gate, plus regular clinician newsletters so the frontline saw their feedback acted on.

Shipped during the sprint

  • Less double-keyingPatient report fields added to the referral process
  • Flexible monitoringOnce- or twice-daily readings, on days patients choose
  • Smoother installsNew checklist and extra training for installation teams
  • Carers includedFamilies trained at onboarding, plus a patient user guide
  • Clearer escalationSecond set of observations requested, without delaying escalation
  • Ongoing supportContinuous patient training and clinician drop-in sessions

Outcomes

  • Sprint continued to Phase 2, focused on the two cohorts where evidence was strongest.
  • A clear scaling gate: 50 patients enrolled before the next go/no-go decision, replacing an arbitrary timeline.
  • Five platform changes prioritised with owners and dates, including out-of-hours escalation and an automated referral workflow.
  • Adoption made an explicit workstream, with ring-fenced capacity for embedding into everyday clinical practice.

What I learned

Test and learn is how cautious teams say yes. With several organisations and cultures involved, real cases and quick feedback built confidence that abstract plans never would.

"We heard you" only counts with visible change. Shipping fixes in days, and fitting around clinicians' time, earned more trust than thorough reporting.

Strong teams beat perfect process. Unexpected cases always come up. Shared accountability across organisations is what absorbs them.