FOUR DATA LAYERS

Patient journey and insights

Events and metrics show what happens. Voice of the Customer and continuous discovery explain why.

Sample dashboard
PatientHCPSurvey · interview · score
1

Enrolment

2

Engagement

3

Care path course

4

Outcomes

5

Care team

Layer 1

North star

Patient

Patient treatment initiation

HCP

Patient invites per HCP

Patient

Active patients

HCP

Active HCPs

Patient

Adherence & retention

HCP

Effective interventions

Patient

PROM improvement

HCP

PROM reviewed within N days

Patient

Reply within 24 h

HCP

Team workload

Layer 2

Metrics

Patient

HCP

Patient

HCP

Patient

HCP

Patient

HCP

Patient

HCP

Layer 3

Qualitative user research

Why it happens

Patient

HCP

Patient

HCP

Patient

HCP

—

Patient

HCP

Patient

HCP

At every stage

Layer 4

Raw events

What happened

Patient

  • Invitation received
  • Invitation opened
  • Invitation accepted
  • Account created
  • Patient segment
  • Care plan started

HCP

  • Invitation sent
  • Care plan created
  • Patient assigned
  • First invitation sent

Patient

  • Message sent
  • Screen viewed
  • Button tapped
  • Session started

HCP

  • Message sent
  • Message replied
  • Diagnosis added
  • Prescription added
  • Note added
  • Session started

Patient

  • Dose reminder sent
  • Dose confirmed
  • Survey completed
  • Reminder opened
  • Education opened
  • Care plan ended
  • Account deleted

HCP

  • Dose plan set
  • Dose plan changed
  • Patient contacted
  • Intervention sent

Patient

  • Baseline PROM completed
  • PROM completed
  • PREM completed

HCP

  • PROM result opened
  • Clinical note added
  • Treatment changed

Patient

  • Message to role sent
  • Team reply received
  • Patient replied

HCP

  • Member joined
  • Member left
  • Patient assigned
  • Session started

How we correlate

One web of cause and effect

Primary hypothesisSecondaryDriver before outcome
1

Enrolment

2

Engagement

3

Care path

4

Platform

5

Care team

Outcome 1

Adherence

Logistic

Outcome 2

Drop-out

Survival

Outcome 3

PROM change

Linear, mixed

How we test, in order

From description to an effect claim

Steps 1–3 · association in timeStep 4 · effect claim
  1. 1Step 1

    Describe by cohort

    Curves by therapy month, split by driver (e.g. adherence: fast vs slow responders).

  2. 2Step 2

    Correlate

    Driver vs target on cohorts and segments, driver measured before the target.

  3. 3Step 3

    Regress with controls

    One model per target, drivers as predictors, controls from the box above.

  4. 4Step 4

    Compare against a control

    Same doctors, patients outside the care path, or staggered start.