PALASH SHARMA / SENIOR PRODUCT MANAGERABTRACE CASE STUDY · 2026

PROACTIVE CARE. SHARED MORE EQUALLY.

Find the gap.
Plan the care.
Close the loop.

Care Planner helps primary care leaders turn unequal care completion into a plan their teams can actually deliver.

Independent proposal built on Abtrace’s existing monitoring, analytics and recall capabilities.

abtraceCare PlannerCONCEPT
PopulationCare planOutcomes
NORTHFIELD PCN · HYPERTENSION

The next 12 weeks

Needs a plan
54
additional people
completing care vs usual care · modelled
Usual care72.7%
With Care Planner78.1%
SYNTHETIC COHORT · TRANSPARENT ASSUMPTIONS
01 / Understand02 / Decide03 / Act04 / Reassess
01 / USERS & PROBLEM

“Who needs more
than another recall?”

A population average tells a leader how the system is doing. It can conceal who is still waiting. The decision is where to focus a finite care team next.

PRIMARY USER PCN = PRIMARY CARE NETWORK

The PCN clinical lead

Accountable for proactive care across practices, with limited capacity and uneven completion.

“I need to explain which group we will help, which appointments we will use, and whether it worked.”
Proposed job to be done · to validate in discovery
DELIVERY PARTNER

Practice manager & recall lead

Turns a priority into staff time, contact attempts and appointments. Needs an executable plan with clear ownership.

BENEFICIARY

Patients with access barriers

May need a phone call, combined visit or reasonable adjustment. Non-response alone does not explain the barrier.

Unseen gap

Aggregate looks healthy

A strong practice average can coexist with lower completion in an underserved cohort.

Unmade decision

Capacity is finite

A leader needs to compare outreach choices with staff time and available appointments.

Unproven result

Activity ≠ completed care

Messages sent are useful operational data. Success here means eligible people complete the care due.

380M+

GP appointments reported in the year to July 2025.[1] System context

Core20PLUS5

NHS England prioritises reducing inequalities, including hypertension management.[2] Policy context

Unverified locally

The size of the planning burden and completion gap requires customer discovery. National data does not establish Abtrace’s baseline.

02 / PRODUCT OPPORTUNITY

Connect the
decision in between.

Abtrace already finds care needs, estimates appointment demand and automates recall. The opportunity is to help a leader decide which response will close a gap within capacity.

EXISTS TODAY

Understand

Population filters, clinical task status, achievement analytics and practice comparisons.

Product documentation [3]
PROPOSED EXTENSION

Decide together

Equity comparison + capacity scenarios + an approved plan + a scheduled outcome review.

Care Planner
EXISTS TODAY

Act & monitor

Automated recall, self-booking and recall analytics, including unsuccessful contacts.

Product documentation [5]

Gap hypothesis: the public documentation reviewed does not describe this integrated workflow. A customer interview and an Abtrace product demo could disprove the gap.

Decision to makeProposed additionOutcome to verify
Who is being missed?Compare completion by approved cohort; show missing attributes.Find a gap worth addressing.
What can we deliver?Model booked slots, attendance and staff effort.Approve a feasible care plan.
Did it make a difference?Compare outcomes with a usual-care group.Scale, adjust or stop.
03 / MEASURABLE HYPOTHESIS

One cohort.
A testable promise.

Use a fixed 1,000-person hypertension cohort over 12 weeks. People are eligible for monitoring at baseline; those already complete remain in the denominator.

PROPOSED PILOT THRESHOLDS · NOT OBSERVED RESULTS

With the same 160 clinical slots, Care Planner will increase completion by at least 5 percentage points and narrow the priority-group gap by at least 8 points versus usual care.

12-week follow-upComparison group decline ≤2 ppOutreach ≤10 staff hours

Here is how the estimate connects.

All cohort values below are assumptions
1. STARTING COHORT650 / 1,000

400 priority-group patients: 200 complete.
600 comparison patients: 450 complete.

65% complete · 25 pp gap
2. USUAL CARE · WEEK 12727 / 1,000

20 priority + 80 comparison slots booked. Assumed attendance: 50% / 84%.

77 completions · 60 slots unfilled
3. PLANNER · WEEK 12781 / 1,000

88 priority + 72 comparison slots booked. Assumed attendance: 80% / 85%.

131 completions · 160 slots used
+54additional completions
781 − 727
+5.4 ppoverall completion lift
78.1% − 72.7%
−16.0 ppgap versus usual care
33.7 pp − 17.7 pp
8.8 hrsassisted outreach effort
88 bookings × 6 minutes
Inspect the arithmetic and the limits

Usual care adds round(20 × 0.50) + round(80 × 0.84) = 10 + 67 = 77 completions. Planner adds round(88 × 0.80) + round(72 × 0.85) = 70 + 61 = 131. Each slot represents one unique person completing one defined monitoring requirement.

Gap = comparison-group completion rate minus priority-group completion rate. Usual care: 517/600 − 210/400 = 33.7 pp. Planner: 511/600 − 270/400 = 17.7 pp. The comparison group is 1.0 pp below usual care, inside the proposed 2 pp guardrail.

Booking demand and attendance improvements are hypotheses, not Abtrace findings. The estimate assumes assisted outreach fills unused capacity and raises attendance. At 60% priority-group attendance, the same plan yields 114 completions, only +3.7 pp over usual care: the completion target would fail. The prototype exposes this sensitivity.

Same clinical capacity does not mean the same total cost: the model adds 8.8 hours of outreach. Pilot value is additional completed care per incremental staff hour; this model makes no claim about hospital admissions, cash savings or clinical outcomes.

04 / INTERACTIVE PROTOTYPE

Make the trade-off.
See what changes.

Three care situations. One decision loop. Choose a scenario, inspect the gap, plan within capacity, launch a simulated programme and review the result.

abtraceCare PlannerNorthfield PCN · Synthetic data
PatientPopulationOptimisationCare Planner Concept
Interactive model · not a clinical forecastEach scenario uses distinct constraints, interventions and synthetic outcomes.
05 / SUCCESS & LEARNING

Measure completed care.
Prove the difference.

Validate the product and the assumptions together. A before-and-after chart cannot tell us how much improvement Care Planner caused.

MeasureWhy it matters12-week pilot rule
Completion liftCompleted eligible people ÷ fixed baseline eligible cohort; planner minus usual care.≥5 pp improvement
Equity gap reductionDifference between groups; compare change against usual-care practices.≥8 pp narrower
Planning effortTime from cohort review to approved plan, using session logs and staff diaries.Median ≤30 min
Discovery baseline needed
Resource feasibilitySlots used and outreach minutes; completed care per additional staff hour.No capacity overshoot
Outreach ≤10 hours
Balanced outcomesComparison-group completion, inappropriate contacts, duplicate outreach and unknown attributes.Decline ≤2 pp; zero serious safety incidents
DESIGN

Compare like with like

Recruit 8–12 practices across two PCNs for a feasibility pilot. Match or randomise at practice level where possible, account for baseline mix and seasonality, and report confidence intervals. Determine powered sample size after observing baseline variance.

DENOMINATORS

Keep the cohort honest

Freeze eligibility and group definitions at baseline. Report deaths, moves, exclusions and unknown attributes separately. A recall marked “success” is not automatically proof of completed monitoring.

Recall status distinction [5]
DECISION

Be willing to stop

Do not scale if uplift depends on extra clinical capacity, if another group falls behind, or if outreach cost outweighs the benefit. If attendance assumptions fail, change the intervention and retest.

06 / DELIVERY & COMMERCIAL THINKING

Start with one decision.
Earn the next release.

The Senior PM’s job is to connect clinical value, reliable data, usable decisions and a proposition a population-level buyer will fund.

WEEKS 1–2

Prove the problem

Observe three population-planning meetings and five recall sessions. Confirm a repeated decision, buyer, baseline gap and current workaround.

Gate: users cannot already solve this well.
WEEKS 3–6

Build the smallest loop

Start with hypertension. Agree event definitions with data engineering, import verified capacity, show assumptions and require owner approval.

Gate: trustworthy data and a feasible plan.
12-WEEK PILOT

Test and expand

Run the agreed comparison. Add other pathways only after the workflow earns repeat use and the impact thresholds hold.

Gate: measurable care value within cost.
WHO PAYS?

A PCN or ICB funding proactive care.

Test a paid population-planning module with existing Abtrace customers. The purchasing case is more completed care per constrained resource, supported by an auditable programme report.

WHAT MUST BE TRUE?

Data access is permitted; the buyer has budget and authority; staff can act on the plan; measured value exceeds implementation, outreach and support costs. Pricing and willingness to pay remain discovery questions.

DATA FOUNDATION

Patient-task eligibility → practice and approved cohort → outreach & booking events → completed care → capacity & staff effort. Show refresh timestamps and missingness. Restrict record-level access, suppress small group counts, and preserve an “unknown” category.

07 / EVIDENCE & ASSUMPTIONS

A proposal you
can interrogate.

Public sources establish the product baseline and NHS context. The impact model, user needs and pilot thresholds are proposals to validate with customers.

  1. NHS England · GP appointments, July 2025System scale only; not a local Abtrace baseline.
  2. NHS England · Core20PLUS5Policy rationale for an equity lens and hypertension focus.
  3. Abtrace · Population ViewExisting patient status, filters and advanced searches.
  4. Abtrace · Staffing and capacity planningExisting appointment estimates and staff task groupings.
  5. Abtrace · Recall AnalyticsExisting recall tracking; success categories differ from care completion.
  6. Abtrace · Analytics in the User Management PortalExisting analytics baseline. Product documentation reviewed October 2026.
  7. Abtrace · Current product interfaceBurgundy, neutral surfaces and clinical status colors inform this concept’s visual language.