Executive summary
Script Lift measures campaign effectiveness with an apples-to-apples comparison. Exposed HCPs are matched to controls who were similarly qualified and similarly likely to prescribe before the campaign started. Multi-stage matching, behavioral calipers, and outlier controls isolate prescribing lift driven by media exposure rather than geography, training, or volume alone. Shorter windows inside the 180-day pre-campaign period capture recent momentum and prescribing cadence, not just total volume.Engineering a statistical twin
Wrango builds the control group through deterministic matching and caliper filtering. Outlier mitigation and backfill keep cohorts balanced so reported lift reflects marketing impact.- Ability to prescribe: Start from a 1:10 pool matched on credentials, state, and specialties so controls are licensed and trained like exposed HCPs.
- Potential to prescribe: Narrow to a 1:2 pool using the 6-month pre-exposure window, matching category prescription volumes.
- Trend shape: 90-day and 30-day category prescription windows and active-month counts help the control mirror both scale and recent trend.
- Precision filtering: A Whale Purge removes the 99.95th percentile of outliers. Remaining candidates are ranked to lock a 1:1 identical twin based on category prescription volume similarity.
Nested pre-campaign windows
| Window | Role |
|---|---|
| 180 days | Primary pre-campaign category Rx volume and active-month counts for matching |
| 90 days | Shows whether an HCP was accelerating, stable, or cooling off before exposure |
| 30 days | Captures immediate pre-campaign cadence |
Phases
Phase 1: Clinical qualification and base pool definition
Phase 1: Clinical qualification and base pool definition
Before prescription behavior is evaluated, Wrango builds a pool of professional peers.
- Exposed audience: HCPs with validated campaign impressions, using a last-touch attribution model.
- Clinical tier matching: Controls are mapped to exposed HCPs using three descending tiers:
- Tier 1 (High): Primary specialty, secondary specialty, state, and credential (exact match).
- Tier 2 (Medium): Broadened slightly when exact matches are scarce.
- Tier 3 (Low): Broadest acceptable baseline to preserve pool volume.
- Initial ratio: Up to 10 control candidates per exposed NPI (1:10).
| Tier | Match criteria | Purpose |
|---|---|---|
| Tier 1 (High) | Primary specialty + secondary specialty + state + credential | Clinical gold standard for peerage |
| Tier 2 (Medium) | Primary specialty + secondary specialty + state | Clinical depth with state-level demographics |
| Tier 3 (Low) | Primary specialty + state | Regional fail-safe for volume |
Phase 2: Behavioral alignment and caliper filtering
Phase 2: Behavioral alignment and caliper filtering
Peers must show similar historical prescribing so they had the same potential to prescribe the target brand.
- Pre-campaign lookback: Category Rx volume over 180 days, plus nested 90-day and 30-day windows.
- Prescribing cadence: Active writing months across 180-day and 90-day lookbacks.
- Caliper match: Control volume must fall within ±15% or ±2 scripts of the exposed NPI (whichever is larger).
- 180-day active-month gap: within 2 months
- 90-day category volume gap: within 30% of the exposed HCP’s 90-day total, with a 1-script minimum buffer
- 30-day category volume gap: within 40% of the exposed HCP’s 30-day total, with the same 1-script minimum buffer
Phase 3: 1:1 twin optimization and recovery
Phase 3: 1:1 twin optimization and recovery
The qualified pool is reduced to a balanced 1:1 exposed-to-control ratio.
- Stage 1 (best twin): Select the single best twin per exposed NPI by lowest absolute variance in pre-campaign category volume.
- Tie-breakers: After the 180-day volume comparison, ties break on 90-day volume gap, 30-day volume gap, 180-day active-month gap, then 90-day active-month gap.
- Stage 2 (unmatched recovery): If an exposed NPI has no match, unused controls are reconsidered when:
- 180-day volume gap within 10% of the exposed HCP’s 180-day total
- 90-day gap within 20% of the exposed 90-day total
- 180-day active-month gap within 1 month
Phase 4: Outlier mitigation (Whale Cap)
Phase 4: Outlier mitigation (Whale Cap)
Extreme prescribers are capped so they do not skew results.
- 99.95th percentile rule: Maximum TRx limit from the 99.95th percentile of total prescriptions in the exposed group.
- Capping: Controls whose post-exposure TRx exceed that boundary are excluded from final aggregation.
- Symmetric pair removal: If either side of a matched pair breaches the whale threshold, the full pair is removed so exposed and control counts stay balanced.
- Pair-integrity guard: Post-match TRx gap between paired HCPs must stay within a 3-script or 1.8x band, whichever is more protective for that pair.
Phase 5: Clinical measurement
Phase 5: Clinical measurement
After matching and capping, outcomes are measured in the post-exposure window from 0 to 180 days after last exposure.
- Total Prescriptions (TRx): Distinct claim numbers in the window.
- New Prescriptions (NRx): Distinct claims flagged as a new fill.
- New patient: No prior claims on record, or gap since prior claim exceeds 180 days.
- New prescriber: No prior claims on record, or gap since last prescription exceeds 720 days.
Phase 6: Statistical significance and confidence reporting
Phase 6: Statistical significance and confidence reporting
Aggregated metrics roll up by campaign, destination, and channel to show lift and clinical impact.
- Lift calculation: Exposed volume minus control volume (difference in differences).
- Confidence intervals: 80%, 85%, 90%, and 95%.
- Margin of error: Applied from sample sizes and variance for both groups.