Gera Prescribing Intensity Index (GPII) — Methodology

1. Data source

Primary:

NHS Business Services Authority (NHSBSA) — English Prescribing Dataset (EPD) with SNOMED Code, March 2026. Dataset: opendata.nhsbsa.net. File: epd_snomed_202603.csv (18,364,409 rows, ~200 MB compressed). Licence: Open Government Licence v3.0.

Denominator (registered patients):

NHS England Quality and Outcomes Framework (QOF) 2024/25 PREVALENCE_2425.csv — HYP (hypertension) listSize = all registered patients per ICB. Published: NHS England Digital. OGL v3.0.

2. BNF chapter selection

Three BNF chapters were chosen as clinically significant markers of prescribing intensity, and because they represent three distinct pharmacological categories where variation between areas is well-documented in NHS literature:

3. Rate computation

# For each ICB and each chapter:
rate = (chapter_items / listSize) × 1,000

# England national rate:
nat_rate = sum(ICB_items across all 42 ICBs) / sum(ICB_listSizes) × 1,000

# Verified England nationals (March 2026):
ab_rate  = 48.8 items / 1,000 patients
ad_rate  = 128.5 items / 1,000 patients
op_rate  = 29.8 items / 1,000 patients
patients = 63,766,671 (QOF HYP listSize)

4. GPII formula

# Raw composite (England-equivalent = 1.000):
raw_i = (ab_rate_i / nat_ab_rate
       + ad_rate_i / nat_ad_rate
       + op_rate_i / nat_op_rate) / 3

# Normalise to 0–100:
GPII_i = round(100 × (raw_i − min_raw) / (max_raw − min_raw), 1)

# Bounds (March 2026):
min_raw = 0.5008  →  GPII 0   (NHS North West London)
max_raw = 1.6545  →  GPII 100 (NHS North East and North Cumbria)
England-equivalent raw ≈ 1.000  →  GPII ≈ 42.6

Higher GPII = higher combined prescribing intensity across the three key chapters relative to the England average. Lower GPII = lower intensity. GPII is not a measure of quality or appropriateness — it is a purely descriptive intensity index. Interpret alongside local demographics, deprivation and disease burden.

5. Coverage and limitations

• EPD covers prescriptions dispensed in primary care in England only. Hospital prescriptions, dental prescriptions and private prescriptions are excluded — this contributes to London ICBs appearing lower (higher rates of private prescribing in London, not captured in EPD).
• The QOF HYP listSize denominator is from 2024/25 (financial year ending March 2025). It is used as a stable annual figure for the March 2026 prescribing period; the mismatch is minor (list sizes change slowly).
• The 42 ICBs in March 2026 EPD match the post-April 2024 ICB structure. Earlier time-series comparisons require mapping to old ICB boundaries.
• GPII bounds (min/max) shift with each monthly update as ICB rates change, so the absolute GPII value for an ICB may shift slightly month-to-month even if its absolute rates are stable.

6. Licence and attribution

EPD data is published under the Open Government Licence v3.0. Contains public sector information published by the NHS Business Services Authority. QOF denominator data published by NHS England under the same licence.