Language: English | Deutsch

Interactive forecast models for pharmaceutical product launches in Germany. Built with Python, Streamlit, and Plotly.

Live Demo: pharma-launch-forecasts.streamlit.app

Five forecasting engines demonstrate the breadth of launch scenarios a Strategic Portfolio Manager faces:

Run the apps locally to explore the interactive dashboards (see Getting Started).

pharma-launch-forecast/
├── models/
│   ├── forecast_engine.py           # Generic Entry engine (Eliquis)
│   ├── brand_competition_engine.py  # Brand Competition engine (GLP-1)
│   ├── rx_otc_engine.py             # Rx-to-OTC Switch engine (PPI)
│   ├── sildenafil_otc_engine.py     # Sildenafil OTC engine (volume-based)
│   ├── sildenafil_patient_engine.py # Sildenafil OTC engine (patient-based)
│   └── ophthalmology_engine.py     # Ophthalmology Portfolio engine (Eye Care Franchise)
├── app/
│   ├── app.py                       # Multi-page entry point (st.navigation)
│   ├── main.py                      # Page: Eliquis Generic Entry
│   ├── glp1.py                      # Page: GLP-1 Brand Competition
│   ├── rx_otc.py                    # Page: Rx-to-OTC Switch (PPI)
│   ├── sildenafil.py                # Page: Sildenafil OTC Switch (volume-based)
│   ├── sildenafil_patient.py        # Page: Sildenafil OTC Switch (patient-based)
│   └── ophthalmology.py             # Page: Eye Care Franchise
├── data/
│   ├── market_data.py               # NOAK/DOAK market data (synthetic)
│   └── glp1_market_data.py          # GLP-1 market data (synthetic)
├── exports/
│   ├── build_excel_model.py                    # Excel generator: Eliquis
│   ├── build_glp1_excel.py                     # Excel generator: GLP-1
│   ├── build_rx_otc_excel.py                   # Excel generator: Rx-to-OTC
│   ├── build_sildenafil_excel.py               # Excel generator: Sildenafil
│   └── build_ophthalmology_excel.py            # Excel generator: Ophthalmology
├── docs/
│   ├── Modell_Volumen.md             # Documentation: volume-based Sildenafil model
│   ├── Modell_Patient.md             # Documentation: patient-based Sildenafil model
│   ├── KNOWLEDGE_BASE.md             # Shared knowledge base
│   └── QUALITY_TESTS.md              # Quality test definitions
└── requirements.txt

Question: What happens when the #1 anticoagulant loses patent protection?

Two perspectives with independent parameter sets:

• Originator (BMS/Pfizer): Revenue-at-Risk quantification. Models market share erosion via exponential decay with configurable floor, authorized generic strategy, and price defense.
• Generic Challenger: Market opportunity sizing. Models uptake via logistic S-curve with a decision cascade that allocates volume by priority:

Key model features:

• Decision cascade: Tender > Aut-idem > Organic (no double-counting)
• Deterministic tender modeling: Per-Kasse checkboxes (won/lost) instead of win probabilities. Tender volume adjusted for aut-idem exclusion (time-dependent, same curve as aut-idem channel) and multi-award contracts (default: 2 main partners, 50% share)
• Tender scenario band: Bear (no tender) / Base (user selection) / Bull (all Kassen won) as transparent risk overlay on the volume chart
• Aut-idem ramp curve tied to Festbetrag (reference pricing) timeline
• Per-Kasse tender modeling (TK, BARMER, DAK, AOKs) with deterministic won/lost and ramp
• Volume decomposition: organic vs. aut-idem vs. tender
• Originator price defense and authorized generic cannibalization
• Analogie-Leitplanken: Benchmark validation against published generic penetration data (Fischer & Stargardt 2016, GaBi Journal, Bayer Q-reports, PMC) with +/-15pp tolerance bands

Question: Will Mounjaro overtake Ozempic in the German GLP-1 market?

Two perspectives with symmetric model structure:

• Eli Lilly (Mounjaro/Tirzepatid): Challenger gaining share with superior efficacy data
• Novo Nordisk (Ozempic/Semaglutid): Defending market leadership despite supply constraints

Key model features:

• Market expansion with logistic dampening (GLP-1 segment growing 25-36% p.a.)
• Indication multiplier system: T2D (base) + Obesity + CV risk + MASH/NASH
• Supply constraint modeling with capacity normalization timeline
• Scenario presets: Base Case, Bull (Lilly), Bull (Novo)
• AMNOG pricing with confidential reimbursement amounts

Real-world data points incorporated:

• 6.33M T2D patients in Germany, 13.9% on GLP-1 (BARMER 2025)
• Wegovy excluded from GKV coverage (G-BA June 2024: Lifestyle-Arzneimittel)
• Mounjaro: first-ever confidential Erstattungsbetrag in Germany (Aug 2025)
• Ozempic supply constraints since 2022

Question: What happens when a prescription drug becomes available over the counter?

Reference case: Omeprazol/Pantoprazol 20mg PPI switch in Germany (July 2009).

Two perspectives:

• Manufacturer: Dual-channel revenue modeling. Rx volume decays as OTC ramps up, but OTC reaches entirely new patients.
• Market: Category disruption. OTC launch cannibalized antacids and collapsed H2-antagonists.

Key model features:

• Dual-channel dynamics: Rx exponential decay + OTC logistic S-curve ramp
• Market expansion: 70% of OTC volume from genuinely new patients (never had Rx)
• Adjacent category cannibalization (antacids, H2-antagonists)
• Consumer awareness S-curve tied to marketing spend
• PPI-typical seasonality (GI complaints peak in autumn/winter)
• Price elasticity: patient pays 100% OTC (vs. GKV copay for Rx)
• Pharmacy margin structure: free OTC pricing vs. regulated Rx (AMPreisV)
• Three scenarios: Konservativ / Base Case / Optimistisch

Real-world data points incorporated:

• OTC market Germany 2024: EUR 10.15B (+7.1%, IQVIA)
• PPI OTC switch 2009: only ~3% of PPI volume moved to OTC (partial switch)
• 11.9% heartburn market expansion in year 1 post-switch
• Antacid cannibalization: -EUR 11M; H2-antagonists: -46%
• OTC PPI max 14 tablets/20mg per BfArM

Question: What is the commercial opportunity if Sildenafil becomes OTC in Germany?

Two complementary models with identical channel/pricing logic but different approaches to market sizing. Reference: UK Viagra Connect (OTC since March 2018).

OTC peak volume is a direct input (tablets/month). Simple, fast for scenario analysis.

OTC peak is computed from an epidemiological patient funnel (prevalence → treatment gap → addressable pool → patients × usage frequency). Best-practice approach for pharma launches.

Both models converge to the same Base Case peak (~2.1M tablets/month) but via different paths. Detailed documentation: docs/Modell_Volumen.md, docs/Modell_Patient.md.

What makes these models unique:

• Dual-channel distribution: Two pharmacy channels (stationaere Apotheke, Online-Apotheke) -- apothekenpflichtig, no Drogerie -- with time-dependent share evolution and channel-specific margin structures
• Discretion effect: Parametrized stigma-driven channel shift (baseline + sensitivity) reflecting that 36% of ED patients cite anonymity as key factor (PMC 2020)
• Rx-OTC consistency: Rx decline is derived from OTC migration, not set independently -- tablets/patients cannot exist in both pools simultaneously
• Rx manufacturer pricing: Both Rx and OTC revenue use ex-factory (manufacturer) revenue for consistency, with rx_manufacturer_share = 52%
• Brand vs. generic OTC: Viagra Connect share erosion with price premium, proportionally distributed across channels
• Tadalafil competitive moat: Cialis/Tadalafil remains Rx-only; some patients migrate to convenient OTC sildenafil
• Treatment gap closure: New OTC patients raise the effective therapy rate -- derived from actual patient volumes, not set independently
• Marketing taper: Gradual 6-month linear transition from launch budget to maintenance (not a step function)
• Profitability KPIs: Breakeven month, Marketing-ROI (5J), separate Rx/OTC gross profit tracking

Key model features:

• Dual-channel Rx/OTC with migration-based Rx decline (no double-counting)
• Logistic S-curve for OTC ramp and Tadalafil migration
• Seasonality (Valentine's, summer peak; autumn/winter trough)
• Price elasticity with compound annual erosion from generic competition
• Patient-based model adds: Rx migration cap (cannot exceed Rx pool), explicit patient funnel table
• Three scenarios: Konservativ (SVA-Auflagen) / Base Case / Optimistisch (BMG erzwingt Switch)

Real-world data points incorporated:

• ~5M men with moderate-to-complete ED in Germany, only ~30% treated (Braun et al. 2000; May et al. 2007; Arnold 2023)
• UK Viagra Connect: 63% new-to-therapy patients (Lee et al. 2021, n=1,162)
• BfArM SVA rejected OTC switch 3x (2022, 2023, 2025) but BMG publicly supports it
• Generic Rx: EUR 0.95-2.30/tablet; Viagra brand: EUR 8-14/tablet
• Online pharmacy OTC share: ~23% general, ~45%+ for stigma categories (ABDA 2024, ecommercegermany.com)
• PMC/Kantar (2020, n=11,456): 36% of ED patients cite discretion/shame as driver for online purchase

Question: How should a specialty pharma company sequence three ophthalmology products into the German market?

Models sequential market entry with shared field force, AMNOG pricing lifecycle, and cross-product synergies. Based on real-world pipeline assets (Oyster Point, Famy Life Sciences acquisitions).

Product Portfolio:

What makes this model unique:

• Sequential portfolio launch: Three products at different times, building on shared infrastructure
• AMNOG pricing lifecycle: Free pricing (6 months) → G-BA Nutzenbewertung → Negotiated Erstattungsbetrag → Annual erosion
• Facharzt adoption S-curve: Logistic diffusion across 6,600+ GKV ophthalmologists
• Field force scaling with synergies: P2 launch costs 70%, P3 costs 60% of standalone (shared reps, KOLs, congress presence)
• MVZ channel growth: Modeling the consolidation trend in ophthalmology practices (+8% p.a.)
• Competitive dynamics: iKervis (Santen), Vevizye (Novaliq/Thea), Vuity (AbbVie), OTC artificial tears

Key model features:

• Per-product patient volume, prescriber adoption, market share, and pricing curves
• Portfolio P&L: revenue decomposition by product, cumulative profit, ROI tracking
• GTM investment: sales reps, MSLs, marketing, congress, KOL programs, digital channels
• Three scenarios: Base Case / Aggressiv (schneller Aufbau) / Konservativ (Tyrvaya verzoegert)

Real-world data points incorporated:

• 8,250+ ophthalmologists in Germany, 6,600 GKV Vertragsaerzte (KBV 2024)
• Dry eye: ~10M affected, 1.5-1.9M diagnosed, only 36% get Rx
• Dry eye market DE: ~EUR 430M (Rx+OTC), CAGR 5.7% (Grand View Research)
• iKervis (Santen): EUR 120-135/month, severe keratitis only
• Vevizye (Novaliq/Thea): EU-approved Oct 2024, moderate-to-severe DED
• RYZUMVI FDA-approved Aug 2023, Tyrvaya FDA-approved Oct 2021
• Public pipeline and investor relations data

Pre-built Excel workbooks are included in exports/ for direct download. Each has professionally formatted sheets:

1. INPUTS - All parameters in editable yellow cells
2. Marktdaten - Competitive landscape and market context
3. Forecast - Monthly forecast with full data table + embedded charts
4. Dashboard - Executive summary with scenario comparison
5. Methodik - Transparency matrix classifying every data point as FAKT / ANNAHME / MODELL / LUECKE

Regenerate them after modifying parameters:

python exports/build_excel_model.py         # Eliquis
python exports/build_glp1_excel.py          # GLP-1
python exports/build_rx_otc_excel.py        # Rx-to-OTC (PPI)
python exports/build_sildenafil_excel.py    # Sildenafil OTC
python exports/build_ophthalmology_excel.py # Ophthalmology Portfolio

# Clone
git clone https://github.com/leelesemann-sys/pharma-launch-forecast.git
cd pharma-launch-forecast

# Install dependencies
pip install -r requirements.txt

# Run multi-page app (all 5 use cases)
streamlit run app/app.py

All data is synthetic and based on publicly available sources:

• BARMER Gesundheitswesen aktuell 2025
• G-BA Nutzenbewertung (Tirzepatid, Semaglutid)
• IQVIA Pharmamarkt Deutschland (public summaries)
• EMA European Public Assessment Reports
• GKV-Arzneimittelindex, Lauer-Taxe (public pricing)
• BfArM Sachverstaendigenausschuss protocols (Sildenafil OTC decisions)
• MHRA / PAGB Frontier Economics (UK Viagra Connect impact)
• PMC Lee et al. 2021 (UK real-world study, n=1,162)
• Cologne Male Survey (Braun et al., Nature)
• Public investor relations and pipeline disclosures
• FDA/EMA European Public Assessment Reports (RYZUMVI, Tyrvaya, Vevizye)
• KBV/BAeK Aerztestatistik 2024 (ophthalmology workforce)
• Grand View Research (DE dry eye market)
• Santen (iKervis pricing), Novaliq/Thea (Vevizye EU approval)

No proprietary IQVIA data, no confidential company data. The models are designed to demonstrate forecasting methodology, not to provide investment advice.

• Python 3.10+
• Streamlit - Interactive dashboards
• Plotly - Charts (dual-axis, stacked areas, grouped bars)
• XlsxWriter - Professional Excel generation with formatting and embedded charts
• Pandas / NumPy - Data processing and mathematical models

MIT