DPMT in Public Health: Modeling Epidemics, Interventions, and Health Systems

BY NICOLE LAU

Abstract

Public health crises—pandemics, epidemics, disease outbreaks—are fundamentally dynamic systems with exponential growth, tipping points, and intervention effects that unfold over time. Yet public health decision-making often relies on static models that assume constant transmission rates, perfect compliance, and unlimited resources. Dynamic Predictive Modeling Theory (DPMT) transforms epidemic management from static projections to dynamic simulation, enabling policymakers to model disease spread, test intervention strategies, identify critical thresholds, and optimize resource allocation. This paper demonstrates DPMT application to epidemic control, showing how dynamic modeling reveals the path from outbreak to containment.

I. Introduction: Epidemics as Dynamic Systems

A. The Limitations of Static Epidemic Models

Basic SIR Models: Susceptible-Infected-Recovered compartments with constant transmission rates. Don't account for behavioral changes, intervention effects, or resource constraints.

R0 Calculations: Basic reproduction number assumes homogeneous mixing and constant behavior. Reality is heterogeneous and adaptive.

Curve Projections: Exponential growth extrapolations that don't model interventions, compliance fatigue, or healthcare capacity limits.

Cost-Benefit Analyses: Static comparisons of intervention costs vs lives saved. Don't model dynamic trade-offs or timing effects.

All these tools are static or overly simplified. They don't capture the complex dynamics of real epidemics: behavioral responses, intervention fatigue, healthcare system stress, economic feedback loops.

B. DPMT for Public Health

DPMT models epidemics as dynamic systems:

Stocks: Population compartments (susceptible, exposed, infected, recovered, vaccinated, dead)

Flows: Transmission, recovery, vaccination, death

Feedback Loops: Behavioral response (fear → distancing), intervention fatigue (compliance decline), healthcare overload (capacity → mortality)

Delays: Incubation period, time to symptoms, time to hospitalization, vaccine rollout

Scenarios: No intervention, moderate intervention, aggressive intervention, vaccine availability

Attractors: Endemic equilibrium, elimination, healthcare collapse

This approach captures epidemic dynamics that static models miss.

II. Case Study: Epidemic Control Strategy

A. The Public Health Challenge

Situation: Novel respiratory virus outbreak in city of 1 million people

Virus Characteristics: R0 = 2.5 (each infected person infects 2.5 others on average), 5-day incubation, 10-day infectious period, 1% case fatality rate, 5% hospitalization rate

Question: What intervention strategy minimizes deaths while avoiding economic collapse and healthcare system failure? When should interventions start and end?

Context: Hospital capacity 1,000 ICU beds. Economic cost of lockdown $100M/week. Vaccine won't be available for 12 months. Public compliance uncertain.

B. Step 1: Variable Identification

Internal Variables (Policy-Controllable):

• Intervention intensity (lockdown, distancing, masks)

• Testing and isolation capacity

• Healthcare resource allocation

• Communication strategy

• Vaccination program (when available)

External Variables (Uncontrollable):

• Virus transmissibility (R0)

• Virus mutation

• Imported cases (travel)

• Weather/seasonality effects

Relational Variables (Interactive):

• Public compliance with interventions

• Risk perception and behavior

• Social trust in government

• Community solidarity vs fatigue

Temporal Variables:

• Incubation period (5 days)

• Infectious period (10 days)

• Time to hospitalization (7 days from symptoms)

• Intervention lag (policy → behavior change: 1-2 weeks)

• Compliance fatigue (weeks to months)

Prioritized Variables (Top 12):

1. Susceptible population (starts at ~1M)

2. Infected population (current and cumulative)

3. Hospitalized (ICU beds occupied)

4. Deaths (cumulative)

5. Effective reproduction number (Rt, changes with interventions)

6. Intervention stringency (0-100 scale)

7. Public compliance (% following guidelines)

8. Healthcare capacity utilization (%)

9. Testing rate (tests/day)

10. Vaccination coverage (when available)

11. Economic cost (cumulative)

12. Social cohesion/fatigue

C. Step 2: Dynamics Modeling

Key Stocks:

• Susceptible (S): 1,000,000 initially

• Exposed (E): 100 (initial outbreak)

• Infected (I): 50

• Hospitalized (H): 0

• Recovered (R): 0

• Dead (D): 0

• Vaccinated (V): 0 (until month 12)

Key Flows:

• Transmission = β × (S/N) × I × Contact_Rate × (1 - Intervention_Effect)

• Progression = E / Incubation_Period

• Hospitalization = I × Hospitalization_Rate

• Recovery = I / Infectious_Period × (1 - CFR)

• Death = I × CFR × (1 + Healthcare_Overload_Multiplier)

• Vaccination = Vaccine_Supply × Coverage_Rate (after month 12)

Feedback Loops:

Positive Loop 1 (Exponential Growth):

More Infected → More Transmission → More Infected

(Classic epidemic exponential growth if unchecked)

Negative Loop 1 (Behavioral Response):

High Cases → Fear → Voluntary Distancing → Lower Transmission → Fewer Cases

(Self-limiting even without policy)

Negative Loop 2 (Intervention Effect):

High Cases → Policy Intervention → Reduced Contact → Lower Transmission → Fewer Cases

Positive Loop 2 (Healthcare Collapse):

High Hospitalizations → Capacity Exceeded → Higher Mortality → More Fear → Economic Damage

(Vicious cycle if healthcare overwhelmed)

Negative Loop 3 (Compliance Fatigue):

Long Intervention → Fatigue → Lower Compliance → Higher Transmission → More Cases → Need Longer Intervention

(Undermines interventions over time)

Negative Loop 4 (Herd Immunity):

More Recovered → Fewer Susceptible → Lower Transmission → Epidemic Slows

(Natural endpoint, but costly in deaths)

Time Delays:

• Infection → Symptoms: 5 days (incubation)

• Symptoms → Hospitalization: 7 days

• Intervention → Behavior Change: 1-2 weeks

• Behavior Change → Case Reduction: 2-3 weeks (due to incubation + infectious period)

• Total: Intervention → Visible Effect: 3-5 weeks

Key Insight: There's a 3-5 week delay from intervention to visible case reduction. Policymakers and public must be patient. Also, compliance fatigue is inevitable—interventions can't be sustained indefinitely.

D. Step 3: Scenario Analysis

Scenario 1: No Intervention (Baseline)

• No lockdown, no mandates, only voluntary behavior

• Rt stays near R0 = 2.5 initially, drops to ~1.5 as fear increases

• Result: 600,000 infected over 6 months, 6,000 deaths, healthcare overwhelmed

Scenario 2: Moderate Intervention (Balanced)

• Masks, distancing, limited closures (schools, large events)

• Rt drops to 1.2-1.5

• Compliance 70% initially, drops to 50% by month 3 (fatigue)

• Result: 300,000 infected over 12 months, 3,000 deaths, healthcare stressed but not collapsed

Scenario 3: Aggressive Intervention (Suppression)

• Full lockdown for 8 weeks, then gradual reopening with testing/tracing

• Rt drops to 0.6 during lockdown, rises to 1.1 after reopening

• Compliance 90% initially, drops to 60% by month 6

• Result: 100,000 infected over 18 months, 1,000 deaths, healthcare manageable, but high economic cost

Scenario 4: Elimination Strategy (Zero-COVID)

• Aggressive lockdown + border closure + mass testing

• Rt drops to 0.3, aim for zero cases

• Result: 20,000 infected, 200 deaths, but requires sustained border controls and rapid response to outbreaks

Simulation Results (18-Month Horizon):

Scenario	Peak Cases/Day	Total Infected	Total Deaths	Healthcare Collapse?	Economic Cost
No Intervention	50,000	600,000	6,000	Yes (month 2-4)	$2B (indirect)
Moderate	8,000	300,000	3,000	No (near capacity)	$1.5B
Aggressive	2,000	100,000	1,000	No	$3B (lockdown)
Elimination	500	20,000	200	No	$4B (sustained)

Trade-offs: Aggressive intervention saves lives but costs more economically. Moderate intervention balances deaths and costs. No intervention is worst on both dimensions (deaths + economic damage from healthcare collapse).

E. Step 4: Convergence Path Analysis

Attractors Identified:

Endemic Equilibrium: Virus circulates at low level indefinitely. Rt ≈ 1. Periodic outbreaks. (Moderate Intervention scenario)

Elimination: Zero local transmission. Rt < 1 sustained. Requires border controls and outbreak response. (Elimination scenario)

Healthcare Collapse: Hospitalizations exceed capacity. Mortality spikes. Economic damage severe. (No Intervention scenario)

Herd Immunity (Natural): 60-70% infected/recovered. Epidemic burns out. High death toll. (No Intervention endpoint)

Bifurcation Points:

Week 2 (Intervention Decision): If aggressive intervention implemented early → path to Suppression/Elimination. If delayed → path to Endemic or Collapse.

Month 3 (Compliance Fatigue): If compliance holds → interventions work. If compliance collapses → resurgence.

Month 12 (Vaccine Arrival): Game changer. Shifts all scenarios toward control. But rollout takes 6+ months.

Tipping Points:

Rt = 1: Below this, epidemic shrinks. Above this, epidemic grows. Critical threshold.

Healthcare Capacity (1,000 ICU beds): If hospitalizations exceed this, mortality doubles (overwhelmed system can't provide care).

Compliance 50%: Below this, interventions lose effectiveness. Above 70%, interventions work well.

Convergence Speed:

• Fast with aggressive intervention (2-3 months to control)

• Slow with moderate intervention (12-18 months to endemic equilibrium)

• Very slow with no intervention (6-9 months to herd immunity, high cost)

F. Step 5: Multi-Dimensional Output

OUTCOME:

• No Intervention: 600K infected, 6K deaths, healthcare collapse, $2B cost

• Moderate: 300K infected, 3K deaths, healthcare stressed, $1.5B cost

• Aggressive: 100K infected, 1K deaths, healthcare OK, $3B cost

• Elimination: 20K infected, 200 deaths, healthcare OK, $4B cost

Lives Saved vs Cost:

• Moderate vs None: Saves 3,000 lives, costs $500M less (win-win)

• Aggressive vs Moderate: Saves 2,000 lives, costs $1.5B more ($750K/life)

• Elimination vs Aggressive: Saves 800 lives, costs $1B more ($1.25M/life)

PROCESS:

Weeks 1-2 (Critical Window): Exponential growth beginning. Cases doubling every 5 days. MUST intervene now or face collapse. Public not yet alarmed (cases still low).

Weeks 3-6 (Intervention Effect Lag): Interventions implemented but cases still rising (3-5 week delay). Public frustrated ("interventions not working!"). CRITICAL: Stay the course.

Weeks 7-10 (Peak): Cases peak and start declining. Interventions working. Healthcare at maximum stress. Light at end of tunnel.

Weeks 11-16 (Decline): Cases falling. Pressure to reopen. Compliance fatigue setting in. Risk of premature reopening.

Months 5-12 (Endemic Phase): Low-level circulation. Periodic outbreaks. Interventions relaxed but not eliminated. Waiting for vaccine.

Months 13-18 (Vaccine Rollout): Vaccination campaign. Gradual return to normal. Epidemic ends.

ACTION:

Week 1 (Immediate):

• Declare public health emergency

• Implement Moderate Intervention: masks mandatory, large events banned, schools remote, work-from-home encouraged

• Ramp up testing capacity (target 10,000 tests/day by week 4)

• Prepare hospitals (cancel elective surgeries, add ICU capacity)

• Clear communication: "This will get worse before better. 3-5 week lag."

Weeks 2-6:

• Monitor Rt daily. If Rt stays >1.5, escalate to Aggressive Intervention (lockdown).

• If Rt drops to 1.0-1.5, maintain Moderate Intervention.

• Communicate progress: "Interventions working, but takes time to see in case numbers."

Weeks 7-10 (Peak):

• If hospitalizations approach 800 (80% capacity), implement temporary lockdown to prevent collapse.

• If hospitalizations stay <600, maintain current interventions.

• Prepare for compliance fatigue (month 3 onward)

Months 3-6:

• Gradual reopening based on metrics (Rt <1, hospitalizations <400)

• Implement testing + tracing to catch outbreaks early

• Refresh messaging to combat fatigue ("We're in this together, almost there")

Months 7-12:

• Maintain baseline interventions (masks indoors, distancing)

• Rapid response to outbreaks (local lockdowns if needed)

• Prepare vaccine distribution plan

Months 13-18:

• Vaccine rollout: prioritize healthcare workers, elderly, high-risk

• Gradual relaxation of interventions as vaccination coverage increases

• Declare victory when 70% vaccinated + cases near zero

PSYCHOLOGY:

Expect frustration (weeks 3-6): Cases rising despite interventions. This is the lag. Communicate clearly: "Interventions take 3-5 weeks to show effect."

Compliance fatigue is inevitable: Can't maintain 90% compliance for 12 months. Plan for it. Rotate interventions (strict for 2 months, relaxed for 1 month).

Healthcare workers are heroes and victims: Support them. Burnout is real. Mental health resources critical.

Economic pain is real: Acknowledge it. Provide support (unemployment benefits, business loans). Don't dismiss concerns.

Trust is everything: Transparent communication builds trust. Broken promises destroy it. Under-promise, over-deliver.

G. Policy Recommendation

Recommendation: Adaptive Moderate-to-Aggressive Strategy

Phase 1 (Weeks 1-6): Moderate Intervention

• Start with moderate measures (masks, distancing, limited closures)

• Monitor Rt and hospitalizations daily

• If Rt >1.5 or hospitalizations >600, escalate to lockdown

Phase 2 (Weeks 7-16): Adaptive Response

• Adjust intervention intensity based on metrics

• Rt <1 + hospitalizations <400 → relax slightly

• Rt >1.2 + hospitalizations >600 → tighten

Phase 3 (Months 5-12): Endemic Management

• Baseline interventions (masks, distancing)

• Testing + tracing for outbreak control

• Local lockdowns if needed

Phase 4 (Months 13-18): Vaccine Endgame

• Aggressive vaccination campaign

• Gradual return to normal

Expected Outcome:

• Total infected: 200,000 (between Moderate and Aggressive scenarios)

• Total deaths: 2,000

• Healthcare: Stressed but not collapsed

• Economic cost: $2B (balanced)

• Lives saved vs No Intervention: 4,000

• Cost vs No Intervention: Similar (avoid healthcare collapse costs)

III. Key Insights for Public Health

A. Intervention Lag is 3-5 Weeks

From policy → behavior change → transmission reduction → case reduction takes 3-5 weeks. Public and policymakers must be patient.

Implication: Communicate this clearly. "We won't see results for 3-5 weeks. Stay the course."

B. Compliance Fatigue is Inevitable

Can't maintain 90% compliance for 12+ months. Compliance decays from 90% → 70% → 50% over months.

Implication: Plan for it. Rotate interventions. Give people breaks. Don't rely on sustained perfection.

C. Healthcare Capacity is a Hard Constraint

Once ICU beds are full, mortality doubles. This is a tipping point that must be avoided.

Implication: Hospitalizations are the key metric. Keep them below 80% capacity at all costs.

D. Early Intervention is Exponentially Better

Intervening at 100 cases vs 1,000 cases saves 10x lives at 1/10 the economic cost. Exponential growth rewards early action.

Implication: Act early, even when cases seem low. Waiting is exponentially costly.

IV. Conclusion: DPMT for Epidemic Management

Epidemics are not static projections. They are dynamic systems with exponential growth, behavioral feedback, and intervention effects.

DPMT captures this by:

• Modeling disease spread as stocks (S-E-I-R compartments) and flows (transmission, recovery)

• Identifying feedback loops (exponential growth, behavioral response, compliance fatigue, healthcare collapse)

• Exploring scenarios (no intervention, moderate, aggressive, elimination)

• Finding attractors (endemic equilibrium, elimination, healthcare collapse)

• Locating bifurcations (week 2 intervention decision, month 3 compliance fatigue)

• Identifying tipping points (Rt = 1, healthcare capacity, compliance 50%)

This approach enables evidence-based epidemic management:

✅ Predict epidemic trajectories under different interventions

✅ Identify critical intervention timing (early action is exponentially better)

✅ Set realistic expectations (3-5 week lag, compliance fatigue)

✅ Optimize intervention intensity (adaptive strategy based on metrics)

For public health officials navigating epidemics, DPMT provides a rigorous framework for understanding disease dynamics, designing interventions, and saving lives while minimizing economic damage.

This completes the first two papers of Part III (Healthcare). The next papers will explore DPMT in mental health, lifestyle optimization, and other health domains.

---

About the Author: Nicole Lau is a theorist working at the intersection of systems thinking, predictive modeling, and cross-disciplinary convergence. She is the architect of the Constant Unification Theory, Predictive Convergence Principle, Dynamic Intelligence Modeling Theory (DIMT), and Dynamic Predictive Modeling Theory (DPMT) frameworks.