6  Health economic evaluation

The health economic evaluation model (03_econ_engine.R) quantifies the financial and utility impacts of the clinical pathway.

The model calculates outcomes across multiple distinct perspectives: operational capacity released, short-term hospital budget impact (Tier 1) and a long-term societal cost-utility analysis (Tier 2).

6.1 Incremental net volumes (Treatment vs. Control)

To ensure the model only attributes value to the intervention itself, the economic model operates on a incremental basis. The model calculates total clinical volumes (such as specialist assessments, pharmacotherapy prescribed, and successful 4-week quits) for both the intervention pathway (pathway_t) and a baseline control pathway (pathway_c).

All subsequent economic calculations—from staffing requirements to cost-utility—are driven by the net difference between these two scenarios, preventing the overstatement of service impact.

6.2 Costing methodology: fixed capacity

Rather than using pure variable costing (e.g. only estimating costs per minute of clinical contact), hospital services require the employment of whole time equivalent (WTE) staff, who carry associated management and administrative overheads regardless of minute-by-minute utilisation.

The model utilises a fixed costing architecture. It calculates the necessary clinical WTEs based on the volume of specialist assessments required, and subsequently scales proportional management (Band 6 and Band 8a) and administrative staff:

# Extract from 03_econ_engine.R

# Calculate base clinical capacity required
wte_required <- total_assessments / capacity_per_advisor
wte_exact    <- max(0, wte_required)

# Scale management and administrative overhead proportionately
wte_band8a   <- wte_exact * get_exact_param("wte_manager_band_8a_per_advisor")
wte_band6    <- wte_exact * get_exact_param("wte_manager_band_6_per_advisor")
wte_admin    <- wte_exact * get_exact_param("wte_admin_per_advisor")

fixed_staff_cost <- (wte_exact * salary_advisor) +
(wte_band8a * get_exact_param("salary_manager_band_8a")) +
(wte_band6 * get_exact_param("salary_manager_band_6")) +
(wte_admin * get_exact_param("salary_admin"))

This ensures the estimated cost of service delivery reflects real-world NHS employment and deployment realities.

6.2.1 Diagnostic variable costing and clinical utilisation

While fixed costing dictates the actual financial burden on the Trust, the model includes a diagnostic module to evaluate service efficiency. It calculates a theoretical “pure variable cost” (representing the cost if staff were paid strictly per minute of clinical contact, such as a 45-minute specialist assessment or a 10-minute post-discharge call).

By comparing this theoretical variable cost against the fixed WTE staff cost, the engine derives a Clinical Utilization Rate. This metric acts as an efficiency footprint, allowing analysts to quantify the proportion of commissioned staff time actively spent delivering direct clinical care versus administrative overhead.

6.3 Operational capacity released: Bed days freed

The epidemiological model’s multi-year Lexis loop calculates the downstream volume of avoided readmissions. To translate this into physical capacity, the model maps the avoided admissions back to the baseline Average Length of Stay (ALOS) derived for that specific demographic and condition matrix. This generates the Net Bed Days Freed metric, providing bed managers with a highly tangible measure of how the intervention clears physical space on the wards over time.

Because this is evaluated chronologically, bed day savings exhibit a natural, multi-year curve. Savings ramp up as successive treated cohorts enter the simulation, reach a saturation peak, and then gradually taper off due to cohort aging and natural background mortality limits over the follow-up period.

6.4 Tier 1 perspective: short-term hospital budget impact

The Tier 1 perspective evaluates financial viability from the viewpoint of the commissioning Trust. It contrasts the fixed service delivery costs against the immediate, cashable savings generated from avoided readmissions.

The volume of avoided readmissions is supplied directly by the Lexis loop. The financial value of these readmissions is calculated using Healthcare Resource Group (HRG) unit costs specific to the prevented disease events. A user-defined proportion_cashable_savings parameter is applied to account for the fact that not all freed hospital bed-days translate immediately into cashable financial returns.

Crucially, the multi-year financial tracking dynamically calculates Net Cost Savings for each individual year of the simulation horizon. It subtracts the ongoing, year-specific intervention costs (such as pharmacotherapy and intervention staff time) from the gross hospital savings generated by prevented admissions in that exact same year.

6.5 Subgroup financial stratification and equity

All Tier 1 financial outcomes—including Net Cost Savings and Bed Days Freed—are fully compatible with the model’s demographic stratification engine. The economic loop calculates parallel financial outputs across Age Bands, Sex, and IMD quintiles simultaneously.

6.6 Tier 2 perspective: long-term societal value

While Tier 1 focuses on acute operational budgets, Tier 2 evaluates the wider systemic value of the intervention, aligning with National Institute for Health and Care Excellence (NICE) methodologies.

This perspective expands the time horizon to the remaining lifetime of the cohort and incorporates community-based costs (such as post-discharge stop smoking services).

6.6.1 Incorporating lifetime cost savings

Successful smoking cessation reduces the lifetime incidence of costly chronic conditions. The model merges age- and sex-specific lifetime cost saving multipliers onto the projected volume of lifetime quitters:

# Extract from 03_econ_engine.R

calc_godfrey <- function(dt) {
temp <- merge(copy(dt), lifetime_costs_dt, 
              by = c(AGE_GROUP_NAME, SEX_NAME), all.x = TRUE)
return(sum(temp$total_quits_lifetime * temp$lifetime_saving, na.rm = TRUE))
}

total_godfrey_value <- calc_godfrey(pathway_t) - calc_godfrey(pathway_c)

6.6.2 Incorporating lifetime QALY gains

To calculate cost-utility, the model must quantify improvements in health-related quality of life and life expectancy. The model maps demographic-specific Quality Adjusted Life Year (QALY) gains to the successful quitting cohort in the same manner as the lifetime costs.

6.7 Calculating the incremental cost-effectiveness ratio (ICER)

The main metric for systemic value for money is the incremental cost-effectiveness ratio (ICER). The model calculates the net societal cost (incremental intervention costs minus lifetime savings) and divides it by the net QALYs gained:

# Extract from 03_econ_engine.R

net_societal_cost <- incremental_intervention_cost - total_godfrey_value
icer_value <- fifelse(net_qalys_gained == 0, 0, net_societal_cost / net_qalys_gained)

This output allows analysts to benchmark the service directly against standard NICE willingness-to-pay thresholds (e.g. £20,000 to £30,000 per QALY).

6.7.1 Net monetary benefit (NMB)

While the ICER provides a standard ratio for cost-effectiveness, the model also calculates the Net Monetary Benefit (NMB) to facilitate easier comparison across multiple service improvement scenarios.

By applying a user-defined willingness-to-pay (WTP) threshold (defaulting to the standard NICE threshold of £30,000 per QALY), the model converts health utility gains directly into a monetary value, subtracting the net societal cost:

# Extract from 03_econ_engine.R

wtp_threshold <- get_exact_param("wtp_threshold", default_val = 30000)
net_monetary_benefit <- (net_qalys_gained * wtp_threshold) - net_societal_cost

A positive NMB mathematically guarantees that the intervention is cost-effective at the specified threshold, providing policymakers with a clear, absolute metric of value.