Table 1

Input parameters for modelling LDCT lung cancer screening in the New Zealand (NZ) setting

VariableBase case (range)DistributionSource in this modelChange with respect to BODE3 model
Performance of CT as a screening test
LDCT sensitivity84.6% (80%–89%)Beta (183, 33)NELSON15Updated from NLST to NELSON evidence
LDCT specificity98.6% (98.4%–98.7)Beta (23238, 330)NELSON15 26Updated from NLST to NELSON evidence
Proportion of population screened per round70% (60%–79%)Beta (56, 24)Based on breast screening target coverage for NZ45Changed to provide equal coverage for Māori and non-Māori
Proportion never screened20% (11%–31%)Beta (12, 48)Bowel screening pilot (Litmus)46Changed to provide equal coverage for Māori and non-Māori
Performance of diagnostic tests
Diagnostic test adherence87.5% (82%–92%)Beta (120, 12.7)NELSON27Updated from NLST to NELSON evidence
Major complication rate from diagnostic test6% (1.1%–15%)Beta (2.4, 37.8)47 48Changed to include rate of pneumothorax from needle biopsy
Incidental findings rate7.5% (5.2%–10.2%)Beta (32, 394)NELSON8No change
Overdiagnosis rate in screening arm8.9%NELSON8Updated to NELSON evidence
Disease/state morbidity
DW for first 5 months lung cancer diagnosis and treatment0.469NilBased on GBD DWs with disaggregation by clinical phase29 49No change
DW for 1 month assumed terminal0.548Nil29 49No change
DW for 5 months assumed preterminal lung cancer0.539Nil29 49No change
DW per annum after diagnosis and treatment (ie, remission)0.315Nil29 49No change
DW for complication from diagnostic test0.0158Nil29 49No change (based on moderate respiratory disease for 30-day duration)
Incidence, stage at diagnosis, mortality and survival rates
Lung cancer stage at diagnosis (%)Local
Regional
Distant
13
23.3
63.7
NilMLCR (online supplemental appendix A, online supplemental file)16Changed from stage distribution imputed from NZ Cancer Registry to the more complete stage distribution of MLCR
Stage distribution with LDCT lung screening (%)
Local
Regional
Distant
66.3
21.6
12.1
NilNELSON27Updated from the proportionate shift observed in NLST to a final distribution that matches NELSON
Lung cancer relative survivalVaried by sex, age, ethnicity and stageNilFrom analyses of linked cancer mortality data by stage and operationalised as log-normal survival probabilities (online supplemental appendix A, online supplemental file).Changed to stage-specific Pohar Perme net survival estimates derived from analysis of NZCR data where SEER extent of disease was available or calculated from TNM values
Background (ie, non-lung) mortality with 1.75 %/2.25% annual decrease for non-Māori/MāoriVaried by sex, ethnicity and ageNilFrom projected life tables by sex, age and ethnicity50No change
Background or expected morbidityVaried by sex, ethnicity and ageNilPrevalent YLDs from NZ BDS49No change
Cessation rateVaried by sex, ethnicity and ageNil‘Pessimistic scenario’ projections from analyses of 2006–2013 census data25‘Best estimate’ projections from census data25
Lead time (years)0.5Nil51No change
Direct costs (all in NZ$)
Cost per person invited30 (19–43)Gamma (25, 1.2)52
Cost per LDCT scan400 (362–440)Gamma (400, 1)Average of three quotesQuotes obtained for this project, middle quote used as base case
Percentage increase in CT costs to account for new nodule monitoring protocol (%)2 (1.1–3.1)Beta (15, 737)NELSON RCT152% additional scans for nodules
Cost per diagnostic test1837 (1214–2622)Gamma (25, 73)52No change
Cost of incidental findings500 (324–714)Gamma (25, 20)17(online supplemental appendix A, online supplemental file).No change
Cost of major complications2835 (1820–4148)Gamma (25, 113)Based on purchasing power parity adjusted53 cost of a pneumothorax54No change
Health system costs
Base cost by sex and age of any citizen, with excess costs* of lung cancer (first year of diagnosis, last 6 months of life if dying of lung cancer and in-between)Applied as multiplier to the costsNormal (1, 0.1) (correlated 1.0 across all sex, age and ethnic groups)See Ref. 32 for methods and dataNo change
  • *Excess to ‘average’ NZ.

  • BODE3, Burden of Disease Epidemiology, Equity and Cost-Effectiveness Programme; DW, disability weight; GBD, Global Burden of Disease; LDCT, low-dose CT; MLCR, Midland Lung Cancer Registry; NZCR, New Zealand Cancer Registry; RCT, randomised controlled trial; SEER, Surveillance, Epidemiology and End Results programme of the National Cancer Institute ; TNM, tumour, node, metastasis cancer staging; YLD, year lost due to disability.