NIHR129713 - Effectiveness and cost-effectiveness of gynaecological surveillance for women with Lynch syndrome - Economic evaluation simulation results - Data files

The results are contained in 5 datasets, each of which is described in further detail below. Each dataset is stored as a tar file, which when extracted will be in Hive/Feather format as:<br> {dataset}/age={age}/genotype={genotype}/data.feather<br> Where {age} is the starting age of the simulated individuals (30, 35, 40 or 45) and {genotype} is the genotype of the simulated individuals (path_MLH1, path_MSH2, path_MSH6, path_PMS2).<br> Each data.feather file covers 500 different parameter sets, and for each of these 4 competing options, and for each of these 1000 simulated individuals, i.e., 2 million simulated individuals per data.feather and 32 million simulated individuals overall. <strong>patient-level-outcomes</strong> This gives patient-level costs, life years and QALYs for conducting the economic evaluation. Each row/observation corresponds to a single simulated patient.<br> Fields: params_uuid - uniquely identifies the parameter set (can be used to join with other datasets) individual_uuid - uniquely identifies the individual (can be used to join with other datasets) competing_option - text description for which of the four competing options applied to the individual total_costs - total costs over the lifetime for the individual, discounted at the rate params['analysis.discount_rate.cost'] (constant 3.5% in the current version) total_life_years - total life years for the individual, discounted at the rate params['analysis.discount_rate.ly'] (constant 3.5% in the current version) total_qalys - total quality-adjusted life years for the individual, discounted at the rate params['analysis.discount_rate.qaly'] (constant 3.5% in the current version) Tip: As each observation corresponds to a single individual it is possible to calculate undiscounted life years lived from total_life_years:<br> drc = log(1 + params['analysis.discount_rate.ly'])total_life_years_undiscounted = - log(1 - drc * total_life_years) / drc <strong>params</strong> The parameters for the simulations. Each row/observation corresponds to a single parameter set. params_uuid - uniquely identifies the parameter set (can be used to join with other datasets) ... - parameter values, which are mostly scalars but some are vectors <strong>cancer-outcomes</strong> Counts of various cancer outcomes. Each row gives the count of a particular cancer outcome for a particular combination of parameter set and competing option. <strong>CAUTION: Any rows which would have n=0 have been omitted from this dataset.</strong> params_uuid - as above competing_option - as above site - colorectal, ovarian or endometrial outcome - Incidence, Recurrence or Mortality stage - Stage of cancer at time of diagnosis: I, II, III or IV (missing for mortality) route - Route to cancer diagnosis (only available if outcome is 'Incidence'): RouteToDiagnosis.SYMPTOMATIC_PRESENTATION, RouteToDiagnosis.SURVEILLANCE, or RouteToDiagnosis.RISK_REDUCING_SURGERY n - Number of times the corresponding cancer outcome occurred (for a simulated population of 1000 individuals) <strong>cancer-free-survival</strong> For each individual, how long did they survive without a cancer diagnosis or becoming censored (principal reason for censoring is death from non-cancer cause) individual_uuid - see above params_uuid - see above competing_option - see above age_event - age of the individual when the diagnosis event or censoring happened event - 1 if a cancer diagnosis happened or 0 if censoring happened cancer - CancerSite.ENDOMETRIUM, CancerSite.OVARIES or CancerSite.COLORECTUM stage - see cancer-outcomes route - see cancer-outcomes age_enter - age of the individual when entering the model (=age) sex - will be Sex.FEMALE for all simulated individuals <strong>cancer-survival</strong> Each row/observation in this dataset corresponds to survival from a diagnosed cancer in a simulated individual.<br> <strong>CAUTION:</strong> For each cancer there are two rows, because users may wish to calculate cause-specific survival or crude survival. Ensure that you filter out the calculation type you do not wish to include individual_uuid - see above params_uuid - see above competing_option - see above survival_type - 'cause-specific' or 'all-cause' (crude) age_event - age at which the individual died or was censored age_diagnosis - age at which the individual was diagnosed with this cancer event - 1 if the individual had an eligible death at age_event (determined by survival_type), 0 otherwise (e.g., died from another cause if survival_type is 'cause_specific', censored) site - see above stage - see above route - see above