Analytical approach:
A decision analytic Markov model was used to assess costs and outcomes associated with the interventions. The model had been previously published but was not used in its entirety; only the core disease progression component of the model was used (see Other Publications of Related Interest). The model estimated the lifespan of women on the basis of participation in one of the screening programmes. The time horizon was 20 years. The economic analysis was reported to adopt the perspective of the public healthcare payer.
Effectiveness data:
Clinical and effectiveness data were derived from previously published studies. Transition probabilities were calculated on the basis of incidence. The incidence rate was assumed to be constant. Prevalence data of different stages was estimated from three studies. The main effectiveness measures used in the model were probability of detecting cervical cancer at different disease stages. These estimates were derived from previously published studies.
Monetary benefit and utility valuations:
Age-specific utility estimates in the general population were derived from the Hungarian National Health Survey in 2000 and from previously published studies. Weights for newly diagnosed cancers at different stages were derived from the original published modelling study undertaken in USA (see Other Publications of Related Interest). Weights for undiagnosed and treated cancer were based on expert opinion.
Measure of benefit:
The summary measure of benefit was quality-adjusted life-years (QALYs) gained. Benefits could be generated over a 20-year time period. Future benefits were discounted using an annual rate of 5%.
Cost data:
Direct costs were pap smear, cytological examination, gynaecological screening, anti-inflammatory treatment, conisation and costs of treating cervical cancer (including in-patient, outpatient, imaging, home care and drugs). Healthcare system reimbursement of travel costs to attend screening and travel costs were included. Costs of treating cervical cancer were derived from individual patient-level data from two Hungarian hospitals. Other costs were derived from tariffs obtained from the National Health Insurance Fund. The price year was 2006. Costs could be incurred over a 20-year time horizon. Future costs were discounted using an annual rate of 5%. All costs were reported in US dollars ($), adjusted for purchasing power parity, using the exchange rate of $1=135 Hungarian forints.
Analysis of uncertainty:
The authors reported that a series of one-way sensitivity analyses were performed by varying model parameters by +/-10%. Probabilistic sensitivity analyses were undertaken by defining distributions for key input parameters and conducted using 5,000 Monte Carlo simulations with sampling from the probability distributions. Results of the probabilistic sensitivity analysis were presented using a cost-effectiveness acceptability curve.