| |
| Cervical cancer screening in Australia: modelled evaluation of the impact of changing the recommended interval from two to three years |
| Creighton P, Lew JB, Clements M, Smith M, Howard K, Dyer S, Lord S, Canfell K |
|
|
| |
Record Status This is a critical abstract of an economic evaluation that meets the criteria for inclusion on NHS EED. Each abstract contains a brief summary of the methods, the results and conclusions followed by a detailed critical assessment on the reliability of the study and the conclusions drawn. CRD summary This study examined the cost-effectiveness of changing the usual cervical cancer screening interval from two to three years, with either reminders for those who did not attend or call-and-recall invitations. Three-yearly screening reduced the health care costs, with no substantial change in the clinical outcomes, compared with two-yearly screening. The methods appear to have been robust and the authors’ conclusions appear to be valid. Type of economic evaluation Cost-effectiveness analysis Study objective This study examined the cost-effectiveness of changing the two-year cervical cancer screening interval to three years, with either a reminder system, for those who do not attend, or a call-and-recall invitation system. Interventions The three screening strategies were the usual practice in Australia of screening every two years for sexually active women aged 18 to 69 years with reminders, screening every three years with reminders, and call-and-recall for screening every three years. The analysis included a human papillomavirus (HPV) vaccination programme, for women aged 12 to 26 years, and a strategy of no screening was considered. Methods Analytical approach:The analysis was based on a discrete-time deterministic Markov model, with a hypothetical cohort of women aged 10 to 84 years, that simulated the progression from HPV infection to cervical cancer or death. A dynamic transmission model was used to simulate the risk of HPV infection. A lifetime horizon was considered. The authors stated that the perspective was that of the health services.
Effectiveness data:The clinical data were from a selection of relevant sources, which included administrative databases, registries, published reviews, and other studies. Most of the epidemiologic data and the inputs for the usual screening strategy were from the Australian population, based on historical data (1997 to 2007) from the Victorian Cervical Cytology Register. The three-year screening strategy data were from New Zealand for the reminder system, and from England for the call-and-recall system. These sources provided the compliance with screening, which was the key input for the model. Some data, such as the accuracy of cytology, were from a systematic review of the literature.
Monetary benefit and utility valuations:Not considered.
Measure of benefit:Life-years were the summary benefit measure and they were discounted at an annual rate of 5%. Other outcomes, such as the incidence of cervical cancer and the deaths from cervical cancer, were reported.
Cost data:The economic analysis included the screening, diagnosis, and treatment costs that were paid by the Australian Medicare system, which was government reimbursed. These costs were from evaluations of cervical cancer screening, in Australia, using costs from the Medicare Benefits Schedule. All costs were in Australian dollars (AUD) and a 5% annual discount rate was applied.
Analysis of uncertainty:One- and multi-way sensitivity analyses were carried out on selected inputs to the model. Most of the alternative estimates were from published studies. Results The three-year screening interval did not increase the number of cases of cervical cancer, nor the cancer-related deaths. The projected discounted lifetime costs per person were AUD 52.1 with no screening, AUD 416.7 with two-yearly screening, AUD 396.4 with three-yearly reminders, and AUD 373.8 with three-yearly call-and-recall. The life-years were 18.8619 with no screening, and 18.8723 with two-yearly screening, three-yearly reminders, and three-yearly call-and-recall.
Three-yearly screening produced similar life-years, but was cost-saving compared with two-yearly screening. Compared with no screening, the incremental cost per life-year gained was AUD 35,000 with two-yearly screening, AUD 33,000 with three-yearly reminders, and AUD 31,000 with three-yearly call-and-recall. The greatest savings were associated with three-yearly call-and-recall.
The most influential inputs were the rates of early re-screening, the compliance rate, the assumptions about pre-cancer natural history, and the human papillomavirus (HPV) infection incidence. The base-case findings were robust to a wide range of parameter variations and the three-yearly screening remained cost-saving in most circumstances. Authors' conclusions The authors concluded that three-yearly screening was associated with a reduction in health care costs, with no substantial change in clinical outcomes, compared with the usual two-yearly screening. CRD commentary Interventions:The comparators were appropriately selected to compare the usual screening practice against the proposed strategies.
Effectiveness/benefits:No systematic review was reported to identify the sources of evidence. The key information on these sources was given and they appear to have been appropriate and relevant to the aim of the analysis. Each source reflected the real implementation of the programme, but in different countries (Australia, New Zealand, and England). Some data were from a systematic review of the literature. The uncertainty in the clinical inputs was extensively investigated in the sensitivity analyses. Life-years were a valid benefit measure, as they capture the most relevant outcome of the screening programme; life expectancy. They can also be compared with the benefits of other health care interventions.
Costs:The cost categories reflected the perspective of the health care payer, as stated by the authors. The sources were also relevant to the perspective, but the resource quantities, prices, and category totals were not provided, limiting the transparency of the analysis and making it difficult to reproduce it for other settings. The key cost items were varied in the sensitivity analysis. Other details, such as the price year and discounting, were reported.
Analysis and results:The projected costs and benefits were clearly presented and were appropriately synthesised using an incremental approach. The uncertainty was satisfactorily investigated and focused on the most uncertain inputs for the model. The results of the sensitivity analyses were clearly illustrated and discussed. The model was extensively described and was a valid choice for cervical cancer. The use of a dynamic model for HPV infection was a particularly good design. The results appear to be specific to Australia and might be difficult to transfer to other settings.
Concluding remarks:The methods appear to have been robust and the authors’ conclusions appear to be valid. Bibliographic details Creighton P, Lew JB, Clements M, Smith M, Howard K, Dyer S, Lord S, Canfell K. Cervical cancer screening in Australia: modelled evaluation of the impact of changing the recommended interval from two to three years. BMC Public Health 2010; 10:734 Indexing Status Subject indexing assigned by NLM MeSH Adolescent; Adult; Aged; Australia; Child; Female; Health Policy; Humans; Mass Screening /classification /organization & Middle Aged; Models, Theoretical; Policy Making; Time Factors; Uterine Cervical Neoplasms /diagnosis; Young Adult; administration AccessionNumber 22011000652 Date bibliographic record published 25/05/2011 Date abstract record published 27/07/2011 |
| |
|
|
