1. Comparative Models of Cervical Cancer Screening in Manitoba Linda DeRiviere et al. Linda DeRiviere et al. Comparative Models of Cervical Cancer Screening in Manitoba
TABLE 1.
Methodological assumptions for models 1–8
a lack of physicians in the healthcare system has resulted in fewer • The rate of specimens that are deemed unsatisfactory for
healthcare practitioners screening women with a Pap smear, evaluation Model Test Implementation Screening Age Range Consumables* Labour Capital Equipment
Characteristics Schedule Intervals Intensity† Intensity‡
as well as a reduced probability that women will be screened • Impact of model changes on the incidence of cervical cancer
frequently. This issue is particularly problematic in remote and requiring hospital treatment 1. Baseline – CC Traditional Pap Baseline model 1–2 years on All women who Low marginal 2 6 – microscopes,
northern communities. Moreover, there is an expected shortage test (cellular test) is the current average present for costs of slide 10-year replacement
screening screening preparation interval
of cytotechnologists (laboratory screeners) in the future, which The importance of this study is its contribution to an empirical strategy
will extend waiting times for results. Currently, Pap tests are literature, which suggests that recent advancements in cervical
2. LBC, using Cellular test Year 1; Same as baseline Same as baseline ThinPrep 3 2 – processors,
performed at seven private and public laboratories, and minimal cancer screening technologies present opportunities for the ThinPrep as primary concurrent consumables stainers, automated
HPV testing is conducted for research purposes only. There is integration of HPV testing as a prevention strategy for the reduc- technology screening testing with CC imaging
no one facility for centralized testing. Similar to other fiscally tion of cervical cancer incidence, as well as enhancing women’s for a period of 6
challenged jurisdictions across Canada, Manitoba Health has health outcomes. In a fiscal climate of doing more with less, the months
been considering the ways in which newer screening platforms analysis has better informed public policy decision-makers of the 3. LBC, using Same as model 2 Same as model 2 Same as baseline Same as baseline SurePath 4 1 – processors,
could impact on the net costs and savings of delivering cervical direct healthcare costs associated with cervical disease detection. SurePath consumables stainers, automated
screening services on a province-wide basis. technology imaging
The objective of this article is to report the findings of a Methods and Eight Comparative Models of 4. LBC, primary Cellular test Same as model 2 Same as baseline Same as ThinPrep 1 3 – processors,
comparative cost analysis for the implementation of new cervical Service Delivery screening; HPV for primary baseline; consumables; stainers, automated
cancer screening technology models. Manitoba is a province Eight proposed models for the delivery of cervical cancer screening triage to assess screening, HPV triage of Qiagen’s hc2 imaging
low-grade results triaged with a low-grade results detection
of 396,635 women aged 15–69 (Statistics Canada 2007). were estimated in terms of assessing the least-cost model. Table viral test technology
Approximately 195,000 cervical screening tests are performed 1 lists the main variables that frame the methodological assump-
5. HPV and LBC Viral and cellular Year 1, HPV 3 years for Age 30+ years: hc2 for HPV 7 5 – processors,
annually, and 45 women are diagnosed with cervical cancer each tions. The study specifically focused on the marginal costs of newer testing as primary HPV-tested HPV testing; testing; ThinPrep stainers, automated
year (CancerCare Manitoba n.d.). technology platforms compared with the baseline model. The screening for women <30 years: LBC for LBC testing imaging
There were two main goals of the costing exercise. The first methodological implication of exclusively focusing on marginal some women
was to develop a framework of models for the delivery of cervical cost changes of direct government expenditures was the omission 6. HPV and LBC Same as model 5 Gradual Same as model 5 Same as model 5 Same as model 5 6 4 – processors,
screening in the province based on a literature review of economic of a societal perspective for the analysis, potentially resulting in implementation stainers, automated
costing models. This exercise generated eight models, which were an understatement of the cost savings of the newer technologies. over a 5-year imaging
subsequently costed out. The second task was to find the least Moreover, the inclusion of additional upstream (e.g., prevention/ period
costly of the eight models, using a practical cost-minimization education costs) and downstream costs (e.g., patient impact/ 7. HPV and CC Same as model 5 Year 1, HPV Same as model 5 Same as model 5 hc2 test; 8 7 – microscopes;
technique that focused on the public budget aspects of health- treatment) was beyond the scope of this analysis. The goal of the as primary conversion kit; 10-year replacement
care resource use. The cost analysis was built on a baseline model study was a narrow one that emphasized a practical and program screening for CC – same as interval
some women model 1
of cervical screening, the traditional Pap smear test. The direct context–specific, value-for-money technique.
healthcare costs in the baseline model were compared with the 8. HPV and CC Same as model 5 Gradual Same as model 5 Same as model 5 Same as model 7 5 8 – microscopes;
implementation 10-year replacement
estimates of two newer technology platforms, LBC (a cellular test)
and HPV testing (a viral test), in eight models. “ he main policy challenge in Manitoba
T over a 5-year interval
period
A substantive literature review identified key variables that is that the current laboratory system
would have an impact on the marginal costs of cervical screening for routine cervical cancer screening is CC = conventional cytology; hc2 = Hybrid Capture 2; HPV = human papillomavirus; LBC = liquid-based cytology; Pap = Papanicolaou.
*Consumables include slides, stains and other sample collection devices. We assumed minimal transportation costs of slides in all eight models. All models assume delivery in a public laboratory system.
platforms compared with the conventional Pap test (Canfell et outdated, deteriorating and unsustainable 1 = highest use; 8 = lowest use.
†
al. 2008; Insinga et al. 2004; Moss et al. 2003; Raab et al. 1998). over the longer term. ” 1 = highest; 8 = lowest.
‡
In gathering data for the cost calculations, the eight models were
estimated with a focus on these key cost drivers:
The cost categories were estimated in 2009 Canadian dollars over
• Length of routine screening intervals a period of five years. We selected this time period for the analysis
• The labour mix, including changes in workload and produc- due to the incremental nature of the implementation plan for Aggregated data were obtained from the provincial database composed of both public and private laboratory services, we deter-
tivity of labour two of the eight models. Annual inflation of 2% was applied to at the Manitoba Cervical Cancer Screening Program, as well as mined that any newly proposed model would be estimated based
• Capital expenditures and marginal costs of consumables capital costs (Bank of Canada 2010). Salary-related cost-of-living the trajectories of care following cervical screening results, which on service delivery in a public laboratory setting. In addition, the
• Impact on repeat cytology from marginal changes in sensitivity allowance was based on collective agreements of technologist and are classified into categories of cellular changes using the Bethesda methodological decisions included the following:
and specificity cytopathologist positions. No inflation factor was applied to fee Classification System. The reader should refer to Boronow (1998)
• Impact on the rate of colposcopy referrals from marginal tariffs or consumables since these costs were expected to remain for more information on the algorithms of care for abnormal • Increased sensitivity of HPV testing (97%) compared with the
changes in sensitivity and specificity (Colposcopy is a more relatively constant over the period of analysis. Depreciation and results. We gathered cost data from published physician fee Pap test (51%)
detailed examination of the cervix using an operative micro- salvage value of capital equipment were excluded from the analysis tariffs. Collective agreements were provided by Manitoba Health. • Reduced frequency of screening of qualifying women for HPV
scope and some simple stains.) since these estimates were immaterial in value. Though the current model of cervical screening in the province is testing from every one to two years to once in three years
80 Healthcare Quarterly Vol.16 No.1 2013 Healthcare Quarterly Vol.16 No.1 2013 81