|HPV testing for cervical cancer screening appears more cost-effective than Papanicolau cytology in Mexico
|Flores YN, Bishai DM, Lorincz A, Shah KV, Lazcano-Ponce E, Hernandez M, Granados-Garcia V, Perez R, Salmeron J
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.
The aim was to determine the incremental costs and effects of different human papillomavirus (HPV) testing strategies, compared with Papanicolaou (Pap) cytology or no testing, in the primary screening for cervical cancer. The authors concluded that HPV testing could be cost-effective, for a large health care delivery organisation, such as the Mexican Institute of Social Security. There were some limitations in the reporting of the methods and results, which makes it difficult to assess the authors’ conclusions.
Type of economic evaluation
The aim was to determine the incremental costs and effects of different human papillomavirus (HPV) testing strategies, compared with Papanicolaou (Pap) cytology or no testing, in the primary screening for cervical cancer.
Five cervical cancer screening strategies were assessed: no screening; Pap test; self-administered HPV test; clinician-administered HPV test; and clinician-administered HPV and Pap tests. Two age groups of women were evaluated for each strategy: 20 to 80 years and 30 to 80 years.
A decision-tree model was used to estimate the cost-effectiveness of the various screening strategies, in a hypothetical cohort of 10,000 women aged 20 to 80 years or 6,680 women aged 30 to 80 years, based on national census data (2000). The time horizon was one year. The authors stated that the perspective was that of the health sector, defined as the Mexican Institute of Social Security (IMSS), a large, national health delivery organisation.
The main effectiveness parameters were the sensitivity and specificity of each of the screening strategies. The base-case estimates and their 95% confidence intervals were from the Morelos HPV study (Salmeron, et al. 2003, see 'Other Publications of Related Interest' below for bibliographic details) and they were reported to be comparable to the findings of two meta-analyses. Other clinical model inputs included the prevalence of high-grade squamous intra-epithelial lesions or cervical cancer, and the percentages of patients receiving colposcopy with or without biopsy, and lost to follow-up. The probabilities for each event were from the Morelos HPV study.
Monetary benefit and utility valuations:
Measure of benefit:
The number of histologically confirmed cases of cervical intra-epithelial neoplasia (CIN) grades two or three, or cervical cancer detected.
The direct costs included those for screening by self- or clinician-HPV and Pap testing, colposcopy with or without biopsy, pathology and laboratory processing, and treatment for CIN 2 or 3 and cervical cancer. The micro-costing (multiplying the unit price by each quantity of resource for each process) used data collected during the IMSS cervical cancer screening programme, in Morelos, Mexico, except for treatment costs, which were published Mexican estimates. The prices (unit costs) were from various sources including the IMSS, the National Institute of Public Health of Mexico (INSP), and market prices. The costs were in 2008 US dollars ($) and a discount rate of 3% over 50 years was used for the equipment costs.
Analysis of uncertainty:
One-way sensitivity analyses were performed on the sensitivity of the screening tests, the cost of the screening tests, and the treatment costs. The ranges of values for the sensitivity and specificity of the tests were based on their 95% confidence intervals, while the costs were varied by ± 25% of their baseline estimates.
The total cost per patient for women aged 20 to 80 years was $110 with no screening, $72 with Pap test, $67 with self HPV test, $52 with clinician HPV test, and $55 with Pap and clinician HPV test. For women aged 30 to 80 years, it was $165 with no screening, $97 with Pap, $94 with self HPV, $67 with clinician HPV, and $67 with Pap and clinician HPV.
The cases of CIN 2 or 3 and cervical cancer detected for women aged 20 to 80 years were 78 with Pap, 93 with self HPV, 122 with clinician HPV, and 128 with Pap and clinician HPV. For women aged 30 to 80 years, they were 82 with Pap, 90 with self HPV, 121 with clinician HPV, and 128 with Pap and clinician HPV.
Screening women aged between 30 and 80 years for cervical cancer using clinician HPV testing or the combination of clinician HPV and Pap was always more cost-effective than Pap alone. HPV testing in women aged 30 to 80 years was more cost-effective than screening women aged 20 to 80 years.
In the sensitivity analysis, the cost per case of CIN 2 or 3 or cervical cancer prevented was most influenced by the sensitivity of the screening tests. Varying the test and treatment costs had greatest impact on the clinician HPV and the combined Pap and clinician HPV strategies.
The authors concluded that HPV testing could be cost-effective in screening for cervical cancer, for a large health care delivery organisation, such as the IMSS.
The authors described the interventions and the selection was appropriate, as a range of strategies were compared, including the existing screening strategy (Pap test only) and no screening. The authors stated that the usual screening practice in Mexico was the Pap test, but the age range and the frequency of screening were not reported.
The details of the clinical study that provided the effectiveness estimates were not reported, but full details of the design, methods, and baseline characteristics were available elsewhere. It was unclear whether a systematic review was undertaken to identify all the relevant evidence, making it uncertain if the best available evidence was used. The source for the prevalence of CIN 2 or 3 or cervical cancer was not reported. The benefit measure only considered the detection of CIN 2 or 3 and cervical cancer, rather than life-years or quality-adjusted life-years, which limits comparison with the benefits of other health care interventions. The potential negative effects for patients from false-positive results were not assessed.
A micro-costing approach was used to determine the costs, but the details of some components were not given, such as the type of personnel and equipment involved and the staff pay rates. The cost estimates were presented separately for each screening strategy, with the associated follow-up and treatment costs, but the resource use quantities were not reported, which reduces the transparency of the study. Details, such as the price year were provided. All costs, except the annual costs for equipment, were assumed to occur in the same year and were appropriately not discounted. The costs for CIN 2 or 3 and cancer treatment were not from the IMSS cervical cancer screening programme in Mexico, but the authors appear to have made efforts to obtain accurate local cost estimates.
Analysis and results:
The model was satisfactorily reported and a diagram was given. The authors stated that they calibrated the model for the sensitivity and specificity of the Pap and HPV tests and they referenced the data source, but they did not explain which parameters were calibrated and how. They described three of the screening strategies as viable, but did not explain this; it appears that they were the strategies that detected the most cases. The authors did not report that one strategy dominated the others, as it was more effective and less expensive. One-way sensitivity analyses assessed if the model’s findings were robust and the results were reported as percentage changes in the incremental cost-effectiveness ratios, but changes in the absolute value of these ratios are more important. A probabilistic sensitivity analysis could have more fully captured the overall model uncertainty. The authors mentioned other factors that should be considered when deciding which strategy was most appropriate, including the existing screening resources and infrastructure, and the role of HPV vaccination in cervical cancer prevention. The authors highlighted the main limitations of their study, such the assessment of HPV testing over the short-term, rather than the long-term.
There were some limitations in the reporting of the methods and results, which makes it difficult to assess the authors’ conclusions.
Supported by the Instituto Mexicano del Seguro Social (IMSS); the Consejo Nacional para la Ciencia y Tecnologia; Instituto Nacional de Salud Publica, the National Institutes of Health, USA; and Digene Corporation.
Flores YN, Bishai DM, Lorincz A, Shah KV, Lazcano-Ponce E, Hernandez M, Granados-Garcia V, Perez R, Salmeron J. HPV testing for cervical cancer screening appears more cost-effective than Papanicolau cytology in Mexico. Cancer Causes and Control 2011; 22(2): 261-272
Other publications of related interest
Salmeron J, Lazcano EC, Lorincz AT, et al. Comparison of HPV-based assays with Papanicolaou smears for cervical cancer screening in Morelos State, Mexico. Cancer Causes Control 2003; 14: 505-512.
Subject indexing assigned by NLM
Adult; Aged; Aged, 80 and over; Cervical Intraepithelial Neoplasia /diagnosis /etiology /virology; Cost-Benefit Analysis; Cytological Techniques /economics /methods; Early Detection of Cancer /economics /methods; Female; Humans; Mexico; Middle Aged; Papillomaviridae /isolation & Papillomavirus Infections /complications /diagnosis; Uterine Cervical Neoplasms /diagnosis /etiology /virology; Vaginal Smears /economics /methods; Young Adult; purification /physiology
Date bibliographic record published
Date abstract record published