Quality assessment of HPV models will be structured according to five domains – three of these will be ‘content-based’ and two will focus on methods. For each of these domains, a working group will be established. Working groups will be chaired by the HPV-FRAME coordinators who will hold regular teleconferences with working group members to facilitate the process.
This working group will focus on framework issues important for the construction of accurate and well-validated models of HPV transmission and vaccination. It will include consideration of all HPV-disease related outcomes (cancers, RRP, warts) in both females and males and of the different existing and emerging prophylactic HPV vaccine types.
This working group will focus on issues important for the construction of accurate and well-validated models of cervical screening. It will include consideration of screening, triage, diagnostic strategies, treatment for CIN and post-treatment surveillance as well as invasive cervical cancer staging and survival.
This working group will focus on issues around the dynamics of HIV/HPV co-infection, including the association between HIV acquisition in people with HPV, and the increase in HPV persistence and progression to cervical cancer in co-infected women.
This working group will focus on methods and data sources for both calibration and validation, considering a number of parameter fitting (calibration) techniques from univariate parameter fitting, to MCMC, to Bayesian methods. Validation will be considered in a framework incorporating ‘orders of validation’ from McCabe and Dixon, i.e. (in order) internal consistency, expert concurrence on model assumptions, external validation with independent data, and comparisons of predictions to real world events post implementation.
This methods-based group will focus on defining appropriate methods for assessing uncertainty in HPV modelling. Particular attention will be paid to the role of scenario and one-way sensitivity analysis as well as to partial and full probabilistic sensitivity analysis techniques.