Information Technology Reference
In-Depth Information
scenarios without vaccination, we also found that it is necessary to stay with
short screening intervals, because it also has a big impact on results. If sexuality
would increase about 75% or more, we could observe a change in trend, and
growing number of new cancer cases in near future. Our feeling might be that
such an increase of sexuality is not too much, because it was estimated that at
this point Poles have less than 2 times lifelong sexual partners than Swedes or
Finns. Moreover aging of Polish society is also involved in growing oncological
problems. The intervention seems to be needed, because we tried to be as much
conservative as possible (model parameters mostly approach expected value from
lower side).
Our approach could be extended to be more realistic. This model takes into
account HPV just as one aggressive strain. There are at least two pathogens: type
16 and type 18 [4], which can cause cancer and existing vaccines are protecting
from both of them. A multistrain model would of course be more realistic, but at
least two times more parameters would be needed. In order to imitate multistrain
reality, we needed to increase the amount of infection, so we multiplied the
partner change rates by 2. By that we assume that properties of both strains
are the same, and that infection or cancer development have to be independent
processes. This is not true in general, but epidemiology of type 16 and type 18
is similar. Also, evolution helps us justify the independence assumption, because
those strains are fighting against each other, and it is very rare to be colonized by
both of them (after the secondary infection only one will survive). We assume
that vaccine acquired immunity is lifelong, but it is already known not to be
true [4]. The trials of this vaccine have not been running for more than 20 years
and no longitudinal studies have been done. However, some of the women have
lost their immunity after only 10 years. As suggested by geographers working in
spatial epidemiology, aspects such as education and whether one lives in urban
or rural area also have a great impact on parameters such as sexual activity,
and it would be more realistic if these aspects were taken into account as well.
However, it is dicult to find data on activity with respect to such factors, and
further disaggregation of the population would lead to a hardly tractable model.
Moreover, the model is already very complex (with more than 100 parameters),
but we did our best to validate it by sensitivity analyses of parameters and
functional of control variables (Figure 7, 6).
Acknowledgments.
We would like to thank to Anna Franzen or Fredrik Lil-
jeros from Stockholm University, Lisa Brouwers and Sharon Kuhlmann form
Swedish Institute of Disease Control and Andrzej Grabowski from National Re-
search Institute in Warsaw for cooperation. AJ thanks to Swedish Institute for
invitation to Sweden.
References
1. Garnett, G.P., Kimb, J.J., French, K., Goldie, S.J.: Modelling the impact of HPV
vaccines on cervical cancer and screening programmes. Vaccine 24S3, S3/178-
S3/186 (2006)
