Prevalence multiple sclerosis has a large spatial variability, both due to environmental and genetic factors. Investigation at a relatively small scale has at least two important advantages: on the one hand, it is
the basis for monitoring disease progression, evaluating cure
strategies, and planning resource allocation; on the other hand, it
can help sort out new etiological agents.
We will provide precise estimates of prevalence of SM in each comune-area level of Italy, for each of several years. We will provide epidemiological estimates at a small spatial scale that are reliable and have been obtained with the same estimation technique (thus making them directly comparable). We will base our efforts on an hospital discharge database.
Our methods will be based on estimates of the number of diseased that have never been hospitalized (even those without a diagnosis of SM) by modeling the number of hospital admissions for each patient with at least one admission.
In pursuing our goals we will develop innovative capture-recapture methods, based on certain covariates, which will lead to additionally reduced bias and, more importantly, to increased precision (shorter confidence intervals).
Capture-recapture methods will be for the first time merged with
small-area estimation methods, as for a few areas only very few
subjects are available. Small-area estimation methods are targeted
at obtaining reliable estimates in an area even when the sample size is small, provided that after aggregation with neighboring areas the sample size is sufficiently increased.
In Kingwell et al. (2013) it is concluded that
"Despite the breadth of the literature on the epidemiology of MS in Europe, inter-study comparisons are hampered by the lack of standardization."
This study will overcome this issue for the nation of Italy, as for
the first time estimates at a relatively small spatial scale will be
provided based on the same method. Hence, they will be standardized
and directly comparable over time and space.